. . . _ . ..J... .39.. ....I........... ...}...
3..........fiu........................._......_.........u... ,..m....... c. .... ... ..

       

           

 

 

  

     

  

  

     

 

 

. . I ..-
......o. ... «.... cu... ,PI. ....
.I ...w‘: ....
. . . 3n”......:..onb
.. .. . .
01.... . ... . . ...?
r .. . [.411’... . ,. .a...‘4fi$3. ...:
.....Cn- 0.. .12. . ... O . .2. s“... . .20 . .4.
o. .. . . . ... D. 9... . . ...: 7.21.. 1...... 0.. ... .... .... .
- .0 ..I. . a: .1. . . ~ 1.8....‘099‘ . .
.. .. ..1. .. .
. ob: ..., ..
. . , ‘o . . .. . .3
. ...-E. . .0 a .

               

 

     

   

         

     

 

      

       

 

     

     

    
       

          
  

 
   

   
     

       
      

 

    

   

 

o 0.... .

 

.
... ... ..
.. . .. .

 

 

. ... . 9... . . ....-.o‘. .
...-0.: .0 . 9",: IV... ...‘...0 .... Q. ....Q. ..
fl .. p.330... .lo..v‘o..l I.vh..v.au..p. . co. '4
.. . . ,3 3.9.. .. 6...... . u . .. ...-0‘01 .... .I... .0... .n .
.. . a...:.. I... . . O. _ \1’9’. In ..“a... ...”N .U.’.Q“.~ O
, . . '9‘... 2...... c A . . . ...: 2 13.9... .....5? ... ..3 .432. ..a.. (J:
... ..«Jf... ... our... ... ...... . ... ...... ... 3.... o .... .9. ...... . ... . ... ... O. 2r.x~n.ls, . I ... m...
. .... o . .vc . I . .... . ...... .o . . q. . . .. (4...... . . . . . on”.
... K:— .... ... ....1. a. . .0"... . . . . 3.0.... I. .o in . _ 0 ..J. .. I. .... ...uhq'; . 0‘: n
... . o .. . .. .. . s . .15.. . 3.0.... . 1.... . . . . 1...: .... .... al. ... . ..
Tu.“ . . OI. .... .. .... .«o 3.... ... ..ou: :- 1.0 i=1"... ova to u.
. . .33.. . ...... .CII..P.-Ou. o..- . Vo‘lan; .. .0 ...... ’ofn?‘r1..o ‘n.‘-II ... I a
...Q .11 ....-. . .. no. a '1’... ...... 1.3.? ...} ..I 0... .0.’..
. h . o. 2... . a... ... 3...... ...; a .
.. . . u .. . .
. - ... ,

  
   
  

  

        

 

 

 

  
         

 

 

 

 

 

                          

  

 

   

 

 

 

.....‘Ilto‘ l I
.... .v.
... u .0. ...... .5:

. 1 9....-. .

  

    

A5.3.... ...
.. .. .—. ... .hoo..la.1.$ .0.
.....L. I (.... .. ..
...»... .......:..-.. ......5 .
. q l 01 D
aarmv . .. 1 ..
0.1 0...... .. . 0". at . ......L. .
. . . .. _. . .
_ . I o. . .. ...... . . . . . . .7... ......th tunings. .. U...
.N..... I... 3 ... . . .u..l...o...l...t 4.
7 .... . 2.. . . .4. .v. . n. G.- ..L ... . . on. no. . .... 5‘}... n. J
.. . .. .c . . .. c 70...... . A. 0.. a 9...... v‘ .u‘ . . 20., .... In ....
.. . - o . 2’. ..I. L 5 ‘.J..Q... . 10... ...-I... 2'..." .
-..}...- . . . . ...... T I.» .. .. .93...‘ 0‘ .—.:. ...-on .I‘.
. . ...! . . . . .93.... Jot! a. ... ...o.
0......; .....I . ..
.. c .0. 9‘. .33. c

. iv... JAiL..£A

 

   

3 o‘.a
3..." I .

   

«wwwié

 
 

        
     

   
     
 

 

 

.02.... .. i. ..zcofi’l.
.. .Wo‘adnflhf: J...‘ i .
.v o 5.4”“. ...”. V"
Then“.

 

 

     

.u.......... .
... I... .1...

 
       

 

....
.... .. .

 

0.: ...

 

    

   

              
 

 

 

    

  

  

 

   

      

     

        

  

       

             

 

 

  

 

            

    

 

.
Q
”I vs. .... .... .4 .
u l v. - t rite-0””: k
. . _ . . .. '0‘. \I: 9'
.....u.....n.!.8. . . ...‘ 2...”: o. ...? ... ”a. .0“ . ... ....Z.l.f8&u$..na 930.....9..~. ..p 0.x“...
.. ....- ..ol...‘ .3... ...... A . ....3... .’~.......... ...n... . .W... ..Q. i... . ...-..."...o. . .... n he“:
.n v . . . ...4}. .¢ ... .. g. n
... ... .... ....rhnbauin.‘ . .. ...“...hve .u . ...}...A». . .. .
9.. .... .... .. . ...di' . . . . . .. . c .. . .
.. .
.. .0 . J09 10...”: .01“. H. I. ...... ...! 3...... . .c . . . . . .
l y . . . .. - . . . .. . . .
"H". O a. “H..Ju-‘qutflh.ountnn.fio.tn1‘ a... .u c :1. o. I.- I‘... 0'0. (G 0. ...I. ...... .Ilh.
. ...‘v' .

 

 

' . I O, I. .‘Q
.0. c. 9..

  

 
     

 

a. . .
....o 0.. in. ....o-z...
...‘u Y. . u a.

.. . I.
.0 . o. I.....
. .

   

     

 

O
.3 .
....

      

    
    

            
      

   

....-...

 

   

 

   

   

 

 

 

     

         

 

 

 

   

 

 

 
   

 

      

     

  

 

 

       

 

 

      

 

 

      

 

        
 

            

0001,10
3’... 3Q. ... 3‘,
...: 43.4. . .4}....Al .... u‘ ... . ..
.5. ...... ...? .. in”... .. . ...... . ...-”aunt”... . ..
.. n .a .p.. {.11.}. . J’a’ ugh- .r..’ s .U
.. c. (a... I f... . _ up”. .«u;b.$«h.swi..“’vgfizuu.«o ...
. v. C... 3..,o o: .. . .o I
. ... ...-.1. . 3 v. . .... ... ......a... . ...... .....lozif. ....luc. .33....
.....32... .00 “...-23”.)... \l‘. .... ..Jlr.
. -5............. . 5.333.... .... .....ue. . ...":
... . .o I. .... I9... .2. ......o. . 0...... ... 5.0%.”.Hrhw
1. ...: . at... J 1.3. . . 1.9.7.11- 05"... .I‘. ... .1 ......K?‘.
.7 . o 3.. . o . ....c 5... . .. , ......1...J‘..'Ie..¢n.: . .‘I... (-
"u.. $7.... 11...... a . .. . . . 123...}: ‘c'5‘550c‘? ‘20...) ‘0
t 2.. . . . -.....5. .... . I . . . . .
. . . . c . ... I .... O .) n .. . ......
“t. . . .. l....-..n. 1..."... 1.03:... ‘30... ....s 0...... .1...“
...1 . . . . ....395.‘ ... 3.1V . .0..!...... . 0!.- JIJ...
V . . . . . . . . 1f‘t".flnaoolc . v .... .
I . 3. .... .I... a... 0.. .. ... .... .. 3?... .0... ... L30.u.0b...0 . .
2. .n .....I

 

        

 

  

. .3.o,....ct
. . . . v .
on}; 0... :1...‘ :v .9.‘ . '1. ... ...-6‘“ .I. h.
. ..
.0...o.’§. ’33..... I" In. 3 .
I .....§.. ..., ...: v ...... 602.931. .
.I o Co v. ...." . Q o 50......2 no

.. v. . ’50....0...
. I .0 .

