I n . . ‘ E... :.Jx..uwfi".fin J“ ‘9! .1 and}; it... .30141: . R. 4.3, 1. . fun figurzmwwfimm.2 flit fink . . v 3 1:1 6 . 5,92 .‘wfi. f... . . . ‘ .25. ‘ .= 3 I Ifl‘l 1"?- ! H i...“ .6 ~ «T; 'm a o . 62"...”04' EV”... Ov- .- : oxhb‘dflr. :1» ’vuuo‘ I Q. t." ’57,, a: tin!!! Yttr-.V.L. .2 u .. Vidri t... in. l o 5.23.311 . . t. :1 m: it: Iii-i 5.0.. , 53:..." gammy a» I 1:: . . This is to certify that the thesis entitled THE EFFICACY OF ELECTROMAGNETIC BONE GROWTH STIMULATORS ON DELAYED AND NON-UNION FRACTURES ON META-ANALYSIS presented by Tara Angela Yeske has been accepted towards futfillment of the requirements for the MS. degree in Kinesiolggy QWW Major Professor’s Signature 05/42 /o?0/o Date MSU is an Affinnative Action/Equal Opportunity Employer LIBRARY Michigan State University be. —- ._.-——. PLACE IN RETURN BOX to remove this checkout from your record. To AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE 5/08 KlProj/AocE-PrelelRC/DateDueJndd THE EFFICACY OF ELECTROMAGNETIC BONE GROWTH STIMULATORS ON DELAYED AND NON-UNION FRACTURES: A META-ANALYSIS By Tara Angela Yeske A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Kinesiology 2010 THE? l..'.. . 1Ca ’3'“ ls‘.‘ 3.“. T3. ABSTRACT THE EFFICACY OF ELECTROMAGNETIC BONE GROWTH STIMULATORS ON DELAYED AND NON-UNION FRACTURES: A META-ANALYSIS By Tara Angela Yeske Purpose: The purpose was to determine the efficacy of electromagnetic bone growth stimulators on healing delayed and non-union fractures. Methods: All eligible studies from MEDLINE and CENTRAL were compiled and reviewed by two reviewers. Inclusion criteria included any randomized controlled or controlled clinical trial comparing a bone growth stimulator to a sham control. All disagreements between the two primary reviewers were adjudicated by a third reviewer. Abstracted data was used to estimate a relative risk. Results: Only four articles met all inclusion criteria and data was extracted for the meta- analysis. The primary finding of the meta-analysis was a summary random effect risk ratio of 2.62, with a 95% confidence interval of 0.78 to 8.78. The test of homogeneity was highly significant (Chi2 = 21.91, 3 d..f., p < 0.0001). Out of the four studies, only Barkers’s favored the control over treatment (RR = 0.91). Conclusion: The primary findings from the random effects method conclude that there was no statistically significant evidence that bone growth stimulators promote healing on delayed or non-union fractures. However the secondary analyses using a fixed effect analysis showed a similar but statistically significant effect of bone growth stimulators on healing compared to sham control. Both analyses have significant heterogeneity and a small amount of included studies. Copyright by TARA ANGELA YISSKE 2010 A successful man is one who can lay a firm foundation with the bricks others have thrtmrn at him. -l)a\-'id Brinkcly ACKNOWLEDGEMENTS I would like to thank my thesis committee Dr. Sally Nogle. Dr. John Powell. and Dr. Mat Reeves. without which none ofthis could be possible. To Dr. Sally Nogle. Associate llcad Football Athletic 'l‘rainer. Your wisdom and patience gives us all something to strive for. You presence and success in this profession sets the standard for all female athletic trainers as you continually push the barrier ofknowledge and professionalism in this career. To Dr. .lohn Powell. Head of Graduate Athletic Training Education and Research Program Director. 1 will forever be at awe of the depths of your knowledge. When faced with intellectual problems its amazing how innovative you can be. Thanks for taking a chance on me. To Dr. Mat Reeves. Associate Professor of Epidemiology. Words cannot express the thanks that l owe you. You were a lighthouse through the storm that was my thesis. Thank you for being such a guiding force through all ofthe Obstacles and chaos of ideas. To Peter .lajou. Medical Student and Second Reviewer. What right-minded medical student would sign on to read over four hundred articles with no compensation. you would. 1 respect your inquisitiveness and eagerness to obtain knowledge wherever you can get it. Thank you for your perseverance. To John (.‘offey. Liaison to the College of Human Medicine and Research Specialist. You truly are the master of your domain. No one can abbreviate like you. Thanks for going above and beyond your job to help me out. l would have been wandering around the library without you. To .loAnn James. Graduate Studies Secretary: Thank you for helping with all the strategic legwork. You are always a positive force encrmraging all grad students. Thanks for that extra little push at the end. To Pete Morano Department Head ofAthletic 'l‘raining. Central Connecticut State University: I look back and I don‘t know how I got here. You educated me when I was as green as they get. You have always been there pushing me through things that I would have never thought I could do in the first place. Thank you for believing in me. even in times when I didn‘t believe in myself. To Virginia. you are one ofthe strongest. most driven. and intelligent people I know. No matter what situation your in. you always strive to make it better. You set by example that through anything. there is no reason why we cannot accomplish things we have set out to do. And yes. for the record. you did close me. To Jamie. only you are given the right to say they understand what l‘ve been through. It’s crazy how we lead parallel iives. Remember the plan - warm place. Thanks for being at the other end of the phone line. I truly would not have made it through this without you. 1 am proud of you. and I can't wait to be there when you finish. To Dustin. ifl lose you tomorrow. I still am the luckiest person in the Uret any decision that l made. C world to have had you in my life. l have never re vi Thanks for always being understanding. taking care of me in hospitals. giving me another brain to bounce ideas ofl‘of. and for being really objective. I wish you all the luck completing law school and remember. you hold the rest ofyour life in your hands. Decide to make it what you want it to be. To Tai. your support and nightly laughs were greatly needed and appreciated. Through the process of writing this l have been fortunate enough to watch you start your own family. Thanks for creating a home that I know I can always come back to. To Mom. we took a big risk in this one. Thank (iod it worked out. But even ifit didn‘t I know you would have always been there. Your support has been the foundation in everything that l have done in my life. Thank you for being with me. every step ofthe way. To Dad. you are one of a kind. It was through your dreams that I learned what I truly wanted to do. Continue to reach for the stars and your dream will come true. . In. '7'“; “—1.- .- TABLE OF CONTENTS LIST OF TABLES ..................................................................... x LIST OF FIGURES .................................................................... xi CHAPTER I INTRODUCTION ..................................................................... I Overview ofthe Problem ..................................................... 1 Significance of the Problem .................................................. 3 Problem Statement and Research Plan ..................................... 4 Need for the Study ............................................................ 5 Research Questions ............................................................ 7 Definition of Terms ........................................................... 8 CIIAPTER 2 LITERATURE REVIEW ............................................................ 10 Anatomy and Physiology of Bone ......................................... 10 The Process ofOsteochondral Regeneration .............................. l2 Factors Affecting Bone Regeneration ..................................... 13 History of Electromagnetics on Remodeling Bones ..................... 15 Bioelectrical Properties in Bone ............................................. 16 Scientific Theories Behind Electromagnetic Stimulators ................ 19 Research on Electromagnetic Stimulators ................................. 20 Performing a Systematic Renew 23 Performing a Meta-Analys1s23 Published Meta-Analyses................................................... 23 .ladad QualityAssessmentTool........................................... .25 C ochrane Quality Assessment .............................................. 26 Radiographic Evidence ....................................................... 28 Review Manager Analysis Program ........................................ 28 CHAPTER 3 METHODS ............................................................................ 30 Search Methods ............................................................... 3] Screening Process .32 Eligibility Criteria ............................................................. 32 Definition of Delayed and Non—Union Fractures ......................... 33 Types ofControl 33 Healthy Population ........................................................... 34 Randomized and Controlled Clinical Trials .............................. 34 INCLUSION CRITERIA Types of Fractures ........................................................... 35 Radiographic Outcome and Diagnosis.....................................35 Types of Intervention ........................................................ 35 Quality Assessment Procedures ............................................ 36 viii Data Extraction ............................................................... 36 Exclusion Criteria ............................................................ 37 Statistical Methods of Data Analysis ...................................... 3 Assessment of Heterogeneity ............................................... 3 " CHAPTER 4 RESULTS .............................................................................. 40 Selection oflncluded Studies 40 Detailed Review: Article Demographics .................................. .41 Detailed Review: Individual Study Methods ............................. 43 Meta-analysis Results 47 Subgroup Analyses .......................................................... 4‘) Sensitivity Analyses ......................................................... 4‘) CHAPTER 5 Statistical Imdmgs . 52 Comparison of Random liffect and Fixed Effects Methods ............. 54 Importance ofBetween Study Heterogeneity..............................5() Meta—analysis Methods ...................................................... .57 Limitations of Inclusion (ntcna58 Comparison by Quality Assessments.......................................5‘) Randomized Trail versus Controlled Clinical Trial ....................... 59 Blinding Outcome Assessors ............................................... .6] Conclusion ...................................................................... 6] APPENDICES A. Quality Assessments ................................................... 77 B. Meta-Analysis Forms ................................................... 82 REFERENCES ....................................................................... ()2 Table 1 Table IQ Table 2 \J Table 4 Table 5 LIST OF TABLES Comparisons of Meta-Analyses using Electroimtgnetic Field Stimulation ................................................................. 63 Final Relevancy Screening Form ....................................... 65 Reasons for Article Exclusion .......................................... 68 Quality Assessment Comparison ...................................... 69 Meta-analysis Statistical Results ...................................... 70 Figure I Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 LIST OF FIGURES Random Effects Anlaysis Risk Ratio .............................. 71 Fixed Effects Analysis Risk Ratio ................................... 72 Sensitivity Analysis Drop Barker .................................... 73 Sensitivity Analysis Drop Scott ...................................... 74 Sensitivity ./\nalysis Drop Sharrard ................................. 75 Sensitivity Analysis Drop Simonis .................................. 76 xi CHAPTER I INTRODUCTION Overview oft/re Prob/cm Delayed union is. "the cessation ofthe periosteal response before the fracture successfully has been bridged." (Marsh. 1908). Fractures that are delayed unions have a retardation ofthe normal process of bony consolidation. Delayed union does not imply that the fracture will never heal. just that the process is slowed. Nonunion fractures are defined as being. "A bone that fails to unite or heal completely. Diagnosis of nonunion is established when a minimum of nine months have elapsed since the injury and the fracture site shows no progressive signs of healing for a minimum ofthree months without a complication from synovial psedoarthrosis." ('Venes & Thomas. l997). Fractures ofthis type are not expected to heal without further intervention such as surgery. Diagnosis ofnonunion or delayed union fractures is made on clinical symptoms including pain. range of motion. and radiological evidence ofcallus formation (Panagiotis. 2006). There are many reasons why fractures fail to heal. These risk factors include inadequate immobilization. comminuted and devascularized bone. poor vaseularity of fracture fragments and surrounding soft tissue. infection. prior irradiation. presence of bony defects. systemic factors such as malnutrition or chronic illness. medical related conditions. and smoking (Saleh & Hak. 2001). Nonunion and delayed union can be treated either surgically or non-surgically. Bone grafts. internal fixation. plate fixation. and intramcdullary nailing are examples of invasive surgical techniques commonly used for fractures (Panagiotis. 2006). Osteoinductive molecules and external electromagnetic bone stimulators are two new. less invasive options for nonunion therapy. Shortly after the first report of using oscillating electromagnetic fields to treat nonunions in 1983. the Food and Drug Administration approved the use of pulsed electromagnetic fields for clinical use for treating nonunion fractures in humans (Otter et al.. 1998). Health insurance companies have already added these devices and apprmed them for billing oftreatment. In a statement put out by Aetna (2006). they cite direct current electrical bone growth stimulators medically necessary for conditions including non-unions. failed fusions. congenital pscudartln'osis. delayed union. and for patients at high risk for spinal fusion. Insurance companies are pushing for the use of bone growth stimulatm's in order to avoid open reductions or bone grafts that would. in consequence. accrue further costs. Clinicians. as a profession. must continually educate themselves on current science by reading various scientific journals. They are bombarded by a plethora of academic articles and must filter through them and judge which ones are truly scientifically sound. Electromagnetic bone growth stimulators are a relatively recent modality and there are conflicting results from various scientific studies disputing their efficacy in regards to bone regeneration. This study will provide answers about bone growth stimulators using only the strongest scientifically based studies. and allow the clinician to make an educated decision of whether to use to modality in practice. Not only will new statistics be generated. but readers can also use the meta-analysis as a tool to quickly review all the scientifically sound articles on the topic. In order to understand the actual efficacy of bone growth stimulators on nonunion and delayed union fractures. a systematic review must be conducted on the current literature. Significance (if/lie Prob/cm Five to ten percent of fractures become delayed unions and approximately one percent becomes nonunion fractures (liinhorn. 1995). With lower extremity fractures. the risk of delayed or non-union increase dramatically. with 29% becoming delayed unions and 109-0 becoming nonunion fractures (Coosemans ct al.. 1988). Delayed and non union rates for open tibial shaft fractures range from 16-60% in lower grade fractures to 43-100% in more severe open injuries (Caudle & Stern. 1987: Sanders et al.. 109-1: Reimer ct al.. 1095). Ifa delayed union or nonunion fracture is left untreated it can result in arthritis in the joint. loss of range of motion. prolonged hospital stays. multiple operations like plates. internal fixation. and bone grafts. long periods of immobilization. increase expenses. and decrease in the quality of life. Fractures that heal in the malunited position may require osteotomy to correct alignment or rotation (Younger & Chapman. 1980). “Particularly severe injuries may become recalcitrant nonunions. requiring multiple operative interventions over months to years to achieve union." (Saleh & Hak. 2001 ). Only one meta analysis looking at the effectiveness ofelectromagnetic Uistered h- kw bone growth stimulators on delayed union and non union fractures was re on the Cochrane Library. and merely few have been published in journals. This study titled [flee/ruinugirelic FieldSIimu/ulirmfor I/zc 'li'culmenl och/rn'crl J Union of'A'on-anion oflong bones by Punt et al. (2004) is an unfinished. ongoing study which only reviews long bones and neglects to address the many predisposed anatomical sites with poor blood supply. There is currently a controversy over whether bone growth stimulators are effective and specifically for what conditions. Effect of'lilectric Stinntlation o/JIttsenloskeletcil SIN/UNIS) .‘f Illeta-Analvsis o/(‘ontrollecl ('linical Trials by Akai and Hayashi (2002) proves there are positive effects when used for tissue healing. Table 1 compares these meta-analyses along with the studies lz'lectromagtzetic l’ielrlsfor the Treatment of ()steoartlzritis by Hulme and associates and Electric Stimulation and ll_\'perl)aric oxygen therapy in the treatment of.\'onanion.v by Karamitros. Kalentzos. and Soucacos (2002. 2006). To date there is not one large definitive trial on the efficacy of bone growth stimulators for delayed and non-union fractures. Because there are so many scientific studies being published on this topic. it is difficult to sift through the contents and discern which studies are truly scientifically sound in order to compound a true result. The studies that have been published are outdated with the most recent collection of studies being from 2001. This gap in literature fails to identify current studies and will be filled by this meta-analysis. Problem Statement and Research Plan The purpose ofthis systematic review is to evaluate the efficacy of electromagnetic bone growth stimulators on delayed union and nonunion fractures at all sites based on randomized controlled trials conducted on humans. For this review delayed union fractures will be classified as no healing at the site with in three or more months and min-unions defined as failure to unite beyond six months. All articles will go through a three step screening process. Article titles found in the search will be initially screened. Ifthe title and abstract appears relevant to the review. the article will then be read in order to identify ifit meets eligibility criteria. Only randomized and controlled clinical trials will be considered. Randomized controlled trials are clinical trials that have a control treatment allocated by a random process. A controlled clinical trial is any study that allocates by a pseudo-random process like coin flips or social security numbers. Ifthe study meets these criteria. data extraction will take place and it will then be included in the review. The quality ofstudies will then be assessed according to the .ladad score and Cochrane approach to quality assessment for inclusion. See Appendix A for the assessment characteristics. ."\'eecl./(')r the Study There have been numerous studies done on stimulating bone growth with clectromagnetics and ultrasound techniques although only a few were performed using the randomized clinical trial design. Ofthe four previous published reviews. only one by Punt and associates (2004) concentrated on bone growth stimulators and their efficacy with delayed and non-union fractures. This same review is being researched on long bones including the tibia. fibula. femur. and radius. Other sites that traditionally have poor blood supply like the scaphoid. are not included. This study involves pulsed and non-pulsed electromagnetic bone growth stimulators and is currently being conducted. A meta-analysis by Akai and Hayashi (2002') used pulsed electromagnetic fields to test efficacy on bone healing and soft tissue healing. Because ofthe criteria. only studies published 'JI "from 1966-1999 were included. This leaves a time gap where the intervention has evolved. More current studies need to be compounded in order to analyze the true efficacy of bone growth stimulators. The methodology ofthis review researches two different cfficacies. which skew the results ofthe true effectiveness on non- union and delayed union fractures (Punt et al.. 2004: Akai & Hayashi. 2002). Out ofthe forty-nine studies reviewed. only twenty-eight were used in the meta- analysis. a limited amount of studies. This study did not disclose their statistics analyzed by ROMA 88. Their reported general findings showed no proofthat bone growth stimulators had a specific effect on health. but had positive findings when used on tissue repair. Another metaamalysis conducted by l-lulme and associates (2002) focused on pulsed electromagnetic fields and their efficacy healing osteoarthritis. All randomized clinical trails and controlled clinical trials published before 2001 were used. Even though controlled clinical trials were included. only three out of 102 studies met the inclusion criteria (Hulme et al.. 2002). Due to the low number of literature. no clinically important results were produced. A review titled lilectric Stimulation anrl lltperharic (Ivygen Them/{v in the Treatment of.\"otnations evaluated studies that used electric stimulation in combination with hyperbaric oxygen (Karamitros et al.. 2006). The results of pulsed electromagnetic fields alone cannot be determined from this study because of its use in combination with hyperbaric chambers. Being the most recent systematic review conducted. no new information was generated because a meta—analysis was never ran. The problem why previous studies have not efficiently addressed the topic includes a wide distribution of various types of injuries. various types of outcome assessments. small number ofincluded studies. and a time gap ofcurrent literature exists. Also many ofthe reviews do not consider selection bias by only searching for journals offelectronic databases and with articles that are only published in English. This study will concentrate on the efficacy electrrmutgnetic stimulation has on delayed union and nonunion fractures at all sites. All randomized and controlled clinical trials published until the present date will be included in the review. Only studies testing bone growth stimulators will be used. This includes direct current. PEMFs. and capacitive coupling. The results ofthe review will provide a scientifically sound consensus on bone growth stimulators and will aid the decision of whether to utilize the modality in general practice for clinicians in the healthcare field. Once the systematic review is completed. an attempt at a meta- analysis to get a summary result will be made. Research Questions With all the conflicting data on efficacy of bone growth stimulators regenerating bone. this meta-analysis will compare only the studies ofthe highest scientific nature by using randomized and controlled clinical trials in order to quantify the true benefit ofthe modality. The statistical analysis will be concentrated on the following questions: I. What is the efficacy ofelectrotmtgnetic bone growth stimulators on delayed union and non—union fractures at all sites? 2. Are there differences in results ofstudies performed as randomized controlled trails compared to clinical controlled trials‘.’ 3. Does the blinding of outcome assessors affect the results? 4. Does the methodological quality of studies affect the outcome when looking at both sensitivity analyses? Definition of'l'erms. Delaved union: an ununited fracture that continues to show progress towards healing or that has not been present long enough to satisfy an arbitrary time standard for nonunion (Phieffer & Gould. 2006). It is usually diagnosed when there is failure to see normal healing ofthe bone on radiographic evidence within three to six months ofthe injury depending on the fracture site (Punt et al.. 2004). Electric Stimulation: there are three distinct forms. direct electric current. pulsing electromagnetic fields (PEMI’s). or capacitive coupled electric energy. The direct electric current uses a generator to deliver electric energy by surgically implanted electrodes into the fusion bed. PFMFs are a time varying current that travels through metallic coils at a certain duration and intensity. Capacitive coupling charges two metal plates that are attached to a voltage source and produces electrical field. Both PEMFs and capacitive coupling utilize electrodes (Oishi & Onesti. 2000). Meta Analvsis: a statistical teclmiquc for combining the results ofa number of individual studies to produce a summary result. It is not synonymous with a systematic review (Khan et al.. 2003). Nonunion: when the normal biological healing process of the bone ceases so that complete healing will not be achieved without further treatment (Pheiffer & Gould. 2006). The United States Food and Drug Administratirm quantities a nonunion as a fracture that has occurred at least nine months previously and has not shown any radiograph signs of progression towards healing for three consecutive months (LaVelle. 1998). Pseudoarthritis: the formation ofa false joint where the fibre cartilaginous cavity is lined with synovium and produces synovial fluid (Panagiotis. 2006). Radiological Criteria: following x-rays. radiological signs as evidence of union of a fracture such as loss ofdistinction at the fracture gap. cortical bridging. and trabecular bridging (Simonis et al.. 2002). Svstematic Review: a method for reviewing and evaluating scientific literature. The review evaluates and interprets all available research relevant to a particular question. It identifies all relevant primary research and standardizes study quality so that only studies of acceptable quality are synthesized (Glasziou et al.. 2001). C I IAPTER 2 LITERATURE REVIEW Non union and delayed union fractures are a common occurrence among bones with poor blood supply. Traditionally in the past only surgical Options were considered for repair. Just recently have modalities like bone growth stimulators and ultrasound been used in an attempt to enhance bone growth and fracture repair. In order to understand the scientific basis behind the modality. the history of bone growth stimulators will be discussed. Also literature will be reviewed regarding the process of using meta-analysis as a research tool. Anatomy and Physiology ofBone Bones are organs that contain several different tissues including osseous tissue. nervous tissue. cartilage. muscle. and epithelial tissue. There are two parts to the bone. the compact bone which is the dense outer layer and the cancellous or spongy bone which lies internal to the compact bone (Marieb. 2001). The entire surface on the outside of the bone is protected by a double layer membrane called the periosteum. Dense. irregular connective tissues make up the first layer ofthe periosteum. The inner ostcogenic layer is made up of osteoblasts. bone forming cells. and osteoclasts. bone reabsorbing cells (\N’atson. 1979). The periosteum is supplied with nerve. lymph. and blood vessels which enter the bone by the nutrient foramen (Marieb. 