STUDY OF THE PLASTIC CARD METHOD FOR THe DETERMINATION OF THE HUMAN BLOOD GROUPS By Erma Marguerite Hill A TKESiS Submitted to the School of Graduate Studies of Michigan State College of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Bacteriology and Public Health 1953 as iHwer it |^| m m I tf femdtaBt* U m IU i «f HiafeigNi tl AfrtilllWI ABMl AfpMit 1b y a r u S I A A U m b I «f Wi n p i f i l i Im p Iki lip w «| MflW JMQNft •F 9 B M 0 Q M O T StapwlMB% tf lniBlillfr bbI PMLIb Im M 115) i n ACKNOWLEDGMENTS The author wishes to express her sincere appreciation to Dr. H. J. Stafseth for his inspiring and thoughtful guidance. Grateful acknowledgment and sincere thanks are extended to Dr. G. D. Cummings of the Michigan Department of Health who stimulated the interest in and provided the opportunity for advanced study, and to Dr. H. E. Cope for his unfailing inter­ est and guidance, and to Dr. M. B. Kurtz for his constant assistance and advice. Sincere thanks are extended to Mr. John Griffin of the Michigan Office of Civil Defense for the opportunity to carry on this investigation ana to use the information obtained. The writer is indebted to Miss Joan Wilhelm, Mr. Robert Gauthier, and Mr. Robert Johnson for their excellent technical assistance in the field study and to Mr. Robert Collins, Mrs. Herbert Davidson, Mrs. Carl Reagh, Mrs. Howard Brown, Mrs. horma Terrill, and Mrs. Pauline La Haie for their assistance in testing the survey specimens in the laboratory. The writer appreciates the Fellowsliip provided through the Michigan Department of Health. •IWHHHHHt ■JHfr-K-ft-fc ■fc-U-K* TABLE OF CONTENTS PAGE ILTKODUCTION......... '............................................ 1 iilSTOrtlCAL REVIEW................................................. 5 GENa h AL PhOCEDOrtES.................. .. 11 Test Tube Methods.............................................. Determination of AoO antigens............................... Determination of ABO agglutinins........................... Determination of ith0(D).antigens............................ Determination of rh*(C) antigens............................ Determination of weaklyreacting kh(Da) anti gens............ 11 11 12 12 12 13 Plastic Card Method............................................ 13 o X T lU l lI* .xji’Mi-- E S —U —'_Li-*S . « . . . • • • 13 • Adequacy of the Plastic Card Method Compared withTest Tube methods for the Determination of AuG and hhQ bloodGroups....... Organization of tne fielo study............................. Required numoer of olood groupdeterminations............. Source of oloou specimens................................ Organization of donor clinics............................ Organization of plastic card typing...................... IlCSUo.uS • Ip 13 13 lt> 17 It 1^ Potentialities and Limitations of u:- PlasticCard Method........ 32 Effect of red cell concentration............................. 32 Effect of albumin content of anti-P.r. serum................... 3b Effect of physical factors..................................* 36 T ine 31 Temperature............................................... 10 Other factors............................................. 13 Sensioiviuy of plastic card method compared with certain laboratory proceo .ires........................................ u3 Eli antibody titration..................................... tU Detection of weakly reacting hit positive^Du ) antigen..... It Coomos anti-glooulin test................................. 30 -I3CL SSIGi, ...................... -1- - - ....................................................................... M Li.TAit--TLT.c< C I T E u cl A STUDY OF THR PLh STIC CARD k FTHOD FOR DiTFRMINATION OF KUwAi. BLOOD GROUPo As a result of the advances in knowledge of red cell antigens and anti­ bodies in human blood, blood transfusion has assumed great importance in the field of therapeutic medicine. Today there are constantly increasing demands for transfusion therapy both for the civilian population and for the armed forces. In addition, it has been anticipated in the current preparation for civil defense tlxat administration of blood would be the ciiief therapeutic measure to oe employed in the event of atomic bombing. While the last decade has seen the development of higlily organized blood procurement systems and transfusion facilities in which little im­ provement is to be expected, the detection of blood groups discovered dur­ ing this period lias required the constant revision of blood typing and cross matching methods and a need for their evaluation exists. The Rh factor was discovered in 19U0 by Landsteiner and Wiener (l), and this anti­ gen was soon recognized as a factor in Dlood transfusion reactions. Con­ tinued investigation resulted not only in subdivision of the Rh blood group but also in the identification of other new antigen-antibody systems which have oeen implicated as the cause of certain transfusion reactions. The Rh blood groups and the other new blood groups were relatively difficult to detect. Furthermore, when it was found that certain Rh antigens, such as D a variants, were weak and that Rh antibodies usually could not be demonstrated in saline diluent, it became necessary to devise new techniques 2 for their detection. With the fundamental changes in blood typing methods has come a constant need for their evaluation in reference to particular conditions. The current preparation for civil defense has created an individual problem in blood typing. The properly identified blood wliich must be ready for immediate use in any bombed area cannot be procured and stock­ piled in advance. Each civilian is both potential donor and recipient, a situation which necessitates the blood typing of large populations as a civil defense measure. It lias been estimated by the Civil Defense Com­ mission of Michigan (2) that it will require the typing for ABO and Rh0 blood groups of 3,000,000 people in tliis state alone to provide adequate transfusion protection. The economic and temporal aspects of the blood typing involved have made it expedient to determine if blood typing tests performed in the field immediate to the individual being typed are as accurate and generally satisfactory as the laboratory methods in use. Sine the safety of blood transfusion is absolutely dependent on proper blood typ ing and cross matcliing, the reliability of a testing method is of paramount importance. The danger from transfusion of incompatible ABO blood groups is immediate and fatalities occur. In an emergency such as would arise in the case of atomic disaster it might be necessary to administer blood with­ out retyping either donor or recipient or resorting to the cross matcliing of bloods which is the safeguard of every transfusion. Furthermore, with potential transfusion of the general population greater emphasis must be placed on the correctness of Hh typing than when all the recipients are to be men. Although danger from immunization by incompatible Kh red cells 3 exists for each sex, it is recognized that the greatest harm comes from the transfusion of females before, or during, the child-bearing age with its attendent possibility of hemolytic disease for the next generation, dany babies have died of hemolytic disease as a result of prior transfusion of the mother with incompatible Rh red cells. The slide tests which have been developed for antigen-antibody re­ actions are applicable to field use and are capable of a high degree of accuracy when correctly used. However, the slide test for ABO blood group­ ing fell into considerable disrepute for mass blood typing as a result of its use for the armed forces in World War II. Comparison of the blood grouping records of men typed at that time and the results of subsequent careful studies of their blood groups have shown discrepancies estimated to be as large as 10 percent. It is recognized that such errors probably arose not only from performance and interpretation of the test itself but also from clerical errors involved in recording x.he results . There are modifications of the slide test, carried out on paper, which provide permanent records and permit re-examination of both the interpretation and records of the results. ABO blood grouping has been found to be satisfactory on various forms of paper, but no adequate survey of its reliability for Rh blood grouping exists. A plastic card technique has been advocated by Levinson and Schlutz (3) for ABO and Rh blood grouping in the field. These workers found it accurate and satisfactory in their own hands and in a limited field study which they initiated. The typing of small groups of individuals in the field was performed by persons skilled in the technique, and the results were compared with those obtained in the laboratory by well h controlled methods, particularly slide methods. Before the method could be considered for use in the civil defense program, an adequate study of its reliability in field tests was necessary. This field study was undertaken at the request of the Michigan Office of Civil Defense. The ABO and Rh blood groups of a statistically signifi­ cant number of people were determined using the plastic card method in actual field conditions. The results were compared with those obtained in the laboratory by the most accurate methods available, and the accuracy of the test was evaluated. An attempt was also made to examine possible sources of error in the performance of the test in the field. In order to determine limitations of the method, a brief laboratory investigation was made of the effect of certain physical conditions and variables which might alter the results obtained by the plastic card method in the field. It did not appear that the plastic card had been employed for the more refined Rh antigen-antibody tests . Although not the primary objective of this study, it seemed that antibody titrations, Coombs anti-globulin tests, and study of the Du antigens on the plastic card might more sharply define the limitations and possibilities of the medium. Therefore, the potentialities of the plastic card medium for tests of this type were investigated in the laboratory. 