. . .... ’

. ou’u . ’

, .00. .2 .u ..3.....i!c...o.3l .. . ... 3 .91.... .3." .
...! ......”‘m; .0 ‘I ..‘.o - .....va a I. (lull

_ . . 3.3.9.1.. 3 ... .... .... ...: I...

. ...!I... ... .. 0 . .

                    

 

  

  

      

   

     

 

 
 

  
 

   

. .'-I.¢’,‘ .I.

         

  
    

 

. ..

 

 
  
      

      

     

. . .. . . . , , . .wu“a1.ufl...)?..l . ..l.\..m...«v. .....mfl.
. o . v n a . . .
.TAII IO. 0.. .

...... 3.. ....
. . ...
L. . L}...

 

                

4fi9>i.o..~...,15 ......3‘ .9. .
OfI‘vaz‘. .4... '1. I‘ ..D.»F
1 . . MK“) 3.... .‘|:.-O¥o‘.u. .. .... ’no... 3
ca.1\. ’5‘. o’I on O. ’0. .‘Oav .nrl
. ...-o. .

. o o u! a. .-
. V ...-...a-ohu . . o. 40.. .10.)...3l
. ‘ ... ....

.9 I .
....

   

. o . . c’ .... _ n .. . .
. .\ .....f.....l.....:...‘.. n . . 3 533...... . 9'33.”
.... .. . .m... 90.1.)...”- . (.....i ...J 553.3“
..o..8.‘\8} . . ...? .. .‘. “
..c. 71‘! .X’....uu 30.3“ .... t . 30.13... . ”Pa

 

 

..o.
\l..... .30
.... ..Io. ....0‘..£...o.

0.0... l . ..

          

 

 

     

 

 

 

 

 

     

      

          

   

  
       
       

        

                    

 

 

1.: 3:33....) 0‘ .
V n I ._’h 0.. '0

 
 

... i...’ ...!
wash}... "..."...
......Q... 0. ... .. . . ..v
. . I ...._......... “v.35... ,. .
n. . . ... . . . . . 1.533.... .. . . ...I..r.l..llo.. ...! I
.02.... ...... .... . ... . .o ....I-.. . .....0... . . , ... , . 1 . .. . . ...... .... \r...’ii.t:. (I
. v.31!!! .....l. .. ... o . 0...... . . . . ‘ .- ... ......mfos a.
n . . . .0 .... IO. . in . 1... o. . ... . . 4 o . .9... v. 5.. 5.‘0 ..
. r5... .. .. ......» s. .....uufiflufi . .i?..«....:. I: ....
. ..a u... ... ... ... 0.. . 4.... o . . ulna. J .
i. J. 0" . 11......0. 7. i ......loa.d.~..\t
v . . ....9¢.nbv..v..c..... ..
1. . .. I". . v. 1...
v). . . I... . ... . . - ...
o .0...’ .0 .leOv...... 0.. .. «.... .. . - . 9.“...
v ., ... . o' . - . . ...... ... 1.11 ... . .. .. . . 5.. . ...... . ‘5.
I . . .10 o. L... .8 . Jr. . . .. .0. ..ll‘......

. no!) .0 .2. ..5 ... ...
”8‘5 0". ‘33...
0

 

 

._.u..u......._.........«. an»...g..,}.......... ......
AD. I. . . 52s ’ h I .
I ....c 3.....‘1’5’.‘~.§.‘A . 2 3m . _ 0.... 1 ‘

          

 

   
 

  

 

   

        
   

 

               

‘...a....0.
o.).-... 919.1,“ .. ‘.. 5v. .1.
. 5.?‘1’...£¢.|b. ...Q. Ill...l.‘b
.02... . t. . . . . . .i .:..‘.....li!..!o.uo.. .035 a!
......o! .... ... a)... . ..X .9? . .u'h.u.m..... . . .‘vd‘...§tt.!.r33..i. l

3.4.... .I... .’.. . . ...... ....) .. .. . . . . .

’ ’3... I... ... . .

.....n.,.lk. 0.. .5.
v 03.. ... . ... ... o ..5. o . ..

 

.\.~. ‘15....) .‘ ‘31: .O‘
I ll. 9...... ... .

 

. ..
. .4. ...o”.u)..’. ...-... .... O.
..X..p..... ....‘o. .06.... ... .... 3.."
....‘Av . ...:.v." .1... . tit-o... 0.5... ..
..., .u . I Y“. o 11.. '8‘: ...»...olt. . ..luwh
.3001! o. I... 0.1 .. ...
0‘ ... .0 w... .n J’... a. ‘3'
. ... .... .... t

a...

     

 

 

 

                   

 
 

  

I .
0.9:.- I ..l-,

  
             

v .
.H.3I..m .. ..n. .

1...... 3- I... .
(Stag-5:... .....Olauzll... n.
......K ......laacur

        

11?;7? - - t7}.

   

 

    

-‘~ c". oIr
C ;‘ I.
.....Q. .1. . . via-‘r'dnur.

  

              

     

. A ls. . . .. . . ......l. .1 .3133... {...i.‘ ..
. 3-...) ..3 ....‘on... .... ... ..‘v .vlvo. ’ J.v.b.‘ I o 5 . . . c
5......0. Q. ....1. ....a ‘1 . . . . . I. . . . u ‘l. .. to
‘u..l. .._ u ..v’tn..-....u~. .‘ 0."... . . .v. .... .MH..Htl.-... v0. ‘3 ‘04 .9

......tI. ......oa. . ... . .- .(

 

 

 

  

.
...... .9

  

 

    

a .‘50. ... ... Q. . . .... .I . Q... .
31. .. 2 3!. 9.1.1.1.»... .
I. . O.

 

          

            

 

  

v.
-..-...... .o '3'.

       

  

. o .0...
. ‘ ...-9......
. ... .00....
..o’. . o.

 

. .I. . (o
Ki...

 

 

‘r'fi

    

    

 

. [.00 Q...) .0.
...-........ ...... ..

   

           
           

 

.. . . t" I I .lu‘t
...! .... a .9 _. ......a..‘oo.§...u....

5...... 2.3.5.; . .....J.....f.. . ....
o. .. . 2. 1.. ...‘.Q."Il.,,
... ......l....l.- f. ...o on
. ...! o. . ‘xli ... u '3 I...‘
ug‘.’ o...0......u .3ov .. .. . I\ \‘10.

     

   

 

 

         

 

.0
.3". .... o...

 

... I .... .... ..g... .11.-
. . . .I?’.. n‘.
. ...-15..
. v

 

 

. .... .’l.. .
.01.... .....Jinoa. .
. . .O ....4 I... 0310. o
3’ . .

  

.... ......n.
0....v-na— . .0..- .-

          

 

 

 

 

.. .
. .' v’ . 33.44..
. ...‘i.......§....c~ ......

     

 

    

 

. ‘. ..Q ....5...

 

I .271. i..-...... .....\ .... .
.QO. v! 0.... .I 13,-. t 1.. . . ‘u ‘o‘... ...... (cs..‘uo
... .a .... ‘9‘: I... .

 

 

    

     

 

- .. .
.0... ' Z .
.Q. 9...... .99....n...w. .

      

 

. . ...,v‘...\ ..
,0 . v.
.7

   

 

......
a... 41......-
. Q. .. ., u .

 

               
      

.9.-

    

. A...tv..0 ..
.‘o. to...- . p03; .... .
0....

 

 

.

 

9... .... ...: . .... . .
.... .15....o..... .
.. -

...u.

 

. . v . .

... ......ntl.
..‘Ouo-....¢t 1.... .. .
. O. . . .o ...

 

                      

.....3... ..
f...) .ln. .