2001 ). The endosteum is a connective tissue that covers internal bone surfaces including the cancellous bone and marrow which also contains osteoblasts and osteoclasts. 10 '32—!!! Bone remodeling is caused by bone deposit and resorption in the periosteal and endostcal surfaces (Marieb. 2001 ). The actual process that triggers calcification ofthe matrix is contrmersial although researchers do know that in order for bone deposits to be made. there are key products needed. When calcium and phosphate reach a certain level. tiny crystals are formed. Another product required for calcification is matrix proteins in order to bind the calcium. Alkaline phosphate which is shed by osteoblasts is a key ingredient for mineralization (Marieb. 2001). During bone resorption. osteoclasts from stem calls in the bone are turned into macrophages. These macrophages move around the bone surface and dig pits called resorption bays as they break down the matrix. The outside ofthe osteoclasts secrete lysosomal enzymes and acid. These two substances aid in digesting the matrix and converting calcium and salt into soluble forms that can be transported through the blood (Marieb. 2001 ). There are two systems in the body that control bone remodeling. These are the negative feedback hormonal mechanism and the body‘s response to mechanical and gravitatiomrl forces on the skeleton. The hormonal mechanism is an interaction between the parathyroid hormone (PTH) and calcitonin. PTII is released when blood calcium levels decrease and stimulate osteoclasts to reabsorb bone which in turn releases more calcium into the system. Calcitonin is secreted when blood calcium levels increase and inhibits bone reabsorption. It also encourages calcium deposits in the bone matrix reducing blood calcium levels. 9 These two elements monitor the body‘s calcium level to keep it at a consistent homeostasis. The body‘s response to mechanical stress and gravitation forces is based offof Woolfs law which states that bones grow or remodel in response to the forces placed on it. In this mechanism the forces are muscle pull and gravity (Marieb. 2001). When bone is loaded it bends and causes compression forces on one side and tension forces on the opposite. Deforming bone causes an electrical current with the compression and tension regions ofthe bone being oppositely charged. This suggests that electrical signals direct the remodeling process. Bone tissue tends to deposit in negatively charged regions while absorption occurs in areas of positive charges. The theory ofthis mechanism is that electrical fields prevent PTH from stimulating osteoclasts which decrease bone absorption at. the site (Marieb 2001). The Process of'()steochomlral Regeneration The healing time for simple fractures can be six to eight weeks or longer for large weight bearing bones and complicated fractures. There are four major phases ofthe healing process. The first phase is hematoma formation. After the fracture blood vessels in the bone. periosteum. and tissue are torn and hemorrhage making a massive clot ofblood at the fracture site. The bone cells around the fracture site are deprived of nutrients and die. This causes the tissue to become swollen. painful. and inflamed. Fibrocartilaginous callus formation is the second phase. Within a few days ofthe fracture a callus from the thrombus is formed out of fibroblasts. collagen. proteoglycans. and chondroeytes (Otter et al.. 1998). Capillaries grow and phagocytes clean the hematoma. Fibroblasts and osteoblasts are sent to the fracture site to rebuild by forming spongy bone and making collagen fibers. The third phase is named the bony callus formation. During this phase new trabeculae bone in the callus starts to harden. This phase initiates three to four weeks after the injury and ends in a firm union in two to three months (Marieb. 2001). Bone remodeling. the last phase. begins during the third phase and continues for several months after the bony callus is remodeled. Excess material is removed and compact bone is laid down to reconstruct the shaft walls. In the end the bone resembles its original structure before the fracture. With normal healing. over time the callus is invaded by blood vessels and the cartilaginous material is removed while the tissue becomes calcified. However this process does not occur with a true nonunion fracture. It is at this time in the bone regeneration process that many clinicians choose to surgically treat the fracture before it becomes a delayed or nonunion. Factors Affecting Bone Regeneration Bones fail to unite not just because they are poorly vascular or were immobilized improperly. There may be underlying factors that inhibit the bone regeneration process itself. Chronological age plays a role in bone regeneration. It is proven that it takes bones longer to heal in the elderly because they have poor circulation (Marieb. 2001). There are many pathological conditions that interfere with bone growth. There have been examples of infected growth plates where the plate is partially or completely destroyed by the infection process (I lall. I990). During the growth phase infections can inhibit or stimulate abnormal intramcmbranous ossification. Children with anemia may have problems with bone growth. Their body has an abnormal increased need ofspace for hematopoietic tissue. This causes an increase in marrow at the expense ofcortical bone and is commonly found in long bones and the skull. The replacement of compact bone leaves it with a porous appearance like dry bone (Hall. 1990). Metabolic processes like hyperthyroidism causes an abnormal pattern in bone growth. In a healthy body. the thyroid hormone stimulates replacement of cartilage in long bones with bone tissue. The amount ofthyroid hormone secreted is inadequate for normal bone growth in a person with hyperthyroidism. The pituitary gland secretes growth hormone which is responsible for the division of cartilage cells in epiphyseal disks. Without this hormone. the long bones fail to develop causing dwarfism (Marieb. 2001 ). Other diseases associated with abnormal bone formation include osteomalacia where bone is inadequately mineralized. osteoporosis where bone resorption out paces bone deposits. and Pagets disease where Paget bone replaces the marrow cavity. Diabetes and malnutrition have been linked to the process of bone growth. For optimal bone growth the body needs proteins and minerals including calcium phosphate. magnesium. and manganese. Vitamin D aids in the absorption of dietary calcium. Vitamin A balances the deposit and removal of bone by being involved in osteoblast and osteoclast activities. Collagen synthesis requires Vitamin C. Without it osteoblasts produce less collagen in the intercellular 14 material of the bone tissue resulting in bones that are abnormally slender and fragile. History of’Electromagnelics on Remodeling Bones Electromagnetism. the physics ofthe electromagnetic field. is responsible for the interaction of atoms. which make up the main foundation for biology and medical field. The 1950's marked an era when the first piezoelectric properties were reported in Japan by Fukada and Yasuda (Ryaby. 1998). It was also during this time that bone tissue was discovered to have electric properties. Other research was also conducted by Bassett and F riedenberg ( 1962) on the ostcogenesis influence of bioeleetrieal properties of bone. F riedenberg (1966) was the first researcher to successfully treat a non-union ankle in their clinic with electromagnetic stimulation. In 1957. Fukada discovered that mechanical loading of the bone generates an electrical potential in the bone tissue (Fukada & Yasuda. 1957). Ile found that when a bone is stressed. there is an electropositive charge on the convex side and an electroncgativc charge on the concave side. This finding coincides with Wolff’s law that bone remodels at areas ofcompression and reabsorbs at areas of tension (Liboff. 2006). Studies prove that walking produces mechanical strains lower then 10 Hz while postural muscle activity produces higher frequencies 20 to 30 Hz (Antonsson & Mann. 1985). Iilectronegative charges can be found at sites with active bone growth including growth plates and epiphysis because osteoblasts are activated by negative charges. Research has shown that bone Ur growth peaks around the cathode. the negative electrode. and decreases around the anode. the positive electrode (Liboff. 2006). Bioelectrical Properties in Bone Potential differences exist in all living tissue. A steady resting potential from microvolts to more then one hundred millivolts has been recorded in biological systems. These differences undergo changes with metabolic processes. injury. illness. mechanical stress. and different states of consciousness (Friedenberg & Brighton. 1966). There are many natural origins ofdirect currents found in tissues including ionic gradients and ion transfers across membranes. structure oftissue with polarized molecules. the semiconductor mechanism in tissues. and the cell metabolism itself. Strained electric potential signals for regulation ofcellular processes such as bone repair and remodeling. Since all tissues are subjected to dynamical mechanical stress. they may use these electrical signals as a regulation component in the rnutinous and repair oftissue function (Ryaby. 1998). For example. bone and cartilage are mechanosensistive. Electric properties from the mechanical load cause streaming potentials produced from fluid flowing through the charged extra cellular matrix. These potentials inform the cells to alter skeletal remodeling due to the changing load (Ryaby. 1998). Friedenberg and Brighton (1966). conducted a study to measure the electrical potential in human bones. Skin potential differences were taken with electrodes from the leg and thigh in thirty six humans. sixteen of which had healing tibial fractures. They found a typical curve pattern in the electricity of healthy bones. In non-fractured bones. the epiphysis was positively charged with respect to the 16 subepiphysis and the metaphysis was negative in respect to the epiphysis. In fractured bones the entire shaft down the epiphysis was electrically negativ ’ and a secondary increase in electric negativity was found over the fracture site. With healing ofthe fracture. the electric curve returns to normal. The fact that electricity at the fracture site is markedly negative supports that either bone growth is stimulates negative charges. or that the electrical negativity at the fracture site was a result ofthe fracture. This curve pattern is consistently similar on the skin and periosteum to the curve on the bone and indexes the magnitude and disposition ofdirect currents in bones. This phenomenon was also observed in mammalian and amphibian bones. Bassett and Becker ( 1962) conducted a study which put two electrodes. one posterior and one anterior to monitor the electric activity ofamphibian bones being bowed. The pressure was applied so the bone would deform concave posterior. When pressure was exerted. immediately the posterior electrode recorded negative electricity that slowly decreased until the force was removed. After the force ceased. the anterior electrode briefly became negative. Deformation in the opposite direction was applied and the same results were observed. The results confirm that these potentials were caused either by piezoelectric properties or a displacement potential. The displacement potential occurs when a number of molecules are bent in the same way displacing fiec charge carriers from in to outside the molecules. Th ‘ amplitude ofthe electrical potential depends on the rage and magnitude of bony deformation. The polarity is determined by the direction of bending (Bassett & Becker. 1962). When bone deforms osteoclasts inundate the surface of tension while osteoblasts swarm the area ofcompression. This happens so bone remodels in a better way to cope with the stress. It is believed that mechanical deformation causes potentials in bone because direct current has been Uenesis. and reUeneration bx v. 9 5a linked with cellular migration. tumor formation. morpho of amphibian limbs (Bassett & Becker. 1962). Since stress potentials from walking or deforming bone influence the activities of bone cells. it is possible that electricity is the underlying factor between mechanical stress and bone remodeling (Watson. 1979 ). Many studies that have been performed prove that electromagnetic fields modify the bone growth process. Norton (1974) completed research on rooster chicks that show bone growth orients to the positive electrode when the induced charge on the bone is negative. In his studies Friedenberg ( 1966) observed that stimulated bone formation surrounds the cathode and radiates out in a manner that coincides with electrical field lines. The release of hydrogen at the cathode gen and makes hydroxyl radicals that contribute k, occurs instead ofconsuming oxy to the alkalinity ofthe tissue making it more favorable to calcification (Watson. 1979). There are various theories behind the origins ofelectric potential in bone. One theory is based on piezoelectricity. This is defined as the generation of electrical current from the stress of bone crystals and is dependent on the rate and magnitude formation (Friedenberg & Brighton. 1966). This theory supports why the concave side of bones are negative compared to the convex. The streaming potential is an alternate theory. This states that electrical potentials are a function ofthe rate of fluid flow in the growth area due to different tissues being comprised of dissimilar corrrponents with inherent surface charges. For example. when a bone bends small channels in the cortex deform which causes movement of liquid towards the surface of tension. Ifthe mobile ions are positive. then the surface under tension will turn positive. This theory explains why measurements of bone potential vary with the rate of load on application. When outside electrical potentials stimulate bone it is hypothesized that one of two things happen physiologically. The electrical field may prevent the parathyroid hormone from stimulating osteoclasts at the fracture site. which decreases bone absorption leading to the formation of more bony tissue. Or the electrical fields induce products ofgrowth factors. which stimulate osteoblasts (Marieb. 2001 ). There are many factors to consider for successful clinical application ofelectric stimulation. Friedenberg believes that current density and voltage or field strength is involved while Lavine and Connolly believe that the passage ofelectrical current across the fracture gap stimulates healing (Watson. 1979). It is believed that a threshold ofenergy needs to be produced. below which no regeneration occurs. Bassett et al. ( 1977) stress the importance of pulse parameters and concluded that the field strength at a nonunion is 1.2-1.6 mV/cm. Scientific llteot‘ics Behind [fleclromagnetic Stimulators Several theories explain the potential bone healing qualities of electromagnetic stimulators. These include strained general electric potentials beinU signals used for regulating cell processes including bone regeneration and C 19 remodeling. This suggests that all tissues that get stressed may use electrical signals in the maintenance and repair oftissues (Ryaby. 1998). Bone growth stimulators are a device made to mimic this electrical signal. There are three different ways to use the stimulators to treat nonunion and delayed union fractures. Direct current stimulation is conducted with an implanted electrode and is the most invasive of the three. Electromagnetic stimulators work by inductive coupling that uses time varying magnetic fields that emit a pulsed electrical current through a coil placed on the fracture site (Punt et al.. 2004). Capacitive coupling stimulators use opposing electrodes that are also placed over the fracture site (Liboff. 2006). Both capacitive and inductive coupling are noninvasive. Electromagnetic stimulators also vary in being pulsed or non-pulsed. These pulsed waveforms have become the standard for the bone growth stimulator industry because the pulses mimic higher frequency potentials that are seen during impact loading during bone tissue (McLeold et al.. 1995). Studies have shown that pulsed electromagnetic currents can differentiate bone cells. reduce osteoclast absorption. increase vaseularity. and increase the rate of osteoblasts in bone formation (Luben. I991 ). A study conducted by Robert Luben (I991) chronicles the hypothetical molecule mechanism that accounts for the effects of low energy electromagnetic fields on bone cell metabolism and it effects on hormone regulation of osteoblast function and differentiation. Research on Electromagnetic Stimulators 20 Research has shown successful results for all three stimulators. Brighton et al. (1981) reported an 84% healing rate with nonunions using direct current. This clinical study initiated at the University of Pennsylvania and expanded through the United States. included 175 patients with 178 nonunions. To be included. the patients had serial roentgerrograms to diagnose their nonunions. Nonunion was defined as the absence of progressive signs of healing for over five months. An implanted direct current stimulator with four cathodes was used for a period oftw'elve weeks. The intensity was set at twenty microampcres of direct current. Ofthe I78 nomrnions. 149 achieved solid bone healing (Brighton et al.. 1981). Along with being an invasive technique. the other negative aspects to using direct current is the by products ofconsumption ofdissolved oxygen and an increase in pH levels at the electrode-tissue surface (Otter et al.. 1998). Pericles Diniz and colleagues (2002) studied how PEMF stimulation affected osteoblasts in different stages of maturation to see ifthe number ofcells or differentiation was changed. Using cell cultures. cell proliferation. differerrtiation. and area of mineralized matrix was measured after being stimulated. The researchers found that PEMF affected the osteoblasts in early stages ofthe culture during cell proliferation and differentiation by increasing bone tissue-like formation. Although the stimulatory effect during the mineralization stage decreased bone tissue-like formation. Overall the stimulatory effect was nrost associated with enhancing cellular differentiation but not increasing the amount of cells (Diniz et al.. 2002). In a study conducted on delayed unions w ith incomplete healing after 16 to 32 weeks. Sharrard (1990) showed that PEMFs provided a substantial benefit than a surgical intervention. This double blind trial of Pl—ilVlFs was performed on patients with tibial osteotomies and resulted in doubling the number of patients at advanced stages of healing within the first 60 days oftrcatment. Sharrard treated forty-five patients with tibial shaft fractures. All were immobilized in plaster and given electromagnetic stimulation units for twelve weeks. Twenty patients had an active unit while twerrty-five patients were fitted with dummy control units. A radiologist‘s assessment concluded five unions. five progress to unions. and ten with no progress to union in the active group. In the control group there was one union. one progress towards union. and no progress in twenty three. An orthopedic surgeon‘s assessment showed union in nine fractures and absence in eleven ofthe active group. The control group presented with three unions. twenty-two without unions. These results were significantly in favor of the active group. p=0.02 (Sharrard. 1990). Even though effective experiments have been reported. there still remain many questions about the science and physiology behind the bone growth stimulators. Currently it is not known the degree that applied magnetic field or induced electrical fields are responsible for the biological response (Punt et al.. 2004). Since the stimulators range on an electrical spectrum from 1 Hz to 1 MI 12. the most beneficial frequency is still debated. (.‘urrently the literature concludes that electric frequencies at or lower then 120 Hz are maximally responsive to bone remodeling activity (McLeod & Rubin. 1990). 7") ~— Performing a Systematic Review Research that is relevant to the health care industry is scattered among a plethora ofjournals. It is the professionals responsibility to determine which journals are scientifically sound and choose the research from those articles to help inform their clinical practice. It is difficult to keep well informed on many topics particularly those issues that are controversial. The systematic review is a technique that attempts to identify all relevant articles on a topic. Relevant studies are categorized according by their study design and quality. Systematic reviews focus on a single question. in this case the efficacy of bone growth stimulators. Ifenough similar articles can be found that present quantifiable data then a meta-analysis. the process of calculating a summary effect estimate. can be generated. Performing a .lleta—Analvsis A meta-analysis combines the results of several studies ofa related research hypothesis. Each study is weighted according to size and sometimes trial quality in order to generate a summary effect estimate. Its purpose extends beyond simply combining the effect from a group of studies. Meta-analyses can also identify important variations between studies i.e.. heterogeneity and can explore origins ofthis variability. Not all systematic reviews will lead to a meta- analysis. Ifthere is insufficient data or studies are too dissimilar then a meta- analysis may not be a logical choice. Published rlleta—A nalr "SC’S In order to create the screening process for inclusion of articles. various published meta-analyses on bone growth stimulators were researched. The one protocol most similar to this study was conducted by Punt. den lloed. and Stijnen (2004). Their study. titled Electronicignetic Field Stimulationfor the Treatment of Delaved (Jilin)? or .’\"on-union o/Long Bones. used pulsed and non pulsed electromagnetic fields on fractures of long bones. Many elements of this study's protocol are derived from it. A review titled Electric Stimulation and llryierharic ().\'_rgen Therapy in the Treatment o/"A'onunions used electric stimulation in combination with other therapy in order to heal delayed and non-union fractures (Karamitros. Kalentzos. & Soucacos. 2006). Pulsed electromagnetic fields. direct current. and hyperbaric oxygen were some of the interventions used. Since this study was a review. a meta-analysis was not conducted and no statistics were gathered. Researches Akai and Hayashi used articles from 1966-1999 in order to test the effects of pulsed electromagnetic fields on bones. soft tissue. and joints (2002). Their study was scientifically sound and masked reviewers during data extraction. In their meta-analysis twelve out of twenty studies were reviewed on bones and sixteen out oftwenty-nine studies were reviewed on soft tissue and joints. Their results show positive findings when PEM F is used on tissue repair although have no specific effect on health or bone healing. lz'lectromagnetic Fields/or the Treatment of'()steoarthritis focused on the use of PEMF to treat osteoarthritis (Hulme et al.. 2002 ). Because (.isteoarthritis is defined as deoencration ofcartilage within a joint. this wearing down of bone & cells can be related to a bone fracture. Out of 102 studies only three met the inclusion criteria for this meta-analysis. No clinically important results were generated due to the low number of literature. .ladad Quality.41ssessment Tool Numerous scales and checklists have been developed to evaluate the quality of randomized clinical trials. The .ladad Scale. originally used to assess the quality of articles on pain relief. has been extensively used to assess study quality in other clinical areas. In a study published by Moja (2005) examining 965 systematic reviews. the most commonly used tool for quality assessment was the .laclad scale. It is a scale that is recommended by the Cochrane Musculoskeletal Group in the preparation oftheir Cochrane systematic reviews (Towheed. 2006). The .ladad Quality Assessment is a seven question scale that measures study design and reporting quality with a numerical score from 0-5. with zero being the weakest and five being the strongest (.ladad. 1996). The first five items are indications of good quality with each counting as one point towards the overall score. These five items are I ) a randomized study. 2) reporting the method of randomization. 3) being double-blind. 4) describing the method ofdouble- blinding. and 5) a description of withdrawals and dropouts. The final two items indicate poor quality. and a point is subtracted for each if its criteria are met. The two items include inappropriate methods of randomization and double-blinding. The .ladad checklist has relative merit because ofits ease and simplistic approach that incorporates the nrost important individual components of methodological quality like randomization. blinding. and handling of patient IQ LII attrition. Allocation concealment is important because its absence has been associated with an exaggeration oftreatmerrt effects (Towheed. 2006). Because of the emphasis of quality reporting instead of actual methodological quality of the trial. .ladad is by no means perfect. To prove the reliability ofquality assessment by multiple raters using the .ladad scale. Clark ct al. performed a study with two groups oftwo independent reviewers who applied the .ladad scale to 76 randomized trials ( l 999). The 76 articles were randomly allocated into four groups and were reviewed during two different time periods. two months apart. The kappa statistic was used to assess inter-rater agreement which ranged from 0.37- 0.39 and improved to 0.53-0.59 with the omission of one item (Clark et al.. 1999). This study also proved that there were high percentages of agreement between ratings. Another study conducted by three surgeon raters on 2169 reports over a ten—year period showed an inter-rater agreement of0.48 which was higher then that ofthe Clark. H. et al. study (Bhandari et al.. 2001 ). After the omission ofthe withdrawal and dropout items it rose to 0.51-0.83. For this meta-analysis assessment of methodological quality will be graded using the .ladad Score. The grade will determine whether the article will be included in the systematic review and meta-analysis. Ifarry article falls below a score ofthree on the .ladad scale. it will be immediately dropped from the review. (J'ochrane Quality Assessment The Cochrane Quality Assessment has a similar goal with the Jadad scale by testing study quality. This is the standard assessment to all Cochrane reviews. 26 Because ofthe length and nature ofthis assessment. the scores derived from it will not be used to exclude any articles. The scores will be compared between studies after the analysis is run. There are several ways to rate ’alidity. One is to rate individual criteria as 'met'. 'unmct'. or 'unclear' and to use individual criteria like adequacy of allocation concealment in a sensitivity analyses. These criteria serve to summarize an overall assessment of how valid the results ofeach study are. Scales with multiple items and complex scoring systems take more time to complete than simple approaches and have not been shown to provide more Ugens & (ireen. 