5 HISTORICAL RiSVIEW It -was the discovery in 1900 by Landsteiner (U) of the ABO blood group antigens in man that gave rise to the concept of the individuality of human blood and provided a basis for the practice of blood transfusion. He found that the red cells of certain individuals were agglutinated by the blood serum of other individuals. Thirty years later he received the Nobel Prize for this discovery. However, it was the finding of the Rh factor, or anti­ gen, in human blood in 19U0 by Landsteiner and Wiener (l) and subsequent developments that clarified many of the hitherto unexplained transfusion reactions. These workers found that rabbits and guinea pigs immunized with the red cells of the monkey Hacacus rhesus developed anti-rhesus antibodies which agglutinated the red cells of 65 percent of the white people of New York as well as the red cells of the monkey. nated as Rh positive or negative. The blood groups were desig­ Wiener and Peters (5) established the importance of the Rh antigen in blood transfusion when they demonstrated that Rh antibody was present in the serum of certain people who suffered hemolytic reactions following transfusion even though blood of the correct ABO group had been administered. Levine and Stetson (6) had reported a case in 1939 in which a mother had been immunized against a fetal antigen that was not of the ABO, Mb, or P blood groups then known, and they called the antigen of the infant a "new" antigen which the mother lacked. The antibody of the mother was later shown to be anti-Rh, and thus the fundamental role of the Rh antigen in 6 erythroblastosis fetalis or hemolytic disease of the newborn was estab­ lished . Continued investigation resulted in the identification of several antigens of the Rh system, designated by the American workers as Rh0 , rh*, and rh* *, and the corresponding Hr antigens. As early as 19U1 Wiener (7), Landsteiner and Wiener (8) , and Levine (9) recognized anti-Rh0 and anti-rh* serums and Levine mentioned the anti-Hr serum. antigenic individuality must be reflected. identified tlie additional antigen rh* *. These workers realized that In 19U3 Wiener and Sonn (10) The Rh0 antigen of Wiener was designated as D, the rh* as C , and the rh* 1 as £ by the English workers. At the end of 19U3 Fisher (11) in England showed that there were six Rh antigens which he theorized fell in three pairs which he called C c , D d , and E e . The theory has been confirmed by the finding of antiserum to each of the antigens. chromosome. Ttiree genes, namely one of each pair, are carried on one The relationship between the members of each pair is one of genetic allelemorphism, that is, one chromosome carries D or d but not both. It follows, since all the nucleated cells in the body except sex cells carry a double set of cliromosomes, that an individual may be either homozygous or heterozygous for each of the three Rh antigens. These six elementary antigens and Cw can be definitively determined by typing with separate specific antiserums. A third C antigen, Cw , allelomorphic to C and c, was described by Callender and Race (12) in 19U6 and its position as such was clearly established by finding examples of specific anti-Cw serum. 7 A group of other antigens allelomorphic to C and D exist for which specific antiserums have not been found. In 19U6 Stratton (13) found D antigens that were not detectable with the usual antiserums and methods of testing. These he designated as Du antigens. They are undoubtedly the same as were described by Wiener (lU) in 19Uli as Eli "intermediates" . While their inheritance is definitely allelomorphic, specific antiserums have not been found. There is no doubt that different grades of Duantigens exist; some can be detected by one anti-D serum and not by others (15,16), and some can be distinguished only by use of the anti-human globulin test. In the anti-globulin test, described by Coombs, Mourant, and Race (17,18), red cells which have been sensitized by anti-D serum, but not agglutinated by it, are exposed to anti-human globulin with resultant agglutination of D positive red cells . The validity of this procedure has not been ques­ tioned for the detection of Du variants; Rosenfield, Vogel, Miller, and Haber (19) were able to elute incomplete Rh antibody from the red cells by a modification of the Landsteiner and Miller method (20) and thus show that absorption of the antibody on the cells had occurred. The Du antigens are of clinical and serological importance, not merely academic, since they are D positive and have been shown by van Loghem (2l) and others (19,22) to be definitely antigenic to D negative persons. They must therefore be excluded when selecting Rh negative blood donors. Few studies of the incidence of Du bloods are available. The incidence of weakly reacting D positive (Du) blood was found by Rosenfield et al. (19) to be approximately 1.6 percent of tlie white population in New York City and about O.U percent of these failed to react directly with any anti-D typing serum. 8 The antibodies, or agglutinins, normally present in human serum that are specific against the A and B red cell antigens were shown by Landsteiner to be capable of causing agglutination in saline diluent. Likewise, the antibodies produced against the M, N, and P red cell antigens, the only other blood groups known before the discovery of the Rh factor, were active in a saline medium. For four years after the discovery of the Rh antigens, the recognition of the Rh antibody was confined to that which was active in saline. In contrast to the other blood groups, the Rh anti­ body was more reactive at 37 C . than at lower temperatures. While it was realized that antibody was not oeing demonstrated in most Rh negative mothers of erythroblastotic infants who were certainly affected by the presence of anti-Rh antibodies in their circulation, it was not until 19itU that Diamond (23) reported that the concentrated globulin of an anti-Rh serum inhibited the effect of the recognized saline anti-Rh agglutinin. In the same year Race (2lj) and Wiener (25) independently observed the same phenomenon. Race termed it an incomplete antibody, and Wiener called it blocking antibody. Many studies were initiated by these findings. In the meantime , al­ though blocking or inhibition of the saline Rh agglutinins showed the presence of a different type of antibody, saline continued to be used as a diluent. Diamond and Abelson (26) demonstrated that incomplete anti-Rh serum agglutinated red cells on a slide if the cell suspensions were very heavyj the test was most successful when the red cells were suspended in their own serum or in albumin. Wiener (27) and Wiener, Hurst, and Sonn- Gordon (28) used plasma as a diluent in the "conglutination" test. In 19U5 9 Diamond and Denton (29) studied various media, particularly those of protein nature, in -which to demonstrate the incomplete antibody, and they selected 20 percent bovine albumin as the most useful. Anti-human glubulin was found (18) to be of great importance for the demonstration of certain types of hh antibodies which do not agglutinate and do not block in saline but are usually demonstrable in albumin. Antibodies of tliis type have been called cryptagglutinoids by Hill, haberman, and Guy (30 ) and described as "blocking in albumin** by Witebsky and Mohn (31) • The latter authors (32) described a fourth type of Rh antibody which can only be identified by anti-human globulin but which does not block the reaction of the other types of Rh antibodies in saline or albumin. On these findings hinge the fundamental changes tliat have occurred in the methods of determination of the Rh antigen-antibody system, that is, use of Rh typing antiserums of the incomplete variety, protein diluents, and especially anti-human globulin. As a result, the methods of detection which were applicable to the Rh sys­ tems were instrumental in the discovery of other blood group systems. Although many studies of the incidence of ABO and Rh blood groups have been made since the discovery of these red cell antigens, few direct comparisons have been published of results obtained by more than one method using blood from the same individual. Chown, Peterson, Lewis and Hall (33) compared the incidence of Rh groups which they found by the capillary method using bloods from 792 persons with the findings of Race, liourant, Lawler, and Sanger (3U) who used the tube method to determine ttie Rh groups of 2,000 persons. Discombe and Meyer (35) tested 1 ,0$9 bloods by the method of Chown for a similar comparison. In a preliminary study 10 of large scale blood typing for civil defense Allen, Diamond, and Madden Ob) determined the AdO and Eh blood groups of 1,029 individuals using both a warm slide technique and test tube methodsj only 215 of the total number were tested independently. They found an error of 0.0b percent by the slide method as compared with an error of O.Ul percent by the tube method . 