 

 

                    
        
 

   

 

 
      

  

  

    

    

  

  
     

 

 

     

   
 

     

    

   
  

    

  

  
 
  

, . .
. ......-..o_..
I. . . . . . . .. on, .I.... .. ...o . .
. I . . ... . . . . ...... . _ . v. .. ...... ..Qr‘.
mi. . .. . I”... ... .. I . .u u o .....u. . o . . . . . . . M . . $4.... :.?.)..0
'. I. . .. . ... L .0 {I ..q o . . - . . . ... ..
l . u h . . . . o. I . .0 . p I . . . . . .
.. . c v .. o . ....
~ . . . I ... .
.. I o . ... O. . . 0
u . . . . . .. . .... . . .. .. .
Q. ... . . .. . . .. . II: . . . .... .12.. .
9. . . , . o v . , I. .... . . . . . ..
l... . ... .. . o . I . . . . . . . . . . . . . .
u . . .n r . ... . o O . ... . . .
.... . ...4 .... o . .
9 . .o ..HH...........“ . . $050.". ...¢ 0
.. .... . . . . . . . . . , . . . ... 11......5... .0 .
” ... .. .I.. . . ., . . . . . . . . ...... \:.II.
. . o v . . . . .. . . . . . .. . . ....
"u . . . . . . . . .. . . . . . . . . , . 1.
.l. . . . . . . . . .. .. . . . . . . . .. . .
O. . . . . . . . . .. Q...O
. . . . ... . .. . .. . . .. . . . a 1 . . t
"C . .c . . . . .. . . .... .... . u l. . . . O t ‘ ... . . .. .. . . .o . . . h... 14-...”
‘ a . . . . - . . . . . 3......t .09 o I . . . ~ o .- u n I: -v.. A . .... .... , . . I ._ .
. . . . . ..... . . .. . .v.. . u .... .. .... Q. ... . c ...-5...}... x. _ v I.
u . 22".. .. v . .... . .. .. .. . .... _ V. .. .. __ . .. .. . .. . .. .. ... ......1 . . 5... .I.
w I . ...-.. 0.1 .8-.. ... .. ... I . ... ...: . . s . . .V. . v. .. . o. .. u! . - .0. . .. . . .... g . o .. ...-...
.o .. .4 . . .... .. n .. 1.. o ....c .. uI. b.“ .) . ... .. . . . . \o .. .~ . ... . . v. .7. . .. .. .....a...l\0. : .. ‘Iroi..)
”a . .. .. ... . . I . u .o I . ... . . . . ... . u .. .. ... . -. ... . .‘II .0 O ..,. \C
. . c. . . . . ... O. o .v ..v . . . ... I... .. . I .... ... .. . . . .. .. . "1"..-
a. o . ... I . 1... ..II 0 I... .... .. .. . . . 3 . .. . . . ...t. .
.. . . . . .. . . 1'...
. . . . . . O . . 6 .u o. r. O
. . . . I. -. . I. .
. I.“ . . . a. . t . . . . o. . o u .. . .o n c I... . .. . ... 4.! ..v.
.. .4 . A ... a . . I . . I t . w .. O c. .- I.) . . p I . fi. . no... .. . s. .. 1“. .L.O..c. I. . . u .
.. . u. u . . . .. . . .0. . . L . . . . o . I . 0 . . . .. . .. .I .. . . o .0 .. . I ... . v .. . . ..I. ......l . ... ...! .0... .I . ... .
. . _. .. ... . . l ... c s . .. I . V n . 4 . n . . O _. .I I I ... .. ......T“ 0.6... .- .., DA ... ... 1...".5.‘ ....
u . . .... .... .. 1. .- . v . Q . a . ... I. .A O . . 1.. . o .. I .. . . .... .1 . d. ......T .. ~. .. . ..-. Q . . «I ... .5 .\¢/. u..\.lr' ‘33....“ 'I
. I. .3 . . .. .. v . .0. . . ...... ..X. . . .. . . . o ._ .. . .... ._ . . 9. I I o . ...) .. .. . .. .. . - .... .. . d..&£. V..hh)..'l.1 .V
.. O. . . . . . .. ... o . . . . . V... .- . ... . . ........ .... .. ... -. . s. . ....... v........ . O 050...... v.
. . . ... I . . . o . . n. . .. . L . . .. I .. . _ . . .. .. . . .. . . ... ,. ... .. ... . 4. . ... s . . ..!.\. P... . . 5. \w.
. ...: O. . . .I .. . I. .. . . -.. o. . . . ... . . . . ... I .. .... . c. .. .. A .. .. ... I. to... -. .
. . .. . . ... ._ .. _ n' .. .. o . I ... .. . I. . .O I C. . .. Q. . .. J x ..I.. .. n.
- . .. _ . .. . ... . . . . . .. . . a. u n . . _ . .w . p ._ . . . .0 O. I.. .. . I _ .. . . . n — . I01. n . .. . .. a I
N . . V. .. ...... . . . . . .. . . .a o . . II o . . . . .... c I o . . . .. .. v. . A . . . .c c Q. .s ... a .. I \
. . pi . . I .
a. . ... .. . . . I . .c . I .. I .... . .. . . . . . 4 o . . .
Q . o o. . :I . o q n. o 4...? . . o .
'u I O . . . . . . . u . on. . . I 0";
I ' n o I
v”. .. . . . . . -
‘ . . . . Q. J L .I ‘ . n
I: . . .. . .. 3...»: ... . .
0‘ . I... ha. \|.O .. m . r . n . . "a. u... .9: .70....
S. . . . . .... . . ..4 ..«t I . I. ...? h w .... £2“... .03. n a calf .» ..mm: 2.0.33.3??? .. ..|.. ......M. .1 . .. ......a ...: a - .3,
... I I. . I . . . ... \ o. . .1 c}... ..t . .. a. 4 ... . ....3'n‘o 0.. t... .. ...: ’2... .. ..J 0.. I Q... 7. ...: .» I. I I . .§ .c- . I. . .
o . ..I c... . ...... < . p . r v... .0 a. i. o-‘. .‘I.’ .0. oil. .5 .. . .I \l ...-D. X ... ... J: .... I... I .. I . .
m... . . .. 5 D. :3... cl mg "3.011;...“ ”—9..“ .u..u.9“.“.‘ . ..m. a. 1...... .....I .- :5... ...w. .... .... c . I!’\~... . . q. .9... I... . .. bl. .
.I. ..l .. , . . . . . . . . ...!2:.»..c.. ... no .. '1. .
umbaw..u.o 0.. . . . Sfio ”...“.n . ......‘t. .O. ..N‘HD‘OZMW.) ¢"ux.".’fi 6 1.....01... inc-.... 1. .... .o 3.... ...}... ...... .09}; . .. 1...“. 7.1”...” 5.... ... .. ..
. I . . o . . | \ . . no .. o... ’1 .. 9. .... ..l V. .. .. inc... 0. . ..v o v. . .. . , . .. . . I . . .
v I.-. ...! ‘5 I .. I..As.| DIN! ‘6‘. V. ..T;. .2" O‘Th _. 3...“. . 0‘. ... .... . ‘cuu- ...... II. 1.. go... to". .. . .... .. ... ......z .... _ . . .. . . .... d o . .
...h. .4...q _ d . .....v‘ . .... I. b .... 7...!- . ......oo? . ...I o o. I II . . . ..
. 30.4.2 .... .. n. o. . I .. A .. .. .
1 . . .