2005). They may carry a |—~ b reliable assessments of validity (Hi greater risk ofconfusing the quality of reporting with the validity ofthe study. These assessments are more likely to include criteria that do not directly measure internal validity. and are less likely to confuse reviewers. For these reasons. it is preferable to use simple approaches for assessing validity that can be fully reported. A study conducted by Lorenzo lvloja and associates assessed the methodological quality of systematic reviews (2005). In this study 965 different systematic reviews were assessed for the methodological quality. Each review data was extracted about the quality assessment of trials included in the systematic reviews. Information extracted included title. authors, type of intervention. and methods for quality assessment. These methods included scales. checklists. components studied. or composite scores. and how they planned to use the quality assessment. either as exclusion criteria or for sensitivity analysis. ()f the 965 reviews. quality assessment was carried out in 88.5% of reviews. more often in Cochrane reviews (Moja et al.. 200] ). This study proves that Cochrane reviews were more likely to include a quality assessment making their protocol of higher quality and their reviews less bias. Rudiogrup/rie [Evidence This meta-analysis requires that both the diagnosis and outcome (i.e.. healing) be based on radiological evidence determined by a clinician or radiologist. This includes images from x-rays. CT scans. or MRI. This process was chosen because some studies use multiple different criteria and because it reduces clinician bias. for example a study reviewing the diagnosis of scaphoid fractures conclude that x-rays. computed tomography. radioisotope bone scanning. or magnetic resonance imaging along with clinical evidence from the physical examination improved the detection of fractures (Schubert. 2000). Biologically. between four and eight weeks new bone begins to bridge the fracture and can be seen on x-rays as a hard callus. This will be viewed during the outcome assessment of all studies. Rer'icrr' :‘lrluncher :ll?(lfl".\'f.\' Program There are several meta-analysis programs on the market. RevMan is the Cochrane Collaboration‘s program for preparing and maintaining Cochrane reviews. This study will use the version RevMan 4.2.10. RevMan not only formats a protocol for the review. but it also keeps track of all references. including ongoing studies. excluded studies. and included studies. It can perform a meta—analysis ofall the data entered presenting the results visually in tables and various graphs. It is the current template used by researchers looking to publish a C ochrane review (RevMan 2003). Chapter 3 METHODS A protocol was devised for the meta-analysis before retrieval of articles in order to establish inclusion criteria used in the search. A medical research specialist was consulted about search terms. The MEDLINE and CENTRAL databases were used to run the search terms. All articles received from both databases were numbered corresponding to a main hit book. The hit book is a formatch excel worksheet that contains each articles relevancy screening form outcome. Each article's title and abstract were then read and screened with the relevancy screening form by two reviewers. The relevancy screening form is a shorter version ofthe data extraction form that contains the minimum standard criteria that studies must have in order to be considered for the meta-analysis i.e. uses bone growth stimula‘rtors. has a control group. is a delayed or non-union fracture. and uses a healthy population. After being assessed by the relevancy screening form. each reviewer input the corresponding article's outcome i.e. yes. no. unsure. into their individual hit book which was later transferred into the main hit book for comparison. This allowed the articles to be independently screened by two reviewers. All articles that received both yes or one unsure and one yes were then put through the data extraction screening lorm. If an article received two unsure ratings or one unsure and a no. a third independent review was consulted. Only one article was disputed and the third independent reviewer approved this study for inclusion. If an article received two no's. then it was excluded from the meta-analysis. To) C The articles that were approved for inclusion were then retrieved in their entirety and evaluated by the data extraction form. The data extraction form is a longer form that lists all criteria the study must posses to be included in the meta- analysis. It is set up into three sections: 1) Inclusion Criteria. 2) Quality Assessment. and 3) Data l‘ixtraction. Once a study made it through inclusion criteria. it was then subjected to two different quality assessments. The scores from the .ladad Quality Assessment determined whether the study was continued to data extraction. Any study that received a score less then three was excluded from the meta-analysis. The Cochrane Quality Assessment score was used in a subgroup analysis and had no affect over determination of inclusion. lfa study made it through all these check points. then its. data was extracted and inputted into the RevMan statistical analysis program. Search IT’IU’INM/X A healthcare research librarian was consulted and the following search strategy was conducted in MEDLINE (Pubmed) and CENTRAL databases. The search was first limited to randomized controlled trials. articles published in English. human subjects. and abstracts accessible online. All articles retrieved were processed through the search strategy previously stated. 1) electric stimulation therapy OR electromagnetic fields OR electromagnetics OR magneties 3) electric capacitive coupling 3) pulsed magnetic freldti‘ 4) (pulsed electromagnetic lield‘t) OR pemtltw| 5) interferential current‘l‘ 6) electri* stimulationltw] 7) #1 OR #2 OR #3 OR #4 OR #5 OR #6 8) fractures. ununited[mesh] 9) pseudoarthros"‘ 10) fracture healingl mesh] 11) #8 OR #9 OR #10 l 2) fracture“< [tw] l3) "fractures. bonel'mesh] l4) non—union[tw] OR delayed unionltw] OR un-united[tw] 15) (#l2OR#l3)/\ND#l—l l6) (#llOR#lS)AND#7 l7) #l6 LIMITS: English. Human. Clinical Trials. Randomized Clinical Trials Screening Process Two reviewers independently screened and selected the studies included in the review. The use oftwo reviewers decreases study selection bias and sets up a conlirmation system for data abstraction. A third independent reviewer was consulted on one article to adjudicate a disagreement regarding inclusion criteria. Only English language articles were used. All relevant articles found in the databases were retrieved. E ligihi/ily Criteria Eligibility criterion is the bare minimum standards a study"s abstract and title must meet in order to pass through the relevancy screening form. The four criteria that are addressed on the form include meeting the designated definition '9.) I J ofa delayed or non-union fracture. having an intervention and a control group. using a healthy population. and being a randomized or controlled clinical trial. The relevancy screening form can be viewed in Appendix B titled l\’Ieta-Anal_\'sis forms. Definition ofDe/iir’ec/ and Non-I 'nion [fractures Delayed union fractures w as defined as no radiographic evidence of healing at the site after three or more months. while nonunion fractures was defined as failure to heal or evidence that the healing process has stopped after six months (Bruser & Gilbert. 1999). 7:\'/)e.\' o/ '( 'ontro/ Studies had to use a bone growth stimulator in comparison to a sham control or placebo control in order to compare the effects ofthe healing process. Studies that use active control modalities such as ultrasound. or surgery like bone grafts exclusively in comparison were not eligible. A sham control is an electromagnetic stimulator that is aesthetically similar to the devices used in the study. although it does not work. This device is set up to conceal from the subject and researchers which units actually work therefore leaving data unaffected by observer bias. The placebo control is another control group where subjects with similar injuries were left untreated to be used as comparisons to the intervention group. Although historically these studies have been conducted. due to the movement of human and animal rights in research. the likelihood of finding recent research with this method is low. A subgroup analysis was attempted with studies that compared bone growth stimulators to other interventions although not enough studies were retrieved. llealt/rt’ Population Eligible studies had to have a healthy population with a medical diagnosis of a delayed union or nonunion fracture. Patients w ho had previous treatment such as bone graphs. other surgeries. or cornorbidities like infection at the fractures site were eligible. Studies were not eligible ifsubjeets had any ofthe following diseases: bone cancer. Insulin Receptor Substrate-l Deliciency. aplastic anemia. osteoporosis. If the study failed to mention specifrcs about the population. the subjects were assumed as being healthy. Randomised ana’ ( 'ontrol/ecl ( 'linieal Trials All eligible studies had to be performed as randomized controlled trials or controlled clinical trials. Randomized controlled trials were defined as a clinical trial that includes at least one test and one control treatment where the treatments administered are allocated by a random process like a random numbers table. computer generated allocation. or by sealed envelopes. Controlled clinical trials (CCT) were defrned as any study that allocates groups according to coin llips. odd-even numbers. patient social security numbers. days ofthe week. medical record numbers. or other such pseudo- or quasi-random processes ( Hi ggens & Green. 2005). The reason for the distinction between the two types of studies is because CCT are deemed to have less scientilic value because of limited allocation concealment. This introduces bias that randomized controlled trials do not have. The subgroup analysis looks at the affect that allocatitm concealment has on the combined statistical data and the results ofthe individual studies. [NC I (.lSIOJ‘v ( "RI TER [.4 Types of. Fractures All bone sites yielding a delayed union or nonunion fracture was used. Any type of bone fracture that resulted from acute or chronic injury was included. including stress fractures as long as they adhered to the preset delinition of delayed and non-union fractures. Radiographic Outcome and Diagnosis Fracture diagnosis using radiographic evidence was required. The diagnosing clinician did not have to be blinded. but this criterion was recorded for further analysis between studies. The primary outcome measure also had to have radiographic evidence of a callus to declare healing ofthe fracture. Radiographic evidence was delined as x-ray. computed tomography. and magnetic resonance imaging. lives of];ztervention Trials of all types of pulsed electromagnetic fields and electromagnetic stimulators. invasive or non-invasive. were included. The latter relies on direct application of an electrical field rather than induced current. Included bone growth stimulator generating units were implantable or external (Aaron et al... 2004). The definition of bone growth stimulator for this study stemmed from the FDA classification of interventions and reads. "A bone growth stimulator provides stimulation through electric and/or magnetic fields to promote ostcogenesis to facilitate the healing of non—union fractures and lumbar spinal fusions. The stimulation may be delivered through capacitive coupling with J J U1 electrodes placed directly over the treatment site. or through pulsed electromagnetic fields." (FDA. 2006). Direct current bone growth stimulators were also included and was defined as direct electrical current applied by surgically-implanted electrodes with the cathode placed at the site of bone repair and the anode placed nearby on soft tissue. Quality A.vsessmenl Procedures We assessed the methodological quality using the .ladad score and by following guidelines from the Cochrane collalmration. Studies with a .ladad score of less then 3 were excluded. The Cochrane Quality Assessment scores were used in a subgroup analysis between studies. Both quality assessments can be viewed in Appendix A. Data Extraction Each reviewer independently read and reviewed the articles and completed the data extraction form. Ifthe studies made it through both quality assessments. results were abstracted onto a data extraction form. Appendix B. The data was then entered into RevMan for analysis. Ifenough information were gathered. then subgroup analyses would have been performed. These subgroup analyses would have compared the differences between randomized controlled trials and controlled clinical trials. how blinding assessors affected results. and whether methodological quality affected the outcome. Because there was not enough information. sensitivity analyses were conducted instead. The first three pages ofthe form detail the inclusion and exclusion criteria for the meta-analysis. Ifthe article fulfilled all inclusion categories. then the reviewer continued onto the fourth through sixth pages for quality assessment. If the study met the minimum score. data was extracted. The data extraction page involved the reviewer to record study specific inlormation and results calculated from the research. This includes the number of participants. mean. and standard deviation for continuous data and dichotomous data containing total number of participants and number healed. livclusion Criteria Any study that was not preformed using a randomized or quasi- randomized process. i.e. CCT. was excluded. Studies that used movement at the fracture site. a pain scale. measurements of mobility. functionality scale. or any other form ofclinical diagnosis as their only outcome measure was not used. Any study that did not adhere to the definitions of delayed and nonunion fractures set by this meta-analysis was excluded. Studies that included subjects having diseases that may impede in the bone regeneration process such as bone cancer. Insulin Receptor Substrate—I Deficiency. or aplastic anemia. was excluded. Fractures caused by medical conditions that weaken the bones. such as osteoporosis was also excluded. Studies that compared a bone growth stimulator to another modality. such as ultrasound. hormones. or surgery were not used. Statistical .l/et/ioa’s of "Data A nalfirxvis The statistical analysis was performed using the computer program RevMan version 4.2. IO (RevMan. 2003 ). The primary analysis was calculated using a random effect risk ratio with a 95% confrdence level using the DerSimonian and Laird method. This method was chosen because ofthe lack of data. low event rates. and small trial size. It uses a diflerent weighting scheme dependent on the risk ratio and has better statistical properties when there are few events. A secondary analysis was conducted using a fixed effect risk ratio with the Mantel- Hacnszel method. Four sensitivity analyses were conducted. omitting one article at a time. to see the affects it had on the overall analysis. Forest plots showing the confrdence interval and the effect estimates for each study were generated. Each block represents a study at the point estimate oftrcatment effect. The horizontal line depicts the confrdence interval while the area ofthe block indicates the weight assigned to the study in the meta-analysis (Higgens & Green. 2005). The confrdence interval totals are represented by a diamond shape. A ssessmei II o f H eterogei re itr ' Heterogeneity. or between study variability is described as any kind of variability between studies in a systematic review (Higgens & Green. 2005). Statistical heterogeneity is defined as variability in the treatment effects being evaluated in the different trials. Ileterogeneity was addressed in this meta-analysis by using randomized and controlled clinical trials. using strict preset criteria. and adherence to the defrnition of delayed union. non-union. and outcome measurement. RevMan tested heterogeneity using a standard chi squared test and I2 test. Ifthe value from the I2 test was > 50%. this signifies that substantial heterogeneity existed. Since only published studies were used in this meta-analysis. publication bias is possible. Published studies generally do not represent all ofthe studies being performed because articles with significant or positive findings are more likely to be published (Glasziou et al.. 2001). Language bias is also a possibility since only articles published or translated in English was used. Negative studies were less likely to be found since it is known that studies without signifrcant results are more likely to be published in non-linglish languagejournals (I liggens & Green. 2005 ). (J) \O CHAPTER 4 RESULTS The primary purpose of this study was to determine the efficacy of electromagnetic bone growth stimulators on delayed and nonunion fractures. For clarity. the results section will be organized into two sections. The first section will provide a detailed review ofthe four included studies with article demographics and individual study methods. The second section will present the results ofthe meta-analysis. Selection oflncluded Sir/(lies Ofthe four hundred and twenty two studies that were retrieved from MEDLINE and CENTRAL. only twenty-tour remained after two reviewers completed the initial relevancy screening that involved examination ofthe title and abstract. The complete article for each ofthe 24 studies were then review ed in full by the two reviewers using strict inclusion criteria contained in the data extraction form. 'I'wenty articles were excluded. The reasons for exclusion are summarized in Table 3. There were many articles excluded for lacking multiple criteria. The majority of articles. 26%. did not use radiographs for diagnosis or outcome. Twenty three percent ofstudies did not have a control group. Other reasons for exclusion were trials not randomized 22%. alternate definition of non-union 18%. other interventit‘ins used %. and use of at risk subjects 3%. The Final Relevancy Screening Form of’l'able 2 details the results from the data extraction form. The two reviewers disagreed on one article. which was examined by a third reviewer and was found to meet inclusion criteria. liour 40 studies therefore met all inclusion criteria and data was extracted for the meta- analysis. This criteria consisted of randomization. defrnition of injury. healthy population. radiographic diagnosis and outcome. and a control group. These studies include Pulsed l/agnetic Field 'l’berapr'for Tibial .\"on-union (Barker. Dixon. Sharrard. & Sutcliffe. I984). .4 Double-Blind Trial of'Pz/lsed Electromagnetic Fields/or Delaved Union of Tibial Fractures (Sharrard. 1990). .-1 Prospect Double-Blind Trial ofE/ectrica/ Capacitive Coupling in the Treatment ofNon—L'nion o/long Bones (Scott & King. I994). and Electrical Treatment of Tibial Non- Union: A Prospective. Randomised. Double-blind Trial (Simonis. Parnell. Ray. & Peacock. 2003). Detailed Review: Article Demographics The four studies included a total of one hundred and sixteen subjects. In Barker's study. published in 1984. sixteen subjects were used with an age ranec from 19-72 years with a mean age of34.4. All subjects were healthy and with fractures that had not healed for at least a year making them min-unions. Nine subjects were randomly allocated to an active stimulator while seven others received a sham unit. The mean age ofthe intervention group was 38 years while the mean age ofthe control group was 29.9 years. In the study by Sharrard. forty-live patients were included in the trial. twenty were randomly assigned to active units while twenty-five were randomly assigned to receive sham controlled units. Only fractures at the site of the tibia were included in this study. The age range of the study was 18-84 years: the mean age of the active group was 34.7 years and for the control group it was 45.4. 4] Scott and King's study contained twenty-one subjects. ten allocated to the active group and eleven to the control group. The age of the subjects ranged from 23-87 years. The average age of the active group was 39.6 years compared to 45.8 years for the control group. In the Simonis study which included thirty-four tibial fractures. the age range ofsubjects was 16-61 years with the mean age of32 years (2003'). Eighteen subjects were randomly allocated to the active group while sixteen were given dummy units for the control group. Every article was evaluated with two different quality assessment measures. the .ladad Scale and Cochrane Quality Assessment (Appendix A). Table 4 compares each studies scores against their meta-analysis weighting. The .ladad Quality Assessment is a simple live question scale that measures study design and reporting quality with a numerical score from (1-5. with zero being the weakest and live being the strongest (.ladad. 1996). No study was excluded because of their .ladad Quality Assessment score. Simonis study was the only one with a perfect score of5. Both Barker and Scott received ratings of4.5 with Sharrard trailing with a score of four. All these scores indicate that the four studies were high quality. however these .ladad ratings differed vastly from the Cochrane Quality Assessment scores. The Cochrane Quality Assessment is similar to the .ladad scale but is a longer form using twelve questions. Answers are given a numeric value based on its scientific strength with the highest possible score being 24 (2 points per question). Scotts study received the top score of22.4. Barker's study followed with a score of21.1. Simonis study received the lowest rating. 18.3. Detailed Review: Individual Sludv Met/rods The Barker study included only tibial non-unions (Barker et. al. 1984). The electromagnetic bone growth stimulators used were developed by Bassett and used coils that fit around the cast ofeach patient. The active machines produced a 1.5 mT peak. 5 ms burst waveform and repeated at 15 Hz. The dummy machine for the control group differed from the active one by an internal connection which diverted their output to an internal load thus ensuring that no electromagnetic stimulation occurred. Both machines housed a concealed clock to check the compliance of each patient with the treatment protocol. Other clinical protocols included immobilization with a full leg plaster cast. non-weight bearing activity. and clinical examinations every 12 weeks. All staff remained unaware of which type of machines patients were allocated to for the full 24 week duration ofthe study. In the study conducted by Sharrard all subjects were diagnosed with a delayed union of the tibial shaft. The 45 cases were enrolled from sixteen study centers using strict admission criteria. Similar to the previous study. all subjects 0IBCS. C? were fitted with a full-leg plaster cast with their knee flexed at 20-30 de They were given either an active or sham stimulator which were indistinguishable in appearance thus blinding was maintained for both the patient and the doctor. The unit consisted ofcopper wire coils positioned on the cast adjacent to the fractures site in a Helmholtz configuration. The signal used for the pulsed electromagnetic stimulation was a quasi—rectangular form set at 15 Hz bursts of20 individual pulses. The patient was instructed to bear no weight on the extremity 43 and to use the unit for 12 hours per day for a period of 12 weeks. Thirty-tour patients with tibial non-unions w ere allocated into two treatment groups during Simonis study . All patients received a unilateral external Iixator with compression. The patients in the active group received a pulsed electrical current from two large external coils placed over the fractures site. The coils wer ‘ attached by telescopic rods so they would be in direct contact with the skin over the non-union site and were positioned with a crepe bandage. The pulse had a 3 ms duration in intervals of40 ms with a peak current of6 A at 150 V passing through the active coils. The electrical device used in the dummy group was similar in appearance and was also applied around the fracture site. However. this device only passed a current into a small secondary coil which was not in contact with the leg. No current was passed through the two larger coils around the fractures site. All study personal were blinded to the assignment status of each case until the conclusion ofthe trial. The patients were instructed to use the devices fourteen hours per day for six months. Iiach device was outlitted with a hidden timer to cheek patient compliance. In the study performed by Scott patients were randomly assigned to receive either an unmodified Orthopak bone-growth stimulator or a modilied device which gave no electrical output for the placebo group. livery patient was managed with a plaster cast or brace with openings for the electrical stimulator to be placed on the skin surface. The active units delivered a live to ten volt peak sine wave at 60 kHz. All units were indistinguishable and were monitored by clinicians during the patient’s visits with the dummy units giving the same signals 44 as the active units upon daily check ofthe battery. Blinding for both patients and physicians was maintained until the conclusion ofthe study. Since most of the patients had previously been encouraged by their doctors to bear weight on the extremity before entering the study. the protocol continued to allow weight bearing while using the bone growth stimulator. Clinical evaluations were performed every three months. All four studies were labeled as randomized clinical trials. but only one study (Simonis) described specilically how the randomization occurred i.e. electromagnetic units were assigned from a randomized predetermined list. The study did not describe how the list was generated. e.g. computer generated. however. the study did describe the allocation concealment process which was done by an independent member ofthe hospital not involved with the study who was responsible for randomly assigning the machines and kept the randomization scheme secure until completion ofthe trial (Simonis. Parnell. Ray. & Peacock. 2003). The only other article that described details ofthe randomization scheme was Barker who described randomizing units by a stratified randomization procedure. The article did not deline the specific factors used in the stratilication however (Barker. Dixon. Sharrard. & Sutcliffe. 1984). Every study stated that it maintained a blinded assessment of outcomes. Two studies. Sharrard and Barker. used separate doctors not involved with the study or managing patients to read the radiographs and determine outcomes. The other two studies. Scott and Simonis. used the same staff involved in treatment and follow up to assess outcomes in the subjects but both articles stated the staff were blinded to treatment assignment (i.e. the codes were not broken until the end ofthe study). Between articles there were also differences in the research methods employed that divided the studies into the two groups. The Scott and Simonis studies whose own researchers judged the radiographs had comparatively more healing in the control subjects 14/23 (60.8%) compared to 136 (30/6) in the other two studies that employed independent clinicians. These results could be from the lack oftruly independent assessors. but are more likely from the different criteria for delining healing. Treatment protocols and duration oftherapy varied over the four studies. In the study conducted by Simonis. subjects were treated for a maximum ofsix months (Simonis. Parnell. Ray. & Peacock. 2003). At monthly evaluations radiographs were taken. If union had occurred. stimulation was stopped and the patient was graduated to weight bearing with an orthoplast gaiter. With Scotts study any patient whose non-union healed within six months was withdrawn from treatment and was monitored until the end of the period. Ifa non-union was still healing at the six month mark. the treatment was continued up to nine months (Scott & King. 1994). The maximum duration oftreatment for Barkers study was one year (Barker. Dixon. Sharrard. & Sutcliffe. 