11 GENERAL PROCEDURES Test Tube Methods The methods of the Michigan Office of Civil Defense (2) for the determination of ABO and Rh blood groups were used throughout this study. Methods similar to these have been described as "modified'1 tube tests in order to distinguish them from tube tests in wiiich saline diluent is used. Antiserums of the incomplete, or hyperimmune, variety are employed in modified tube methods. These serums have been designated as "slide test" serums since anti-Rh serums containing saline-active agglutinins had previously been designated as "tube test" serums. The antiserums used were commercially availaole serums of the incom­ plete or "slide test" variety which conformed to the potency standards of the National Institutes of Health C37,33). Each lot of all the antiserums used was tested in the laboratory for specificity and suitability for the methods with bloods of predetermined blood groups. Antiserums of the same manufacture ana lot numbers were used for the tube procedures in the laboratory as were used for the corresponding plastic card tests in the field . Determination of ABO sntigens (direct typing). One drop each, 0.05 ml., of anti-A and anti-3 serum was placed in 3 x 3/t inch test tubes. To each was added an equal volume of a 2 percent suspension of the red cells to be tested in their own serum. The mixtures were thoroughly shaken 12 allowed "bo stand at room temperature for 60 minutes, and examined grossly for agglutination. The results were recorded by + or - signs. Determination of ABO agglutinins (reverse typing). Two drops, 0.1 ml., of serum from the blood to be tested were placed in each of two 3 x 3/S inch test tubes and heated for 15 minutes at 56 C . One drop, 0.05 ml., of a 2 percent suspension of known group A, Kh negative red cells in 0.9 percent sodium ciiloride solution was added to one tube, and an equal volume of a known group B, Kh negative red cell suspension was added to the second tube . The mixtures were shaken thoroughly and centrifuged for 2 minutes at 1,000 rpm. Each of the tubes was examined for agglutination, and the results were recorced. The results of the direct ABO typing and the confirmation test were recorded without reference of one to tne other. The results of the in­ dependent tests were correlated and summarized as the blood group A, B, A b , or 0. Determination of RhQ antigens. One drop, 0.05 ml., of anti-Rh0 (anti-D) serum was placed in a 3 x 3/8 inch test tube. An equal volume of the 2 percent suspension of red cells in their own serum was added. The red cell-antiserum mixture was thoroughly shaken, incubated at 37 C . in a water bath for 60 minutes, and centrifuged for 1-2 minutes at 1,000 rpm. examined grossly for agglutination, and the result Each tube was was recorded by a + sign or the abbreviation neg. Determination of rh1 (C) antigens. All specimens which were found to be Rh0(D) negative in the aforesaid test were retested in the same manner 13 both with a second anti-Rh0 (anti-D) serum of different manufacture and with an anti-Rh0*(CD) serum. If the second anti-D serum gave a positive result when the first test was negative, it was necessary to justify the existence of contrary results. Since opposite results may be a reflection of inherent differences in the strength of the antiserums or may be due to the presence of a weakly reacting Rh positive (Du) antigen, the re­ actions with additional antiserums were studied, and the anti-human globulin test was applied to apparently negative bloods. If the results with anti-CD serum were positive and if the Ooombs antihuman globulin test with anti-D serum proved the cells to be D negative, the results were recorded as rh'(C) positive. Determination of weakly reacting Rh(Du) antigens. Equal volumes, 0.1 ml., of a 2 percent suspension of red cells in 0.9 percent sodium chloride solution and of incomplete anti-Rh0(anti-D) serum were placed in a test tube, mixed, and incubated for 60 minutes at 37 C . Control tests were prepared in the same manner using 20 percent diagnostic albumin in­ stead of anti-Rh0(anti-D) serum. The cells were washed three times with approximately 5 ml. of pliysiological saline. Two drops of anti-human globulin serum were added to the 2 percent suspension of washed cells and mixed. After incubation at room temperature for 15 minutes, they were centrifuged at 1,000 rpm. for 1 minute and examined for agglutination. Plastic Card Method The plastic card used was recommended by Dr. Sidney Levinson and his co-workers. The cards were of cellulose acetate, 8.5 x 6 c m . and 0.015 mm. Hi thick, with properly labelled printed circles for each of the blood grouping tests. Commercially available anti-A, anti-B, and anti-Rh0(anti-D) "slide test11 serums were used which conformed to the potency and specificity standards of the National Institutes of Health and which had been retested in the laboratory for this procedure. One drop of each serum was placed at one edge of its designated circle. One small drop of blood each for the tests with anti-A and anti-B serums and two full drops for the Rh test were placed within the circle near the antiserum. The blood and anti­ serums were immediately and thoroughly mixed by means of clean flat-sided toothpicks, always mixing the blood and anti-Rh serum first. The card v&s allowed to remain flat on the table for 2 minutes with no intervening stirring or motion. At the end of tldLs time the card was lifted to a vertical position, the mixtures allowed to drain to the bottom of their respective circles , and the excess material was absorbed by wiping around the base of the circle with a cotton-tipped applicator. The results of the tests were read and recorded at once on the card, the International blood group as A, B, AB, or 0 and Rh. group as Positive or Negative . 15 iiXPiiuIiuiiVlAL STUDIES Aaeguacy of the Plastic Card Method Compared with Standard Tube Methods for the Determination of ABO and fthQ Blood Qroups This field study was undertaken at the request and under the auspices of the Michigan Office of Civil Defense which was currently engaged in a blood typing program in Michigan. In order to obtain data concerning the plastic card method it was determined tliat parallel tests using the plastic card method and the standard Office of Civil Defense tube methods for the determination of the AnO and ith0^,D) blood groups were to oe performed on bloods from a statistically significant number of individuals. The tests by the plastic card method were to be performed at the location of the procurement of oloou oy a testing team wliich was to consist of a super­ visor, one technician, and two other individuals trained in the procedure. The numoer of donors whose blood was to oe tested by one card-typing team was not to exceed 2>0 per day. a blood-procurement team was to operate in conjunction with the plastic card-typing team to obtain blood specimens by venipuncture for the standard tube tests. Organization of the Field Study hequired number of blood group determinations. This study was designed to oe sufficiently comprehensive to be statistically significant in evaluatin : the worth of the plastic card method for establistiment of the Inter­ national AbG blood group and the hh blood group. Because the two methods 16 of blood typing were to be performed with blood from the same donor and the one method of typing was assumed to be correct, the problem was one of checking a new procedure against an accepted correct procedure rather than a statistical evaluation of two random samples or two undetermined methods. In considering the volume of tests required if the accuracy was to be maintained within 0.3 percent error, the problem was approached from the standpoint of group A3. Since ttiis group has the lowest frequency, con­ stituting only h percent of all ABO groups, the volume of tests required to prove accuracy for the AB group would be more than adequate for all the other groups. In order to obtain a measurable error of group Ab (one incorrectly determined AB olood) it would be necessary to have approximately 10,OCX) tests. From tiiis number there should oe approximately JiOO cases of group A3, either hi positive or negative; a G.3 percent error among these would allow 1.2 incorrect A3 tests. Thus, with 10,000 pests it would be possible to obtain a measureable error of group A3. however, on phis same oasis with 10,000 tests ana a frequency distribu­ tion of ul percent group A., 10 percent group B, and percent group 0, 12.3 incorrect group A, 3.0 incorrect group 3, and 13.3 incorrect group 0 pests would be within the limits of 0.3 percent degree of accuracy. Source of ulood specimens. The initial study was conducted in the l'tiree hi vers area and the final study, with which this report primarily deals , was conducted in the Traverse City area since each was a community in wnich approximately 10,000 persons could be expected to report for blood typing. 