 

 

 

TEA. LIBRARY
3 Michigan State
2 at, University

This is to certify that the
thesis entitled

A RADIOGRAPHIC ASSESSMENT OF PEDIATRIC
FRACTURE HEALING AND TIME SINCE INJURY

presented by

CHRISTINA A. MALONE

has been accepted towards fulfillment
of the requirements for the

MS. degree in FORENSIC SCIENCE

 

 

7/Q WMW

:/ /Major Professor’ 8 Signature
//' 42 0 m/

Date

MSU is an Affirmative Action/Equal Opportunity Employer

 

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

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5108 KzlProj/Aoc8PrelelRC/DateDuerindd

 

A RADIOGRAPHIC ASSESSMENT OF PEDIATRIC F RACTURE HEALING AND
TIME SINCE INJURY

By

Christina A. Malone

A THESIS

Submitted to
Michigan State University
in partial fitlfillment of the requirements
for the degree of

MASTER OF SCIENCE
Forensic Science

2008

ABSTRACT

A RADIOGRAPHIC ASSESSMENT OF PEDIATRIC FRACTURE HEALING AND
TIME SINCE INJURY

By

Christina A. Malone

Although the physiological process of fracture healing has been well studied,
there is little information available on the radiographic assessment of the rates of
pediatric fracture healing. The aims of this thesis are to extrapolate data on fracture
healing from radiographs, evaluate rates of fracture repair in young children, and
compare how these rates vary with the age of the individual and the skeletal element
involved. The schedule presented may enable forensic anthropologists or radiologists to
supply information on the timing of injuries to radiographed remains that exhibit varying
degrees of healing.

This study examines 294 radiogaphs of tibial and radial fractures in 107 infants
and young children. A series of stages were used to describe and measure the typical
bone repair process for these individuals. The sample was segmented into age groups (0-1
years, 2-3 years, and 4-5 years) and fracture location (forearm and leg). The variation in
fracture healing rates among age groups was examined using AN OVA with Bonferroni
correction. The variation in fracture healing between the tibia and radius was determined
using Welch’s t-test. The results determined that age affected the healing rate at the
beginning and end of the healing process, with younger individuals healing at a faster
rate. Fracture location influenced the healing rate at the beginning to middle of the

healing process, with forearm fi'actures healing faster than leg fractures.

ACKNOWLEDGEMENTS

I would like to acknowledge my thesis committee, Dr. Sauer, Dr. Fenton and Dr.
DeJong, for all their help and advice during the research and writing process of this
thesis. Their input was invaluable, and my research was enhanced by their contributions.
I would also like to thank Dr. Edward SIadek and Sabrina for their assistance in obtaining
my sample. I would not have been able to complete this research without them. My
fellow graduate students also deserve thanks. Their ideas and support have been priceless.
I would also like to thank Tony for his words of encouragement along the way. Finally, I
would like to thank my parents for their unconditional support throughout my struggles

and accomplishments in my education and in life in general.

iii

TABLE OF CONTENTS

LIST OF TABLES ....................................................................................... vi
LIST OF FIGURES .................................................................................... ix
CHAPTER 1

INTRODUCTION ..................................................................................... 1

CHAPTER 2

BACKGROUND — FRACTURE HEALING ..................................................... 6

CHAPTER 3

BACKGROUND — INFLUENCE OF AGE ..................................................... 14
Child Abuse .................................................................................. l9
Radiography of Pediatric Fractures ..................................................... 24

CHAPTER 4

BACKGROUND — INFLUENCE OF F RACTURE LOCATION ........................... 28
Forearm ...................................................................................... 28
Leg ............................................................................................ 29

CHAPTER 5

QUESTIONS AND HYPOTHESES ............................................................. 32
Specific Aims ................................................................................ 32
Hypotheses ................................................................................... 33
Expected Results ............................................................................ 35

CHAPTER 6 ,

MATERIALS AND METHODS ................................................................. 36
Materials ...................................................................................... 36
Methods ....................................................................................... 37
Statistical Analyses ..................................................................... . ..... 38

Stages .................................................................................... 38
Age ....................................................................................... 39
Fracture Location ...................................................................... 40

CHAPTER 7

RESULTS ........................................................................................... 42
Radiographic Atlas of Pediatric Fracture Healing ....................................... 42
Stage Analysis ................................................................................ 47
Age Analysis ................................................................................. 49
Fracture Location Analysis ................................................................. 53

iv

TABLE OF CONTENTS CONTINUED

CHAPTER 8

DISCUSSION ........................................................................................ 56
CHAPTER 9

CONCLUSION ...................................................................................... 63
APPENDICES

Appendix A ........................................................................................... 67
Appendix B ........................................................................................... 78
BIBLIOGRAPHY .................................................................................... 98

LIST OF TABLES

Table 1. Summary of Sample ..................................................................... 36
Table 2. Stages and Healing Times ............................................................... 47
Table 3. Analysis of Variance with Bonferroni Correction for Six Stages ............... 79
Table 4. Analysis of Variance with Bonferroni Correction for Three Age Groups with

Six Stages: Stage 1 ................................................................................. 80
Table 5. Analysis of Variance for Three Age Groups with Six Stages: Stage 2 ........... 80

Table 6. Analysis of Variance with Bonferroni Correction for Three Age Groups with
Six Stages: Stage 3 .................................................................................. 81

Table 7. Analysis of Variance with Bonferroni Correction for Three Age Groups with
Six Stages: Stage 4 .................................................................................. 82

Table 8. Analysis of Variance with Bonferroni Correction for Three Age Groups with
Six Stages: Stage 5-6 .............................................................................. 83

Table 9. Summary of Analysis of Variance for Three Age Groups with Six Stages and
Mean Healing Times ............................................................................... 83

Table 10. Analysis of Variance with Bonferroni Correction for Three Age Groups with 3
Stages: Stages 1-2 ................................................................................... 84

Table 11. Analysis of Variance with Bonferroni Correction for Three Age Groups with 3
Stages: Stages 3-4 .................................................................................. 85

Table 12. Summary of Analysis of Variance for Three Age Groups with Three Stages
and Mean Healing Times ......................................................................... 85

Table 13. Analysis of Variance with Bonferroni Correction for Three Age Groups with
Two Stages: Stages 1-3 ............................................................................. 86

Table 14. Analysis of Variance with Bonferroni Correction for Three Age Groups with
Two Stages: Stages 4-6 ............................................................................ 87

Table 15. Summary of Analysis of Variance for Three Age Groups with Two Stages and
Mean Healing Times .............................................................................. 87

vi

LIST OF TABLES CONTINUED

Table 16. T-test for Two Age Groups with Six Stages: Stage 1 ............................. 88
Table 17. T-test for Two Age Groups with Six Stages: Stage 2 ............................. 88
Table 18. T-test for Two Age Groups with Six Stages: Stage 3 .............................. 88
Table 19. T-test for Two Age Groups with Six Stages: Stage 4 ............................. 89
Table 20. T-test for Two Age Groups with Six Stages: Stage 5 ............................. 89
Table 21. T-test for Two Age Groups with Six Stages: Stage 6 .............................. 89
Table 22. Summary of T-tests for Two Age Groups with Six Stages and Mean Healing
Times ................................................................................................. 90
Table 23. T-test for Two Age Groups with Three Stages: Stages 1-2 ....................... 90
Table 24. T-test for Two Age Groups with Three Stages: Stages 3-4 ....................... 90
Table 25. T-test for Two Age Groups with Three Stages: Stages 5-6 ...................... 91
Table 26. Summary of T-tests for Two Age Groups with Three Stages and Mean Healing
Times ................................................................................................ 91
Table 27. T—test for Two Age Groups with Two Stages: Stages 1-3 ........................ 91
Table 28. T-test for Two Age Groups with Two Stages: Stages 4-6 ........................ 92
Table 29. Summary of T-tests for Two Age Groups with Two Stages and Mean Healing
Times ................................................................................................. 92
Table 30. T-test for Fracture Location with Six Stages: Stage 1 ............................ 92
Table 31. T-test for Fracture Location with Six Stages: Stage 2 ............................. 93
Table 32. T-test for Fracture Location with Six Stages: Stage 3 ............................. 93
Table 33. T-test for Fracture Location with Six Stages: Stage 4 ............................. 93
Table 34. T-test for Fracture Location with Six Stages: Stage 5 ............................ 94

vii

LIST OF TABLES CONTINUED

Table 35. T-test for Fracture Location with Six Stages: Stage 6 ............................ 94
Table 36. Summary of T-tests for Fracture Location with Six Stages and Mean Healing
Times ................................................................................................. 94
Table 37. T-test for Fracture Location with Three Stages: Stages 1-2 ...................... 95
Table 38. T-test for Fracture Location with Three Stages: Stages 3-4 ...................... 95
Table 39. T-test for Fracture Location with Three Stages: Stages 5-6 ...................... 95
Table 40. Summary of T-tests for Fracture Location with Three Stages and Mean
Healing Times ....................................................................................... 96
Table 41. T-test for Fracture Location with Two Stages: Stagesl -3 ......................... 96
Table 42. T-test for Fracture Location with Two Stages: Stages 4-6 ....................... 96