1984). For the first six months the patients were casted and used the bone growth stimulator. Evaluations were made every six weeks. The patients were kept casted and stimulation continued for the six months even if union occurred in prior weeks. Sharrard patients received treatment for three months (Sharrard. 1990). No evaluations were made until the end ofthe three month period. The patients had the ability to opt out the trial and would not be included in the study. Even though inclusion criteria consisted of radiographic diagnosis and outcome. each articles definition of healing diflered slightly. Sharrard and Scott used clinical assessments along with the radiograph to deline union. These assessments included measuring mobility ofthe fracture with a goniometer. a visual analogue pain scale. and rating ofdischarge ifan infection was present. A non-union was defined as healing ifthere was less pain. less motion at the site of fracture. and a definite increase ofcallus and trabecular bridging radiographically in comparison to findings at the previous visit. Barker's study also used clinical examinations along with stress radiographs to define union. If both observers where unable to detect movement on imaging when the tibia was stressed. then the fracture was defined as clinically united. Simonis based their definition of healing strictly on three radiological signs. loss ofdistinction at the fracture gap. cortical bridging. and trabecular bridging. No clinical assessments were taken. These differences could contribute to between study variability. By choosing the criteria of radiograph diagnosis inter-clinician differences was introduced. There is a possibility ofdetection bias with the variability ofclinicians and their diagnostic experience reading the radiograph and evaluating clinical signs of fracture healing. illetc'i-analysis Results All results were input as dichotomous data (i.e.. intervention (intervention vs. control) and healed (yes vs. no)). Table 5 is a summary ofthe meta-analysis results using a random effect and fixed effect model. It also includes the results 47 ofthe sensitivity analyses w here each individual study was removed and the results re-calculated. Each sensitivity analysis was calculated using the random effect model. The primary finding ofthe meta-analysis was a summary random effect risk ratio of2.62. with a 95% confidence interval of 0.78 to 8.78. The 7. statistic test ofthe overall effect and was 1.57 (P=0.12). Figure 1 displays the forest plot ofthe meta-analysis based on the random effects model. It shows the percentage weight assigned to each study in the analysis and includes their individual risk ratios and confidence intervals. There was significant heterogeneity between the results ofthe four studies. The test of homogeneity was highly significant (Chi3=21.9l. 3 d.f.. p < 0.0001 ) and the I2 statistic showed substantial between study variability (86.3‘.’/b). Out of the four studies. only Barker‘s favored the control over treatment (RR=0.91). The other three studies all favored the intervention group but the RR estimates varied substantially from 1.78 to 14.18.(‘lig 1) A secondary analysis using the fixed effect method is shown in figure 2. The summary fixed effect risk ratio was calculated at 2.36. 95% Cl 1.57 to 3.53 with a Z score of4.16 (P<’.0.0001 ). The fixed effect analysis results in substantial changes in the weights given to the individual studies when compared to the random effect model. This is expected as the weights calculated in the fixed effect method include only with-in study variability and are based only on the sample sizes ofthe individual studies. whereas the random effect model includes an additional term for the between study variability Despite the overall 48 statistically significant summary RR with the fixed effect analysis there was again significant between study heterogeneity. The test of homogeneity was Chi2 :2 l .9. 3 d.f.. p<10.00()1 ). and the I2 statistic again showed substantial betw een study variability (86.3%). Subgroup :1 nal r ives Initially it was planned to run three separate subgroup analyses addressing the differences between randomized and controlled clinical trials. the blinding of outcome assessors. and methodological quality. Since only four studies ended up being included in the analysis. there was not sufficient information to run these subgroup analyses. Sensitivitv .4 nalvses In order to examine how much the overall meta-analysis results were influenced by each individual study we undertook a sensitivity analysis in which each study was dropped and the analysis was repeated. The first sensitivity analysis dropped Barker‘s study leaving 100 study subjects from the other three studies. This study was given a 34.6 % weight in the original analysis and had a risk ratio of 0.91. After removing this study the risk ratio increased to 5.48 with a non-significant 95% CI 0.72 to 41.48 (Z scorc=l .65. P=0.10). The tests for heterogeneity (Chi228.55 2 d.f.. P=0.()1) and 13:76.6% again revealed significant heterogeneity. After dropping Barker‘s study this sensitivity analysis has the highest risk ratio which suggests a much higher healing rate with bone growth stimulators. However the presence of heterogeneity means that this summary estimate should be interpreted with caution. 49 The second sensitivity analysis dropped Sharrard's study. In the original meta-analysis this study heavily favored the bone growth stimulator. Individually this study had a 12.50 risk ratio and was weighted 18.5394) ofthe overall rneta- analysis. After removing this study the sensitivity analysis had 77 subjects. The risk ratio decreased to 1.62 with a 959;. CI 0.06 to 4.33 (Z score: 0.96. P: 0.34). Heterogeneity was calculated as (Chi3:10.37. 2 d.f.. P=0.006.) and 13:80.7%. This analysis had the lowest risk ratio ofthe four but also a smallest confidence interval making the RR value more precise. We can conclude that there is little effect of healing from bone growth stimulators. Although this analysis shows moderate heterogeneity. In the meta-analysis Scott's study favored the bone growth stimulator with a risk ratio of 14.18 and was weighted 12.5894). Once removed the sensitivity analysis had 95 participants. the risk ratio decreased to 1.94 with a 95% Cl 0.63 to 5.92 (Z=I.I6. P=0.25). Heterogeneity was calculated (.Chi3=14.83. 2 d.f.. P:0.0006). and 13:86.5‘3/6. This analysis shows little benefit of healing from bone growth stimulators. The last sensitivity analysis dropped Simonis study. The study contributed 34.23% weight to the meta-analysis with a risk ratio of 1.78. This sensitivity analysis had 82 participants and the calculated risk ratio increased to 4.99 with a CI 0.13 to 190.28 (7:086. P=0.39). Heterogeneity was calculated (Chi2=28.45. 2 d.f.. P<0.00001) and 13:93.0%. This analysis had the second largest risk ratio suggesting a healing effect from the bone growth stimulator but it also had an exceedingly wide confidence interval and highest heterogeneity 50 nullifying the conclusion. 51 CHAPTER 5 DISCUSSION Statistical Findings Since the meta-analysis included only four studies. I chose to vary the statistics in order to gain more insight into the effect of bone growth stimulators. Both the fixed effect and random effect methods are presented because one proved statistically significant while the other did not. The meta-analysis uses the random effects risk ratio for its primary findings because this method is more conservative with wider confidence intervals. The random effects meta-analysis showed a non-significant risk ratio 012.62 (95% CI: 0.78-8.78 p=0.12). The random effects method assumes that each study is a sample from a larger population of studies. and takes into account between-study variability. The study specific weights are more evenly balanced in a random effects model compared to the fixed effects model. Although this meta-analysis was not statistically significant. the summary point estimate did find that fracture healing was 2.62 times more likely with a bone growth stimulator than without. Limitations of the study include small number of studies each with a sample size and significant heterogeneity. Because of these limitations the summary findings need to be interpreted with extreme caution. There are many reasons why significant heterogeneity was present in the meta-analysis. This can be attributed to the various methodology across studies. First. the varying protocols used in the different studies may have contributed to study-to-study variability. We cannot be certain whether the unit itselfor the additional protocols (termed cointerventions) facilitated healing ofthe fractures. For instance in the Simonis study. their patients were advised to be strictly non- weight bearing while the Barker and Sharrard studies immobilized patients by placing them in a long leg plaster cast. This differed in the Scott study where they allowed weight bearing: they reported. "Similarly. even when we considered that a period of non-weight bearing was advisable at the beginning ofcapacitive coupling. thisjudgrnent was not enforced. Most ofthe patients had previously been encouraged to bear weight on the extremity. and we thought it inappropriate to change this behavior because ofthe risk of introducing an additional variable." (Scott & King. 1994). Other variations in study methods include different lengths of follow-up. For example. Sharrard followed his subjects for only three months while the duration other studies lasted for nine months up to a year. Every study used a different bone growth stimulator unit with varying duration and dosage. Simonis was the only study to assess bone growth stimulators on other long bones besides the tibia. In his discussion. Sharrard mentions the wide spectrum of bone and soft tissue injury that may occur. The different injuries sustained by individual subjects can vary between simple and severely displaced fractures where simple fractures do not need supplemental aid to achieve union. Sharrard also commented that his two treatment groups differed significantly in age distribution. When looking at the statistics. Simonis and Barkers studies have Ur w more frequent healing rates in the control group. Simonis treatment groups had an imbalance of prior operations with the control group having I 1 operations prior to the study. Barker contributed the high healing rate in their control group to the fact that the additional protocol resulted in long term immobilization and non- weight bearing. l-Ie comments that because ofthe amount oftime the control group was required to use the unit. it further reduced the amount of limited activity on the affected limb which then promoted healing. There were important differences in baseline healing rates between studies. In the Sharrard and Scott studies only a minority ofthe control group healed i.e. 4% and 0% respectively. In et_)mparison in the Simonis study where half healed and 85% of Barker‘s control group. This can be explained by Barker having the longest duration of treatment of one year compared to Sharrard and Scott with three and nine months. When reducing Barker's duration of treatment. at three months the study had only a 28% healing rate. It is also interesting comparing publication between the four articles. Two studies. Sharrard‘s and Scotts. were published from The .Iournal of Bone and .loint Surgery. Barkers was published in Lancet and the Simonis study was in Injury. This is relevant because this could signify publication bias where journals only publish studies with significant findings. Both Sharrard and Scott‘s studies guest a beneficial effect of 98— were published in the same journal with results that su bone growth stimulators. Comparison o/Random Effect and Fixed Effects t’llL’l/llfl/S The random effect and fixed effect model analyses generate different 54 results primarily because of the differently ways the two methods calculate study variance. Both meta-analyses use the same studies although their contributing weights are significantly different which results in dissimilar risk ratios. 7. scores. and P values. The primary difference between the two is the inclusion of between study variance in the random effect model. In RevMan the calculated study weights are dependent on both the sample size and the event rate. Because estimations are more precise when there are more events. trials with high event rates get weighted more. This explains why the Barker study. that has the least amount of participants but the largest healing rates (i.e. events) was given heavy weighting. At the other extreme. a trial with little or no even rates gets little or no weight. This is illustrated by the Scott study which is given the least weight in both meta-analyses because ofthe controlgroup having no events. The weighted averages differ between statistical methods because of how the mathematical models interpret standard error. confidence intervals. and variance. The Mantel-Hacnszel method operates by giving studies with less variance. or standard error. more weight while assuming that every study is evaluating a common treatment affect. The DerSirmonian and Laird or random effects method does not assume that only one treatment effect exists. i.e. bone growth stimulator. This method factors in other variables and estimates the mean and standard deviation ofdifferent effects when giving the risk ratio. The fixed effect method seen in Figure 2 estimates study specific weight only on the basis ofthe size ofeach study i.e. within study variability. The fixed effect analysis weighted the four studies very differently from the random effects model for example it significantly reduced the weight ofthe Sharrard and Scott study from 31.1 I 0o to 8.24%. Importance oflietrreen Studv Heterogeneitr‘ Despite the debate as to whether the results are statistically significant or not. both meta-analyses show considerable heterogeneity. Even though the meta- analysis was set up with strict criterion for scientifically sound results. there were still major differences between studies that we could not control for. These differences included patient selection factors. design of bone growth stimulator units. dosage and duration oftreatment. differences in the definition of healing. use of adjunct protocols with the units. patient compliance. and time between follow up appointments. All of which could have contributed to the differences in outcome between studies. Most ofthe four articles did not address the problem of patient compliance with the bone growth stimulator unit. Because the intervention is marketed as a portable unit and results are dependent on patient use over a prolonged period of time. results can be affected from inadequate instructions or non-compliance. Two studies used a concealed clock that recorded the activity ofthe stimulator (Barker. Dixon. Sharrard. & Sutcliffe. 1984. Simonis. Parnell. Ray. & Peacock. 2003). One patient from the control group of Barkers study was non-compliant with treatment protocol and dropped from the study. All subjects from Simonis adhered to protocol. In Scotts study the devices issued a readout that showed the number of days of usage to assess compliance. Although the unit could not record 56 the duration ofeach use. One patient from each group were excluded due to failure to comply with the use ofdevice. The units in Sharrard's study could not monitor patient compliance although it was found that one patient tampered with his device and broke the code dropping him from the study. Another source of between study variability was the fact that one study included different fracture sites. Barkers. Simonis. and Sharrard‘s articles designated tibial fractures as inclusion criteria while Scott's study researched the effect on other long bones (Scott & King. 1994). Nineteen percent of fractures in Scotts study were on the femur while 9% were ulnar fractures. The majority of fractures. 71% occurred over various sites on the tibia. This introduces heterogeneity because some bones are innervated with a bigger blood supply then others. thus allowing more chance to heal. :I/eta-anal)A’sis I let/rods The main purpose of this study was to evaluate the efficacy of bone growth stimulators on delayed and nonunion fractures using randomized controlled trial design. This research question proved to be a difficult one to evaluate because of the small number of highly variable studies published on the subject. Out of the twenty-four articles that made it through the initial screening. only four met final inclusion criteria which included RCT or CCT. definition of injury. healthy population. radiographic diagnosis of injury. comparison against a bone growth stimulator. and a control group . The majority of studies were excluded because there was no radiographic outcome or diagnosis and many studies did not have a control group. Even when we adjusted the inclusion 'J r \l criteria to include only randomized studies with a control group we found only an additional 5 studies. After reviewing the articles there was not a significant amount to run another separate meta-analyses. A subgroup analyses was planned to compare the bone growth stimulator to alternate modalities like surgery or ultrasound. Only five out ofthe twenty four articles addressed this issue and so there were not enough to warrant a subgrorlp analysis. In hindsight only two arm trial studies with a control group were included. Many studies were excluded because they did not include a control group ofsubjects whose fractures would receive no interventions to heal. Limitations oflnclusion ( 'riteria Only four out of the twenty-four studies passed the data extraction form. The main reason why most studies did not make it to data extraction is because they did not have a diagnosis or outcome radiograph identifying healed fractures. Seventeen out ofthe twenty articles did not specify how the diagnosis was made or used clinical tests to prove healing. Even ifthis criterion was eliminated. twelve out of these seventeen studies had no control group or randornization. two important criteria that cannot be omitted. These articles got past the initial relevancy screening with abstracts that stated they were randomized controlled trials. but after retrieving and reading the full article. they did not meet the definition ofa randomized trial. Ofthe other five studies that were excluded from the meta-analysis. none ofthem met the definition ofdelayed and non—union fractures. These studies used acute fractures and subjects with osteoarthritis. two extremely different injuries with prognoses that are documented to more easily 38 return to healing without intervention. They also used other controls included surgery. osteogenic proteins. or ultrasound. Because there are many restrictions on human research. experiments with a sham or control where subjects are not guaranteed to get a treatment are getting harder to conduct or find. To run another study by adjusting any one ofthcse criteria would be accepting articles that no longer addressed the initial research question. Com/unison by Quality Assessment When establishing the methods there were two forms used for quality assessment. The .ladad score. a table of five items with scores ranging from zero to five. were calculated. Any study with a score less then three was excluded from the meta-analysis. Only one study. Electrical Treatment of Tibial .\'on- Union: A Prospective. Ramionu'sed. Double-Mimi Trial. received a perfect score of five while the rest of the studies received 4.5 or 4 (Simonis. Parnell. Ray. & Peacock. 2003 ). Since the initial inclusion criterion for the meta-analysis were strict. every study that made it through the initial criterion passed the .ladad quality assessment. The Cochrane Quality assessment tool was added to ensure the methodological quality ofthe included articles. No studies were excluded because oftheir scores. Randomized Trials versus ( 'ontrolled ( 'linica/ Trails The difference between a randomized controlled trial and a controlled clinical trial was that the former uses a formal randomization scheme generated from a computer or a random numbers table while the latter uses a non- randomized or quasi-random process such as organizinU groups according to date h La 59 ofbirth or social security number. All four articles included in this analysis were designated as randomized controlled trials. Only one article. Electrical 'Ireatment o/‘Tibial .\'on-l.7nion: .21 Prospectire Randomi:cd. Double-blind Trial. was randomized by allocating the coded electromagnet units from a randomized predetermined list. (Simonis. Parnell. Ray. & Peacock. 2003). Barkers study used a minimization procedure of randomization to keep the groups as even as possible (Barker. Dixon. Sharrard. & Sutcliffe. 1984) Sharrard and Scotts studies stated that the study was randomized. but they did not reveal the procedure as to how. We cannot know if they were truly randomized because the methods of random assignment were not revealed. This is why randomized and controlled clinical trials were included in the meta-analysis in order to keep any articles that may be quasi-randornized. It is important for studies to address allocation concealment or randomization will be lost and bias will be introduced. Ifthe allocation of units are not concealed then researchers can deduce the subjects grouping despite randomization and preferential treatment can be given. Scotts study kept allocation concealment by stating that the manufacturer who kept the randomization code took no part in allocating the units to patients. nor was he involved in the study or informed of the outcomes after the code was broken. At the end ofthe study. the researchers tested the units to ensure the information was correct. Barkers study did not state measures to keep allocation concealment. but multiple times declared that all staff were unaware ofthe machine type during the study. 60 Blinding ()utcome Assessors Another research question posed at the beginning ofthe study asked ifthe blinding of outcome assessors affect the result. ()ne article stated that the junior member involved in patient care also performed clinical outcome tests that were used for outcome diagnosis (Scott & King. 1994). This would be a clear example ofbias ifthe assessor was not blinded. None of the four articles reported a lapse in blinding for any oftheir trials. ( 'onclusion This meta-analysis was limited by the small number oftrials which were highly variable in their design. Based on the data available there is insufficient evidence to support the use of bone growth stimulators for nonunion fractures. Most studies identified either were not randomized. had poor methodology. or used bone growth stimulators for other means. With the abundance ofcvidence on direct stimulation with spinal surgery it would be easy to run a meta-analysis on this subject. Another more accessible way to research bone growth stimulators would be comparing its affects on acute fractures. Since many acute fractures are known to heal on their own. you could ethically use a control group for comparison to retain scientific integrity. Besides a sports medicine setting where athletes do not have the luxury oftime for their fractures to become delayed or non-unions. it would not be clinically relevant to the general populations. But the results ofsuch a study could give insight to the effects of healing bone growth stimulators to aid the healing of fractures. In conclusion I encourage and urge researchers to run more scientifically sound 6| studies in order to advance the uncharted technology and usage of electromagnetic bone growth stimulators. ta: .3 Em: co 38:: 2: >65 5: moo—o- Embmcm SE: a E: .32.)? a _,._:c. <_.. Z < Z 232:2? 2:55am? 959.8 pacing? Egzzu 825- 1122mm- Acowxxo 2.598%5 .322: 8:3 £3» 2022558 E :ozflsEcm 2.820 new... 680 .m £8828 we ..> .mowEeEv. ..< .moEEEmv— chm .’.:Q.2:~2Q=. C E.:::.§.: 22 E .5935 tome/.3 “2.23?ng 35 29:22:22.4 bezofim $253.68; ._2 52 8.52% EU- _m>._w.~:_ our—0U: CCU .WGWOI 8% 25:52.8 .5; 88.5.En :82- 84a :0: new 822m- $83 9 Scam e .m .80: :8 ..m .Ezm ”km 3:3 :3 Sat 35: £392 \e 5:57:02 .6 no.5: 80:53:86 e8 ii 952%- Ezm- £2223; .SEfl .Esnc EseEexe 28:28.: 888:8 wEen £55 meson wco_ 2: Ex zctetgew <...Z 3:528 beam- .83 2.2on 252 >mm- LEmE- co nezoememoe 3:0- EEK etmzu§=oé8m 22:33:..— 3381 mix—55838355 558.23:— mEanEeoA—m 555:0 20:23::m 20E camcmeobuoE wEm: mom..»._m:<-m82.fio m20mthEoU _ eBmH mm <_>_O~_ Est 363.. 6622:“: 30.86 E: 66- 86:7. 25.28 co mar/o :90. BE? 32.6.. E 8622: 86:7. 60:52- 6:: :88 E mEogEu .2.._:.E_.E wees: ._,.:_§e,___s- mEuEmmomme 2:83: .6 89: 80:8. WEE? 6 89).: .6 :oc_5_.:m6 22.3- E38 6:3: :0 63: 8:3 mwc6cc 363;. 5:8; :c 83.6 2.28% we; LEmE :2: .62: o: 3an 2.2:- 621%: com so 832?: 36:; :20 :5 o _ - meson :0 8.52.5; 86:3. 020 :5 m_- 8.48m .3 22.528 .0625- 282.52: 00:03: :50 camo- 269: 88.3 E065: 2:3 88.5.:6 6:: 6203:. 60692 6.63 gm :E:o:.__m._oQ- 8m: Ememca mm 520m- mEmm- 688?“: E32 6% 9:6 wE.:6 m._m.$u_.6._ xmmE- .mzocuwoeefi; ob 5:66 :8 .36 860:. mE_o_...o:mm: tom e8 6cm 8:09 :o “:2?— .6 .8859 2: 38: 63m- 32 -02: E2... 83:3 68:- 889 2 .533: a. .2 a? U.3 Emtcétfiuxc innate: W axe}... T ...£=2.§m. 636538ax: 2.83ng >8me 3:93 _ 2an 5"! If, P l o_ 35 6.5.66 3.58.3 666.6 :06: 6.9366 :8 .66: utocmeEBoE 686: .6 :26 6226-06.56 < S 6_ M: _< .3 .ma .mEoém _n::6::6 66:66 63:66:”: 628386 a “66:60: :66 .6 66:68.; 68685 c _< .3 F< REV—Lam 62: 6:6-2666 : .6 8686 6:85 .66: .6: EB: 68 .3326 66c 6:0:me 63:5 c 6 o— _< .6 O :65 3:06 w:o_ 6 :06: .666 62586 66 E 966:8 36638 626866 6&6 6E6 66:06 4625386 < 66662 33.5 6:55 66k 3:20 6:50 66662 hum EEK .wbm u a S: v6 63.5 SEE/x 606205 E6“— wEEBQm 566.56% 6:5 m 2an (35 636.065: 6: :06: 69866666 .::0..w 66:8 02 _< .6 < .6o>.m 68¢. 8:38: 6:0: .6 625.6: 2: 6.: 66: 26366686 662:: 6:536 65.60065. 0: .m:m....:::o£m0.6 33 _< .6 IO .60.:- m:_::-.m8:mo E 66: 0:6:meot86 602:: 6 flout“. 606.6 66.6 :26 6:26-66:06 < 66606:: 2:830 0: .8...:0:.c $9.: 6 6:: dub _:_:...66N_E06::.. .0: 430:0. 6:68 02 _< .6 L .0NN::6m 86:6: E 3.56:: 36.5. E m66: 2686 666:8 92:66:06 65 6:..m06m. 6.08:0 0: .8: 6...”. 6:0: .::0..m 66:8 0: 686065: 62 _< .6 PU .:0:_.u._..m twin. 6:06 6 6:29:00 6.66.68 .UQ .63, 638.: 66:6: 66; 8:26.... :8: .6 6w: .m_w0:m:6 66.11665 0: .3868 6...:6..6.:..6EC0 3.: _< .6 «A .250... 3:668 6:86t8E mEm: 6% 65: 66.6 8.20:... .56: E .1366: 6:066 :0::.:E:m 66:66:08.. .0: .6:.03..0.£m2.w:_6 :85063 0: 5.6:; 0: .866: 6: .m:.6._....0 _:_E6..6.:..BE.6 3: .< .5 E .250: :5: 2: .6 5:85: E :06: 0. 6:56 6:062: 6.: 6.: 660E < 6:...00065 A: m..w0:.u.:6 0: :06: 63.566.70.66 96...». 66:8 0: 6056065: .02 _< .6 < 6.1m 6.3m 858:: 2.06.6 666:6: 6:. 6.: 66: 0:2.me0...86 601:: 66m: £va 0: $506066 6.: 69:066.... 0: ._8N_E06:S 6: .966). 66:8 02 _< .6 A: 59:0: m6. :::0: .6 66:58.: 6> :3 .4005 6669.86.56 36:66? I 660m E36 6:: ._A.6.n...:m.6 6m: .966 66:00 0: .66N_:.06:E 62 _< .3 :0 4:8: 8.6.5 605 6:: 8.28.: 6:0: w:o_ 6 606.666 62:66.:- 66N_E06::.. 6: .65m062 2:860 0: .3: 66825: .9660. 66:8 02 _< .6 .3 >3 .m:oEE_m “—35: 6:: Snug—mm 63:26:06: .66.: oEgm 66:5. 6:: £65.66:qu 3.56:: :3: $60566 6566:. o: .:0:MNE.06::.. 6:68 0: .68: >86; dab 6:: _< .6 03 .663 28602:: .8360 _:_xm-:..:: E :0:£:E:m 66:8 .026 66:62:5— =0~§6km xe-xagneux 655x 6:: 3:000 866% 6663.0zm - E6: wE868m 60:66: 6E: m 666-: 66 55060.66 66.0.0066. 0: .966). 6.600 0: 606665: .02 _< .3 < .O._vCOhUm—>_ 6:60.. 06.03.. 626..-:0: 5:60. E 0626.. 606.006 6.0... 0606662806 6056: 0:0...0. 66:00 0: 60.26066. .0: 36.06.... 66.60: 6 60.90.06 .02 _< .3 P 02.3.36 3:0: :0 56.0... 0.6606982“. 6:: 2.0.6.: .0666 .3 605:8 6.6.5 60.0....0:0....6 6 0.00....m .mom .000 0: ..0.E_:0:._ 56.066 66.0666. 0: .6660: 6.66 .0:0...0. 6.600 0: 26:5 6600263. 330 6:0 6:5 6600666 :a000..:m 6.0... 