17 The location and hours for each blood typing clinic were scheduled in advance, by the representative of the Office of Civil Defense, to attract as many of the population as possible. Schools, factories, state institutions, and public locations for those individuals who were not members of special groups were selected as sites for bleeding clinics. Thus it was possible to obtain a test population that was representative of all age groups and which included infants, preschool children, elemen­ tary and high school children, and adults . Organization of Donor Clinics . Blood typing clinics were so arranged t’ nat a first clerk prepared the informationblank, required by the Office of Civil Defense, on which was recorded the name, address, and signature of the donor. The information olanks were assigned consecutive, identify­ ing accession numbers. The donor carried the blank to a second clerk who recorded the identical number and name on a plastic blood-typing card. The plastic card and information blank were handed to a hostess who es­ corted the donor to the table where the blood types were determined oy the plastic card method after the donor was properly identified with the information on the d a n k and on the plsstic card. The donor was then directed to the venous deeding station where the name on the information d a n k and the donor were again identified. Tne consecutive, identifying numbers of the information blanks had been paralleled on adhesive tape and firmly affixed to the sterile 3 inch becton-Dickinson vacutainers which were used for the collection of venous d o o d specimens. Tne vacutainers had been placed in numerical order in 16 50—place metal racks and were withdrawn from the racks only in numeri­ cal order to accompany the corresponding information blank carried by the donor. After the venous Dlood specimen was and correspondingly numbered vacutainer, it rack in numerical order. was drawn into theidentified returned to themetal The original metal rack and its numbered vacu­ tainers of olood were sent directly to the laboratory. Organization of plastic card blood typing. After identification of the name of tlie individual presenting himself for typing with that on the plastic cara, one member of the typing team placed the antiserums in their respectively laoelled circular areas on the plastic card. ^ sec­ ond member of the team prepared the donor's finger by spongingit with VO percent ethanol and punctured the finger tip with a sterilebard- Parker olaae . Without necessarily waiting for a free-falling drop, drops of olood were expressed and touched to the circular areas on the cara withort touching the antiserun. After the performance of the test as descrioed under General Proceuures ana the recording of tne results on the card, the card was placed vertically in a proven ooard which held 23 cards, bach card was placed in numerical order. When the serum- olooo mixture was completely dry, each series of 25 cards was reviewed, packaged together with an outer label oearin.r the series numoers, date of examination, ana the initials of the supervising bacteriologist. All the card typings were sent each day to the laboratory where they were reexaainrd for clerical and technical errors oefore comparisons were made with tne results of corresponding tube tests which were performed and 19 recorded independently in the local testing laboratory in the Tliree hivers study and in the Laboratory of the Michigan Department of health, Lansing, in tiie Traverse City study. Results The initial study of the plastic card method for blood typing in the field was undertaken tc determine whether or not the plastic card method could be employed in a large scale typing program using locally recruited and trained personnel of average ability anc training. The initial instruction in the use of the method was demonstrated by Dr. Sidney Levinson. The supervisory personnel of the card-typing teams were persons with a college background and some training and experience in laboratory procedures, including that of olood typing in small hos­ pitals. A group of o,tyi individuals was tested in the Three Rivers area during a 6-week period . The parallel blood typing by the standard Office of Civil Defense methods was performed in a local laboratory which was established as were the other laboratories for the Michigan Office of Civil Defense ulood tyring program. Two venous olood specimens were ootained from 10 percent of the first hundred donors and from 2 per­ cent of all subsequent ones; the duplicates ware submitted to the Michigan Department of health Laboratories for checking of accuracy. A summary of the results obtained in the study conducted in the Throe Rivers area is presented in Taole 1. Tne results of 0,115 blood typings performed by the plastic card method were in agreement with the results obtained by the standard tube methods. It may oe noted in the .u PitaLx^i.Jlra 31'IL'i Ur' I ■;yi .iLOOij ukuop Dij‘i1iiiu.Ilvj.iIOi,D b l T h a PLASTIC CAiaJ itJiliob AS PEiU,'Ciu iD AVKaAGa SUF^iiVI&IC* i.umber i f. X V1'3 V" .L3 II c ^ .ica U - ,91 s ; ViGi‘ao'lo:\. ~ jx i-o ft/ correct r / s u l t s T e c ln ic a lly •insatisfactury ana inconcLjsiv* oasts -r r c r s i e c l c r i c a l tran scrip tion ai.0 V03 J J •J ! *:S % .0 3>cl 30 0.3:- jC is * 0.01 yrounin • 1 - s t s if a T c r j ii. nij0 .^rouciri'-xrrorj i:; Percent 0.2 ro O 21 table ttiat 3&S typings by the plastic card method were discarded. Since these were classified with the errors , it constituted an imme­ diate error of U.J4 percent. 'I'nree hundred and twenty-five of these results were discarded because incorrect tecimique was used on two successive days when a technician experienced only in the slide method of olood typing was allowed to substitute on one team. The technician adamantly followed a technique used in slide methods for blood typing; the olood and antiserums were mixed by a rotary motion and allowed to dry on the card instead of draining off the excess material. sults were somewhat obscured. The re­ In most instances the results could be read ana were recoraea. A clerical error of O,3U percent was found; with these, the results were evident ana satisfactory out incorrectly recorued on the card. The error due to an actual discrepancy oetween the results founa uy the two methods was 0.21 percent. Vfhen the resxALts found oy the plastic card method were analyzed ana the total serformance of the test in the field was examined, it was eviuent ooth that the plastic earn method possessed the potentialities of a desirable method ana that there were inherent dangers in the metnod unless the olood typing was conuucted by adequately trained and experienued personnel. It was also recognized tiiat parallel tuioe testing of the xiighest degree of accuracy would be more readily obtained in a permanent laboratory prepared to perform the more refined hh antigenanticody studies. study of this nature was onaertaken in the Traverse City area wuere tne olood specimens of g,6b3 individuals war-.: tested both by the plastic card mothoa and by the standard Office of Civil Defense 22 procedures. Well qualified bacteriologists had oeen carefully trained in olood typing procedures and in the plastic card teclinique oy the author. The tests performed on the plastic cards were examined and the results were recordeo at the place of olood procurement oy testing reams under the direct supervision of the responsicle oacteriologists. i:.e standard Office of Civil Defense olood taping was performed in the Laboratories of the Hichigan Department of Health under the direct supervision of the author. In addition, all the plastic card test results wire re-examined oy the author, upon submission of the cards each day, for verification of the interpretation of the results, and the results were then compared with tus tuoe test results which Lad seen independently teste^i and recorded. L sur-mary of t».e results Ootaineu in the Traverse City area is cre­ setitad in Table 2. The results of y,olO of the ole oh tynings performed tne plastic card method were in agreement with the results obtained :y tue standaru tuoe r.ethoi. The recorded results of 1;, olood typings serf armed ~y the plastic card n.etnod ail not agree wi. t'.. taose obtained :..y the tuoe i,ethos . The results of 2l of the rtn tests on the cards were inconclusive; therefore, no results were recorded in the field for bloods. Tnrcc clerical errors were observed on the cards; these •,;er= incorrect records of tiis evident results. In every case where discrepant or inconclusive results existed, trie sloou typings were immediately repeated from the venous specimens in tj.o laboratory. At least tbr-to dif .Vr-~nt ant* -D serums arm. anti-CD serums were used for tries-! ret-eat tests. The uloocs were also tested with clij A C C u iu -C I Oi- i, ^ 5 3 C 'iAX'i.' >.ai0up J i i i i a i l A ' i I G : . j Y!:« P L A 3 1 IC C n n j i.nTHCD nuriuer Individuals tested Percent d,od3 9 ,-10 W Ju r e s u l t s inconclusive 2c 0.29 Results discrepant 15 0 .1 6 Error in Ano r;rouDin~ 1 0.01 Error in nhc proupinj lii o .i5 3 C .03 Results in ccr.plete a n\ Results with c l e r i c a l error -r.t r\j 2k nnti-C and anti-E serums since the weak Du antigen is frequently associ­ ated with C and E antigens. If indicated, the anti-human globulin test was performed using several different incomplete Ah antiserums. Of the 21' inconclusive results in which the plastic card tests were not considered definitely negative or definitely positive, 10 bloods were ih-j^D) negative in the tuioe test and also negative when tested with the Coomos anti-human glooulin test. oe idi0 positive by the ouoe test. Eighteen of the oloods were proved to however, three of these were D a variants . in the discrepant results, w'nich are sujnmarized in Table 2, there ’ .jus one olood grouping which was recorued as group AB on the card but was founc to be group d by the test tube procedure. This single AriO .loco grouping discrepancy represented an error in judgment in the field. Tno plastic card test was unsatisfactory in appearance, and it should ..avo been repeated iru.eaiately in the field. The other lU discrepancies between the results obtained in the field by the plastic card method and in the laoorator,/ oy the test tube method wore discrepancies in An olood types. live of the results on tne plastic card were rear! as Ah0(.d) positive . The test tuoe procedures showed that two of these bloods were Ah0CD) and rh*(C) negative and three of them were AhoC^) negative, r h ’(C) positive. *.inti of the lU eloocs wit'h which discrepancies occurred in Ah typing were recorded as u:i negative on the plastic card, but they proved to be ith positive when tested by laboratory tube procedures, however, six of these nine bloods were "low grade" Du variants which were detected, in the laboratory only 25 the use of more than one antx—D and anti-CD serum. definitely