Table 43. Summary of T-tests for Fracture Location with Two Stages and Mean Healing
Times ................................................................................................. 97

viii

LIST OF FIGURES

Figure 1. Fracture Event. Reproduced from Hufnagl (2005) .................................. 10
Figure 2. Granulation. Reproduced from Hufnagl (2005) ..................................... 11

Figure 3. Mature Callus. Reproduced from Hufnagl (2005) .................................. 11

Figure 4. Partial Bridging. Reproduced from Hufnagl (2005) ................................ 12
Figure 5. Almost Complete Bridging (tibia) and Complete Bridging (fibula). Reproduced

from Hufnagl (2005) ................................................................................ 13
Figure 6. Individual #10: Forearm fracture, Stage 1, 0 days healing ......................... 42
Figure 7. Individual #10: Forearm fracture, Stage 1, 3 days healing ........................ 43
Figure 8. Individual #10: Forearm fracture, Stage 2, 17 days healing ....................... 43
Figure 9. Individual #10: Forearm fracture, Stage 3, 30 days healing ....................... 44
Figure 10. Individual #10: Forearm fracture, Stage 3, 46 days healing ...................... 44
Figure 11. Individual #107: Leg fracture, Stage 1, 6 days healing ........................... 45
Figure 12. Individual #107: Leg fracture, Stage 2, 11 days healing .......................... 45
Figure 13. Individual #107: Leg fracture, Stage 2, 28 days healing .......................... 46
Figure 14. Individual #107 : Leg fracture, Stage 3, 49 days healing ........................... 46
Figure 15. Individual #8: Age 1, Forearm fracture, 0 days healing ........................... 68
Figure 16. Individual #17: Age 4, Forearm fracture, 5 days healing ......................... 68
Figure 17. Individual #92: Age 1, Leg fracture, 1 day healing ............................... 69

Figure 18. Individual #109: Age 5, Leg fracture, 0 days healing ............................ 69
Figure 19. Individual #14: Age 4, Forearm fracture, 27 days healing ........................ 70
Figure 20. Individual #74: Age 1, Forearm fracture, 7 days healing ......................... 7O

ix

Figure 21
Figure 22
Figure 23
Figure 24
Figure 25
Figure 26
Figure 27
Figure 28
Figure 29
Figure 30
Figure 31

Figure 32

Figure 33

LIST OF FIGURES CONTINUED

. Individual #67: Age 2, Leg fracture, 54 days healing ............................ 71
. Individual #75: Age 5, Leg fracture, 39 days healing ............................ 71
. Individual #47: Age 5, Forearm fracture, 50 days healing ....................... 72
. Individual #86: Age 1, Forearm fracture, 75 days healing ....................... 72
. Individual #98: Age 3, Leg fracture, 42 days healing ............................ 73
. Individual #100: Age 5, Leg fracture, 40 days healing ........................... 73
. Individual #12: Age 3, Forearm fracture, 51 days healing ....................... 74
. Individual #35: Age 5, Forearm fi‘acture, 40 days healing ....................... 74
. Individual #26: Age 5, Leg fracture, 63 days healing ............................ 75
. Individual #7: Age 0, Forearm fracture, 49 days healing ......................... 76
. Individual #49: Age 3, Forearm fi'acture, 111 days healing ...................... 76
. Individual #40: Age 5, Forearm fracture, 418 days healing ...................... 77
. Individual #115: Age 2, Leg fracture, 188 days healing ......................... 77

CHAPTER 1: INTRODUCTION

While once a relatively minor aspect of forensic anthropology, the assessment of
trauma in forensic cases is of great importance when determining the injuries and physical
abuse that may have occurred to an individual. One of the most significant aspects of such
trauma is the timing of the reported injury with respect to death (Sauer 1998). Skeletal
trauma can usually be differentiated among antemortem, perimortem, and postmortem. While
perimortem trauma may be more crucial in the medico-legal death investigations,
antemortem trauma, the focus of this study, can provide valuable information that may aid in
the identification of an individual or in the history and timing of past traumatic events. Past
trauma of an individual may be an indictor of future injury or cause of death. Additionally,
the amount of healing indicated may be able to reveal if inaccuracies are present in the
explanations or timing of questionable injuries. Therefore, the analysis of antemortem trauma
and a complete understanding of the fiacture healing process may enable forensic
anthropologists to apply information on the timing of injuries to skeletonized remains that
exhibit some degree of healing.

The occurrence of trauma before death is indicated by the presence of bone
remodeling. When antemortem fractures occur, there is a schedule of the healing process,
indicating what occurs within the bone after a certain amount of time. The manifestation of
this healing implies that the individual’s tissue continued to live after the trauma. In adults,
rates have been established for what can be expected for a normal healing process. The
healing process consists of an inflammatory response to the injury, repair of the bone, and
additionally remodeling and recontouring. As part of the healing process, a callus will form

to strengthen the compromised injury site. Frost (1989) states that significant callus

formation may take four to sixteen weeks to occur in adults. Sledzik and Kelley (ms) state
that an osseous reaction of healing may be present in as few as seven days, with an average
of thirteen days required to indicate healing when vieng dry skeletons. Therefore, any
evidence of skeletal healing on dry-bone suggests that the injury in question occurred at least
a week before death.

Unfortunately, such a schedule is not fully understood for pediatric cases. As children
are still in the formative phase of bone growth, healing of traumatized bone may occur at a
faster rate (Hobbs 1989; Kempe et a1. 1962). Ogden suggests initial ossification may occur
by 10 days after the injury (1984), while Patton generally states that healing in children
occurs in half the time as healing in adults (1992). It has also been suggested that a periosteal
reaction may be present in children within four days of the fracture event (Prosser et al
2005). Since these rates do vary between children and adults, the assessment of fracture
healing in children at the time of death can only be made in very general terms. Even though
the rates of healing are not well understood in pediatrics, such information is useful when
assessing the timing of injuries and whether the trauma is consistent with the story presented.
Through analyzing the timing of fractures, one may be able to demonstrate whether an
accidental event occurred or intentional abuse caused the trauma.

In the present study, radiographic stages are applied to the fracture healing rates in
young children to determine a range of time-since-injury that would be consistent with the
trauma. From the use of these radiographs, the age of young children and the location of the
injury are examined as factors in the rate of the healing process. While radiologists can
readily determine if a fracture is recent or older, radiologic dating of fractures in children is

often considered an “inexact science” and the current methods available for this analysis are

sparse (Prosser et a1 2005). Radiologists often base their estimates of the timing of injuries on
their personal clinical experience, not on primary research. The concerns inherent in the
radiographic dating of fractures demonstrate the “urgent need for research to validate the
criteria used in the radiologic dating of fractures in children younger than 5 years” (Prosser et
a1 2005:1282).