0_.0...0.::.0..00.0 .0656: >0 66.3.0: 6 606.00.; 63.6065: .0: 6:06;. 66:00 02 _< .7. a 5.02.0 6.6.66.5 6.50.61. 06: 606066. :6 6.: 66.60: 6:: :06: 606.066 606.6066 £060.. .m0>.-:0. < 559.906 66606.... 0: 5:67.066 .966). 66:00 0: 6026066.. .02 _< .5 <2 .803 9.4.6.6.. E6.-w:0_ 6...: 6:62.20 06066060. 6:0 606:0 03.000506 < .:0_.5:0E.:5E 606...: :05... 6:65 _0>0_.._:E E 6:06.65 UG 06666666 05: 0:... m50:.0.:6 66606.... 0: .3660... .50.... 05: _< .6 01. 560.. 55060.5... :36... 6.0.0.0650: 68:06:63. :0 0.2mm 6:6 66066.5 60.6020 Um. 6 60.6 066 65:00:60 02.000505 0:060. 6.600 0: 026006.00. 5.6.02.5 605: .6526. 6.. 665066... OZ .< :0 < .6:0xmm 6:5 660063: 6 ”2:06.: 05.: :wE E 02:: 6:6 60.. 066 3360.66 6.. 6:23:25 0:00 60:80.0 0606666. 6.600 6.. :0..:0>..0.E .060 5:00.55 .0605 E86 0m: .6536. 6.. 66.0006... 02 _< .6 .0603 .z...:.0...z0 .030. 066 36.0.5 0:02 m:0_ 66:0 -60.... 2:0: :0 06.5.0055 60.09.000.20 6 5.00.6 0.; 560.5 E 62 _< .0 P ...06_0..:0.0.0.N 3:06 :0 566.... 0.003.660.606 6:: 6:65 53.55 60.0....0.0....6.6 300.cm 60.01660: 0608.6 0: .:0.:::0:..0..06 .0660. 6.600 02 _< .0 6 .8206... “CUP—:0 “out—u ~:m~m~.Oo 5:3 EEC—7:0: we usurp—“no.5 .0;qu .AUSHm Luacou_~_.._:._ < 5.3.506. 6.660060% 65:! 0:: :80. N 2.5 67 mvzqs? m........ 35:? m_...... 35:7. :3... 36:7. #_...... 35:; m_::.. .0273? N22. o..o-oouo..ooo..otuou...- fiur: 23:29:92: .3320 AG $2353.... 5.25.55 .5 so 9.33.3 :QCECUU 9 552:5 E: 392 Am. 0 o . o o o o o . - :.:u::5:5 .5 $3355 u_;9~._w2_:~._ oz 3. .c....-o-o.o-oooon ..:..EBEEBEE E: Euczbgzm i. ...........................Q:Av._w _Fu.::OU OZ AI. o-uguco...oouooo. ......:.:...5529; 1.x._..:~.3 mm: 2 :o_m:_uxm 23:6. S; 233% m Bank ()8 7”: r. \Efi. 3:32.3ch $3.223E3m ..g.,.:..s:.\.,.:.i w. ..:c.:.\._.~-:o../.. 3.5K x: EEE§C 32.:qu -mEoE_m wdm v .,..u.::.:.V.E 3.5.5? 29:5 EESmQ .E\..€3E ...:m:m:=5.:um\m \Sézk x: \Efi \E:M\-&£EQ T 65:95 235% E23 1mm m4. .3 ~E.E...,.~-222»? EEEEC 2: E usczzcu «0...:3339 \Extyixa \Et \EQQ-..._m:< >.:>Em:om we we m 250m no.5: m_m>._n:< .9_>Em:om me E m :g.:..._.:..;m no.5: m_m>._§_< .£>Emzum we 2: m A._...,_._..m gen: 3:2,... __c_>_._m=om 8 E _ w xx 3&5 Eoncmz mo 3 _ w mm 53.5 33E 5 5.550 .m> :ocflsEcm u:2_wm:_c.:uu_m _o 3:332?— mo:::m «83:5 .5 Satan—=50 353m Eocmsfim margaész m Bank 70 7:259 W53“. :_o:_Eu._._- WEE; Am _ dug: mm; NV ”69:9 :595 ._2 36% c n: H E m m _ 2 no 2 .A 333.: We .8. _ mu -25 3553.22.fiwwh 22:53 2 $5582: ow ”3:95 E59 a; .32 S.“ 8.2: :u some 33 a 3 :3.“ .8: w: 2.2 Ba 3...: .5 .meoém :28 .21: 2.: 32 + :5 E3 28m 5%. .E. : on: $2. _ mm; 3.2 BEE; 3 _ R 2 a c 2.: to g 23m 5 :3 .x. 5 some 22 Z: .ccwaséa a 253:5: xx 5.0.6.5 25—55.: mm _o._:5,u 525.5; Assam m_m.h_m:< NE Swim 553:5. we 5:5220 55:09 E55 .3 :ocflzEcm ougmeobouE 5 “cemtngoU 35.8.5 5:5 :02 95 :05: 3.2.300 :0 29.2555 5.520 25m ocuzmmthoo_m.5 .8855 3533* 28m vim $9.226. mwoutm Eons—Na _ 253m 71 35:00 ESE E2239; Egg 2 _ ooodve c _ .v HN ”auto 298.5 Em up; gnaw”; E m. m _ 2 no 2 iooodvemu,:v._o._mun:_u”éaaweaetotmfi 22:53 m._ .CcoEEEC 9m ”$296 730% :3. i: 2a 8.2: :u some :33 * ow mm $3 8.: w: B. _ m Ea w :2 E. .2555 $38 .82 5.: 3m _ III _ .5 SE scum :0..on .E. : We 9% 4. my; 2.5. 22:25 w: is :3 $2. lwl E 3 33m 5 £2. 3 C so? 2...: 2...: ..A._ow28-e; a 2.22:“: NE 5903 255.: mm 3.550 E9680; >655 max—Ea mm tutu box?— mo ”2:830 _c.::o.u 595 .m.» :ocflzEcm ozocmeobooE S ”cemCmQEoU mutzuai :25. :02 E5 :cE: @9330 :o m._o:w__:::m 5520 25m ococmeotoo—m ,3 “835.5 ”.523x 2:». fix 235% magi 85 m 0.5%; 75:93 2.3:”. :3. _ v at: 2%. 23:53; 30.8,.— md _.o Sfonn: m3. TN 30th :395 :8 $3. 23.8”: .2 _ if: in .mmwnmzu ”20:80:22. 5: 33+ 22:53 a .CEEEEC mm MEPQ :20 :1 8.2: Ill-T :3 2.2: .98 mm m: :23 .8. : w: 3: Q u 28 2...: .5: .2855 5.8m 8.2 2.: 3.8 .T :8 23 38m 29% .35: OWE E. _ m AIIIII mm: 305— 33:95 5 came c\c C as? Z3: 2...: .mequbzm .5 E5355 mm E933 953:5: am 75:93 25:55.; >.3_:m £33m QEDV m_m>._m:< brizmcom mo 660850 _o.::ou :Esm .m> :ocmEEtm ocucmeEUBm _o :523an0 3.565.; :oE: :oZ 3:: :oED 3.1290 :c m._8:__::3w 5.50.5 ocom 25:w§:o.:oo_m .8 zowowtm ”32.51 Svtmm QED m_m_.m_m:< bmizmcmm q 2:me r‘ 5:50 m5; .1. EuEEu...r $5.25 3 m m _ mto No _.c H Ammdumv cZHN ”Unto 25.55 .5; “mp; =3?»qu .Eoccdnmv Hutu .mwe—HMEU ”£25538; .5; 5.3% 22:53 2 .CEEEPFV mm ”3:95 35% no... .33 I: 8.2: :5 :35 :33 fi 2 :3.“ .8: x: 2.3 + 0; SE .5 .255; :93 .3: Ed 0:: TI mm: 52 22:23 :1 .33 :2 3:. lil 3 K Exam 5 $3 a c 5 £8 E: Z: >..5w28-€m S 253:5: mm E393 253:5: mm 5.550 EQEEPF roam 350m no.5: $372.4. .€._,:_m:om mo ”258:0 5:59 EEm sf :ocssEcm ucucmeobouE S Eo£._8:_ou 3565; 5:5 :52 cc... :oE: Egon :o £95255 5:80 25m ucgmmEotooE .5 38:5 EoEom seam QED mix—act. _E>Em:mm V 23.2; 74 5:50 $5.32 50:03:. 25.3; 2 m m _ We no .8 idneooanuse._2_§.58,:EH .58”: .68.?.: T8 .R.2u~:_0 ”22.89.55; 5: 53 20:50.3 3 .CEEEEC on ”8:96— 55% E4 .82 8: 8_ all! :0 2.8: .5: 3. E. :3... 8.: E: 2.2:. n 28 w :2 5: .88.; :38 .82 2.: 2.2 A :8 20 88 5.: .32 :8 2.2.. i E o...» 5.5: 5 88 2. C 88 2...: 2...: _,.._c.ué_..ué; .5 250:3: mm 2.11.5.5 2:00:05 am 53:00 52500.; .€_:m statmzm 00:: 22:05“: b_.€_m:0m we “2:00:50 53:00 Eacm .m> cocflsEcm 2838:0585 S :0mtw0E0U 3:522; :02: :02 0:0 :05: 00.375 :0 £05.55 6.50.5 25m 0:0:w9:0:00_m,5 385.2 ”32.51 03:85 00.5 £2283 22../£85m m 0.592 75:00 $0.52 E255; 20.2““. E m m _ We m.c _.o Acmdnn: cwdHN gouty :50; :02 7.021 $0.3”: .A 508.0an mntv .mwwmunEU 2:259:20: :02 03% 20.5.03 n .CEEEEHV mm “3:95 2302. 5302 .m _ .3 00.: 8.2: ~ I :U $30 30% 3. Om 2.3% .32 w _ .3 mm 2 :5 2.0 :ch :93 4...: _ 3.2 3.3 ‘| mm: 00.2 03.525 5: .32 :20 ES to o: 3.50 C 23.0 fix. C 93.0 Z0: 2...: 25330.03 :0 E595: mm 220.)» 2.00:0: am 6:50 E25022. .€_:m 3:0:zm 00.53 23.7212 .c_.,:_m:0m 0: ”2:00:00 _0::0U E55 .5 :0_:~__:::m 0g:w::_0:00_m _o ”:0m_._09:0u 3:505; 52:: :02 0:0 :25: 00230: :0 m._0§__:::m 5.50:0 2.0m 0:0:mmE0.:00_u .3 20:05.5 ”52.52 mEoEmm QED 292052 >.:>Em:um o EswE 76 APPENDIX A Quality Assessments 77 JADAD QUALITY ASSESSMENT l) Jadad-11' the article has a score less then three. stop here and check not usable at the top. SCORE: A Jadad score is calculated using the seven items in the table bel ow. The first five items are indications of good quality. and each counts as one point towards an overall quality score. The final two items indicate poor quality. and a point is subtracted for each if its criteria are met. The range of possible scores is O to 5. .-- Was the studv described as double blind? ladad Score ( aleulatlon w..- -._-...—.-.-._..._. as randomly. random. and randomization)? -...._..-....1 Was the method used to generate the sequence ol randomi/ation described and app10p1iate (table 01 1andom numbers. computer- generated. etc)? vm—ww —. - .__.—..~__.fi.,_- -._ .... _..,._.__ -._.. m-.- _....... --. ‘ L. l - — .- _...—..._..-__.._.__.—~ . Was the study described as randomized (this includes words such Score O'I OYl OVI W as the method of double blinding1 deselibed and a1111it1111 iate O/l (identical placebo. actne placebo dummy. ete)‘. Deduct one point it the method used to generate the sequence ol 1andomization w as described and it was inappropliate (patients were allocated alternately. or according to date of birth. hospital number. etc). Deduct one point if the study was described as double blind but the method ol‘blinding was inappropriate (e.g.. comparison of tablet vs. injection with no double dummy). 78 . _ - —_._ .__._.__.__.._....._._..w.____- " " 1 Was there a desc1 1ption (1t w 1thdravx als and dropouts’ l . "‘""'l' (11 1 (1-1 -_.-._.._. _. ._ _._..__..___._4 0 1 COCHRANE QUALITY ASSESSMENT This score will be used for a sensitivity analysis. No studies will be excluded because of this score. SCORE: / a Items Scores Notes A. Was the assigned 2 = method did not allow Cochrane treatment adequately disclosure ofassignment code (see concealed prior to allocation? 1 = small but possible change of Handbook) disclosure ofassignment A = clearly unclear yes 0 = quasi-randomised or open B = not sure list/tables C = clearly no B. Were the outcomes of 2 : withdrawals well described patients who withdrew and accounted for in analysis described and included in the l = withdrawals described and analysis (intention to treat)? analysis not possible (1 = no mention. inadequate mention or obvious differences and no adjustment C. Were the outcome 2 = effective action taken to assessors blinded to blind treatment status? I = small or moderate chance of unblinding ofassessors 0 I mentioned or not possible D. Were the treatment and 2 : good cmnparability ofgroups control group comparable at or confounding adj ustcd for entry? in analysis I = confounding small; mentioned but not adjusted for O 2 large potential for confounding or not discussed E. Were the subjects blind to 2 = effective action taken to assignment status after blind subjects allocation? 1 = small or moderate chance of unblinding ofsubjects () —= not possible or not mentioned (Sunless double- blind) or possible but not 79 done F. Were the treatment providers blind to assignment? 2 = effective action taken to blind providers 1 = small or moderate chance of unblinding of providers (l : not possible or not mentioned (unless double- blind) or possible but not done (1. Were care programmes other than the trial options identical? 2 = clearly identical 1 = clear but trivial differences 0 = not mentioned or clear and important differences in care programmes H. Were the inclusion and exclusion criteria clearly defined? [J = clearly defined 1 = inadequately defined 0 = not defined 1. Were interventi011s clearly defined? IQ = clearly defined 1 = inadequately defined 0 not defined ll .I. Were the outcome l J l — clearly defined ltem score = measures used clearly I = inadequately defined total defined? 0 = not defined score/numbe r of Outcomes: outcomes 1: Clinical consolidation 2: Radiographic consolidation 3: Pain 4: Function 5: Complications due to stimulation K. Were diagnostic tests used 2 : optimal ltem score = in outcome assessment 1 : adequate total clinically useful? 0 : not defined. not adequate score/numbe r of Outcomes: outcomes 1: Clinical consolidation 2: Radiographic consolidation 3: Pain 4:Funcnon 5: Complications due to stimulation L. Was the duration of 2 = optimal ltcm score = surveillance active and l 4' adequate total clinically appropriate? 0 = not defined. not adequate score/numbc r of 80 Outcomes: outcomes : Clinical consolidation : Radiographic consolidation : Pain :Funcuon : Complications due to stimulation U1 4::- t4) ls) 1— (Punt et alt.. 2004) 8] APPENDIX B Meta-Analysis Forms Relevancy Screening Form Hit number Inclusion Criteria Injury~ delayed andzbr nonunion fractures a! any .vilc __Delayed union- failure of fracture to unite after three months ____Non union-failure of fracture to unite after six months Looking for fractures at all sites. This includes all injuries that lead to non-union like past failed surgeries or stress fractures Does studv meet this criteria? (Circle one) Yes No Unsure Intervention- bane grmrl/z stimulator compared 10 place/1a or sham can/ml (m/j' _Experimental (Jroup- use one ofthree types of bone growth stimulators. Direct Current. Capacitive Coupling. or Pulsed Electromagnetic Fields (PEMF) _Control Group~ Bone growth stimulator must be compared to a sham control (fake BGST) or placebo control (no treatment given). Studies that use active control modalities such as ultrasound. or surgery (e.g. bone grafts) are excluded. Does study meet this criteria? (Circle one) Yes No Unsure Population- herd/11v human pupa/alinn it'll/mu] cancmnilanl disease,”Candi/inns __Uses healthy humans _Subjects cannot have any ofthe following diseases: bone cancer. Insulin Receptor Substrate-l Deficiency. aplastic anemia. osteoperosis. If nothing is mentioned it will be assumed that the population is healthy. Does studv meet this criteria? (Circle one) Yes No Unsure Trial Type- randamixd can/rolled trial or quasi-ramlamizc'd contra/led Iria/ _Randomized Controlled Trial- allocate treatment using random numbers table. computer generated allocation. or sealed envelope _Controlled Clinical Trail (quasi-randomized)- allocate treatment using coin flip. alternative (odd/even number) assignment. patient social security number. days ofthe week. medical record number. etc. Does studv meet this criteria? (Circle one) Yes No Unsure Should the article be included in the Meta-Analysis Yes (all the criteria above marked as Yes) __ No (at least one criteria marked as No) Unsure (at least one criteria marked as Unsure) Data Extraction Form Useable __ Not Useable: Explain why ArticleTitle: Analysis Date: Reviewed bv: Revman ID: INCLUSION CRITERIA- Ifa boxes from each section is not checked/circled, STOP here, check not useable at top, and do not move to Quality Assessment 1) Type of Trial Randamixd Comm/led Y'rial— Studies where the treatments administered are selected by a random process such as the use ofa random numbers table. computer generated allocation. random number generator (ERNIE). or a sealed envelope. D Can/rolledClinical Trial- Treatment allocations using coin fiips. odd-even numbers. patient social security numbers. days ofthe week. medical record numbers. or other such pseudo- or quasi-random processes. 2) Definition ofInjury D Delaved Union- failure to see normal healing of the bone on radiographic evidence within three to six months ofthe injury depending on the fracture site. Cl Nonunion- failure to unite beyond six to nine months. 3) Population C] Healthy human population. no subjects with diseases that may impede in the bone regeneration process (bone cancer. Insulin Receptor Substrate-1 Deficiency. aplastic anemia. osteoperosis') Fill in the following information ifgiven 0 Age range of population 0 Mean age 0 Age specific subgroup results 84 4) Diagnosis of Injury Radiogrcmlzie Evidence (x-ray. bone scan. CT scan. MRI) Type used: 5) Interventions (circle one used, if the study does not fit any of the definitions, stop here and check not usable at the top) 0 Uses Direct Current- uses a generator to deliver electric energy by surgically implanted electrodes into the fusion bed 0 Pulsed electromagnetiefields (PE.\IF)- time varying current that travels through metallic coils at a certain duration and intensity. uses electrodes. does not require surgery 0 Capacitive Conpling- charges two metal plates that are attached to a voltage source and produces electrical field by using electrodes. does not require surgery D 6) Control Group/Type of Placebo (circle one used, if the study does not fit any of the definitions, stop here and check not useable at the top) 0 Placebo Control (control group receives no treatment) 0 Sham Control (fake BUST. no emit energy" emits low levels ofelectrieity proven not to stimulate osteogenesis). 7) Outcome Radiographic Evidence (x-ray. bone scan. CT scan. M Rl) Type used: 'I'ime l’rame (circle one) 0 Finite date to healing set as 0 Serial measurements taken 8) Blinding of Outcome Assessors (check which one applies) Were the outcome assessors blinded to treatment status Outcome assessors were blinded (2) Outcome assessors were not blinded ( l ) Blinding of outcome assessors was not mentioned (0) EXCLUSION CRITERIA- check all that apply, if one or more are checked, stop here and check not usable on top of first page. No type of randomization ofthe experimental or control group Study uses pain scale. mobility measurement. or return to activity as outcome Assessment 85 Study does not use bone growth stimulators Delayed union defined as failure to heal before three months Nonunion defined as failure to heal before six months Study does not use delayed or nonunion fractures Study uses animals Study uses humans with bone cancer. Insulin Receptor Substrate-l Deficiency. aplastic anemia. osteoperosis __ There is no control group in the study ___ Control group treated with ultrasound. hormones. bone grafts. or other types of surgery 86 QUALITY ASSESSMENT I) Jadad-Ifthe article has a score less then three. stop here and check not usable at the top. SCORE: o A .ladad score is calculated using the seven items in the table below. The first five items are indications of good quality. and each counts as one point towards an overall quality score T he final two items indicate poor quality. and a point is subtracted for each if its criteria are met. The range of possible scores is 0 to D. Jadad Score ( alculatron I Item 1 Scorel NN as the study dcscnbcd as randomi/cd (this includes 0” words such as randomly. random. and randornr/atron) ’ Vas the method used to generate the sequence of 0” randomization desc1ibed and appropriate (I table of random " numbers. computer- -generated. etc)? NN as the study described as double blind) ‘ 0” I Was the method of double blinding described and 1 0/1 appropiiate (identical placebo. active placebo. dummy. mete)? NN as there a dcsci iption of withdrawals and dropouts 0/1 I Deduct one point if the method used to gene1 rate the 0/-1 sequence of randomization was described and it was % inappropriate (patients were allocated alternately. or accordin11 to date ofbirth. hospital number. etc). I -. __. fi.fi+4 .1-4 Deduct one point if the study was described as double i Ol-1 blind but the method of blinding was inappr‘opr iate (e.g.. comparison of tablet vs. injection with no double dummy). 87 2) Cochrane Quality assessment tool-This score will be used for a sensitivity analysis. No studies will be excluded because ofthis score. SCORE: / 24 Items Scores Notes A. Was the assigned 2 = method did not allow Cochrane treatment adequately disclosure of‘assignment code (see concealed prior to l = small but possible change of“ Handbook) a“ocadon? disclosure of‘assignment A = clearly yes B = not sure C = clearly no B. Were the outcomes of‘ patients who withdrew described and included in the analysis (intention to treat)? unclear () : quasi-randomised or open list/tables 3 : withdrawals well described and accounted for in analysis 1 = withdrawals described and analysis not possible 0 = no mention. inadequate mention or obvious differences and no adjustment C. Were the outcome 2 it effective action taken to assessors blinded to blind treatment status‘? I = small or moderate chance of unblinding of‘assessors ('l = mentioned or not possible D. Were the treatment and control group comparable at entry? 2 = good comparability of‘ groups or confounding adjusted for in analysis 1 = confounding small; mentioned but not adj usted for O = large potential for confounding or not discussed 16. Were the subjects blind to assignment status after allocation'.’ [J = effective action taken to blind subjects l : small or moderate chance of‘ unblinding of subjects = not possible or not mentioned (unless double- blind) or possible but not ( V 88 done F. Were the treatment providers blind to assignment? 2 = effective action taken to blind providers 1 = small or moderate chance of unblinding of providers 0 : not possible or not mentioned (unless double— blind) or possible but not done G. Were care programmes other than the trial options identical? [J = clearly identical = clear but trivial differences 0 = not mentioned or clear and important differences in care programmes # H. Were the inclusion 2 I clearly defined and exclusion criteria 1 = inadequately defined clearly defined? 0 = not defined 1. Were interventions 2 3 clearly defined clearly defined? 1 = inadequately defined 0 = not defined .1. Were the outcome measures used clearly defined? 2 = clearly defined 1 = inadequately defined 0 = not defined Outcomes: 1: Clinical consolidation : Radiographic consolidation : Pain : Function : Complications due to stimulation 'uJ I») U142. ltem score = total score/ number of outcomes K. Were diagnostic tests used in outcome assessment clinically useful? 2 = optimal l = adequate O : not defined. not adequate Outcomes: : Clinical consolidation : Radiographic consolidation : Pain :Funcfion : Complications due to stimulation Ul-I-‘n'aJlQ-d Item score = total score/num ber of outcomes L. Was the duration of surveillance active and clinically appropriate? 2 = optimal l = adequate 0 = not defined. not adequate Item score = total score/ number of outcomes 89 Outcomes: : Clinical consolidation : Radiographic consolidation : Pain :Funcnon : Complications due to stimulation Q.) [Q— '-J1 4:. (Punt et alt.. 2004) 90 DATA EXTRACTION l) Diehotomous Data- Binary data where each individuals outcome is one of only two possible responses: healed v. not healed Bone Growl/7 S'Ir'mu/u/or ('(m/m/ (iron/1 Number llealed (11) Number of Participants (N) Number llealed (11) Number of Participants (N) 2) Continuous Data- where each individuals outcome is a measure of numerical quantity Bone (from/1 .S'Iinm/umr (. '(mI/‘o/ (iron/1 Number of Participants (N) Mean Standard Deviation (SD) Number of Participants (N) Adean Standard Deviation (SD) 9] REFERENCES Aaron. R., Ciombor. D.. & Simon. B. (2004). Treatment ofNonunions with Electric and Electromagnetic Fields. Clinical ()rllio/meclics, 419. 21-29. Aetna. (2006). Clinical Policy Bulletin: Bone Growth Stimulators [Announcement posted on the World Wide Web]. American Medical Association. Retrieved September 5. 2007 from the World Wide Web: http :/ /'w11-w.aetna.comr’cpbl medical (data/.3 00 3 991/03 43 .htm l. Akai. M.. & Hayashi. K. (2002). Effect ofElectric Stimulation 011 Musculoskeletal Systems: A Meta-Analysis ofControlled Clinical Trials. Bioelecu'(1nu'ignelic.1'. 23(2). 132-143. Antonsson. E.K.. & Mann. R.W. (1985). The frequency content of gait. Journal of Biomechanics. 18. 39-47. Bassett. C.. & Becker. R. (1962). Generation ofelectric potentials by bone in response to mechanical stress. Science. 137. 1063-4. Bassett. C.. Pilla, A.. & Paw'luk. R. (1977). A non-operative salvage of surgically- resistant pseudarthroses and non-unions by pulsing electromagnetic fields. Clinical ()rllmpaeclics and Related Research. 124. 128-143. Bhandari. M.. Richards. R.. Sprague. S.. & Schemitsch. E. (2001). Letter to the Editor: Quality in the Reporting of Randomized Trials in Surgery: Is the Jadad Scale Reliable? Controlled Clinical Trials, 22. 687-688. Brighton. C.T.. Black. J.. & Friedenberg. Z.B. (198] ). A multicenter study ofthe treatment of non-union with constant direct current. Journal ofBone and Joint Surgery. 63. 2-13. Bruser. P.. & Gilbert. A. (1999). Finger Bone and Join! Injuries. London: Martin Dunitz, Ltd. Clark. H., Wells. G.. Huet. C.. McAlister, F.. et al. (1999). Assessing the Quality of Randomized Trials: Reliability of the Jadad Scale. Con/rolled Clinical Trials, 20. 448-452. Clark. 0.. Castro. A.. Filho. J.. & Djubelgovic. B. (200] ). Interrater Agreement of Jadad’s Scale. Proceedings" from the 9m Internalional Cochrane Colloquium, 31. Retrieved November 12. 2007. from BioMed Central. Caudle. R..l.. & Stern. P.J. (1987). Severe open fractures ofthe tibia. ./ Bone Joint Surg Am, 69. 801-807. 93 Coosemans. W.. Rommens. P.. Broos. P.. & Gruwez. J. (I988). Specific problems in the management of segmental fractures of the tibial shaft [Specifieke behandelingsproblemen bij de bifocale tibiaschaftfractuur]. .~fcta ('lzirurgica Belgica. 88(5). 347-53. Diniz. P.. Shomura. K.. Soejima. K.. & Ito. G. (2002). Effects of Pulsed Electro111agnetic Field (PEMF) Stimulation 011 Bone Tissue Like Formation Are Dependent on the Maturation Stages ofthe Osteoblasts. Bioeleclrontagnetics. 23. 398-405. Einhorn. T. ( I 995). Enhancement of fracture-healing. Journal (if/Jone & Joint .S'urgert’. American I'olume. 77(6). 940-56. Friedenberg. 7... & Brighton. C. (I966). Bioelectric potentials in bone. Journal of Bone (fr Joint Surge/11'. xllHUI'lL'CHI l’olume. 48(5). 915-23. F ukada. Ii. & Yasuda. I. (I957). 011 the piezoelectric effects of bone. .N'ippon .1'eit'igaltu :asslzi. [Journal oft/7c l’li'taviological Society of Japan / . 12(1 1 ). 58-69. Glasziou. P.. Irwig. L.. Bain. C.. & Colditz. G. (2001). Stavte/natic Reviews in Health Care. A Practical Guide. Cambridge: Cambridge University Press. Hall. B. (I990). Bone. Nova Scotia: CRC Press. Higgins .l.. & Green S. (2007). ('ocltrane Handbook/Or Suvtematic Reviews of Interventions 4.2.X. Retrieved August I3. 2007. from The Cochrane Collaboration Web site: 111.111.2211;11:11:...11111‘hlane.dkr’cocltran 1 h and book/7111111.11 I1_1_1_11k._|11111. Hulme. J.. Robinson. V.. DeBie. R.. Wells. G.. .Iudd. M.. & Tugw'ell. P. (2002). Electromagnetic fields for the treatment ofosteoarthritis. Cochrane Database o/Sj'stetnatic Reviews. Issue I. Art. No.: C D003523. DOI: 10/1002/14651858.CD003523. Jadad. A. R.. Moore. A.. & Carroll. D. (1996). Assessing the Quality of Reports of Randomized Clinical Trials: ls Blinding Necessary? (on/rolled Clinical 7rials. l7, l—l2. Karamitros. A.. Kalentzos. V.. & Soucacos. P. (2006). Electric stimulation and hyperbaric oxygen therapy in the treatment ofnonunions. Injury. 37(1 ). 863-73. Khan. K.. Kun7.. R.. Kleijnen. .l.. & Antes. G. (2003). .S:1'.1'teimttic Reviews to Support Evidence-based .‘I/eclicine. How to Review and Apply Findings of Healthcare Research. London: The Royal Society ofMedicine and Press. 114 LaVeIIe. D. ( I 998). Delayed union and nonunion of fractures. Campbell ’s operative orthopaedics. 9. 2579 -629. Liboff. AR. (2006). Encyclopedia ofleedical Devices and h2.1'trutnentation. Rochester. MI: John Wiley and Sons. Inc. Luben. R. (1991). Effects oflow-energy electromagnetic fields (pulsed and DC) on membrane signal transduction processes in biological systems. Health Physics. 61 . 15—28. Marieb. E. ( 2001 ). Human Anatomy and l’lnzviology. San Francisco: Benjamin Cummings. Marsh. D. (I998). Concepts of Fracture union. delayed union. and nonunion. (i'linical Orthopedic Related Resources, 355. 22-30. Masarni. A.. & Hayashi. K. (2002). Effects ofEIectrical Stimulation on Musculoskeletal Systems; A Meta-Analysis ofControlled Clinical Trials. Bioelectromagnetics. 23. I 32—143. McLeod. K..l.. & Rubin. CT. (1990). Frequency specific modulation of bone adaptation by induced electric fields. Journal ofTheoretical Biology, 145. 385-396. McLeod. K..l.. Rubin. C.T.. & Donahue. HI. (1995). Electromagnetic fields in bone repair and adaptation. Radio Science. 30. 233-244. Moja. P.. Telaro. E.. D’Amico. R.. Moschctti. I.. et al. (2005). Assessment of methodological quality of primary studies by systematic reviews: results ofthe metaquality cross sectional study. Bill]. 330. 1053. Norton. I .. (I974). In vivo bone growth in a controlled electric field. Ann. A". I'. Academy Science. 238. 466-477. Oishi. M.. & Onesti. S. (2000). Electric Bone Graft Stimulation for Spinal Fusion: A Review. Neuro.1'zu'gery. 47(5). 1041-1056. Otter. M.. McLeod. K.. & Rubin. C. (1998). Effects ofElectromagnetic Fields in Experimental Fracture Repair. Clinical Orthopaedics and Related Research, 355. 890-8 1 04. Panagiotis. M. (2006). Classification of non-union. Injury, 37(9). 927. Phieffer. L.. & Gould. J. (2006). Delayed Unions of the Tibia. The Journal of Bone and Joint Surgery (American). 88. 205-216. 95 Punt. B.. den Hoed. P.. & Stijnen. T. (2004). Electromagnetic Field stimulation for the treatment ofdelayed union or non-union of long bones. Cochrane Database o/Ststematic Reviews, Issue 4. Art. No.: C D004960. DOI: I 0. I 002/] 465 I 858.CD004960. Review Manager (RevMan) [Computerprogram]. Version 4.2 for Windows. Copenhagen: The Nordic Cochrane Centre. The C ochrane Collaboration. 2003. Riemer. B.L.. DiChristina. D.G.. & Cooper. A. (1995). Nonrearned nailing of tibial diaphyseal fractures in blunt polytrauma patients. J Orthop Trauma. 9, 66-75. Ryaby. J. (1998). Clinical Effects ofElectromagnctic and Electric Fields on Fracture Healing. Clinical ()rtlmpaedics and Related Research. Fracture Healing Enhancement. 335. 205-2l5. Saleh. K.. & Hak. D. (2001. August). Socioeconomic Burden of Traumatic Tibial Fractures: Nonunion or Delayed Union. II-"ehrI/D, Retrieved September 9. 2007, from Medscape database. Sanders. R.. Jersinovich. I.. Anglen. J.. DePasquale. T.. & Herscoviei. D. (I994). The treatment of open tibial shaft fractures using an interlocked intramedullary nail without reaming. J Orthop Trauma. 