established as They were variants by the Coombs anti—globulin test. It had been observed in the Three reivers study that many aggluti­ nations on the plastic card were not optimum although the correct results could usually be arrived at oy careful observation. In establishing standards for the plastic cara testing wliich was to be performed at Traverse City, it was determined that every test must appear satisfactory, the criteria being either adequate, well-defined agglutination or definite lack of agglutination. releasing the donor. to :,e recorded. Otherwise the test was to no repeated before If the test was still unsatisfactory, no result was In an endeavor to meet this standard, the plastic card testing teams repeated lii'J ltd tests and m AuO tests . As mag be ooserveu ir. Taole 3 only lc examinations , all of which were ith tests, were incon­ clusive after the second test. There wore 10 additional uh tests which w* re considered inconclusive on the first test and si.ould nave been re­ peated. The number of tests repeated in the field corresponds to 22 specimens from the same series of bloods wl.icn wer re-examined in the laboratory in oraer to verify results or to attain completely satisfactory tests . It would have been desirable to repeat the blood typing of all the individuals in which inconclusive or discrepant results were ootnined. .towover, a seconu card typing and a second venous olood specimen for laboratory testing could only ue ontainec from 10 of the 13 individuals whose blooo tests Lad resulted in discrepant results oy the two methods. It may nt noted in Tables l ,e fir. ,ar. of o n . u n a . oese a .3 - ■JJ 1.32 O.Oo .^0 yroupx:^ o.iy Care t a s t s inconclu sive a f t e r re per. tea te s t Caru wCsfs inc-Ducf <.SxVoj Percent >3 '; v un0 .roupm- -xXc1.’* no^ o^'t'unf: 10 c,io re i-" _ .}_i_-'j vfJ ha ..XjOC’jJ ItllO'-' i-ji *J OivX vJX: L .,01U- opeoirit ..uroei X7 u X37 :vx kOS i l t nocorded on t i c Cara -X0 (D) + ( + ; + , ,172 l,23u 1 ,5oC C' J ,1b") 17I 3 X3'L,0u3 c ,060 ■,X3 i,?'l ftli0(D) + ?? ?•? -ea O riginal rxardr.ation r e s u lt Reported R e - a n a lv of V sis P la s t ic Card nesul t 07 Laboratory -■og. or. s a t . l iis a t , + r.sat. ;insat. + J n s a t. .,O i"* ; :* /.iXi •'•*. i. 'X •*—t_.• Unsat. ^r * i»6f. ivJ 'jt-i JJ -... D :ui0(L) •■e d* *'t^v ,~t • neg. i.e . rh*;C; Repeat Lxar.ination r l a s t i c Card n !io(L) «eg *'(j + .,eg. + + ■eg. + + o rsa t, ■Jijsat. 1,eo • —-xjI *dX lOl/iJ P. J»ji 111.. lil'it i i',.- . . I(ba) +(2JU) i.L 3 , H - u) ..eg. ■oc-r >* +(L11) ‘■62. + + Laboratory fiRo(3) not repeated ‘■eg. ..eg. ‘■eg. -■eg. not repeated -iot repeated i.ot repeated +(DU) +(du) +(DU) +(Wk) +(DU) not repeated rii (C) ■eg, + + + AivALi jI S OF x.iiO ...jjOOD i^ullF RESULTS DIoCkiiFANT oY PLASTIC CAnl) .kI.i, L/._>OiN Repeat Lxanination Ori ;in a l Cjxar.inaclor Specimen nuinuer n e s u lt itecorded or. P la s t ic Card ri.a—s na 1 c sis oi P la s t ic Card n e su lt reported oy Laboratory ^ •'l a s t l c CarU Laboratory ..•3s u i t lie 2 ursSo . 1 0 d ro co 29 plastic cards, the results of five of the Kh tests and the one AoO test, 'which was teclmically unsatisfactory on the first examination, did not o.-rcc with the original results obtained oy the plastic card method. The,,' uici agree ooth with the original ana the duplicated results obtained on lood from the same individuals oy laboratory tube procedures. These five results were classified as teclmical errors in the performance of the original card test. When the other ii of the 10 bloods were retyped by the plastic card methoa, the same results were obtained as haa Deen found with the original plastic card test; all bloods appeared to be Ah no -alive . These four results on the plastic card were in disagreement with the Ah0(D) positive results obtained oy the laboratory tuoe procedures, uowever, the correct Ah0(D) positive typings were ascertained only after application of several antiserums arid the use of the anti-human globulin test, Table 6. If these four results had oeen compared with the results from only one tube test or from one of lower sensitivity, they would likely have been found to be in agreement. ..evertheles s , these four resulus were considered as errors inherent in the plastic card method. A review of all the parallel results showed that of the IS' blood typ­ ing results determined by the plastic card method which did not agree with the tube method results, 11 had been considered of questionable value before comparison of the parallel results. slood groups had oeen recorded on five plastic cards which the field Lacteriolc^jsts annotates as tests of dubious value. An acioitional six results were considered unsatisfactory oy the author on review of the card tests prior to com­ parison with the laboratory results. These 11 plastic card typings itii T'iPiAG Oi1 lu DLOOD ISIAG AULTIrLfc TUS& TiSid Decree o f A gglutin ation Specimen lo'dtine ieSo Incomplete Serums 1Vi Ill - 191 2 |. 1,23d 1,360 2,7 Ay 3 ,h y A,0A3 9 , a97 : ,9b2 ' ,060 1 ]. - inti-D 2 2 Li u 2 e - - 1 + + i ,659 - + 7,672 ,967 - 2 f- , 6 3 3 A n tiCD Supplementary Tests Incomplete Serums S alin e A gglut. Serums /•.ntx— b A n ti- A n ti- ^ n t i a L i D 9 3 . — - - - A i u A A i, + 3 A 2 + 2 A 1 3 A + L: n n :i 1 - + 1 u - 2 + + i. + ii - - L - - 2 A 2 3 - A n ti-G looulin T ests 1 Anti-D serums e A 2 3 0 ~ - M l — — _ - - 3 A A M 3 2 - A i A > A A 3 - A it A i M 2 3 - 2 - 3 + - - - - - - - 3 3 A A h 3 2 3 - 1 - 1 a - - - 'i + I - - 3 - - i a - - - - - A - - - - - u - A A A 2 1 + 2 * Serum used for p l a s t i c card t e s t , fc* Weakly rea ctin g id. p o s itiv e (Du) uloods. i /! n 1. A 3 u A ii 3 3 A A 3 Kli0(D) r h ‘ (C) ^eg. +£* + + *eg. L - A itn Groups j, —* +W beg. -eg. neg. + +irZ Aeg. Aeg. -HHi + + 31 should not have been recorded as conclusive on the basis of a single sample. Among the plastic card tests of dubious value were two recorded as Kh0CD) positive, bloods 3697 and 167, which were entirely negative in laboratory tests. When olood '.,697 was retested oy the plastic card method, it appeared negative. The results with blood 167 were not con­ sidered discrepent when tne original test was re-examined so the test was not repeated; it appeared to the reviewer to be Ah negative on the original card, and the error was thus one of interpretation and record­ ing of the results. Of the four discrepent results wliich appeared to be satisfactory on the original plastic card, one olood (no. 7 ,65>9) was designated as kn0(D) positive on the card but it was persistently Rh0(D) negative, ri*(C) positive on laboratory testing; repeated tests on tliree plastic cards were Ah0(D) negative . The tiiree remaining discrepancies in results were from bloods that appeared to be satisfactory Rh0(D) negatives by the plastic card method and also gave negative or weak reactions with the same antiserums in t.he first tube procedure. Tne correct results, ith0(D) positive, were only recognized because all oloods which appeared to ue ith negative ware routinely suojected to additional tests in the laboratory procedure . In this study the determination of Au>0 and hh0(D) blood groups by the plastic card method exhibited a technical error of 0.16 percent and a total error of 0.69 percent, when inconclusive results and clerical errors were included, as compared with results found oy the laboratory test tube methods. 32 Potentialities and Limitations of the Plastic Card Method It is well known in the field of serological testing that such factors as temperature , time, diluent, ratios of reacting substances and even the method of combination of reagents frequently have a significant influence on the results . Therefore, in order to test more completely Lb'..- potentialities and. define the limitations of the plastic card as a vehicle for the antigen-antibody reactions of olood grouping, laboratory studies were made of certain of the variables which might influence the results of the plastic card tests. In view of tne fact that agglutination in the Rh system is more difficult to analyze, emphasis was placed on ex­ amination of the Mil antigen-antibody reaction. Effect of Red Cell Concentration Levinson and Schlutz (3) , who advocated the plastic card for blood eroupine, suggested in a personal communication that two drops of fresh blooc and one drop of anti-Kh0 serum ;:e used routinely for the deter­ mination of xth types oy the plastic card method and that tliree drops of the olood of anemic individuals might be required for optimum aggluti­ nation. On the other hand, they recommenced less than a. drop of olood, a small amount on a fist toothpick, as a suitable volume to add to one Iron of anti-A ana anti-a serums. Allen, Diamond, and Madden (3c) used a nvery small" amount of olood to determine ABO groups oy the slide method. It was observed in the preliminary field study at Three Rivers that when a larger volume of blood than that recommended was inadvertent­ ly added to the serum for AuO grouping, the resulting agglutination 33 appeared more distinct. It seemed desirable to determine the optimum amounts of antigen and antibody with as quantitative a procedure as the elastic card medium permitted. Rh positive red cells were centrifuged until there was no change in volume of the packed cells. Suspensions of the packed red cells ranging from 15 to 60 percent were prepared in group specific serum. A commercially prepared incomplete anti-R'n0(anti-.D) serum which con­ formed to the National Institutes of Health requirements was used for the tests. It was necessary to pipette one amount of serum and one amount of red cell suspension at a uime and to proceed with the mixing, timing, and draining of the serum-cell mixture, because of the size