While forensic anthropology cases normally center on skeletonized remains of an
individual, radiographs of such dried bone may reveal the timing of injuries that took place
shortly before death and elucidate more information on the trauma that occurred. The
importance of such fracture healing rates with pediatric radiographs lies with many child
abuse cases. Child abuse has become a relatively recent research topic, with focus growing
on it within the later half of the twentieth century (Brogdon 1998; Kerley 1978). Trauma in
pediatric cases remains an imperative topic, as within the United States alone, there are
approximately 2,000 children dying per year due to child abuse; furthermore, many of these
children are under the age of 4 (Brogdon 1998). Additionally, fractures that occur in children
under age three are more likely to be associated with child abuse than the skeletal injuries of
older children (Kleinman 1987). While not all child abuse can be documented skeletally,
there are skeletal injuries that may indicate its occurrence, such as appendicular trauma, rib
fractures, and injuries to the shoulder girdle region (Brogdon 1998). Additionally, multiple
fiactures at different stages of healing are often considered indicative of child abuse (Walker
et a1. 1997; Kerley 1976; Gwinn et a1. 1961). When such skeletal trauma occurs, radiographs
are the typical method employed for examining the injury of the child, with CT or MRI

added for head injuries.

While typically forensic anthropology cases are often limited to skeletal remains,
there is also the possibility of examining radiographs that indicate the patterns of trauma to
the bone in a living subject or in skeletonized remains. The aims of the present study are to
extrapolate data on fracture healing from radiographs, evaluate rates of fracture repair in
young children, and compare how these rates vary with the age of the individual and the
skeletal element involved. Very few radiographic studies have focused on pediatric fractures;
thus, through the present research, an outline for what can be expected in the fracture healing
process based on pediatric radiographs has been developed. The goal of this project is to
construct and describe stages, based on radiographs, that indicate the amount of healing that
is expected by a certain time and to demonstrate how soon there may be radiographic
evidence for fracture healing in pediatric cases.

The importance of the suggested study lies with the forensic anthropological analysis
of pediatric skeletal cases. In such an instance, a forensic anthropologist must “note carefirlly
the stages of healing of different fractures in remains of children who have multiple fiactures
that occurred prior to time of death as well as any fractures that are unhealed and may have
occurred at the time of dea ” (Kerley 1976: 337-338). Additionally, when assessing such
skeletal remains, radiographic examination can aid in the confirmation of antemortem versus
perimortem injuries, as the antemortem injuries will display a radiopacity associated with
healing bone (Kerley 1978). Thus, through radiological data on the healing rate of pediatric
fractures, the forensic anthropologist will be given another tool to assess the timing of
traumatic injuries in the assessment of skeletal fractures. Additionally, within child abuse

cases, the estimation of the timing of injuries is often complicated by the abuser’s delay in

seeking medical care (Walker et a1. 1997; Brown 1976). The present study will be an attempt

to assist in confirming or refuting the proposed timing of injuries.

CHAPTER 2: BACKGROUND — FRACTURE HEALING

Fracture healing is a unique process, and while the rate healing occurs may vary,
it is important to understand the physiological reaction of the bone after fracture
occurrence. Fracture healing is normally broadly grouped into three stages: inflammatory,
reparative, and remodeling (Jones 2003). At the time of trauma, the fracture event causes
blood vessels to rupture and a hematoma to form, indicating the inflammatory stage.
Afier the initial trauma, bone reacts by osteoprogenitor cells differentiating to form
fibroblasts and other cells. These cells will then form a soft, fibrous, granular tissue at
either end of the fractured bone and extend in to and replace the forming hematoma, thus
signaling the beginning of the reparative stage. While no new bone is actually formed at
this point, it represents the first structuralizing that occurs at the fracture site. Next, a
primary woven bone callus forms. increasing the stability of the weakened fracture site.
Within this primary callus there are three different types of calluses connecting the bone:
a sealing callus (joining the broken ends), an endosteal callus (joining the opened marrow
cavity), and a periosteal callus (bridging over the entire site). A bulging shape is typical
of the periosteal callus. A third step in this process is the primary woven bone callus
being replaced by a secondary lamellar bone callus. The woven bone, and eventually any
excess callus, will be resorbed (Ortner 2003; Hendrix 2002; Frost 1989). Finally, during
the remodeling stage, the bone is recontoured and restructured internally, a process which
can take place concurrently with lamellar bone remodeling (Frost 1989).

During life, the osteogenic properties of bone allow healing to transpire. While
the healing of bone is normal, the rate at which this process occurs can vary among

individuals and among different bones and locations within the body (Ortner 2003). The

medical literature addresses the stages of bone healing and the remodeling rates, yet often
these articles suggest subtle differences, making an exact timescale a complicated task
(Hufnagl 2005).While these stages may be considered an arbitrary assignment, such a
categorization may be useful when studying the expectations of fracture repair (Hendrix
2002).

The process of healing can be understood at a number of different levels: clinical
radiographic, biochemical, or histological (Hendrix 2002). The radiographic assessment
of fracture healing, while relatively absent in the forensic anthropology literature, has its
foundation in the medical realm. Both Toal and Mitchell (2002) and Hendrix (2002) have
examined radiographic characteristics of fracture healing. Immediately following
fracture, bone edges will appear clearly outlined, with a blurring of this line within five to
fourteen days. Following this, a “fluffy” callus appears and becomes radiopaque. An
initial bridging of the callus may be present around the one month mark. Several months
after the initial fracture, the fracture line may become obliterated, after which remodeling
takes place (Hendrix 2002; Toal and Mitchell 2002).

The use of radiographs to evaluate the stage of fracture healing is somewhat
controversial and the research has been limited by small samples of human subjects or
animal studies. In 1991, De Palma and colleagues carried out a study of radiographic
images, examining adult femoral fractures in Italy. The authors examined 15 radiographs
of femoral fractures before and at one, two, four, and six months after surgery. From such
research, the authors concluded that radiographs can be used to assess the stage of
remodeling in fracture repair (De Palma et al. 1991). While this study indicates the

promise of radiographic assessment of fracture healing. there are also studies indicating

its weaknesses. In 2001, Blokhais and colleagues (2001) carried out a study using
radiographs to assess the healing process in mid-shaft tibial fractures of goats. In this
instance, the authors felt that radiography was a poor indicator of the healing process
(Blokhais 2001).

While there are investigations indicating conflicting results when using
radiographs to assess fracture healing, there are also analyses that have used radiographic
assessment of fracture healing to determine the effect of various factors on the healing
process. For instance, Kline and colleagues (2005) used radiographic assessment to
determine the impact of BMP-14 deficiency in mid-shalt femoral fractures of mice. In
this study, the authors stated that such radiographic grading was able to indicate fracture
healing within a range of one to four days (Kline et al. 2005). Other studies have also
employed radiographic stages of fracture healing, specifically in non-human samples
(Risselada et al. 2005, Heybeli et al. 2002).

Researchers have also developed methodology for the actual assessment of
fracture repair from radiographs. Tiedeman and colleagues (1990) devised a scoring
system for assessing fracture healing in adult dogs, based on several features: the amount
of bone formation, the amount of the fracture line present, and evidence of remodeling on
the intramedullary canal and cortex (Tiedeman et al. 1990). Also, in 2003, Grigoryan and
colleagues examined radiographs and CTs of 39 adult men and women with radial or
tibial fractures at one, two, four, eight, twelve, and sixteen weeks post-fracture. In this
analysis, the authors were able to determine a time frame of eight weeks for radiological

union of the fracture. based on the presence of radiographic features, such as lines.

fracture gaps, callus appearance. callus-to-cortex ratio, and radiological union (Grigoryan
et al 2003).