8. 504-510. Schubert. H. (2000). Scaphoid Fracture Review ofDiagnostic Tests and Treatment. Canadian Family Physician. 46. 1825-1832. Sharrard. W. (1990). A double-blind trial of pulsed electromagnetic fields for delayed union 011 tibial fractures. Journal ofBone and Joint Surgerv, 728. 347-355. Towheed. T. (2006). Is the .ladad Score the Proper Evaluation of Trials? The Journal otheumato/ogt’. 33(8). l7I0-I7I I. Venes. D., & Thomas. C .I.. (I997) Taher ’s ( 'yclopedic lIv/edical Dictionary. Philadelphia: PA. Davis Company. Watson. .l. (1979). The Electrical Stimulation of Bone Healing. Proceedings of the IEEE, 67(90). 1339-1351. Younger. E.M.. & Chapman. M.W. ( I989). Morbidity at bone graft donor sites. .I Orthop Trauma. 3. 192. (2006). Proposed Reclassificationfor Bone Growth Stimulator Devices. 96 Questions/Or the ()rthopeadic and Rehabilitation Devices Panel. Retrieved October 28. 2007. from The US Food and Drug Administration Website: http://www.fda.govldefault.l1tm. REFERENCES TO STUDIES Included Studies Barker. A.. Dixon. R.. Sharrard. NN".. & Sutcliffe. M. (I984). Pulsed Magnetic Field Therapy for Tibial Non-Union. Interim Results ofa Double-Blind Trial. Lancet. 8384. 994-6. Scott. G.. & King. .I. (1994). A Prospective Double-Blind Trial ofEIectrical Capacitive coupling in the Treatment ofNon-Union of Long Bones. Journal o/Bone and Joint Surgery. 76(6). 820-6. Sharrard. W. (I990). A Double-Blind trial ofPulscd Electromagnetic field for Delayed Union ofTibiaI fractures. Journal oj'Bone and Joint Surgerv. 72(3). 374-55. Simonis. R.. Parnell. E.. Ray. P.. & Peacock. .l. (2003). Electrical Treatment of tibial non-union: A Prospective randomized. double-blind trial. Injury. 34(5). 357-62. Excluded Studies Abdel-Salam. A.. Eyres. K.. & Cleary. .l. (199] ). Internal Fixation ofclosed Tibial fractures for the Management of Sports Injury. British Journal o/‘Sports .l/edicine, 25(4). 213-7. Adams. C.. Robinson. C.. Court-Brown. C.. & McQueen. M. (2001). Prospective randomized controlled trial of an Intramedullary nail versus Dynamic screw and plate for Intertrochanteric fractures of the Femer. Journal of ()rthopeadic Trauma, 15(6). 344-400. Adolfsson. L. Lindau. T.. & Amer. M. (2001). Actruak screw fixation versus Cast immobilization for Undisplaccd Scaphoid wrist Fractures. Journal of 1 land Surgery, 26(3). 192-5. Aldophson. P.. Abbazadegan. H.. Boden. H.. & Salernyr. M. (2000). Clodronate increases Mineralization ofCaIlus after C ollcs‘ Fractures: a Randomized. Double-Blind. Placebo-Controlled. Prospective Trial in 32 Patients. Acta Orthopaedic Sectndim’tvian, 7 I (2), I 95-200. 97 Agorastides. 1.. Sinopidis. C.. E] Meligy. M.. & Yin. Q. (2007). Early versus Late Mobilization after Hemiarthoplasty for Proximal Humeral Fractures. Journal ofShou/der and Elbow Surge/1v, l6(3). S33-8. Ahl. T.. Dalen. N.. Lundberg. A.. & NN'ykman. A. (1994). :lcla Orthopaedic Scandanavian. 65(2). 166-70. Ahrengart. L.. Tornkvist. H.. Fornander. P.. & Thorngren. K. (2002). A Randomized Study ofthe Compression Hip Screw and Gamma Nail in 426 Fractures. Clinical Orthopaedic and Related Research. 401. 209-22. Ahuja. S.. & Chatterji. S. (2002). The Mennen Femoral Plate for Fixation of Periprosthetic Femoral Fractures following Hip Arthoplasty. lnjurv. 33( I ). 47-50. Akesson. K.. Vergnaud. P.. Delmas. P.. & Obrant. K. (I995). Serum Osteocalcin increases during Fracture Healing in Elderly Women with Hip Fractures. Bone, l6(4). 427-30. AI_Belasy. F. (2005). A Short Period of h/Iaxillon1a11tlibt1la1‘ Fixation for the Treatment of Fractures ofthe Mandibular tooth-bearing area. Journal of Oral and .llaxillo/acial Surgery 63(7). 1456-8. AI-Sukhun. J.. & Lindgvist. C. (2006). A Comparative Study of Two Implants Used to Repair Inferior Orbital Wall bony Defects: Autogenous Bone Graft versus Bioresorbable Poly-L/DL-Iactidc Plate. Journal oj'Oral and .‘Ilaxillofacial Surgery 64( 7). l038-48. Aladin. A.. & Davis. T. (2005). Dorsal Fracutre-dislocation ofthe Proximal Interphalangeal Joint: A Comparative Study of Pereutaneous Kirschner Wire Fixation versus Open Reduction and Internal Fixation. Journal of 1 land Surgery, 30(2 ). 120-8. Alho. A.. Benterud. J.. Ronningcn. H.. & Hoiseth. A. (I992). Predication of Disturbed Healing in Femoral Neck Fracutre. Radiographic Analysis of 149 Cases. Acta Orthopaedic Scandartavian. 63(6). 639—44. Altay. T.. Ozturk. H.. Us. R.. & Gunal.. 1. (I999). Four-part Posterior Fracture- Dislocations of the Shoulder. Treatment by Limited Open Reduction and Pereutaneous Stabalization. A cta Orthopaedic Trauma and Surgery, I99. l-2. An. II.. Simpson. J.. Glover. J.. & Stephany. J. (1995) Comparison between Allograft plus Dcmineralized Bone Matrix versus Autograft in Anterior Cervical Fusion. A Prospective Multicenter Study. Spine, 20(20). 221 l-6. 98 Andermahr. J.. Elsner. A.. Brings. A.. & Hensler. T. (2006). Reduced Collagen Degredation in Polytraurnas with Traumatic Brain Injury causes Enhanced Osteogenesis. Journal of.N'eurotraurmr. 23(5). 708-20. Antich. A.. Androver. P.. IN/Iarti-(i21ri11. D.. IN/Iurias-Alvarez. J.. & Puente-Alonso. C. (I997). Iixternal Fixation and Secondary Nailing ofOpen Tibial Fractures: A Randomized. Prospective Trial. Journal ofb’one andJoint Surgery. 79(3). 433-7. Asazuma. T.. Masuoka. K.. INI’lotosuneya. T.. & Tsuji. T. (2005). Posterior lumbar Interbody Fusion using Dense Hydrozyapatite blocks and Autogenous Iliac Bone: Clinical and Radiographic Examinations. Journal of‘Spinal Disorder Technology. I8. 4I-7. Atkin. D.. Bohay. D.. Slabaugh. P.. & Smith. B. (I995). Treatment of Ulnar Shaft Fractures: 21 Prospective Randomized Study. Orthopedics, 18(6). 543-7. Atkins. R.. Sudhakar. J.. & Porteous. A. (1998). Use of modified Ilizarov olive wires as pushing wires. Journal ofOrthopedic Trauma, 12(6). 436-8. Audige. L.. Griffin. D.. Bhandari. M.. & Keliam. J. (.2005). Path Analysis of factors for Delayed Healing and Nonunion in 416 Operatively Treated Tibial Shaft Fractures. Clinical Orthopedics and Related Research. 438. 22l-32. Bach. A.. & Hansen. S. (I989). Plates versus External Fixation in Severe Open Tibial Shaft Fractures. A Randomized Trial. (.‘linical Ort/mpedics and Related Research, 24 l . 89-94. Bain. G.. Zacrest. A.. Paterson. D.. & Middleton. J. ( I997). Abduction Strength following Intramedullary Nailing of the Femer. Journal o/‘Orthopedic Trauma. I 1(2). 93-7. Bansal. R.. & Craigen. M. (2007). Fifth Metacarpal Neck Fractures: is follow-up required? Journal of '1 land and Surgery Europe. 32(1). 69-73. Bar-On. E.. Sagiv. S.. & Porat. S. (I998). External fixation or flexible intramedullary nailing for Femoral Shaft Fractures in Children. Journal of Bone and Joint Surgery. 80(4). 891-2. Barquct. A.. Francescoli. I... Rienzi. D.. & Lopez. L. (2000). Intertrochanteric- subtrochantcric Fractures: Treatment with the long Gamma Nail. Journal ofOr/hopedie Trauma, 14(5). 324-8. Barrios. C.. Brostrom. I... Stark. A.. & Walheim. G. (I993). Healing 99 Complications after Internal Fixation of 'I‘rochanteric Hip Fractures: the Prognostic value of Osteoporosis. Journal ofOr/hopedic Trauma. 7(5). 438-42. Bartonicek. 1.. & Dousa.. P. (2002). Prospective Randomized Controlled Trial of an Intramedullary nail versus Dynamic screw and plate of Intertrochanteric Fractures ofthe F emur. Journal o/'()rthopedic Trauma. 16(5). 363-4. Bast. S.. Hoffer. M.. & Aval. S. (1998). Nonoperative Treatment for Minimally and Nondisplaced Lateral Condyle Fractures in Children. Journal of Pediatric Orthopedics. 18(4). 448-50. Baumgaertel. F.. Dahlen. C.. Stiletto. R.. & Gotzen. L. (1994). Technique of using the AO-femoral Distractor for Femoral Intramedullary Nailing. Journal of ()rtl‘zopedic Trauma. 8(4). 315-21. Behrens. F.. & Searis. K. (1986). Iixternal Fixation ofthe Tibia. Basic Concepts and Prospective livaluation. Journal of'Bone andJoint Surgery, 68(2). 246-54. Benazzo. F.. Mosconi. M.. Becarisi. G.. & Galli. U. (1995). Use of Capacitive coupled Electric Fields in Stress Fractures in Athletes. (.‘linical Orthopedics and Related Research, 310. 145-9. Biachut. P.. O'Brien. P.. Meek. R.. & Boekhuyse. H. (1997). Interlocking Intramedullary Nailing with and without Reaming for the Treatment of closed Fractures ofthe Tibial Shaft: a Prospective. Randomized Study. Journal of'Bone andJoint Surge/1r. 79(5). 640-6. Bibbo. C.. & Patel. D. (2006). The Iiffect of Demineralized Bone Matrix-Calcium Sulfate with Vancomycin on Calcaneal Fractures Healing and Infection Rates: A Prospective Study. Foot and Ankle International. 27(7). 487-93. Billie R.. Simic. P.. Jelic. M.. & Stern-Padovan. R. (2006). Osteogenic protein-l (BMP-7) accelerates healing of Scaphoid Non-Union with Proximal Pole Sclerosis. International Orthopedic, 30(2). 128-34. Bischoff. R.. Buechler. [J.. De Roche. R.. & .lupiter. .1. (1994). Clinical Results of Tension Band fixation ofAvuIsion Fractures ofthe Hand. Journal oflland .S'urgerj’, 19(6). 1019-26. Blum. J.. .lanzing. H.. Gahr. R.. & Langendorff. II. (2001). Clinical Performance ofa new medullary humeral nail: Antegrade versus Retrograde Insertion. Journal of'Urthopedic Trauma. 15(5). 342-9. 1 ()0 Blum. J.. Rommens. P.. & .lanzig. H. (1997). The Unreamed Humeral nail- a Biologial Osteosynthesis of the Upper Arm. .‘lcta ('hir Belg. 97(4). 184-9. Bombaci. 11.. Gereli. A.. Kucukyazici. 0.. & (iorgec. M. (2005). A New Technique ofCrossed pins in Supracondylar Iilbow Fractures in Children. Orthopaedics. 28( 12 ). 1406-9. Bond. C.. Shin. A.. McBride. M.. & Dao. K. (2001 ). Pereutaneous Screw Fixation or Cast Immobilization for Nondisplaced Scaphoid Fractures. Journal of'Bone and Joint Surge/1v. 83(4). 483-8. Bong. M.. Capla. E.. Egol. K.. & Sorkin. A. (2005). Osteogenic Protien-l (bone morpohogenic protein-7) Combined with Various Adj uncts in the Treatment ofHumeral Diaphyscal Nonunions. Bull Hosp .lt Dis. 63(1-2). 20-3. Boriani. S.. De lure. F.. Bettelli. G.. & Specehia. L. (1994). The results ofa multicenter Italian study on the use ofthe Gamma nail for the treatment of pertrochanteric and subtrochantcric fractures: a review of 1 181 cases. ('hir Organi .‘l/()\'.79(2). 193-203. Borrelli. .l.. Ricci. W.. Anglen. .l.. Gregush. R.. 8; lingsberg. .l. (2006). Muscle strength recovery and its effects on outcome after open reduction and internal fixation of acetabular fractures. Journal of Orthopedic Trauma. 20(6). 388-95. Borsalino. G.. Bagnacani. M.. Bettati. E.. & Fornaciari. F. (1988). Electrical stimulation ofhuman femoral intertrochanteric osteotomies. Double-blind study. ('linical Orthopaedics and Related Research. 237. 256-63. Braakman. M.. Oderwald. E.. & Ilaentjens. M. (1998). Functional taping of fractures of the 5th metacarpal results in a quicker recovery. ln/urt'. 29( 1 ). 5-9. Brara. H.. & Fessler. R. (2000). The role of anterior lumbar interbody allograft bone dowel fusion as an adjunct to posterior segmental lumbar fixation. Clinical Neurosurgery. 47. 528-33. Braten. M.. Helland. P.. Grontvedt. T.. & Aamondt. A. (2005) External fixation versus locked intramedullary nailing in tibial shaft fractures: a prospective. randomized study of78 patients. .-lrchires ofOr/hopedic Trauma and Surge/"1’ 125(1). 21 -6. Brighton. C.. Shaman. P.. Heppenstall. R.. Esterhai. .l.. Pollack. S.. & 101 F riedenberg. Z. (1995). Tibial nonunion treated with direct current. capacitive coupling. or bone graft. Clinical Orthopedic and Related Research. 321. 223-34. Brighton. C.. Black. J.. Friedenberg. K.. Esterhai. J.. Day. I... 8; Connolly. J. (1981). A multicenter study of the treatment of non-union with constant direct current. Journal ofBone and Joint Surgery’. 63(1 ). 2-13. Broos. P.. & Rcyndcrs. P. (1998). The unreamed AO femoral intramedullary nail. advantages and disadvantages ofa new modular interlocking system. A prospective study of67 cases. Acta Orthopedic Belgium. 64(3). 284-90. Broos. P.. & Vanderspeeten. K. (1997). Our first experiences with the Unreamed Femoral Nail (UFN). Acta ('hir Belg. 97(1). 27-32. Buciuto. R.. & Hammer. R. (2001 ). RAB-plate versus sliding hip screw for unstable trochanteric hp fractures: stability of the fixation and modes of failure--radiographic analysis of218 fractures. Journal of Trauma. 50(3). 545-50. Buciuto. R.. Hammer. R.. & Herder. A. (1997). Spontz‘tneous subcapital femoral neck fracture after healed trochanteric fractures. ('linical Orthopedic and Related Reseat'cl‘t. 342. 156-63. Buckley. R.. Tough. S. McCormack. R.. & Pate. G. (2002). Operative compared with nonopertive treatment of displaced intra-articular calcaneal fractures: a prospective. randomized. controlled multicenter trial. Journal ofBone and Joint Surge/:1", 84(10). 1 733-44. Burd. T.. Hughes. M.. & Anglen. .l. (2003). Heterotopic ossification prophylaxis with indomethacin increases the risk of long-bone nonunion. Journal of Bone and Joint Surgery. 85(7). 700-5. Cakirgil. G.. Saplakoglu. A.. & Yazar. T. (1989).le compared effect ofa four- coiled system in pulsed electromagnetic field stimulation. Orthopedics. 12(11).1481—4. Calhoun. J .. Henry. S.. Anger. D.. & Cobos. J. (1993). The treatment ofinfected nonunions with gentamicin-polymethylmethacrylate antibiotic beads. Clinical Orthopedics and Related Research. 295. 23-7. Canadian Orthopaedic Trauma Society. (2003). Nonunion following intramedullary nailing of the femur with and without reaming. Results ofa multicenter randomized clinical trial. Journal ofBone and Joint Surgery. 85(11). 2093-6. 102 (2007). Nonoperative treatment compared with plate fixation ofdisplaced midshaft clavicular fractures. A multicenter. randomized clinical trial. Journal (rf'Bone and Joint Surgery 89(8). 1866-7. Capanna. R.. Donati. D.. Masetti. C.. & Manfrini. M. (1994). Effect of electromagnetic fields on patients undergoing massive bone graft following bone tumor resection. A double blind study. Clinical Orthopedics and Related Research. 306. 213-21. C arragee. E. (1997). Single-level posteriolateral arthrodesis. with or without posterior decompression. for the treatment of isthmic spondylolisthesis in adults. A prospective. randomized study. Journal of'Bone and Joint Surgery. 79(8). 1 175-80. Caruso. G.. Lagalla. R.. Derchi. L.. & Iovane. A. (2000). Monitoring of fracture calluses with color Doppler sonography. Journal o/‘Clinical (_fltrasouml. 28(1). 20-7. Cassinelli. E.. Young. 8.. Vogt. M.. & Pierce. M. (2005). Spica cast application in the emergency room for select pediatric femur fractures. Journal of Orthopedic Trauma, 19(10). 709-16. Castillo. R.. Bosse. M.. MacKenzie. E. & Patterson. B. (2005). Impact of smoking on fracture healing and risk of complications in limb-threatening open tibia fractures. Journal o/‘Orthopea’ic Trauma. 19(3). 151-7. Ceffa. R.. Bombelli. M.. Boero. S.. & Marre lBrnenghi. G. (1997). Extensimetric monitoring of healing in the treatment with the llizarov apparatus. A multicenter clinical trial. Bull Hospital and Joint Disease. 56(1). 41-5. Chan. Y.. Ueng. S.. Wang. C.. & Lee. S. (2000). Antibiotic-impregnated autogenic cancellous bone grafting is an effective and safe method for the management of small infected tibial defects: a comparison study. Journal ofTraurna, 48(2). 245-55. Chapman. J.. Henley. M.. Age]. J.. & Benca. P. (2000). Randomized prospective study of humeral shaft fractures fixation: intramedullary nails versus plates. Journal ofOrthopedic Trauma. 14(3). 162-6. C hapman. M.. Bucholz. R.. & Cornell. C. (1997*). Treatment of acute fractures with a collagen-calcium phosphate graft material. A randomized clinical trial. Journal of'Bone and Joint Surgery. 79(4). 495-502. Charalambous. C.. Siddique. 1.. Zenios. M.. & Roberts. S. (2005). Early versus 103 delayed surgical treatment ofopen tibial fractures: effect on the rates of infection and need of secondary surgical procedures to promote bone union. Injury: 36(5). 656-61. Chen. C.. Chen. W.. 8: Shih. C. (2002). Surgical treatment for distal clavicle fracture with coracoclavicular ligament disruption. Journal ofTraurrra. 52(1). 72-8. Chen. A.. Hou. J.. Bao. J.. & Guo. S. (1998). Comparison ofbiodegradeble and metallic tension-band fixation for patella fractures. 38 patients followed for 2 years. Acta Orthopedic Scarularurvia. 69(1). 39-42. Chevalley. F.. & Gamba. D. (1997). Gamma nailing of pertrochanteric and subtrochantcric fractures: clinical results ofa series of63 consecutive cases. Journal ofOrthopedic Trauma. 1 1(6). 412—5. Chiu. F.. Chen. C.. Lin. C.. & 1.0. W. (1997). Closed humeral shaft fractures: a prospective evaluation ofsurgical treatment. Journal o/Trauma. 43(6). 947-51. Chiu. F.. Lo. W Chen. C.. & Chen. T. (1996). Unstable closed tibial shaft fractures: a prospective evaluation ofsurgical treatment. Journal of Trauma. 40(6). 987-91. Chu. 1 Lo. W.. Chen. C.. & Chen. T. (1996). 'l‘reatmcnt of unstable tibial fractures with interlocking nail versus Ender nail: a prospective evaluation. eronghua l'i Xue Za Zhi. 57(2). 124-33. Cho. T.. Chol. 1.. Chung. (7.. & Yoo. W. (2006). Isolated congenital pseudarthrosis ofthe fibula: clinical course and optimal treatment. Journal ofl’ediatr'ic Orthopedics. 26(4). 449-54. Choi. 8.. Kim. 11.. Kim. M.. & Han. S. (2005). IVIanagement of mandibular angle fractures using the mandibular angle reduction forceps. lnternatiorurl Journal of‘Oral and .lla.\‘r'llo/acial Surgery. 34(3). 257-61. Choi. 8.. Huh. J.. & Yoo. J. (2003). Computed tomographic findings of the fractured mandibular condyle after open reduction. lnternatiorurl Journal of'Oral and rlrlaxillotacial Surgery. 32(5). 469-73. Christie. J.. Howie. C.. & Armour. P. (1988). Fixation ofdisplaced subcapital femoral fractures. Compression screw fixation versus double divergent pins. Journal o/Bone ant/Joint Surgery. 70(2). 199-201. Clatworthy. M.. Clark. D.. Gray. 1).. & Hardy. A. (1998). Rcarned versus 104 unreamed femoral nails. A randomized. prospective trial. .lourrurl ot'Bone and Joint Surgery. 80(3). 485-9. Clinkscales. C.. & Peterson. H. (1997). Isolated closed diaphyseal fractures ofthe femur in children: comparison ofeffectiveness and cost of several treatment methods. Ortholwdies. 20( 12). 1 131—6. Cohen. A.. & Shaw. D. (2001). Focused rigidity casting: a prospective randomised study. J R Coll Surg lidinh. 46(5). 265-70. Cole. R.. Bindra. R.. Evanoff. B.. & Gilula. 1.. (1997). Radiographic evaluation of osseous displacement follrming intra-articular fractures of the distal radius: reliability ofplain radiographys versus computed tomography. Journal ofllandSurge/11'. 22(5). 792-800. Collinge. C.. Devinney. S.. llerscovici. D.. & DiPasquale. T. (2006). Anterior- inferior plate fixation of middle-third fractures and nonunions of the clavicle. Journal ot'Orthopedic Trauma, 20( 10). 680-6. Colteriohn. N.. & Bednar. D. (1997). Procurement of bone graft from the iliac crest. An operative approach with decreased morbidity. Journal of'Bone andJoint Surgery. 79(5). 756-9. Coupland. R.. Hoikka. V.. Sjoeklint. 0.. & Cuenod. P. (1992). Assessment of bullet disruption in armed conflicts. Lancer. 339(8784). 35-7. Court-Brown. C.. Walker. C.. Garg. A.. & McQueen. M. (1999). Half-ring external fixation in the management oftibial plafond fractures. Journal of Orthopedic Trauma. 13(3). 200-6. Crutchfield. E.. Seligson. 1).. Henry. S.. & Warnholtz. A. (1995). Tibial pilon fractures: a comparative clinical study of management techniques and results. Orthopedics. 18(7). 613-7. Cundy. P.. & Paterson. D. (1990). A ten-year review of treatment ofdelayed union and nonunion with an implanted bone growth stimulator. Clinical Orthopedic and Related Research. 259.216-22. Dazinger. M.. Caucci. D.. Zecher. S.. & Segal. D. (1995). Treatment of intercondylar and supracondylar distal femur fractures using the GSH supracondylar nail. .»‘frnerican Journal ofOrthopedics, 24(9). 684-90. Dc Ridder. V.. de Lange. S.. Koomen. A.. & lleatley. F. (2001). Partridge osteosynthesis: a prospective clinical study on the use ofnylon cerclage bands in the treatment of periprosthctic femoral shaft fractures. Journal of ' Orthopedic Trauma. 15(1). 61-5. 105 Deshmukh. R.. Lou. K.. Neo. C.. & Yew. K. (1998). A technique to obtain correct rotational alignment during closed locked intramedullary nailing ofthe femur. Injury. 29(3). 207-10. Dhakai. A.. & Awais. S. (2003). l\rlanagement of femur shaft fracture with close locked intramedullary nailing: comparison ofVariwall and predrilled Kuntscher nail. Nepal Med Coll Journal. 5(2). 73-8. Dias. J.. Wildin. C.. Bhowal. B.. & Thompson. .1. (2005). Should acute scaphoid fractures be fixed? A randomized controlled trial. Journal (rt/Jone and Joint Surgery. 87( 10). 2160-8. Dickson. K.. Friedman. J.. Buchholz. J.. & Flandry. F. (2002). The use of BoneSource hydroxyapatite cement for traumatic metaphyseal bone void filling. Journal of Trauma. 53(6). 1 103-8. Doetsch. A.. Faber. J.. Lynnerup. N.. 8; \\'-"at_ien. I. (2004). The effect ofcalcuim and vitamin D3 supplementation on the healing of the proximal humerus fracture: a randomized placebo-controlled study. Calcification Tissue lnternatiorurl. 75(3). 183-8. Doi. K.. & Sakai. K. (1994). Vascularized periosteal bone graft from the supracondylar region ofthe femur. rllicrosurgerjt. 15(5). 305-15. Domb. B.. Sponseller. P.. Ain. M.. & Miller. N. (2002). Comparison of dynamic versus static external fixation for pediatric femur fractures. Journal of Pediatric Orthopedic. 22(4). 428-30. Donati. D.. Capanna. R.. Campanacci. D.. & Del Ben. M. (1993). The use of massive bone allografts for intercalary reconstruction and arthrodeses after tumor resection. A multicentric European study. ('hir Organi _l/or. 78(2). 81-94. Dujardin. F.. Bencz. C.. Polle. G.. & Alain. J. (2001). Prospective randomized comparison between a dynamic hip screw and a mini-invasive static nail in fractures ofthe trochanteric area: preliminary results. Journal of Orthopedic Trauma, 15(6). 401 -6. Durrnus. A.. Cakmak. A.. Disci. R.. & Muslumanoglu. L. (2004). The efficiency ofelectrornagnetic field treatment in Complex Regional Pain Syndrmne Type 1. Disability and Rehabilitation. 26(9). 537-45. Earnshaw. S.. Aladin. A.. Surendran. S.. & Moran. C. (2002). Closed reduction of 1 06 colles fractures: comparison ofmanual manipulation and finger-trap traction: a prospective. randomized study. Journal o/Bone and Joiru .S'urgerv. 84(3). 354-8. Ebinger. T.. Roeseh. M.. \N’achter. N.. & Kinzl. L. (2001). Functional treatment of physeal and periphyseal injuries ofthe metacarpal and proximal phalangcal bones. Journal o/l’ediatric .S'urgert’. 36(4). 61 1-5. Egol. K.. Su. E.. Tejwani. N.. & Sims. S. (2004). Treatment ofcomplex tibial plateau fractures using the less invasive stabilization system plate: clinical experience and a laboratory comparison with double plating. Journal of Trauma. 57(2). 340-6. Egol. K.. Dolan. R.. & Koval. K. (2000). Functional outcome of surgery for fractures of the ankle. A prospective. randomized comparison of management in a cast or a functional brace. Journal of'Bone and Joint Surge/1v. 82(2). 246-9. Ekelund. I-.. Hagberg. I... Horberg. L.. & .lorgsholrn. P. (2007). Imaging of four- corner fusion (SLAC arthrodesis) of the wrist with 64-slice computed tomography. A cta Radiol. 48(1). 76-9. El-Mowafi. 11.. Abulsaad. M.. & El-Adi. W. (2005). Prophylactic surgical correction ofCrawford's type II anterolateral bowing ofthe tibia using llizarov's method. Acta Orthop Belg. 71(5). 577-81. El-Said. N. (1999). Osteotomy ofthe tibia for correction ofcomplex deformity. Journal of'Bone and Joint .S'urgerjt 81(5). 780—2. Ernarni. A.. Larsson. A.. Petren-Mallmin. M.. & Larsson. S. (1999). Serum bone markers after intramedullary fixed tibial fractures. ('linical Orthopedic and Related Research. 368. 220-9. Epstein. N. (2002). Anterior dynamic plates in complex cervical reconstructive surgeries. Journal ofSpir-ral Disorder Techniques, 15(3). 221 -7. Epstein. N. (2000). Anterior cervical diskectomy and fusion without plate instrumentation in 178 patients. Journal of Spinal Disorders. 13(1). 1-8. Etchebehere. E.. Caron. M.. Pereira. J.. & Lima. M. (2001). Activation of the growth plates on three-phase bone scintigraphy: the explanation for the overgrowth of fractures femurs. European Journal ofNuc/ear Medicine. 28(1). 72-80. Eyres. K.. Saleh. M.. & Kanis. J. (1996). Effect of pulsed electromagnetic fields 107 on bone formation and bone loss during limb lengthening. Bone. 18(6). 505-9. Eyres. K.. Saleh. M.. & Kanis. .l. (2004). Effect of pulsed electromagnetic fields on bone formation and bone loss during limb lengthening: a preliminary report. Journal o/Bone andJoint Surgery. 76. 92. Faria. M.. Lu. Y.. Heaney. K.. & Uthamanthil. R. (2007). Recombinant human bone morphogenetic protien-2 in absorbable collagen sponge enhances bone healing oftibial osteotomies in dogs. l'eterinarian Surge/1r. 36(2). 122-31. Ferguson. J.. & Nicol. R. (2000). Early spiea trcatgment of pediatric femoral shaft fractures. Journal ofl’edia/r'ic Orthopedics. 20(2). 189-92. Fernandez. A.. & Masliah Galante. R. (2001). Osteosynthesis ofdiaphyseal fractures ofthe radius and ulna using an internal fixator (PC-Fix). A prospective study. lru'urjr. 32(2). 44-50. Fernandez Fairen. M.. Guillen. J.. Busto. J.. & Roura. .l. (1999). Fractures ofthe fifth metatarsal in basketball players. Knee .S'urgerv. Sports Traurmrtologt’ and Arthroscopj'. 7(6). 373-7. Finkemeier. C.. Schmidt. A.. Kyle. R.. & Ternplernan. D. (2000). A prospective. randomized study of intramedullary nails inserted with and without reaming for the treatment of open and closed fractures of the tibial shaft. Journal ofOrt/ropedic Trauma. 14(3). 187-93. Fischer. G.. Pelka. R.. & Barovic. .l.(2005). Adjuvant treatment ofknee osteoarthritis with weak pulsing magnetic fields. Results ofa placebo- controlled trial prospective clinical trial. Orthopedic und ihre Grenzgehiete. 145(5). 544-50. Fleming. 8.. Krag. M.. Huston. D.. & Sugihara. S. (2000). Pin loosening in a halo-vest orthosis: a biornechanical study. Spine. 25(1 1). 1325-31. Flinkkila. T.. Nikkola-Sihto. A.. Raatikainen. T.. & .lunila. J. (1999). Role of metaphyseal cancellous bone defect size in secondary displacement in C olles' fracture. Archives ofOrthopedic and Trauma Surgery. 1 19(5-6). 319-23. Foead. A.. Penafort. R.. Saw. A.. & Sengupta. S. (2004). Comparison oftwo methods of percutaneous pin fixation in displaced supracondylar fractures ofthe humerus in children. Journal o/‘Orthopedic Surgery. 12(1). 76-82. Fourie. J.. & Thompson. M. ( 1998). A model for the prediction oftime to union in 1 08 fracutres ofthe tibia. l’hj'siotherapr Research lnternatiorurl. 3( l ). 27-36. Fourie. J.. & Bowcrbank. P. (1997). Stimulation ofbone healing in new fractures of the tibial shaft using interfcrcntial currents. l"h_t«'siorherapr Research International. 2(4). 255-68. Franck. W.. Ollvieri. M.. Jannasch. O.. & llennig. F. (2003‘). lixpandable nail system for osteoporotic humeral shaft fractures: preliminary results. Journal ot'Tr'auma. 54(6). 1 152-8. Friedlaender. G. (2004). Osteogenic protein-1 in treatment oftibial nonunions: current status. .S'urgical 'l‘echnologt' International. 13. 249-52. Friedlaender. G.. Perry. C.. Cole. J.. & Cook. S. (2001). Osteogenic protein-1 (bone morphogenetic protein-7) in the trearntne oftibial nonunions. Journal o/‘Bone andJoint S'urger'j: 83. 151-8. Fritz. T.. Weiss. C.. Kriegistein. C.. & Quentmeier. A. (1999). The classic nail in the therapy of trochanteric fractures. A prospective. controlled study. Archive ofOrt/urpedic and Trauma Surgery. 199(5-6). 308-14. Fritz. T.. Hiersemann. K.. Kriegistein. C.. & Friedl. W. (1999). Prospective randomized comparison of gliding nail and gamma naili in the therapy of trochanteric fractures. A rchives o/‘Or'thopedic and Trauma .S'urgert’. 199(1-2). 