of the defined area witliin the circles on the plastic card, a total volume of* 0.2 ml, was the feasible limit of quantity that could oe used. To oacn of tliree defined areas on plastic cards was added 0.05 *"1. -■nti-hh0 serum ana to a fourth was added C.l ml. of the serum. The red cell suspension of a given concentration was added to the serum using O.G'5, 0.10, and 0.15 ml. with the 0.05 ml. amounts of serum and using 0.10 ml. of red cell suspension with the 0.10 ml. of serum. cells and serum were thorouglily mixed, and a time interval minutes was The red of exactly 2 allowed oetween the mixing and draining of the excess material, from the card. The experiment was repeated several times. The effect of red cell concentration on ABO reactions on the plastic card was examined in a similar manner. Representative data are presented in Tables I ana t . It was observed that a considerable latitude in the proportions of red cells and serum was possible, nowever, 0 .Op ml. of anti—Rh serum T/.iILL '{ rilr'SCT Or’ RiiD CuLL CoNCLdTie.IICvi IOr* rL. iUiU.0 iICu«S Ou iiL. PLASTIC CArtD Volume o f Red Volume of C ell Susoension , A n ti -Hi ^ Serum, nil. ml. degree of A gglutination in 2 ;minutes Percent ned C ell Concentration in Autologous Serum 00 50 50 30 23; 20 L h 3 1 1 i 3 3 1 O.C5 0.0p h 0.10 0 .0 5 3* 0.15 0.05 3- 5 h n 3 1 0.10 0.10 3* n 5 n c 1 ± V isc o u s. 15 + + + + % i i ’ridCT Oj jusD C^LL CUNCSl.TRATICN ON ADO udACTICNs 0 ., THn PLASTIC CARD Anti serum Volume of Red C ell Suspension » Sp eci­ Volume fic ity ml. Degree o f A gglutin ation in 2 minutes Percent Red C e ll Conce n tratio n in Autologous Serum 50 10 20 oO* 30 25 15 l Anti-A Anti-P 0.05 0.05 0 0 0 .05 Anti-A A nti-o 0.05 0.05 0 0 0.10 Anti-A Anti-u 0.05 0.05 0 0 0.025 * 2 2 1 b 3 3 2 2 b h h L 3 3 a 3 3 2 2 h h U A b u U a a a h i h 3 h \ 1 ..o t e s t . UJ 36 and 0.10 ml. of U0-50 percent red cell suspension was the optimum ratio for the interval of 2 minutes. Red cell suspensions of more than 50 percent concentration with this amount of serum were viscous and the mixture did not drain from the cards easily to give well separated agglutination particles. With red cell suspensions of 25 and 30 per­ cent the agglutination was satisfactory but less visible . With red cell suspensions of 15 and 20 percent the agglutination was unsatisfactory. Wllen 0.10 ml. of anti-Rh serum and 0.10 ml. of 60-50 percent of red cell suspension were used, satisfactory results were obtained, but it unnecessarily increased the quantity of antiserum required. Since normal whole blood contains approximately hb percent of red cells , the ratio of one part of anti-Rh serum to two parts of U0-50 per­ cent red cell suspension which was found to be optimum in the quantita­ tive study corresponds to the amounts recommended oy Levinson and Schlutz for routine Rh testing oy the plastic card method. However, it appeared in the present studies tliat tne recommendation of Levinson and Sciilutz and of Allen et al. that a very small volume of olood be used for ABO olood group determination was not only unnecessary but even un­ desirable for the plastic card method for the determination of ABO blood groups. Effect of Albumin Content of Anti-Rh Serum The usefulness of albumin as a diluent in the detection of Rh anti­ gens with incomplete Rh antibody was originally demonstrated oy Diamond and Denton (29). The commercially available incomplete anti-Rh serums 37 in present use are usually diluted with albumin in their preparation. It had been observed that certain anti-Rh serums used for the plastic card method of typing were more viscous and dried more readily than others. When the albumin concentration of five samples of commercially available anti-Rh serums was determined and found to range from 21-30 percent, the question arose as to the effect tliis variation might have on the reactivity of the Rh testing serums used in the method under study. To examine the effect of albumin concentration on reactivity, the titers of individual serums were to be determined after dilution of the serum with albumin of several concentrations. Commercially available anti-Rh serum could not be used because it already contained albumin. Therefore, tliree anti-Rh serums, each of which had been produced by immunization of Rh negative male individuals in the course of experimental production of diagnostic antiserums , were selected for study. The serums ■were .first absorbed free of A BO antibodies using Rh negative red cells. Twofold serial dilutions were then prepared in 10, 15, 20, 25, and 30 percent albumin wiiich had been tested and found suitable for use as a diagnostic reagent. Each serum in each concentration of albumin diluent was tested both Dy the plastic card method and by tube titration pro­ cedures . While the reactivity on the plastic card was maximum when the antiserums were diluted in 30 percent albumin, the antiserum and blood dried quickly at warm temperatures. The reactions were obscured by failure of the fluid to drain readily from the card. The tube titrations were more 36 sensitive when dilutea with 25 than with 30 percent albumin. The re­ activity. of a typical serum in the several concentrations of aloumin ufluent is recorded in Table y. Progressively decreasing reactivity was ocserved with user-asir.g albumin content, the titration endpoint of tne serum shown in 'i'aele y was c ,ly2 wnsn diluted in 30 oercent al. u-iin as contrasted with an endpoint of 2 5o whcr. diluted in 10 percent =1;umin. Tne same relationship existed between sensitivity ana albumin content in each of the a:.r serums stufic i . The commercially prepared ldnLo ) £ .i10i< 0.. r.o -L i l V l i i Or r.i.i ± —i*. SmtUi'i Pereeni Alcuriiin 2 ; ), 30 2-; 1 i i: 20 13 10 L u u u id \ u 30 23 20 h L: 13 *4 10 L : h i i i .i }* u u 11 ii L i 32 u 4i i l\ i\ a u >4 n u u \ It; i a h ii i 4. ■4 a ii /: n L ii a a Degree oi' A gglutination D ilu tion of Serun in albumin 2 ^ 6 4 12 c 312 1 ,0 2 1 2 ,0ir n ,096 . ,1P2 P la s t ic Card 1 1 >f . 2 3 3 + I .1 1 1 3 3 i L 2 t 3 3 2 3 1 a H u i \ L i* 2 1 C i T est tu ai a i , !. i' ii i. <: 3 1: + - - - - - — - 3 i ii 2 2 2 2 1 3 3 3 0_ < c 'J r L 1 X I A. ± Control 1 6 ,3 4 32,76.: - - - - - - - - - /•v c - - - 2 2 1 + 1 - - - - - - - - vo VO hO oup B cells and two parts of hh positive red cells were used with e correct antiserums . The antiserum and rea cell suspension was 30=d togetner for exactly 30 seconds, and the r:r:‘.oc wc s timed until mixture was drained from the card . The maximum time tested was itei oy drying of the cell-serum mixture on the card, ana it was de:e.',nt on temperature and nomiuiby to a considers_le extent. The interval of time tna t elapseu wr.c no*, an important factor with - ;roup A arc -roup d , effect with su~,,t o u u red cells tested. red cells. .c attempt was mace to study Periods of time less chan 2 minuses r-. not adequate for maximum, or optimum, ar gLutiiiation of rd. positive cells oy anti-in. st-roji, -r-rglutinatj on was satisfactory with inter­ im of time of two i.inutes or lon.tr, to the time limited by drying. ..refore, the safest time for ottim-i agglutination was an interval of 3 minutes cetween completion of stirring arc draining the cell-serum xture from the plastic care. T:ds corr_sconced well with the interval time suggested oy Levinson and Sciilutz . u cessary to determine the highest dilution of antiserum which agglutit-wd red cells to the same degree at given temperatures , or the effect tr.nerature on the titers of the antiserums. Commercially prepares anii-:u.0 Ca m i-u) serums oi' the incomplete , "slid': test" , variety were studio-'- . Serial twofold dilutions of six r.r.s were v repared in 2 j p^rcor.t albumin and ti.en tested at It , 25 and C. The laboratory was r.aintaino at t.i--- respective temperatures , and a.j materials used in ti:e procedure were allowed to attain these •• ;,r. natures bef ox*e they w .re usou . Tne tecrutiquc c 1 testing described Je moral Procedures was stricil\ onservcJ fox' oac. test. . ,r tne results of representative titrations at the uiff m l r o are presented in Ta^lc; 1C. 1 .Is tn« leinperatur-. eteretocJ from 3i w . .rantis-iru". .e o -■’... u-^^s r --■0 l r -. .-i.ui*;-.:; ■- ~:*t.a .113>.i an a -*v a.utinr 0 .oi. :t^er tat r .no ^ i + ^ bU at 3: c., tenpera- 1:32 at 23 C . , an, 1:1 at 1 C. j, r-r*. na o n) o bit liar decreases in : e .iv t.* as the t u p uatur . uecronscu Were fount w ‘th tne other cotv*ei ai a:.ti-nl.Q teru.s. I, m s oencluu.1 .tat aevtio ,r 1; 3 for use in It test reus t oe of s .iff i.cien tiy high titer to permit tne maintenance .isfactory r-.'activity ir. the rsr.rp of torr.terat ,ro ti.ai m.ieht oe enj. . . o' .ret . The six rh. anti serums stall::; exiiioiteu complete a gglutinati or. ■iiutionc of 1:2 au.t 1:,. e ve n wher the t er.j ora turn- was maintain-:? i at j . itviu-ii.tuy mh ar.ticcrums wit,', titers teat com'omoo to tne ^.att-one 1 L . .tec of ..emit., sttmarbs w sv 3 caps .-.I., of adequate reactivity and i„ r.ot .j a sourct of error under normal vt riot Lons of te:tp erature . 1 a:.ou tit rt;er,c:n - it 01 01 ;;: eratiC'’t or• - v i n titrr lions, it v: s somewhat countjractea by the a over s.: effect of 10 Ll KiivCTlVri'x Oi- iiiCQi'llLrlT^ I-aii, Sc«iui-*S A T D u vr‘£in.i2lNT Tii-iPiihATUnfiS T e m p e r - __________________Degree of A gglu tin ation __________________ Serum ature D ilu tio n o f Serum In Albumin " I* umoer u lo 2y6 2 611 127 °C. 32 312 1' 2;.) 