Additionally, Langley-Hobbs examined the fracture healing process in small
animals. She identified two types of healing that may occur: primary/direct and
secondary/spontaneous. In the former case, clinical union may be present prior to
radiographic union, whereas in secondary healing, clinical union may not be achieved
until weeks after radiographic union (Langley-Hobbs 2003). Clinical union commonly
refers to the point in the healing process at which fixation devices may be removed,
radiographic union, on the other hand, refers to the point in the healing process where
bridging is present over the fracture gap (Langley-Hobbs 2003). Secondary healing.
occurring afier external stabilization of the region, is indicated radiographically by
resorbtion (indicated by the loss of sharp edges), callus formation (indicated by callus
visibility), healing (indicated by fracture lines and the callus increasing in opacity), and
remodeling (indicated by the cortical shadow becoming visible). Primary healing,
associated with internal stabilization of the fracture and a very minimal gap at the fracture
site, may be indicated by reduced density at the fracture site, seen as a “smudging” at the
interface (Langely-Hobbs 2003: 32). Thus, each stage of fracture healing has certain
characteristics associated with it.

In 2005, Hufnagl conducted a study to devise a method of determining age-since-
trauma based on radiographic images. He examined 62 sets of radiographs (each set
being comprised of a minimum of four radiographs), coming from individuals aged two
to 93. Information was included on the sex of the individual, the ancestry of the

individual, the type of fracture, and the date of the fracture. Additionally. fractures were

examined from radius, tibia, and femur. From this sample, Hufnagl presented six stages
of fracture healing that are demonstrated radiographically: fracture, granulation, mature
callus, partial bridging, almost complete bridging, and complete bridging. The time of
healing was also determined for each of the sets of radiographs. This data allowed
Hufnagl to examine the role of age, sex, ancestry, and weight placed on the bone in the
healing rates. Of these factors, it was found that only age was significantly correlated
with the stages of fracture repair (Hufnagl 2005).

The phases of radiographic healing used by Hufnagl are of particular importance
to the present study and will be described further here:

1: Fracture Event — absence of observable bone healing

 

Figure 1: Fracture Event. Reproduced from Hufnagl (2005).

10

2: Granulation — beginning of resorption along the fracture fragments and the
initial indicators of callus formation; widening of the fracture gap due to
resorption, blurring of the fragment edges, and appearance of faintly mineralized

buds of callus (fluffy callus).

 

Figure 2: Granulation. Reproduced from Hufnagl (2005).

3: Mature Callus - radiopaque and appears as dense as regular bone with the
exception of its bulging out and over the fracture gap; fracture line still clearly

defined.

 

Figure 3: Mature Callus. Reproduced from Hufnagl (2005).

11

4: Partial Bridging — refers to point in time when the callus is connected across
the fracture gap in some areas; fracture line is beginning to blur and is not clearly

evident along some portions of the fragments.

 

Figure 4: Partial Bridging. Reproduced from Hufiragl (2005).

12

5: Almost Complete Bridging — points in fracture healing when fractures are
classified as clinical union; fiacture line is almost completely blurred
6: Complete Bridging — fracture line is completely eradicated in the x-rays and

evidence of the line is not longer observable (Hufnagl 2005).

 

Figure 5: Almost Complete Bridging (tibia) and Complete Bridging (fibula).

Reproduced from Hufnagl (2005).

13

CHAPTER 3: BACKGROUND — INFLUENCE OF AGE

While similarities abound in the healing process between children and adults, it is
important to realize children are not merely miniature adults; children have their own
unique skeletal and physiological characteristics. Sub-adult bone is undergoing growth,
and as such, may heal at a faster rate (Hobbs 1989; Kempe et al. 1962). Studies have
indicated that in children it is possible for ossification or even for full strength to be
regained after 10 days (Lewis 2007; Ogden 2000). In an examination of 275 femoral
fractures in 265 children under the age of 18, Skak and Jensen found that the time for
fracture union varied between three weeks and two years (1988). The researchers
discovered that side, sex, type of trauma, or type of fracture did not affect the amount of
time required for union, but age was demonstrated to influence healing time based on a
log-normal distribution with an increment in mean time to union of 0.7 weeks per year
(Skak and J enson 1988). This study and others demonstrate that the rate of healing has
been directly related to age — the younger the individual, the faster the healing (Glencross
and Stuart-Macadam 2000; Jones 2003). For instance, a femoral fracture in an infant may
heal in 3 to 4 weeks, whereas a similar fracture in an adolescent may take 12 to 16 weeks
to heal (Jones 2003). The variations in healing time between children and adults
necessitate further examination of pediatric fracture healing.

The rapid healing in children can contribute to a loss of any deformity that could
have resulted from trauma, masking injuries that would have been more dramatic in their
adult counterparts (Lewis 2007). While faster healing is acknowledged in pediatric
fractures, there is no radiologic evidence that has been shown to substantiate this tenet

(Prosser et al 2005). Due to variations in bone trauma and healing that are associated with

14

age, anyone who is assessing pediatric fractures should be aware of the mechanisms of
injury, the causes, the acute responses, the treatments, and the long-term effects of
fractures in children (Ogden 2000). Particularly, the differences between adult and
pediatric skeletal injury should be described based on variations in anatomy, physiology,
and biomechanics (Ogden 2000).

The anatomy of pediatric bone differs from the mature bone of adults. The
ossification and union patterns of the sub-adult skeleton create differences in the
likelihood of fracture and in the healing time of certain regions of bone. For instance,
consider the dense bone of a diaphysis compared to the spongy bone of the metaphysis
and epiphysis (Ogden 2000). Young children, and more specifically infants, are afforded
increased defense against skeletal injury through their unmineralized bone. As
ossification occurs, the thinning areas of the physes become increasingly prone to injury
(Glencross and Stuart-Macadam 2000).

In addition to the ossification and union patterns of the sub-adult skeletal, the
increased vascularity and periosteal characteristics will cause bone to heal at a faster rate.
Ogden (2000) gives an eloquent review of the impact of such features on the fracture
healing process. Ogden notes that at birth fracture healing is “remarkably rapid,” but
immediately the repair process begins to decrease through childhood and adolescence
(2000:42). The thicker, stronger periosteum of pediatric bone tends to resist displacement
and is more biologically active, decreasing the healing time. Pediatric periosteum, while
thicker, is more likely to separate from the bone. Ogden (2000) states that this
characteristic aids in preventing complete rupture of the periosteum generating an

“intrinsic stability” due to less displacement of fracture fragments. As an individual ages,

15

the periosteum thins, decreasing in osteogenic activity and lengthening the healing time
required for skeletal injury (Jones 2003).

Pediatric physiology implies that a child will respond differently to the metabolic
and physiological stresses of skeletal trauma than an adult. Adults will have an increased
metabolic demand after trauma, whereas in children there is little or no difference
detected (Ogden 2000). Children tend to have an already high metabolic rate so less of a
change is necessary when trauma occurs, perhaps accounting for the increased survival
rate after severe trauma when compared to adults. While afforded some physiological
protection after a traumatic occurrence, children may be at a higher risk for hypovolemic
shock or hypothermia due to blood volume and surface area ratios (Ogden 2000). Such
physiological reactions may influence the repair process and rate of healing of the
traumatized bone.

Due to the differences in bone structure and bone composition, the skeleton of a
child has unique biomechanical properties. Sub-adult bone is less brittle, more porous,
contains larger haversian canals, contains more water, and has a lower mineral content
(Auringer 2002). These characteristics lend to an increased elasticity and plasticity in
children’s bone (Auringer 2002). Generally, adult bone has ceased apposition and is only
responding to the stresses placed on specific skeletal elements (Ogden 2000).
Additionally, adult bone tends to fail in tension, whereas pediatric bone may fail in
tension, compression, or both (Ogden 2000). The biomechanical properties of pediatric
bone, creating an increased capacity for plastic and elastic deformation, give rise to an
increased tendency for tensile failure to be deficient, leading to incomplete fractures. such

as plastic deformation and greenstick fractures (Ogden 2000; Auringer 2002). The

16

porosity of bone and rough surface that results from pediatric bone structure contribute to
the need for increased energy and time before bone will fail, aiding in the prevention of
complete fracture (Jones 1994). As an individual ages, the plastic phase of bone
deformation is shortened, increasing the vulnerability for complete fracture.