1-6. Furian. J.. Perrin. R.. Govender. P.. & Petrenko. Y. (2007). Use of osteogenic protien-l in patients at high risk for spinal pseudarthrosis: a prospective cohort study assessing safety. health-related quality of life. and radiograph fusion. .\'eurosurgical Spine. 7(5). 486-495. Gaebler. C.. Berger. U.. Schandelmaler. P.. & Greitbauer. M. (2001 ). Rates and odds ratios for complications in closed and open tibial fractures treated with unreamed. small diameter tibial nails: a multicenter analysis of467 cases. Journal o/‘Orthopedic Trauma. 15(6) 415-23. Gaetani. P.. Almar. E.. Panella. L.. & Levi. D. (.2006). Functional disability after instrumented stabalization in lumbar degenerative spondylolisthesis: a follow-up study. Functional Neurology. 21(1). 31-7. Galatz. L.. Williams. G.. Fenlin. J.. & Ramsey. M. (2004). Outcome of open reduction and internal fixation of surgical neck nonunions ofthe humerous. Journal ofOrthopedic Trauma. 18(2). 63-7. Gerlach. K.. & Schwarz. A. (2002). Bite forces in patients after treatment of 109 mandibular angle fractures with miniplate osteosynthesis according to Champy. lnternatiomrl Journal of'Oral .lla.\'illo/acial Surgery. 31(4). 345- 8. (‘iicqueL P.. Giacornelli. M.. Karger. C.. & Calvert. J. (2005). Treatment of epiphyseo-metaphyseal fractures in children vs ith an orginal implant. The threaded pin with an adjustable lock (F ixano). Operational Orthopedic 'l'raumatologt’. l7( 1 ). 51-65. Ginandcs. C.. & Rosenthal. D. (1999). Using Hypnosis to accelerate the healing of bone fractures: a randomized controlled pilot study. .»1lrernarive Therapy Health and .lledicine. 5(2). 67-75. Giordano. N.. Battisti. E. Geraci. S. & Fortunato. M. (2001). Effect of electromagnetic fields on bone mineral density and biomechanial markers of bone turnover in osteoporosis: A single-blind. randomized pilot study. ('urrent Therapeutic Research. 62(3). 187-193. Gogus. A.. Ozturk. C.. Tezer. M.. & Camurdan. K. (2007). "Sandwich technique" in the surgical treatment of primary complex fractures ofthe femur and humerus. International Orthopedics. 31(1 ). 87-92. (_ioldhagen. P.. O'Conncr. D.. Schwarzc. D.. & Schwartz. E. (1994). A prospective comparative study of the compression hip screw and the gamma nail. Journal o/‘Orthopedic Trauma. 8(5). 367-72. Gomar. F.. Orozco. R.. Villar. J.. & Arrizabalaga. F. (2007). P-15 small peptide bone graft substitute in the treatment of non-unions and delayed union. A pilot clinical trial. lnternatiorurl Orthopedics. 31(1). 93-9. Goodwin. C.. Brighton. C.. Guyer. R.. & Johnson. J. (1999). A double-blind study ofeapacitively coupled electrical stimulation as an adjunct to lumbar spinal fusions. Spine, 24(13). 1349-56. Gopal. S.. & Giannoudis. P. (2001'). Prospective randomized study of reamed versus unreamed femoral intramedullary nailing: an assessment of procedures. Journal o/‘Orthopedic Trauma. 15(6). 458-60. Govender. S.. Csimma. C.. Genant. H.. & Valentin-Opran. A. (2002). Recombinant human bone morphogenetic protein-2 for treatment of open tibial fractures: a prospective. controlled. randomized study of four hundred and fifty patients. Journal ofBone and Joint Surgery, 84(12). 2123-34. Graham. .1. ( 1968). Early ofdelayed weight-bearing after internal fixation of 110 transcervical fracture ofthe femur. A clinical trial. Journal o/Bone and Joint Surgery. 50(3). 562-9. Gregory. P.. Sullivan. J.. & Ilerndon. W. (1995). Adolescent femoral shaft fractures: ridged versus flexible nails. ()r'tlr()}_)eclic.s'. 18(7). 645-7. Grill. F.. Bollini. G.. Dungl. P.. & Fixsen. .l. (2000). Treatment approaches for congenital pseudarthrosis oftibia: results of the EPOS multicenter study. European Paediatric Orthopaedic Society (EPOS). Journal o/‘l’ediatric Orthopedics. 9(2). 75-89. Grose. A.. Gardner. M.. Hcttrieh. C.. & Fishman. F. (2007). Open reduction and internal fixation oftibial pilon fractures using a lateral approach. Journal otOrt/ropedic Trauma. 21(8). 530-7.SO:J Orthop Trauma Grubb. S.. Lipscomb. 11.. & Suh. P. (1994). Results ofsurgical treatment of painful adult scoliosis. Spine, 19(14). 1619-27. Gruen. T.. Poggie. R.. Lewallen. D.. & llanssen. A. (2005). Radiographic evaluation ofa monoblock acetabular component: a multicenter study with 2- to 5-year results. Journal ofArthroplastv. 20(3). 369-78. Gunal. 1.. Ozcclik. A.. Gokturk. E.. Ada. S.. & Demirtas. M. (1999). Correlation of magnetic resonance imaging and intraoperative punctate bleeding to assess the vaseularity of scaphoid nonunion. Archives ofOrthopedic and Trauma Surgery. 1 19(5-6). 285-7. Gupta. R.. Rahcja. A.. & Modi. U. (1999). Colles' fracture: management by pcrcutaneous crossed-pin fixation versus plater of Paris cast immobilization. Orthopedics. 22(7). 680-2. Gustilo. R.. Gruninger. R.. & Davis. T. (1987). Classification oftype 111 (severe) open fractures relative to treatment and results. Orthopedics. 10(12). 1781- 8. Haas. N.. Hauke. C.. Schutz. M.. & Kaab M. (2001). Treatment ofdiaphyseal fractures ofthe forearm using the Point Contact Fixator (PC-Fix): results of387 fractures ofa prospective multicentric study (PC-Fix II). Injurv. 32. 851-62. Haidukewych. G.. Serns. S.. Huebner. D.. Horowitz. D.. & Levy. B. (2007). Results of polyaxial locked-plate fixation of periarticular fractures ofthe knee. Journal o/‘Bone andJoint Surgery. 89(3). 614-20. Haidukewych. G.. & Berry. D. (2003). Salvage of failed internal fixation of 111 intertrochanteric hip fractures. (i'linical Orthopedic and Related Research. 412. 184-8. Hammacher. E.. van Meeteren. M.. & van der Werken. C. (1998). Improved results in treatment of femoral shaft fractures with the unreamed femoral nail? A multicenter experience. Journal ot'l'r'auma. 45(3). 517-21. I-Iandolin. L.. Kitjunen. V.. Arnaia. 1.. 8; Kiuru. M. (2005) No long-term effects of ultrasound therapy on bioabsorbable screw -fixed lateral malleolar fracture. .S'carulanavian Journal of‘Surgert'. 94(3). 239-42. Ilandolin. L.. Kiljunen. V.. Arnala. 1.. & Kiuri M. (2005). Effect of ultrasound therapy on bone healing of lateral malleolar fractures ofthe akel joint fixed with bioabsorbable screws. Journal ot'Orthopedic Science. 10(4). 391-5. k I’landolin. L.. Kiljunen. V Arnala. 1.. & Pajarinen. .l. (2005). The effect of low intensity ultrasound and bioabsorbable self-reinforced poly-L-lactide screw fixation on bone in lateral malleolar fractures. .-frchives of Orthopedic and Trauma .S'urgery. 125(5). 317-21. Ilanft. J.. Goggin. J.. Landsman. A.. & Suprenant. M. (1998). The role of cobined magnetic field bone growth stimulation as an adj unet in the treatment of neuroarthrtmathy/Charcot joint: an expanded pilot study. The Journal of Foot and Ankle .S'rrrgery. 37(6). 510—5. Hansen. P.. & Hansen. T. (1998). The treatment of fractures ofthe ring and little metacarpal necks. A prospective randomized study of three different types of treatment. Journal oflland Surgery. 23(2). 245-7. Ilarding. I.. Parry. D.. & Barrington. R. (2001 ). The use of a molded metacarpal brace versus neighbor strapping for fracture of the little finger metacarpal neck. Journal «tr/Hand.S'urgery. 26(3). 261-3. Hardy. D.. Dcseamps. P.. Krallis. P.. & Fabeck. L. (1998). Use of an intramedullary hip-screw compared with a compression hip-screw with a plate for intertrochanteric femoral fractures. A prospective. randomized study ofone hundred patients. Journal o/Bone and Joint Surgery. 80(5). 618-30. Harrison. W.. Lewis. C.. & Lavy. C. (2004). Open fractures ofthe tibia in HIV positive patients: a prospective controlled single-blind study. Injury. 35(9). 852-6. Heck. B.. Iibraheim. N.. & Foetisch. C. (1997). Direct complication of trochanteric osteotorny in open reduction and internal fixation of acetabular fractures. American Journal of"Orthopedics. 26(2). 124-8. Hcetvcld. M.. Raaymakers. E.. van Eck-Smit. B.. 8; van Walsurn. A. (2005). Internal fixation for displaced fractures of the femaoral neck. Does bone density affect clinical outcome? Journal o/Bone and Joint Surgery. 87(3). ’5 ‘9 367-73. Helfet. D.. Howey. T.. Dipasquale. T.. & Sanders. R. (1994). The treatment of open and or unstable tibial fractures with an unreamed double-locked tibial nail. Orthopedic Review. Supp. 9-17. Henie. P.. Zimmermann. G.. & Weiss. S. (2005'). Matrix metalloproteinases and failed fracture healing. Bone. 37(6). 791-8. I—Ienley. M.. Chapman. J.. Agel. J.. & Harvey. E. (1998). Treatment oftype 11. 111A. and IIIB open fractures ofthe tibial shaft: a prospective comparison of unreamed interlocking intramedullary nails and half-pin external fixators. Journal ofOrthopedic Trauma. 12(1). 1-7. Hernefalk. L.. Granstrom. P.. & Messner. K. (1997). Sequential scintigraphy and orthoradiographic measurement of femoral shortening after femoral neck fractures. Archives o/Orthopedic and Trauma Surgery. 116(4). 198-203. Hernefalk. L.. & Messner. K. (1996). Rigid osteosynthesis decreases the late complication rate after femoral neck fracture. The influence of three different osteosynthesis devices evaluated in 369 patients. Archives of Orthopedic and Trauma Surgerv. 1 15(2). 71 -4. I-Ierscm'ici. D.. Ricci. W.. McAndrews. P.. DiPasquale. T.. & Sanders. R. (2000). Treatment of femoral shaft fractures using unreamed interlocked nails. Journal of Orthopedic Trauma, 14(1). 10-14. Hertel. R.. Eljer. H.. Meisser. A.. & Hauke. C. (2001).Biomechanical and biological considerations relating to the clinical use of the Point Contact- Fixator--evaluation of the device handling test in the treatment of diaphyseal fractures of the radius and/or ulna. Injury. 32. 10-14. Hertel. R.. Lambert. S. Muller. S.. Ballmer. F.. & Ganz. R. (1999). On the timing of soft-tissue reconstruction for open fractures of the lower leg. Archives o/Orthopedic and Trauma Surgery. 199(1-2). 7-12. Hinsenkamp. M.. Ryaby. J.. & Burny. F. (1985). Treatment of Non-Union by Pulsing Electromagnetic Field: European Multicenter Study of308 Cases. Reconstruction Surgery T raumatologv. 19. 147-151. 113 Ilofmann. G.. Conschorek. O.. & Buhrcn. V. (1999). Segment transport employing intramedullary devices in tibial bone defects following trauma and infection. Journal of'Ort/ropedic Trauma. 13(3). 170-7. IIull. J.. Sanderson. P.. Rickrnan. M.. & Bell. M. (1997). External fixation of children's fractures: use ofthe Orthofix Dynamic Axial Fixator. Journal ofpediatric Orthopedics 6(3). 203-6. lluusko. T.. Karppi. P.. Kautiainen. 11.. & Suorninen. H. (2002). Randomized. double-blind. clinically controlled trial ofintranasal calcitonin treatment in patients with hip fracture. CalcifTissue Interrurtiorurl, 71(6). 478-84. Iannacone. W.. Bennett. F.. Delong. W.. & Born. C. (1994). Initial experience with the treatment ofsupracondylar femoral fractures using the supracondylar intramedullary nail: a preliminary report. Journal of" i Orthopedic Trauma. 8(4). 322—7. - lida. Y.. Kuroda. T.. Kitano. T.. & Mizuno. K. (2000). Dexa-rneasured bone density changes over time after intertrochanteric hip fractures. lv'ohe Journal ofilledical Science. 46(1-2). 1-12. lnan. M.. E1 Rassi. G.. Riddle. E.. & Kumar. S. (2006). Residual deformities following successful initial bone union in congenital pseudoarthrosis of the tibia. Journal ot'l’ediatric Orthopedics. 26( 3). 393-9. lngrnan. A. (2000). Pereutaneous intramedullary fixation oftrochanteric fractures ofthe femur. Clinical trial ofa new hip nail. Injury. 31(7). 483-7. Iseda. T.. Nakano. S.. Suzuki. Y.. & Miyahara. D. (2000). Radiographic and scintigraphic courses of union in cervical interbody fusion: hydroxyapatite grafts versus iliac bone autografts. Journal ot'.\'uclear .l/edicine. 41( 10). 1642-5. Isiklar. 7... Demirors. II.. Akpinar. S.. & Tandogan. R. (1999). 'I‘wo-stage treatment ofchronic staphylococcal orthopaedic irnplat-related infections using vancornycin impregnated PMMA spacer and rifampin containing antibiotic protocol. Bull Hosp Joint Disease, 58(2). 79-85. Jacobson. J.. Gorman. R.. Yamanashi. W.. Saxena. B.. & Clayton. L. (2001 ). Low-amplitude. extremely low frequency magnetic fields for the treatment ofosteoarthritie knees: a double-blind clinical study. .r'flternative Therapies in I lealth and .lledicine. 7(5). 54-64. Jamdar. J.. Rao. B.. Netke. S. & Roomi. M. (2004). Reduction in tibial shaft 114 fracture healing time with essential nutrient supplementation containing ascorbic acid. lysine. and praline. Journal ofi-llternative Complement .lledicine. 10(6). 915-6. Jazrawi. I... Kummer. F.. Simon. J.. & Bai. B. (2000). New technique for treatment of unstable distal femur fractures by locked double-plating: case report and biomechanical evaluation. Journal ot'Trauma. 48(1). 87-92. 802.1 Trauma .lenis. I... An. 11.. Stein. R.. & Young. B. (2000). Prospective comparison ofdirect current electrical stimulation and pulsed electromagnetic fields on instrumented posterolateral lumbar arthrodesis. Journal of Spinal Disorders. 13(4 ). 290-6. .lenis. I... An. H.. Stein. R.. & Young. B. (2000). Prospective comparison ofthe effect ofdirect current electrical stimulation and pulsed electromagnetic fields on instrumented posterolateral lumbar arthrodesis. Journal of'Spinal Disorders. 13(4 ). 290-6. Jensen. C.. & Jensen. C. (1996). Biodegradable pins versus Kirschner wires in hand surgery. Journal oflland Surgery. 21(4). 507-10. .leyarn. M.. Andrew. J.. Muir. I... & Mcgovern. A. (2002). Controlled trial of distal radial fractures treated with a resorbable bone mineral substitute. Journal off/and .S‘urgery. 27(2). 146—9. Johnson. L.. l\/Iorrison. K.. & Wood. D. (2000). The application ofarthroseopic principles to bone grafting ofdelayed union of long bone fractures. :1rthroscopy. l6(3 ). 279-89. Jones. A. (2005). Recombinant human bone rnorphogenic protein-2 in fracture care. Journal of'Orthopedic Trauma. 19(10). S23—5. Joseph. B.. & Mathew. G. (2000). Management ofcongenital pseudarthrosis of the tibia by excision ofthe pseudarthrosis. only grafting. and intramedullary nailing. Journal o/I’ediatric Orthopedics. 9(1). 16-23. Jupiter. (1997). Repair of five distal radius fractures with an investigational cancellous bone cement: a preliminary report. Journal of'Orthopedic Trauma, 1 1(2). 1 10-6. Kabak. S.. Halici. M.. Tunccl. M.. & Avsarogullari. L. (2004). Treatment of midclavicular nonunion: comparison ofdynarnic compression plating and low-contact dynamic compression plating techniques. Journal ofShoulder and Elbow Surgery, 13(4). 396-403. Kaewpornsawan. K. (2001 ). Comparison between closed reduction with pcrcutaneous pinning and open reduction with pinning in children with closed totally displaced supracondylar humeral fractures: a randomized controlled trial. Journal ot'l’ediactric Orthopedics, 10(2). 131-7. Kanai. 11.. Igarashi. M.. Yamamoto. S.. 8: Oda. 11. (1999). Spontaneous subcapital femoral neck fracture complicating a healed intertrochanteric fracture. Archives ofOrthopedic and TraumaSurgery. 1 19(5-6). 271-5. Kankare. J. (1998). Opertive treatment ofdisplaced intra-articular fractures ofthe calcaneus using absorbable internal fixation: a prospective study of twenty—five fractures. Journal ofOrt/ropedic Trauma, 12( 6). 413-9. Kankare. J. (1997). Tibial condylar fracture fixed with totally absorbable. self- reinforced polyr'glycolide screws. A preliminary report. Archives of Orthopedic Trauma Surgery. 1 16(3). 133-6. Kaplan. B.. Hoard. M.. & Park. S. (2001). Immediate mobilization following fixation of mandible fractures: a prospective. randomized study. Larvngscope. l 1 1(9). 1520-4. Kapukaya. A.. Subasi. M.. Necmioglu. S. & Arsian. H. (1998). Treatment of closed femoral diaphyseal fractures with external ‘fixators in children. .llr'chives of'Orthopedic Trauma Surgery. 1 17(6-7). 387-9. Karladani. A.. Granhed. l-I.. Edshagc. B.. & Jerre. R. (2000). Displaced tibial shaft fractures: a prospective randomized study ofclosed intramedullary nailing versus cast treatment in 53 patients. Acta Orthopedic Scar-ulirurvian, 71(2). 160-7. Kawaguchi. S.. Sawada. K.. Nabeta. Y.. & llayakawa. M. (1998). Recurrent dorsal angulation ofthe distal radius fracture during dynamic external fixation. Journal of'llandSurgery. 23(5). 920-5. Keating. J.. O'Brien. P.. Blaehut. P.. & Meek. R. (1997). Locking intramedullary nailing with and without reaming for open fractures ofthe tibial shaft. A prospective. randomized study. Journal of'Bone and Joint Surgery. 79(3). ‘1" 334-41. Keener. J.. Parsons. B.. Flatow. E. & Rogers. K. (2007). Outcomes after percutaneous reduction and fixation of proximal humeral fractures. Journal of Shoulder and EIDUH‘ Surgery. 16(3). 330-8. Kennedy. J.. Ilarty. J.. Casey. K.. & .lan. W. (2003). Outcome after single technique ankle arthrodesis in patients with rheumatoid arthritis. ('linical Orthopedics and Related Research. 412. 131-8. 116 Kenwright. J.. Richardson. J.. Goodship. A.. & Evans. M. ( 1986). Effect of controlled axial micrornovernent on healing oftibial fractures. lancer. 22(2). 1185-7. Keret. D.. Bollini. G.. Dungl. P.. & Fixsen. J. (2000). The fibula in congenital pseudoarthrosis ofthe tibia: the EPOS multicenter study. European Paediatric Orthopaedic Society (EPOS). Journal ot'l’ediatric Ortlunredics.9(2). 69-74. Kershaw. C.. Cunningham. J.. 8: Kenwright. .l. (1993). Tibial external fixation. weight bearing. and fracture movement. ('linica/ Orthopedics and Related Research. 293. 28-36. Khanduja. V.. Lirn. C.. Vemulapalli. K.. Lee. C. (2006). I’)etachable functional focused rigidity cast for metatarsal fractures. British Journal oj'A'ursing. 15(5). 282-4. Kitoka. H.. Schaap. E.. Chao. E..& An. K. (1994). Displaced infra-articular fractures of the calcaneous treated non-operatively. Clinical results and analysis of motion and ground-reaction and temporal forces. Journal of Bone ant/Joint Surgery. 76(10). 1531-40. Knapp. D.. Jones. E.. Blanco. J.. & Flynn. J. (2005). Allograft bone in spinal fusion for adolescent idiopathic scoliosis. Journal o/Spinal Disorders and Technology 18. 73-6. Konrad. G.. Markmiller. M.. Lenich. A.. & Mayr. E. (2005). 'I‘ourniquets may increase postoperative swelling and pain after internal fixation ofankle fractures. Clinical Orthopedics and Related Research. " 33. 189-94. Kopylov. P.. Aspendberg. P.. Yuan. X.. & Ryd. 1.. (2001). Radiostereometric analysis ofdistal radial fracture displacement during treatment: a randomized study comparing Norian SRS and external fixation in 23 patients. A cta Orthopedic Scam/into'ian. 72(1 ). 57-61. Kopylov. P.. Runggvist. K.. Jonsson. K.. & Aspenberg. P. (1999). Norian SRS versus external fixation in redisplaced distal radial fractures. A randomized study in 40 patients. Acta Orthopedics Scam/inavian. 7()(1 ). 1- ,- 5. Kornblum. M.. Fischgrund. J.. Herkowitz. 11.. 8; Abraham. D. (2004). Degenerative lumbar spondylolisthesis with spinal stenosis: a prospectiv long-term study comparing fusion and pseudarthrosis. Spine. 29(7). 733-4. Koshirnune. M.. Kamano. M.. Takarnatsu. K.. & (_)hashi. H. (2005). A 117 randomized comparison of locking and non-locking palrnar planting for the unstable Colles' fractures in the elderly. Journal (rt/land .S'urgery. 30(5). 499-501. Kosygan. K.. Mohan. R.. & Newman. R. (2002). The Gotfried pcrcutaneous the fixation ofintertrochanteric fractures ofthe hip. Journal o/Bone and Joint Surgery, 84(1). 19-22. Koval. K. (2001 ). Preoperative skin treaction was not useful for hip fractures. Journal (rt-Bone andJoint .S'urgery. 83(2). 303. Kozin. S.. & Bishop. A. (1994). Tension wire fixation ofav'ulsion fractures at the thumb metacarpophalangeal joint. Journal of’lland Surgery. 19(6). 1027- 31. , Kreder. H.. Agel. J.. McKee. M.. & Schemitsch. E. (2006). A randomized. controlled trial ofdistal radius fractures with metaphyseal displacement but without joint incongruity: closed reduction and casting versus closed reduction. spanning external fixation. and optional pcrcutaneous K-wires. Journal ofOrt/ropedic Trauma. 20(2). 1 15-21. Krettek. C.. Konemann. B.. Miclau. T.. 8: Schandelmaier. P. (1997). A new technique for the distal locking of solid AO unreamed tibial nails. Journal o/‘Orthopedic Trauma. 1 1(6). 446-5 1. Kuklo. T.. Potter. 13.. Lenke. I... & Polly. D. (2007). Surgical revision rates of books verses hybrid versus screws versus combined anteroposterior spinal fusion for adolescent idiopathic scoliosis. Spine. 32(20). 2258-64. Kuklo. T.. Bridwell. K.. Lewis. S.. & Baldus. C. (2001). Minimum 2-year analysis of sacropelvie fixation and 1.5-S1 fusion using SI and iliac screws. Spine. 26(18). 1976-83. SOzSpine Lamberg. T.. Remes. V.. Helenius. 1.. & Schlenzka. I. (2005). Long-term clinical. functional and radiological outcome 21 years after posterior or posterolateral fusion in childhood and adolescence isthmic spondylolisthesis. European Spine Journal. 14(7). 639-44. Larsen. 1... Madsen. J.. Hoiness. P.. & Ovre. S. (2004). Should insertion of intramedullary nails for tibial fractures be with or without reaming? A prospective. randomized study with 3.8 years' follow-up. Journal of Orthopedic Trarrrna. 18(3). 144-9. Lee. F.. Sineopi. S.. Lee. F.. & Vital. M. (2006). Treatment ofcongenital 118 pseudarthrosis ofthe tibia with recombinant human bone morphogenetic protein-7 (thMP-7). A report of five cases. Journal of'Bone and Joint Surgery. 88(3). 627-33. Lei. H.. & Yi. 1.. (1998). One-stage open cancellous bone grafting ofinfected fracture and nonunion. Journal o/‘Orthopedic .S'cience. 3(6). 318-23. Leung. K.. Lee. W.. Tsui. H.. & Liu. P. (2004). Complex tibial fracture outcomes following treatment with low-intensity pulsed ultrasound. (Tltrasourul .lledicinc Biologv. 30(3). 389-95. Leung. F.. & Chow. S. (2003). A prospective. randomized trial comparing the limited contact dynamic compression plate with the point contact fixator for forearm fractures. Journal ofBone and Joint Surgery. 85(12). 2343—8. Lin. .1. (2006). Effectiveness of locked nailing for the displaced three-part proximal humeral fractures. Journal ofTraurna. 61(2). 363-74. Lindahl. O. (1984). Survey of orthopedic surgical research in university center. ('linical Orthopedic and Related Research. 191. 136-8. Lindquist. S.. l\/Ialmgvist. B.. & Ullrnark. G. (1989). Fixation of femoral neck fractures. Prospective comparison of von Bahr screws. Gouffon screws. and Ilessel pins. Acta Orthopedic .S’carulinavian. 60(3). 293-8. Linovitz. R.. Pathria. M.. Bernhardt. M.. & Green. D. (2002). Combined magnetic fields accelerate and increase spine fusion: a double-blind. randomized. placebo controlled study. Spine. 27(13). 1383-9. Lonner. J.. Koval. K.. Golyakhovsky. V.. & Frankel. V. (1995). Posttr'aumatic nonunion ofthe distal tibial metaphysis. Treatment using the llizarov circular external fixator. .‘trnericanJournal ofOrthopedics. Supplement. 16-21. Luna Gonzalez. F.. Lopez ArEvalo. R.. Meschian Coretti. S.. & Urbano Labajos. V. (2005). Pulsed electrornagentic stimulation of regenerate bone in lengthening procedures. A cta Orthopaedica Belgica. 71(5). 571-6. Lyritis. G.. & Johnell. O. ( I993). Orthopaedic management of hip fracture. Bone. Suppl. 11-17. Mandronero. A.. Pitillas. 1.. & Manso. F. (1988). Pulsed electromagnetic field treatment failure in radius non-united fracture healing. Journal of Biomedical Engineering. 10(5). 463-6. Madsen. J.. Naess. I... Aune. A.. & Alho. A. (1998). Dynamic hip screw with 119 trochanteric stabilizing plate in the treatment of unstable proximal femoral fractures: a comparative study with the Gamma nail and compression hip screw. Journal ofOrt/ropedic Trauma. 12(4). 241-8. Madsen. F.. Linde. E.. Andersen. E.. Birke. 11.. & llavass. I. (1987). Fixation of displaced femoral neck fractures. A comparison between sliding screw plate and four cancellous bone screws. .1 cIa Orthopedic Scarulirtavian. 58(3). 212-6. Mammi. G.. Rocchi. R.. Cadossi. R.. & I\t"lassari. L. (1993). The electrical stimulation oftibial osteotomies. Double-blind study. (,"linical Orthopedics and Related Research. 288. 246-54. Maneerit. J.. Meknavin. S.. & Hanpanitkitkan. S. (2004). Pereutaneous versus open bone grafting in the treatment oftibial fractures: a randomized prospective trial. Journal ofJ/edicine .Alssociation Thai. 87(9). 1034-40. Marchetti. P.. Vicenzi. G.. lrnpallorneni. C.. & Landi. S. (2000). The use of elastic nails for intramedullary fixation of humeral fractures and nonunions. Orthopedics. 23(4). 343-7. Markmiller. M.. Tjarksen. M.. Mayr. F. & Ruler. A. (2000). The unreamed tibia nail. Multicenter study ofAO/ASIF. Osteosynthesefragen/Association for the Study oflnternal Fixation. Langenlwclts .-‘lrchive of'Surgery. 385(4). 276-83. Marsh. J.. Muehling. V.. Dirschi. D.. & Ilurwitz. S. (2006). Tibial plafond fractures treated by articulated external fixation: a randomized trial of postoperative motion versus nonrnotion. Journal ofOrthopedic Trauma. 20(8). 536-41. Marsh. J.. Bonar. S.. Nepola. J.. & Decostcr. T. (1995). Use of an articulated external fixation for fractures of the tibial plafond. Journal of Bone and Joint Surgery. 77(10). 1498-501. Mattsson. P.. & Larsson. S. (2003). Stability ofinternally fixed femoral neck fractures augmented with resorbable cement. A prospective randomized study using radiostereometry. Sccrrularurvian Journal of'Surgery. 92(3). 215-9. Mazda. K.. Khairouni. A.. Pennecot. G.. & Bensahel. II. (1997). Closed flexible intramedullary nailing ofthe femoral shaft fractures in children. Journal of Pediatric Orthopedics. 6(3). 198-202. McAndrcw. M.. Gorman. P.. & Lange. T. (1988). Tricalcium phosphate as a bone 120 graft substitute in trauma: preliminary report. Journal ofOrthopedic Trauma. 2(4). 333-9. McGuire. R.. & Amundson. G. (1993). The use of primary internal fixation in spondylolisthesis. Spine. 18(12). 1662-72. I\/’1c1_.auchlan. G.. Cowan. B.. Annan. I.. & Robb. J. (2002). Management of completely displaced metaphyseal fracture of the distal radius in children. A prospective. randomized controlled trial. Journal of'Bone and Joint Surgery. 84(3). 413-7. McQueen. M.. Hajducka. C.. & Court-Brown. C. (1996). Redisplaced unstable fractures ofthe distal radius: a prospective randomized comparison of four methods of treatment. Journal of Bone and Joint Surgery. 78(3). 404-9. Merloz. P.. Tonetti. J.. Pittet. L.. Coulomb. M.. & Lavcllee. S. (1998). Computer- assisted spine surgery. Computer Aided Surgery. 3(6). 297-305. Mills. W .. Chapman. J.. Robinson. L.. & Slimp. J. (2000). Somatosensory evoked potential monitoring during closed humeral trailing: a preliminary report. Journal ofOr/hopedic Trauma. 14(3). 167-70. Moed. B.. Subramanian. S.. van Holsbeeck. M.. & Watson. J. (1998). Ultrasound for the early diagnosis of tibial fracture healing after interlocked nailing without reaming: clinical results. Journal ofOrt/ropedic Trauma. 12(3). 206-13. Mologne. T.. Lundeen. J.. Clapper. M.. & O'Brien. T. (2005). Early screw fixation versus casting in the treatment of acute Jones fractures. .-1rnerican Journal ofSports illedicine. 33(7). 970-5. Mooney. V. (1990). A randomized double-blind prospective study of the efficacy of pulsed electromagnetic fields for interbody lumbar fusions. Spine. 15(7). 708-12. Moorcroft. C.. Thomas. P.. Ogrodnik. P.. & Verborg. S. (2000).A device for improved reduction of tibial fractures with external fixation. Proc Inst 11..-led Eng. 214(5). 449-57. Moore. J.. Shank. J.. Morgan. S.. & Smith. W. (2006). Syndesmosis fixation: a comparison ofthree and four cortices of screw fixation without hardware removal. Foot and Ankle International. 27(8). 567-72. Moroni. A.. Faldini. C.. Hoang-Kim. A.. & Pegreffi. F. (2007)).Alendronate improves screw fixation in osteoporotic bone. Journal of Bone and Joint Surgery. 89( 1 ). 96-101 . Moroni. A.. Faldini. C.. Pegreffi. F.. & Giannini. S. (2004). IIA-coated screws decrease the incidence of fixation failure in osteoporotic trochanteric fractures. Clinical Orthopedics and Related Research. 425. 87-92. Muldoon. M.. Padgett. D.. Sweet. D.. & Deuster. P. (2001). Femoral neck stress fractures and metabolic bone diseases. Journal ofOrt/ropedic Trauma. 