1 el Ji ii i1 A, ,LV: 10 2o 3; * Ii n 23 33* *i i t4 237 u 1, 1. /J Ji /; li i i. /; 3 3 1 1 2 3 1 + h 3 h i 2 Ji 1, 1 3 U + 1 2 Control _ -f - - - - 1 X - - - _ _ + I 3 - - - - r\ c 1 - - - + _ _ _ - - - - + + + 1 3 2 1 _ Test mixtureo drying. frIV) hi rapid drying or the cell-serum mixture at the higher temperatures. inis readily leads to false interpretations of the results and consti­ tutes a possible source of error. Other factors. examined. The effects of certain other nhvsical factors were It was readily observed that a thoroug.. mixture of anti serum :-.r.' food was necessary if uniform sensitization and agglutination of th r .- cells was to occur. th If was found tnat the suesequent treatment of . s -run—cell roixluro on the plastic card had marked effects on the r .-suits. The agglutination was not apparent uni ;js the excess fluid was r .r.oveu from the carl. VTuon the cloou ana antis: r um, were rotated, or slowly rocked , and allowed to stand on the card without removing ths ■-xv. us, tne r. suits wo r. cor.ylo to uy unsatisfactory. Th-^ observation of r .-s hits outair.eu sy the latter tec uni quo , wnich is usee in slide methods is uopendent on the agglutination being visiole tiirough a transparent meuiur.. Observation of tae a gglutinatior. of red cells on the plastic mr-: de ends ur.or. an aunersnt pattern of agglutination. Sensitivity of the Plastic Card method Compared with Certain Laboratory Procedures dh- detection and quantitative aetermination of xh antimonies in c erums of uersor.s wuo have teen sensitizes by r.h antigens and the taction oi Weekly reacting kh positive ^Du) antigens , either by sin vie eta--- tti'- •- or -,y th-. ar -•!' cation of tne Cos m s anti-vlotul in procedure . m. */ ■* require rol? tively refined techniques as compared wit:, those necessary tu it =mine the more reauily typed ku -loot rrour s . M i l e it was not Wi expected ■that- the plastic card method would be suitable or desirable as 3 standard method for tliese determinations, they afforded an opportunity to study the capabilities and limitations of the plastic card method. lho plastic card method had proved unexpectedly sensitive in quantitative tests involving albumin diluent concentrations , and tliroughout the lanoratory study it had been noted that there was little difference be­ tween the results obtained by the plastic cara method and by the slide method. Since the warm sliae method had been proposed for the typing of large populations, particularly in civil defense preparation, by Elliott and Driffitts (3H) ana by Allen, Diamond, ana madden (36), it was included in the parallel comparison of test tube method , plastic card method, and warm slide method . hh antloody titration. Six anti-hn0 typing serums of the incom- lr to variety' that were prepared oy several commercial raanufacturers 3no 20 serums from pregnant women in which anti-hh anti nodies had oeen .;t-cted were subjected to parallel testing. Twofold serial dilutions of ti.e s aruns were prepare a in 1-ml. or 2-r.l. volumes of 2y percent diagnostic aloumin since this percentage approximated the average albumin content of representative commercial serums . A separate pipette was used for the preparation of each dilution of serum. ib) red cells and pooled group specific serum, a U S Using washed hh0 percent suspension of red cells was preoared for use in the plastic cara and warm slide tests and a 2 percent suspension for use in the tube test. Each series of serum dilutions was tested oy the tuoe method of tne national institutes of Health for the standardization of anti-kh serums. For tuis method 0.1 ml. of each serum dilution and 0.1 ml. of the 2 percent rod cell suspension were combined, thoroughly mixed, incubated in a water oath at 37 C . for oO minutes, and centrifuged for 1 minute at 1000 rpm. Each tube was examine;: over a fiourescent light for gross agglutination, taking care to handle the tube and contents gently. The titer was recorded as the last serum dilution in which 1-plus aggluti­ nation remained for 15 minutes after the first reading. Each dilution of serum was also tested by the plastic card method descrioed under Jeneral Procedures and by the warm slide method described by Griffitts, nlliott, and Cox (lO). The results of the tliroe comparative tests are shown in Tables 11 and 12. The six commercial antiser’ims , which usually represent combined serums, exhioited titers of 1:61 to 1:12£ when tested oy the National institutes of health tube procedure. While little difference was evident Lr. the titers obtained by the plastic card ana warm slide methods, the titers were one or two twofold dilutions lower than those obtained with c.i.. cube procedure. The same relative sensitivity was observed in the titers founc for serums of individuals. Those serums in which a titer of only 1:2 or 1:1 was found with the tube test were negative oy Doth .lastic card and warm slide methods. The finding that antibody determinations on the plastic card were • 11'/ onlv one or two twofold ablutions less than with the tJ positive '' v standard tube test verifier the relatively .ji~h sensitivity of the plastic care method which has oeen indicated elsewhere. That the results with the plastic card method were comparable to those witn the warm slide ii .jvtCi’IVlTi OF A^TJ-Hii SKutl.S AS BaTaHtfLU) or Two FLASTIC O AitD, WAKM SLI^L ALL LULL .laTK G D S D'2;y’o of Sor or; i*TOtj Card S lid <2 L’uoe jf'lutdnation D ilu tio n of Sfaruin in liluuinin .art nod 2 ii oy< L\ i, !. - i 3 I it L. U i.. 1, + _ 1r 32 2 3 3 1 1 •d4 . — - — _ _ _ - - - - + ] Card n ii Li i, 3 3 1 *, >i i . h j ! l+ 1 3 0.1 i<;.,‘ T me a 1,02U . 3 - < jl2 . Card 51 L'<: Tuue f 256 . 1 C 12 a Control . _ - - - - t L A- _ _ + C 1 - + - — hi lil.3LE 12 COrAA-i'iATr/E h ii ANTIBODY T lT n A T IG N A- LUn.S 0 3 6 3LOOJJ CkCOUPii'.G SKhlh-.s ANu 2 0 I r J u I V lJ lA .L SE idA ,S DETERMINED uX T n u P L A o lIC C A itD , WAibi SLIDE &U~> TUBE 4-.ETK0DS Sorurc Plastic Card b--tlo? 29 i 1,1 u: 1, 01 l.tjOZ 0 ua 1 3a * U'. Lo 32 32 32 32 _L • ^ 1i 12 c u ,09' 312 cL I. 32 32 - a L....1 5- 128 va 123 1 .. w Antibody Titer Warn Slide 32 32 I'd 32 32 32 12 _ A,09:. 2> J' ^ > U i i. Lu 2 ,0b. i,02L — - 1-1 'M -U 12 L ou 32 u >12 .,192 >12 Id *■- lr. j>2 2 2 3>■ 121 io >12 _ i-r Ev Tube 2 ,0b': 2 ,0bb 2 - 2 ' 12 ■: --4-i 2>> 2;.c. 32 >12 12 c ,09c -,0,b .. U8 method also showed "that, "the test was capable of a high degree of sensi­ tivity. Although the warm slide method has not been recommended as a substitute for the test tube teclmique for antibody titration, it has been considered reasonably accurate , particularly for the preliminary estimation of titer, and is purported to have the advantage of eliminat­ ing zone phenomena in high-titered serums (Uo). Detection of weakly reacting Rh positive (Du) antigen. Tne plastic card method is designed for use of fresh capillary olood, but since it •.ab been possible to demonsti'ate good reactions with bloods submitted for laboratory testing, even with diluted antiserurrs , an attempt was made to compare a series of bloods designated as D variants by the blood (rouping laboratory. The first series of 22 bloods were of the low grade Du group. They hac been found negative by two different anti-Rh0lanti-D) serums, and they had been tested by the anti-human glooulin procedure when they were found to be positive with an anti-Rh0 1(anti-CD) serum. consisted of free cells in its own serum. Each blood sample As is customary for warm slide testing in the absence of oxalatecl blood or fresh capillary blood, tne sample was centrifuged ana a portion of serum was removed to prepare a nO-^O percent suspension of the red cells. Each blood sample was testeo both by the plastic card and by tne warm slide methods, neither procedure was effective in detecting any reaction with the most potent commercial anti-Rh0 serum available, as may be seen in Table 13. The same anti-Ri:0 serum had yielded negative results in the one-stage tube test. TARLE 1} i'bj(D) TYFldG 0/ VJEAKLY REACTING ku POSITIVE JLOODS ■r of routine Tube Sneciuer.s ^ n t i____________ id.j^D) 22 30 - V*7:-: + Plastic Card ’'/anu Slide A n tiA nti P o s itiv e negative InconP o sitiv e negative incona.ij'^CD) g lo o u lin _______________________ e lu siv e ___________________________e lu siv e + + 0 22 0 0 22 0 + + 3u y 7 36 7 7 cVO 50 A second series of 50 bloods was of the weakly reacting Rh0 group. One or both of the tests with two different anti-Rh0 serums had been doubtfully positive, or weakly reacting, and the oloods had been con­ firmed as rtho(D) positive bv the Coombs anti-globulin procedure . They wire tested Dy both the plastic card and the warm slide methods. Of the 50 oloods tested by the plastic card method, 31 were Rh3CD) positive, v were R'n0CD) negative, and 7 were inconclusive . Wien these same bloods were tested by the warm slide method, 3b were Uh0(D) positive, 7 were ro.^D) negative, ana 7 wore inconclusive. Again the results Obtained with the plsstic card and warm slide methods were similar, but only 70 percent of the weakly reacting nh0(D) positive bloods were detected by those methods . Coomos anti-glooulln test. The coomos anti-globulin tests performed luring the field study describee, were mace with an anti-human serum rea -*ont which was standardize a onl;, for tube testing. A variety of the reagent standardized for use eitner in test tuoe or slide tests has since ..ecoine available. j s s of this reagBnt confirmed the opinion of some workers that the detection of the antioody coated on the red cell out not nrotucing agglutination, as is the case with the D''1 variants of Ah posi­ tive oloods, was setter determined on a slide or tile t'nan in the tuoe. A comparison was made in the laboratory of the warm slide and plastic cars methods for the detection or confirmation of the presence of B a _iooc tyres using tnis new anti-globulin reagent. invariants of tne "low grade" group us;n Ten bloods which were in the previous experiment were 51 tested. Duplicate preparations of O.h ml. of 2 percent suspension of the red cells in saline were sensitized with an equal volume of in­ complete anti-Rh0(anti-D) serum by incubation in test tubes in a water oath at 37 C. for bO minutes, Red cell controls ox' each ulood were included in which. 20 percent albumin was used instead of the anti-Rh0 serum. The cells were washed tliree times with 0.9 percent sodium chloride solution. After the last centrifugation the saline was removed as com­ pletely as possible to leave a 140-50 percent suspension of red cells. Insteaa of placing the drop of reagent serum separately on the slide as ilr^cted by tne manufacturer, 0.05 ml. of tiie anti-human globulin serum was adaed to the sedimented cells in each of tne duplicate test tubes a n a to the control tube. After thorough mixing, the duplicate mixtures were nlaced on the plastic card and on the slide. The mixture on the s'ii-ee was warms a on a viewing oox and roekea rently for 2 minutes wiiils that on the plascic card was allowed to remain undisturoed for 2-2 1/2 ..inutes ana the excess fluid was drained from the card, as were other plastic card tests. The reactions with the two methods were not significantly different. It was found tiiat adherence to the optimum heavy red cell suspension of LO-50 percent of red cells was more necessary ’'or satisfactory results with the plastic card method than with the warm slide method. When tnis was maintained, the results ootainea by the two rothods were the same (Ta: le In) . . 1 ■Li-j L < Cu.r>.nJoU» t v r.I'C 1 rl.OLdJLi i. 1 ,-b l o , -i l l v - S l i C u [ ; f . ' C l ; r r . 11 Ii coninlntt. ..icxor j.fill-in .31 3-> 'Ad 323 A2 >;w 1 ,030 i , 21- -1,31.'.* 1 ,;.0A i o r i l i v e C ontrol ..'j a l i v o C o n t r o l )(lj) AInslie Card Si *. i i ;i(/^o S lia 3 1; 52 DISCUSSION The attempt to evaluate a method for determining blood groups of the A d G and hho specificities leads one to a consideration of the results in the light of the objective. For the blood typing of large numbers of people for the formation of a living olood bank and for the creation of a centralized file of blood types which would be almost immediately avail a d e in case of disaster in any area, the object is determination of the most clinically significant olood groups with the highest degree of accuracy compatible with present knowledge and the reasonable accomplish­ ment of that objective. It is generally recognised triat laboratory methods of slood typing from venous specimens, which may be resorteo to repeatedly for recheckeng, permit a high degree of accuracy . The methods generally preferred arr test tuue tecimiques. Tiie objections to test tube procedures for mass .Hood typing are usually cited as expense, time consumption, require mont of !„or:; apparatus and equipment, am: the danger of clerical errors. In contrast, the slide test metnods are less expensive and the results may oe determined immediately, but great objection to the slice method .as arisen as a result of inaccuracies that occurred in olood typing slic>; metnod in World War II. In the present s tudy by the determination of AoO and iih0(B) blood -tro u t ) 3 by t}ie plastic csra ms thoc , a : odifice tion of the slide technique , exhibited a. technical error of 0.1' percent and a total error of 0.U9 53 percent, when inconclusive results and clerical errors were included, as compared with results found by a laboratory test tube method. Tne laboratory method involved the additional use of agglutinin determina­ tions , or reverse typings, in the ABO system and the use of multiple antiserums and Coombs anti-globulin testing in the Ith system. With the test tuoe methods tiiere was one known error which was due to incorrect i .•■.ratification of one ABO olood group in the 9 ,653 olood samples , an error of 0.01 percent. These results appear to be the reverse of those resorted oy j-llen, Diamond, and ..adder (36) in a stuay of the warm slide method compared with tne tuoe test methods for olood typing using a s;r.all scries of floods. It is evident that the plastic card method rooorted herein would nave appeared even more accurate had. it ceen com­ pare:. with a less comprehensive method of tuoe testing. The accuracy of tnn plastic caru method for ABO grouping, even with out recourse to reverse typing, was well illustrated in this series. T ;:s one error in AnO grouping was due to an inadequate plastic card test as was reauily d-naonstranle by reoetition of the test. Such an error ir. any method can only be avoided by critical ana painstaking attention to detail. The permanence of the plastic card determination readily re­ vealed the nature of the error. It is possible that factors contriouting to the slide testing in 1..U el cod *grouping early in World War II were multiple. The present higi: stanbarus of potency for olood grouping serums were first proposed oy tne hational Institutes of health in 19U6. The commercial serums now available must possess an adequate minimum of potency against the 5U sabgroups of A as well as against the A and B antigens. The subgroups of A, which are the- most difficult to detect of the ABO group, could ..ave oeen readily overlooked in the olood typing as it was performec .'.urine the war with no definite time requirements for the test. The presence of the more critical rih determination in the series of tests imposes a time requirement on the testing wiiich is advantageous. An advantage of tne test tuoe method in any serological testing lies in the fact that results can oe re-examined over a longer period of time tv .u-ii is us-tally possible v;hon slice techniques are used, and, t..-.rufor-: , ..ore adequate control of clerical error can he incorporatou. •;.is respect, since the result on tne plastic card is not further altered by oryi.ng or a •;!r.g, the plastic car a method possesses remarkable essicilities for tn-.; control of clerical error. l:t • ..Is error is crotacly urdinown. In overact- slide test— In t.xis study an error of 0.03 •r? :.t occurro :• where throe incorrect recordings were ■r-.3 tits t.ut w-cr-.- evident on tne car.'.. ..." ; „ I*. I. - ;C O r of ti.e 1his iii.*stra tes tne value of ^ 1 jT cx Iil> critical review of ti.e technical a rue clerical ac •curacy oefore an immedi­ ately recorded result ue used as a permanent donor card or je given to t;.-_ donor upon completion of ths test. It is known that any one Ah test may fail to identify some of th-_ v •'if-- r v M rr: D ^ variants of the rfr. positive group. In tiiis series nine rIcons, or C.l percent, were considered to os of this group. must It _ at r.rociated that tne- number of '..foods placet in this category .e sans:.tivit.- art accuracy of tne tests used 55 and on the alertness and exj erience of persons ;erforrring the tests. In addition, it is necessary to determine that the anti serums used do not contain traces of antibodies other than the requir?d ar.ti-hh0(D) j inc - antisT-runs proJ ac-.e a,, irnr.unira tior. of Lunar, suojects, as are rurrer.t use, man- contain traces of infrequently occurring :n. Individuals caps ole of produc­ es anair.st r.any ..loo., eroun antirens, so.. of which are undcuhtouly still unknown as can he surr.ised by the 11 ci:araot•_ri2*»d ir.eer.-o-r.a^rt s y s tons (hi) and !.. -r-.-cjir.r list of a.r.trocs others of ran--. occurrence. It follows that the Co ones anti- -le _ ulir. test coul- or. occasion c a u s t h e a oriutir.at ion of cells s-iti lisel a;, an unk novae anti tody sine*-- it t s: .oific o n l y for hirean tio.-alir.. In ad ii tier, to the routine us -- of rultipl Io r.ot yrour sye h.fic .rut is to ..e no ;r.tive, iC wi-ich w.r-j s irens on all oloods positive were tested wit:, the ar.ti-;;Io._td.ir. test since the wear: If" naiin-n is frequently associate ^ v:itn t.e r h 1 ; ar.ti/en. In an_.ti.er 3 ..ro -^s of 12,. 1. con- c-cutrve ..loot t-sts the author found 2 ..orcent of those- which would have of .rtf e . . t f. considers - rh*(f) '.-loots to f. hh0(f~) positive t-y the Coo: .ts anei--lo.uI.in test, .loo .s ori.*inail„. ty. 1;) found (Cl; . as rh * (C) 00 wh0 , ana nosenf ield et a l . .7 rere-out of sIcons originally typed as rh* (C) to be oloO'.S. 'wentv-two s: r i f f n (u2) found 32 percent of a series of , w r linens, or 0.22 percent of the oloods tested in thiss eortonol v :ly studio • with nultirlc s 5 runs and anti-^looulin 56 tests to prove or confirm their classification at Hh0(I>) positive. All tnese oloods might have been considered low grade Du , or even ith0 nega­ tive , had they been tested by less sensi tive methods and the results would have thus compared even more closely with those of the card test. It is interesting that of the bloods studied because of discrepant results or doubtful elastic card tests ana classified uy laooratory t sting as woakl.v reacting iui0^Du ) positive oloods, eight were rh*(C) positive. hr: adaitional tirrec uloocs were hh0Ci>) negative, rh* (C) posi•.iv-. . ..one was rh* *(l) positive . It is perhaps more significant that -.a jo of three bloods wnich were originally tp-e:.. by the card method as positive :ut considered negative on repeat determination had one countful.Ly positive test among the cattery of tube tests used. i.one of s-• oloods was positive with the Coombs anti-glo.,.ulin tea a following a t i.-o tea sensiti '/ation of the colls with five different serums . They worj therefore considered Kh0^d) negative. It is entirely possible that I:....- uso of auditional anti-human globulin serums night nave altered tne cl css if:’ca bioi: of these tests. •races are apparent in w d i yen Lo -hen l o ) It has sii.ee b =en ocs-u'vea that uiffcr- s ba ndaraized anti-human globulin s e m i s , recenll;. demonstrated bn?t a definite prozone phenomenon is t.-xl.ici t