15(3).181-5. Munk. 13.. Fokjaer. L... Larsen. C.. & .lohannsen. H. (1995). Diagnosis of scaphoid fractures. A prospective multicenter study of 1.052 patients with 106 fractures. A cta Orthopedic Scandirtavian. 66(4). 359-60. Musialek. J.. Fillip. P.. & Nieslanik. J. (1998). Titaniurn-nickel shape memory clamps in small bone surgery. Archive of Orthopedics and T raurna .S'urgery. 1217(6-7).34l-4. SO:Arch Orthop Trauma Surg Nand. S.. Sengar. G.. Nand. S.. Jain. V.. & Gupta. T. (1996). Dual use ofsilver for management ofchronic bone infections and infected non-unions. Journal of'lr‘tdian Medical Association. 94(3). 91-5. Nassif. J.. Gorczyca. J.. Cole. J.. & Pugh. K. (2000). Effect of acute reamed versus unreamed intramedullary nailing on compartment pressure when treating closed tibial shaft fractures: a randomized prospective study. Journal ofOrt/ropedic Trauma. 14(8). 554-8. Noordecn. M.. Lavy. C.. Shergill. N.. & Tuite. J. (1995). Cyclical micrornovement and fracture healing. Journal ofBone and Joint Surgery. 77(4). 645-8. Oda. 11.. Nakamura. K.. Matsushita. T.. & Yamamoto. S. (2006). Clineial use ofa newly developed calcium phosphate cement. Journal ofOrthopedic Science. 1 1(2). 167-74. Odvina. C .. Zerwekh. J .. Rao. D.. & Maalouf, N. (2005). Severely suppressed bone turnover: a potential complication ofalendronate therapy. Journal of Clinical Emlocrinology Metabolism, 90(3). 1294-301. Oh. C.. Park. 8.. Inh. J..& Kyung. H. (2000). Fracture separation ofthe distal humeral epiphysis in children younger than three years old. Journal of Pediatric Orthopedics, 20(2). 173-6. ('f)hnishi. 1.. Sato. W.. Matsuyaman. J.. & Yajirna. H. (2005). Treatment of congenital pseudarthrosis ofthe tibia: a multicenter study in Japan. Journal ofl’ediatric Orthopedics. 25(2). 219-24. 122 OlaOlorun. D.. Oladiran. 1.. & Adeniran. A. (2001). Complications of fracture treatment by traditional bonesetters in southwest Nigeria. Family Practice. 18(6). 635-7. Olcrud. C.. Rehnberg. L.. & Hellquist. E. (1991). Internal fixation of femoral neck fractures. Two methods compared. Journal ofBone and Joint Surgery. 73(1). 16-9. Onuminya. J. (2006). Performance ofa trained traditional bonesetter in primary fracture care. South Africa :1 lea’ical Journal. 96(4). 320-2. Oransky. M.. & Gasparini. G. (1997). Associated Iumbosaeral junction injuries (LSJIs) in pelvic fractures. Journal ofOrthopedic T raurna. 1 1(7). 509-12. Ostermann. P.. Ekkernkamp. A.. Henry. S.. & Muhr. G. (1994). Bracing of stable shaft fractures ofthe ulna. Journal ofOrthopedic T raurna. 8(3). 245-8. Panozzo. A.. Manfrini. M.. & C apanna. R. (1996). Effect ofelectromagnetic fields on patients undergoing massive bone graft following bone tumor resection. Journal o/Bone and Joint Surgery, 78(2-3). 133. Pape. H.. Rixen. D.. Morley. J.. & Husebye. E. (2007). Impact of the method of initial stabilization for femoral shaft fractures in patients with multiple injuries at risk for complications (borderline patients). Annals Surgery. 246(3). 299-501. Parker. M.. Raghavan. R.. & Gurusamy. K. (2007). Incidence of fracture-healing complications after femoral neck fractures. Clinical Orthopedics and Related Research. 458. 175-9. Partio. E.. Tumpo. P.. Hirvensalo. E.. & Bostman. O. (1997). Totally absorbable fixation in the treatment of fractures of the distal femoral epiphyses. A prospective clinical study. Archives Q/Orthopedic and Trauma Surgery, 116(4), 213-6. Partio. E.. Hirvensalo. E.. Bostman. O.. & Rokkanen. P. (1996). A prospective controlled trial of fracture of the humeral medial epicondyle--how to treat? A nnales chirurgiae et gvnaecologiae, 85(1). 67-71. Patel. A.. Ricci. W.. McDonald. D.. & Borrelli. J. (2006). Treatment of periprosthetic femoral shaft nonunion. Journal ofArthroplasty. 21(3). 435-43. Poeisson. A.. Vavruch. I... & Hedlund. R. (2007). Long-term randomized comparison between a carbon fiber cage and the C Ioward procedure in the cervical spine. European Spine Journal, 16(2). 173-8. Petersen. M.. Lauritzen. J.. Schwarz. P.. & Lund. B. (1998). Effect of nasal salmon calcitonin on post-traurnatie osteopenia following ankle fracture. A randomized double-blind placebo-controlled study in 24 patients. Acta Ortlurpedic Scandinavirrn. 69(4). 347-50. Pfeiff'er. B.. Nubling. M.. Siebert. H.. & Schadel-Hopfner. M. (2006). A prospective multi-center cohort study of acute non-displaced fractures of the scaphoid: operative versus non-operative treatment. BJIC .1Itrsculoskeletal Disorders. 1 1. 41. Pflugrnacher. R.. Beth. P.. Schroeder. R.. & Schaser. K. (2007). Balloon kyphoplasty for the treatment of pathological fractures in the thoracic and lumbar spine caused by metastasis: one-year follow-up. Acta Radiol. 48(1). 89-95. Plecko. M.. & Kraus. A. (2005). Internal fixation of proximal humerus fractures using the locking proximal humerus plate. Oper Orthop Traurnatol. 17(1). 25-50. Popovic. N.. Gillet. P.. Rodriguez. A.. & Lemaire. R. (2000,). Fracture ofthe radial head with associated elbow dislocation: results of treatment using a floating radial head prosthesis. Journal of Orthopedic T raurna. 14(3). 171- 7. Porter. K.. & Davies. J. (1986). The control of pain after Keller's procedure-~a controlled double blind prospective trail with local anesthetic and placebo. Ann R Coll Surg Engl, 67(5). 293-4. SO:Ann R Coll Surg Engl Prince. R.. Sipos. A.. Hossain. A.. & Syverson. U. (2005). Sustained nonvertebral fragility fracture risk reduction after discontinuation of teriparatide treatment. Journal ofBone and Mineral Research. 20(9). 1507-13. Priehett. J. (1993). Rush rods versus plate osteosyntheses for unstable ankle fractures in the elderly. Orthopedics Review, 22(6). 691-6. Prokop. A.. Helling. H.. Hahn. U.. & Udomkaewkanjana. C. (2005). Biodegradable implants for Pipkin fractures. Clinical Orthopedics and Related Research. 432. 226-33. Rajasekhar. C.. Ray. P.. & Bhamra. M. (2001 ). Fixation of proximal humeral fractures with the Polarus nail. Journal ofShou/der and Elbow Surgery, 10(1). 7-10. Ramamurthy. C.. Cutler. I... Nuttal. 1).. & Simison. A. (2007). The factors affecting outcome after non-vascular bone grafting and internal fixation :for the nonunion of the scaphoid. Journal o/Bone and Joint Surgery. 89(5). 627-32. Randelli. P.. Landi. S.. Fanton. F.. & Hoover. G. (1999). Treatment ofipsilateral femoral neck and shaft fractures with the Russell-Taylor reconstructive nail. Orthopedics. 22(7). 673-6. Raschkc. M.. Rasmussen. M.. Govender. S.. & Segal. D. (2007). Effects of growth hormone in patients with tibial fracture: a randomized. double- blind. placebo-controlled clinical trial. European Journal of Erulocrirurlogv. 156(3 ). 341 -5 1 . Rebuzzi. E.. Pannone. A.. Schiavetti. S.. & Santoriello. P. (2002). IMHS clinical experience in the treatment of peritrochanteric fractures. The results ofa multicentric Italian study of98l cases. Injury. 33(5). 407-12. Rehnberg. L.. & Olerud. C. (1989). Fixation of femoral neck fractures. Comparison ofthe U ppsala and von Bahr screws. A eta Orthop Scand, 60(5). 579-84. Reilly. T.. Sasso. R.. & Hall. P. (2003). Atlantoaxial stabilization: clinical comparison of posterior cervical wiring technique with transarticular screw fixation. Journal ofSpinal Disorders and Techniques, 16(3). 248- 53. Richards. P.. Kurta. 1.. Jasani. V.. Jones. C.. & Rahrnatalla. A. (2007). Assessment ofCAOS as a training model in spinal surgery: a randomized study. European Spine Journal. 16(2). 239-44. Richter. D.. Ostermann. P.. Ekkernkamp. A.. & Muhr. G. (1998). Elastic intramedullary nailing: a minimally invasive concept in the treatment of unstable forearm fractures in children. Journal of’l’edr'atric Ortlurpedics. 18(4). 457-61. Riedel. G.. & Valentin-Opran. A. (1999). Clinical evaluation of thMP-2lACS in orthopedic trauma: a progress report. Orthopedics. 22(7). 663-7. Riemer. B.. Miranda. M.. Butterfield. S.. & Burke. C. (1995). Nonreamed nailing ofclosed and minor open tibial fractures in patients with blunt polytraurna. Clinical Orthopedics and Related Research. 320. 199-24. Rikli. D.. Businger. A.. & Babst. R. (2005). Dorsal double-plate fixation ofthe distal radius. Oper Orthop Traumatol, 17(6). 624-40. Ring. D.. Jupiter. J.. Brennwald. J.. & Buchler. U. (1997). Prospective multicenter 125 trial ofa plate for dorsal fixation of distal radius fractures. Journal of Hand Surgery. 22(5). 777-84. Ristinierni. J.. Flinkkila. T.. Hyvoncn. P.. & Lakovaara. M. (2007). RhBMP-7 accelerates the healing in distal tibial fractures treated by external fixation. Journal ofiBone and Joint .S'rrrgery. 89(2). 265-72. Rodriguez-Merchan. E.. & Galindo. E. (1999). Intra-articular displaced fractures of the calcaneus. Operative vs. non-Operative treatment. Intermrtional Orthopedics. 23( 1 ). 63—5. AU:Rodriguez-Merchan. E. (1995). Compression plating versus hackethal nailing in closed humeral shaft fractures failing nonoperative reduction. Journal ot'Or'thopedic Trauma. 9(3). 194-7. Roland. N.. Hughes. J.. Daley. M.. & Cook. .1. (1993). Electromagnetic stimulation as treatment oftinnitus: a pilot study. Clinical Otolary*ngologv and Allied Sciences, 18(4). 278-81 . Rornanus. B.. Bollini. G.. Dungl. P.. & Fixsen. J. (2000). Free vascular fibular transfer in congenital pseudoarthrosis of the tibia: results ofthe EPOS multicenter study. European Pediatric Orthopedic Society (EPOS). Journal o/I’ediatric Orthopedics. 9(2). 90-3. Rompe. J.. Rosendahl. T.. Schollner. C.. & Theis. C. (2001 ). High-energy extracorporeal shock wave treatment of nonunions. Clinical Orthopedics and Related Research. 387. 102—1 1 l. Saarenpaa. 1.. Haikkinen. T.. & .lalovaara. P. (2007). Treatment ofsubtrochanteric fractures. A comparison ofthe Gamma nail and the dynamic hip screw: short—term outcome in 58 patients. Interrurtiorurl ()rtlunredics. 31(1 ). 65- 70. Sadowski. C.. Lubbeke. A.. Saudan. M.. & Riand. N. (2002). Treatment of reverse oblique and transverse intertrochanteric fracture with use ofan intramedullary nail or a 95 degrees screw-plate: a prospective. randomized study. Journal ofBone and Joint Surgery, 84(3). 372-81. Safoury. Y. (1999). Use of reversed-flow vascularized pedicle fibular graft for treatment of nonunion of the tibia. Journal ofReconstruction ll licrosurgery. 15(1 ). 23-8. Sakamoto. K.. Nakamura. T.. Hagino. H.. & Endo. N. (2006). Report on the Japanese Orthopaedic Association's 3- year project observing hip fractures at fixed-point hospitals. Journal ofOr/hopedic Science. 1 1(2). 127-34. 126 Sarda. I-.. Crernicux. A.. Lebellec. Y.. & Meulernans. A. (2003). Inability of 99mTc-ciprofloxacin scintigraphy to discriminate between septic and sterile osteoarticular diseases. Journal ol'.\'uclear .lledicine, 44(6). 920-6. SattarSyed. A.. Islam. M.. Rabbani. K.. & Talukdcr. M. (1999). Pulsed Electromagnetic Fields for the Treatment of Bone Fractures. Bangladesh .lled Res Counc Bull. 25(1). 6-10. Satter Syed. A.. Islam. M.. Rabbani. K.. & Talukder. M. (1999.). Pulsed electromagnetic fields for the treatment of bone fractures. Bangladesh .lled Res Counc Bull, 25(1). 6-10. Saxena. A.. DiDomenico. L.. Widtfeldt. A.. & Adams. T. (2005). Implantable electrical bone stimulation for arthrodeses ofthe foot and ankle in high- risk patients: a multicenter study. Journal ofl‘oot and Ankle Surgery. 44(6). 450-4. Scheerlinck. T.. & Ilandclberg. F. (2002). Functional outcome after intramedullary nailing of humeral shaft fractures: comparison between retrograde Marchetti-Vicenzi and unreamed AO antegrade nailing. Journal o/‘Tr'aurna. 53(2). 60-71. Schildhauer. T.. Lcdoux. W.. Chapman. J.. & Henley M. (2003). Triangular osteosynthesis and iliosacral screw fixation for unstable sacral fractures: a cadaver and biomechanical evaluation under cyclic loads. Journal of Orthopedic T rautna. 17(1). 22-3 . Schmalholz. A. (1989). Bone cement for redislocated Colles' fracture. A prospective comparison with closed treatment. A cta ( )rthop Scand. 60(2). 212-7. Schmiedt. C.. Lu. Y.. Ileaney. K.. & Muir. P. (2007). Comparison of two doses of bone healing in dogs. Am J l'et Res. 68(8). 834-40. Schuind. F.. Haentjents. P.. Van Innis. F.. & Vandcr Maren. C. (1999). Prognostic factors in the treatment ofcarpal scaphoid nonunions. Journal ofHand Surgery. 24(4). 761 -76. Schwarz. N.. Posch. E.. Mayr. J.. & Fischmeister. F. (1998). Long-terrn results of unstable pelvic ring fractures in children. Injury. 29(6). 431-3. Segur. J.. Torner. P.. Garcia. S.. & Cornbalia. A. (1998). Ilse of bone allograft in tibial plateau fractures. Archives ofOrt/ropedics and Trauma Surgery. 1 17(6-7). 357-9. 127 Scide. K.. Weinrich. N.. Wenzi. M.. & Wolter. D. (2004). Three-dimensional load measurements in an external fixator. Journal of 'Biornechanics. 37(9). Selvakumar. K.. Saw. K.. & Fathirna. M. (2001). Comparison study between reamed and unreamed nailing ofclosed femoral fractures. ."lr/edical Journal otll-lalavsia. 56. 24-8. Scrnbo. 1.. Holmguist. 11.. Redlund-Johnell. 1..& .lohnell. O. (1994). Radiographic prediction of fialure after fixation ofcervical hip fractures. Acta Orthop Scand. 65(3). 295-8. Shepherd. L.. Shean. C.. Gelalis. 1.. & Lee. J. (2001). Prospective randomized study of reamed versus unreamed femoral intramedullary nailing: an assessment of procedures. Journal of Orthopedic Trauma, 15(1). 28-32. Siambanes. D.. & Miz. G. (1998). Treatment ofsymptomatic anterior cervical nonunion using the Rogers interspinous wiring technique. American Journal of(_)rthopedics. 27(12). 792-6. Sicbenrock. K.. Gautier. E.. Ziran. B.. & Ganz. R. (1998). Trochanteric flip osteotomy for cranial extension and muscle protection in acetabular fracture fixation using a Kocher-Langcnbcck approach. Journal of Orthopedic Trauma. 12(6). 387-91. Simmons. J.. Mooney. V.. & Thacker. I. (2004). Pseudarthrosis after lumbar spine fusion: nonoperative salvage with pulsed electromagnetic fields. American Journal ofOrthopedics. 33(1 ). 27-30. Siris. E.. Harris. S.. Easteli. R.. & Zanchetta. .1. (2005). Skeletal effects of raloxifene after 8 years: results from the continuing outcomes relevant to Evista (C ORE) study. Journal ofBone and :llineral Research. 20(9). 1514-24. Skoff. H. (1995). Bone marrow/allograft component therapy. A clinical trail. American Journal ofOrthopedics. 24(1). 40-7. Smith. W.. Ziran. B.. Agudelo. J.. & Morgan. S. (2006). Expandable intramedullary nailing for tibial and femoral fractures: a preliminary analysis of perioperative complications. Journal of(.)rthopedic Trauma. 20(5). 310-4. Sochart. D. (1998). Poor results following internal fixation ofdisplaced subcapital femoral fractures: complacency in fracture reduction. Archives of Orthopedic and Trauma Surgery. 1 17(6-7). 379-82. Sornrner. C.. Gautier. E.. Muller. M.. & Ilelfet. D. (2003). First clinical results of 128 the Locking Compression Plate (LCP). Injury. 34(2). 43-54. Sonne-Holm. S.. Nordkild. P.. Dyrbye. M.. & Jensen. J. (1987). The predictive value of bone scintigraphy after internal fixation of femoral neck fractures. Injury. 19(1). 33-5. Sorgcr. J.. Kirk. P.. Ruhnke. C.. Bjornson. S.. & Levy. M. ( 1999). Once daily. high dose versus divided. low dose gentarniein for open fractures. Clinical Orthopedics and Related Research. 366. 197-204. Stancic. M.. Gregorovic. E.. Nozica. E.. & Penezic. L. (2001). Anterior decompression and fixation versus posterior reposition and semirigid fixation in the treatment of unstable burst thoracolumbar fracture: prospective clinical trial. Croatia Medical Journal. 42(1). 49-53. Steffcnsen. B.. Caffesse. R.. Hanks. C.. & Avery. J. (1988). Clinical effects of electrornagentic stimulation as an adjunct to periodontal therapy. Journal ofI’erimlontologv. 5 9(1). 46-52. Stoffeien. D.. & Broos. P. (1998). Minimally displaced distal radius fractures: do they need plaster treatment? Journal ofTr'aurmr.44(3). 503-5. Strohrn. P.. Muller. C.. Boll. I & Pfister. U. (2004). Two procedures for kirschner wire osteosynthesis ofdistal radial fractures. A randomized trial. Journal o/Bone and Joint Surgery. 86-12. 2621 -8. Sweet. F.. Lanke. L.. Bridwell. K.. & Blanke. K. (2001). Prospective radiographic and clinical outcomes and complications ofsingle solid rod instrumented anterior spinal fusion in adolescent idiopathic scoliosis. Spine. 26(18). 1956-65. Swiontkowski. M.. Aro. H.._ Donell. S.. & Esterhal. J. (2006). Recombinant human bone morphogenetic protien-2 in open tibial fractures. A subgroup analysis of data combined from two prospective randomized studies. Journal ofBone and Joint Surgery. 88(6). 1258-65. Szivek. J.. Roberto. R.. Slack. J.. & Majeed. B. (2002). An implantable strain measurement system designed to detect spine fusion: preliminary results from a biomechanical in vivo study. Spine. 27(5). 487-97. Tejano. N.. Puno. R.. & Ignacio, J. (1996). The use of implantable direct current stimulation in multilevel spinal fusion without instrumentation. A prospective clinical and radiographic evaluation with long-term follow-up. Spine. 21(16). 1904—8. Tengstrand. B.. Cederholm. I Sodergvist. A.. & Tidermark. J. (2007). Effects of 129 protien-rieh supplementation and nandrolone on bone tissue after a hip fracture. Clinical Nutrition. 26(4). 406-5. Thalgott. J.. Sasso. R.. Cotler. H.. & Aebi. M. (1997). Adult spondylolisthesis treated with posterolateral lumbar fusion and pedicular instrumentation with A0 DC plates. Journal o/‘Spinal Disorder, 10(3). 204-8. Tidermark. J.. Ponzer. S.. Carlsson. P.. & Sodergvist. A. (2004). Effects of protein-rich supplementation and nandrolone in lean elderly women with femoral neck fractures. Clinical Nutrition, 23(4). 587-96. Tidermark. J .. Ponzer. S.. Svensson. 0.. & Sodegvist. A. (2003). Internal fixation compared with total hip replacement for displaced femoral neck fractures in the elderly. a randomised. controlled trial. Journal of Bone and Joint Surgery. 85(3). 380-8. Tidermark. J.. Zethraeus. N.. Svensson. 0.. Tornkvist. H.. & Ponzer. S. (2002). Quality of life related to fracture displacement among elderly patients with femoral neck fractures treated with internal fixation. Journal of’Trauma. 16(1). 34—8. 'l‘omaino. M.. King. J.. & Pizillo. M. (2000). Correction oflunate malalignrnent when bone grafting scaphoid nonunion with humpback deformity: rationale and results ofa technique revisited. Journal of Hand Surgery. 25(2). 322-9. Tomic. S.. Bumbasirevic. M.. Lesic. A.. & Bumbasirevic. V. (2006). Modification of the Ilizarov external fixator for aseptic hypertrophic nonunion of the clavical: an option for treatment. Journal of'Orthopedic Trauma.20(2). 122-8. 'I‘ornetta. P.. & Tiburzi. D. (1997). The treatment of femoral shaft fractures using intramedullary interlocked nails with and without intramedullary reaming: a preliminary report. Journal of'Orthopedic Tratrrna, 1 1(2). 89-92. Tornetta. P.. Bergman. M.. Watnik. N.. & Berkowitz. G. (1994). Treatment of grade-IIIb open tibial fractures. A prospective randomised comparison of external fixation and non-reamed locked nailing. Journal o/iBone and Joint Surgery, 76(1). 13-9. Trock. D.. Bollet. A.. Dyer. R.. Fielding. L.. Miner. W.. & Markoll. R. (1993). A double-blind trial ofthe clinical effects of pulsed electromagnetic fields in osteoarthritis. Journal o/Rheumatolog 2, 20(3). 456-60. Tu. Y.. Lin. C.. Su. J.. & Hsu. D. (1995). Unreamed interlocking nail versus 130 external fixator for open type III tibia fractures. Journal oj'Traur-na. 39(2). . 361-7. Tucker. M.. Schwappach. J.. Leighton. R.. Coupe. K.. & Ricci. W. (2007). Results of femoral intramedullary nailing in patients who are obese versus those who are not obese: a prospective multicenter comparison study. Journal of Orthopedic Trauma. 21(8). 523-9. Tudisco. C.. Bollini. G.. Dungl. P.. & Fixen. .l. (2000). Functional results at the end of skeletal growth in 30 patients affected by congenital pseudoarthrosis ofthe tibia. Journal ofl’ediatric Ortlurpedics. 9(2). 94- 102. Uhlin. B.. & Hammer. R. (1998). Attempted unreamed nailing in tibial fractures: a prospective consecutive series of 55 patients. A cta Orthopedic Scandcrnavian. 69(3). 301—5. Upadhyay. A.. .lain. P.. Mishra. P.. & Maini. L. (2004). Delayed internal fixation of fractures ofthe neck of the femur in young adults. A prospective. randomized study comparing close and open reduction. Journal ot'Bone andJoint .S'urgert'. 86(7). 1035-40. Van der Poest Clement. E.. Patka. P.. Vandormael. K.. & Haarman. H. (2000). The effect ofalendronate on bone mass after distal forearm fracture. Journal o/Bone and .llineral Research. 15(3). 586-93. Van Laarhovcn. C .. Meeuwis. J.. & van der Werken. C. (1996). Postoperative treatment of internally fixed ankle fractures: a prospective randomised study. Journal o/Bone and Joint Surgery, 78(3). 395-9. Van Vugt. A.. Oosterwijk. W.. & Goris. R. (1993). Predictive value ofearly scintirnetry in intracapsular hip fractures. A prospective study with regard to femoral head necrosis. delayed union and non-union. .-lrchives of Orthopedics and Trauma Surgery. 1 13(1). 33-8. Verheyen. C.. Smulders. T.. & van Walsurn. A. (2005). High secondary displacement rate in the conservative treatment of impacted femoral neck fractures in 105 patients. Archives of'Orthopedic and Trauma Surgery. 125(3). 166-8. Vinnars. B.. Ekenstam. F.. & Gerdin, B. (2007). Comparison ofdirect and indirect costs ofinternal fixation and cast treatment in acute scaphoid fractures: a randomized trial involving 52 patients. A cta Ortlurpedics, 78(5). 672-9. Wada. T.. Usui. M.. Isu. K.. & Yamawakii. S. (1999). Reonstruction and limb 131 salvage after resection for malignant bone tumor of the proximal humerus. A sling procedure using a free vascularised fibular graft. Journal o/‘Bone and Joint Surgery, 81 (5). 808-13. Waikakul. S.. Harnroongroj. T.. & Vanadurongw an. V. (1999‘). Immediate stabilization of unstable pelvic fractures versus delayed stabiliztation. Journal o/Jledical .-~1ssociation Thai. 82(7). 63 7-42. Waikakul. S.. Sakkarnkosol. S.. & Vanadurongwan. V. (1998). Vascular injuries in compound fractures of the leg with initially adequate circulation. Journal o/Bone and Joint Surgery. 80(2). 254-8. Wallny. T.. Sagebicl. C.. Westerman. K.. & Wagner. U. (1997). Comparative results of bracing and interlocking nailing in the treatment of humeral shaft fractures. Intermrtiorurl Orthopedics. 21(6). 374-9. Wang. C.. Chen. H.. Chen. C.. & Yang. K. (2001). Treatment of nonunions of long bone fractures with shock waves. Clinical Orthopedics and Related Research. 387. 95-101. \RF'ang. C.. Liu. 11.. & Fu. T. (2007). The effects of extracorporeal shockwave on acute hi gh-energy long bone fractures ofthe lower extremity. Archives of Ortlurpedic and Trauma Surgery. 127(2). 137-42. Watson. J.. Moed. B.. Crarner. K.. & Karges. D. (1998). Comparison of the compression hip screw with the Medoffsliding plate for intertrochanteric fractures. Clinical Orthopedics and Related Research. 348. 79—86. “Weber. 1).. & Peter. R. (1999). Distal femoral fractures after knee arthroplasty. International Orthopedics. 23(4). 236-9. Wei. S. Born. C.. Abene. A.. Ong. A.. & Hayda. R. ( 1999). Diaphyseal forearm fractures treated with and without bone graft. Journal ofTr'auma. 46(6). 1045-8. Weiss. S.. Zimmermann. G.. Baumgart. R.. & Kasten. P. (2005). Systemic regulation of angiogenesis and matrix degradation in bone regeneration- distraction osteogenesis compared to rigid fracture healing. Bone. 37(6). 781-90. \K’elch. W.. Willis. S.. & Gerszten. P. (2004). Implantable direct current stimulation in para-axial cervical arthrodesis. Advanced Therapv. 21(6). 389-400. Wentz. S.. Eberhardt. C.. & Leonhard. T. (1999). Reconstruction plate fixation with bone graft for mid-shaft clavicular non-union in semi-professional athletes. Journal o/‘Orthopedie Science, 4(4). 269-72. Whorton. A.. & Henley. M. (1998). The role of fixation of the fibula in open fracture ofthe tibial shaft with fractures ofthe ipsilateral fibula: indications and outcomes. Orthopedics. 21(10). 1 101-5. \X-""ililborg. O. (1990). Fixation of femoral neck fractures. A four-flanged nail versus threaded pins in 200 cases. Acta Orthopedics .S'camlinavian. 61(5). 415-8. Wilkins. R.. Kelly. C.. & Giusti. D. (1999). Bioassayed demineralized bone matrix and calcium sulfate: use in bone-grafting procedures. Anna/es Chirugiae et Gynaecologiae. 88(3), 180-5. Wittwer. (3.. Adeyemo. W.. Yerit. K.. & Voracek. M. (2006). Complications after zygoma fracture fixation: is there a difference between biodegradable materials and how do they compare with titanium osteosynthesis? Oral Surgery Oral ll/ledicine Oral Pathology Oral Radiology, 101(4). 419-25. Wolfe. S.. Mih. A.. Hotchkiss. R.. & Culp. R. (1998). Wide excision of the distal ulna: a multiecenter case study. Journal of'Hand Surgery-z 23(2). 222-8. Wong. T.. lp. F.. & Yeung. S. (2006). Comparison between pcrcutaneous transverse fixation and intramedullary K-wires in treating closed fractures of the metacarpal neck of the little finger. Journal of Hand Surgery. 3 l ( 1). 61 -5. Wood. CL. Boyd. R.. Carothers. T.. & Mansfield. F. (1995). The effect of pedicule screw/plate fixation on lumbar/Iumbosaeral autogenous bone graft fusions in patients with degenerative disc disease. Spine. 20(7). 819-30. Wright. T. (1997). Treatment of humeral diaphyseal nonunions in patients with severely compromised bone. Journal ofSou/h Orthopedic Association. 6(1). 1-7. Wu. 1-.. Renucci. J.. & Song. D. (2004). Rigid-plate fixation for the treatment of stemal nonunion. Journal of Thoracic Cardiovascular Surgery, 128(4). 623-4. Wu. C. (2003).Rcaming bone grafting to treat tibial shaft aseptic nonunion after plating. Journal o/‘Orthopedic Surge/y, 11(1). 16-21. Wu. C.. & Chen. W. (1997). Treatment of femoral shaft aseptic nonunions: comparison between closed and open bone-grafting techniques. Journal of Tramna. 43(1 ). 1 12-6. Wu. C.. Shih. C.. Chen. W.. & Tai. C. (1999). Effect of reaming bone grafting on treating femoral shaft aseptic nonunion after plating. Archives of Orthopedic and Trauma Surge/y, 1 19(75-6). 303-7. Yang. X.. Fei. J.. Wang. Z.. Yu. H.. & Sun. .1. (2005). Experimental study and clinical observation of minimum-contact plate in long bone fracture. ("hinese Journal ofTraumatologv, 8(2). 105-10. Yoo. M.. Cho. Y.. Kim. K.. & Khairuddin. M. (2006). Treatment of unstable peritrochanteric femoral fractures using 95 degrees angled blade plate. Journal ofOrthopedic Trauma, 20(6). 687-92. Young. C.. Nanu, A.. & Checketts. R. (2003). Seven-year outcome following Colles' type distal radial fracture. A comparison of two treatment methods. Journal oflland Surgery, 28(5). 422-6. Yue. J.. Sontich. J.. Miron, S.. & Peljovich. A. (2001). Blood 110w changes to the femoral head after acetabular fracture or dislocation in the acute injury and perioperative periods. Journal ofOrtlwpedic Trauma, 15(3). 170-6. Ziegcnfeldcr, T.. Hubschcr, J.. Abendroth. K.. & Syrbe. S. (2001). Effects of differentiated stress stimuli and electromagnetic fields on bones. [)eutsche Zeitschrift fur Sportmedizin, 52( 7-8). S51 . Ziegentelder. T.. Hubscher. J.. Abendroth. K.. Syrbe. S.. Ehrlich, S.. Ndinda. T.. & Gabriel. H. (2002). Effects ofdifferentiated stimuli caused by putting weight and electromagnetic fields on bones. International Journal of Sports .r‘lledicine. 238. 1468. Ziegenfelder. T.. Hubscher. J.. Abendroth. K.. Syrbe. S.. Ehrlich. S.. Ndinda. T.. & Gabriel. 11. (2001). Effects of differentiated stress stimuli and electromagnetic fields on bones. Deutsche Zeitsehri/t'fur Sportmedi:in. 52(7-8). 51. Zoega. B.. Karrholm. J.. & Lind. B. (1998). One-level cervical spine fusion. A randomized study. with or without plate fixation. using radiostereometry in 27 patients. Acta Orthopedics Scandinavian, 69(4). 363-8. Zollingcr. P.. Tuinebreijer. W.. Kreis. R.. & Breedervcld. R. (1999). Effect of vitamin C on frequency of reflex sympathetic dystrophy in wrist fractures: a randomised trial. Lancet, 354. 9195. 134 3 1293 01063 6801 " u N. H Ti ill I“ “II n I" SH Illiil mu HI