When skeletal trauma does occur in children it tends to present as physeal
fractures, torus fractures, greenstick fractures or even as plastic deformation without
fracture (Lewis 2007). Physeal fractures, the only complete fracture mentioned, involve
the growth plate of the bone. Such injury may lead to angular deformity, irregular
shortening, or joint incongruity. Torus fractures occur when there is a failure of cortical
diaphyseal bone causing the membranous metaphyseal bone to be affected resulting in a
slight bulging at the diaphyseal/metaphyseal junction. Greenstick fractures occur when
bending on the compression side allows the cortex and periosteum to remain intact, while
the force causes fracture on the tensile side. Plastic deformation, which may occur in a
fall where the hands are outstretched. results in a broad curvature over the length of the
bone (Glencross and Stuart—Macadam 2000). As evident from the types and frequencies
of injuries, youth contributes to some intrinsic protection from harm.

The types of fracture and injuries also differ between children and adults. Due to
their small size, children may be more susceptible to fatal soft tissue injuries. Altemately,
the forces of falls are not as great due to the small size of children, preventing more
serious injuries from occurring. Essentially, most childhood injuries tend to be low-
impact “trip, stumble and fall” occurrences (Auringer 2002). The risks and typical
mechanisms of injury also vary with age. According to modern clinical data, each age

group will have 3 characterized pattern of fracture varying in location and nature. Each

17

age may even have a “typical injury” (Jones 1994). For instance, today, children under
the age of one are most susceptible to physical abuse, children ages five to 10 are
vulnerable to falls and traffic-related injuries, and older children are more likely to incur
sports-related injuries (Glencross and Stuart-Macadam 2000). Due to non-mobility,
children under the age of two are likely to have sustained injuries by having someone else
injure them (Jones 2003). While the hazards children encounter have undoubtedly
changed over the years, it is likely there are similarities in fiacture causation. To
illustrate, consider the increase in activity of children over their lifetime. All children will
encounter falls, but infants are unable to protect the more vital areas of their bodies. In
non-ambulatory children, falls are a common traumatic event that may result in fractures
to the clavicle or skull. Ambulatory children will be more likely to suffer fracture to the
limbs due to attempts to break their falls (Glencross and Stuart-Macadam 2000).

The characteristics of sub-adult bone continue to influence the traumatized region
after a fracture has occurred. Due to the increased vascularity of pediatric bone, the
inflammatory process occurs at a faster rate and plays a greater role in the healing
process. This increased response time shortens the early stages of fracture healing in
children, when compared with adults (Jones 2003; Ogden 1982; Ogden 1963; Rang
1983). Furthermore, the subperiosteal callus that forms during pediatric fracture healing
contributes an increased stability to the fracture site, which may allow the fracture to be
considered healed by the end of the reparative stage (Jones 2003). While pediatric
fractures may heal at a faster rate, they may not always remodel. Remodeling capacity in
children is based on the age of the child, the distance of the fracture from the end of the

bone, and the amount of angulation (Ogden 2000; Jones 2003). Satisfactory remodeling

18

tends to occur in children with two years of growth ahead, fractures near the ends of the
bone, and injuries not associated with the movement of a joint (Ogden 2000). On the
other hand, delayed union or non-union are uncommon in children due to the increased
healing capacity of pediatric bone (Ogden 2000).

In addition to the biomechanical properties of bone itself, the child, as a patient,
differs significantly from his or her adult counterpart. Of concern for pediatric patients is
the potential for lack of patient history or erroneous data, especially problematic in
abused children. A child’s account of the traumatic incident tends to be inaccurate or
incomplete, whereas the parent or guardian may not have witnessed the occurrence
(Ogden 2000). Follow-up care may also be problematic. While often recommended to
continue follow-up until six months to a year after injury, most parents or guardians will

discontinue medical assessment when their child appears to be better (Ogden 2000).

Child Abuse

Child abuse, often considered a modern clinical phenomenon, should be
considered in all pediatric skeletal trauma cases (Jones 2003). Furthermore, any
profession dealing with sub-adult skeletal injuries should be familiar with the indications
and trauma associated with child abuse (Kleinman 1987b). Skeletal trauma, while not
often life-threatening in child abuse cases, can leave evidence of recent trauma as well as
well-healed injuries. Additionally, the high frequency of skeletal trauma in child abuse

cases suggests that further examination is necessary in pediatric fracture cases (Kleinman

1987b).

l9

The documentation of child abuse, while having a significant history, has only
been identified within the past century. In 1946, Caffey examined long bone injuries in
six infants with subdural hematomas. Through the correlation of these injuries, he
recognized the reality of child abuse and brought it to the attention of others in the
medical field. Caffey’s catalyst prompted other medical professionals to examine
children for abuse through clinical and radiographic evidence (Kleinman 1987b). In
1962, “Battered Child Syndrome” (BCS), coined by Kempe and colleagues, indicated
that certain skeletal injuries including appendicular trauma, rib fractures, and injuries to
the shoulder girdle region may be suggestive of child abuse (Brogden 1998). Battered
Child Syndrome identified such fundamental injuries as a clinical condition in which
serious physical abuse is directed at children. The key to identifying Battered Child
Syndrome is multiple fractures in various stages of healing (Walker et al. 1997; Kerley
1976; Kempe et al 1962; Gwinn et al. 1961).

The skeletal surveys, including radiographs, of abused children may demonstrate
multiple injuries in various stages of healing, findings inconsistent with the history
provided by a parent or guardian, metaphyseal lesions, and fractures of ribs, sterna,
vertebrae, and scapulae (Auringer 2002). The most common injuries in abused children
tend to occur on long bone diaphyses, except in infancy, when skull, metaphyseal, and rib
fractures are more common (Auringer 2002). Eighty percent of the fractures in abused
children occur when the child is under the age of 18 months; in nonabused children 85%
of fractures occur after the age of 5 years. In children under one year of age, fractures are
relatively uncommon, yet multiple severe fractures during the first postnatal year may

indicate a problem (Auringer 2002).

Evidence for abuse may seem unquestionable, yet such a conclusion is still
problematic. While the identification of such features may quickly prompt action from a
medical practitioner, before child abuse may be addressed “metabolic disorders or
skeletal displasias must be considered.” (Ogden 43:2000). After the elimination of such
conditions, the “diagnosis” of child abuse requires either indisputable radiological
evidence of fractures or fractures that were not consistent with the given history
(O’Connor and Cohen 1987). While the mechanism of injury would obviously be one
manner in which to dispute the history of injury, the ability to date fractures would
further enable a radiologist or anthropologist to confirm or refute the stated occurrence.
For instance, consider a child brought to the hospital. The parents have stated that the
child fell off a bicycle landing onto an outstretched hand earlier that morning. When the
child is examined there is indeed a fracture to the forearm, but instead of the fresh injury
that would be expected, the fi'acture seems to have been healing for several days. In this
case, the story given by the parents may have been erroneous. While the mechanism of
injury was consistent with the fracture, the timeframe given was not. This inconsistency
may indicate that the parents where unwilling to give a complete and true account of
what occurred.

While child abuse has only recently been identified clinically, it does merit a
historical and anthropological examination. Beating children in the past has been socially
sanctioned in some groups of people and is presented in historical documents (Walker
2001a; 2001b). For instance, consider this Old Norse rhyme dating from the 17th century:

The child will not be quiet, the child will not be quiet,
Take it by the leg and hit it against the wall,

The child will not be quiet.
Cited in Lewis (2007:175)

21

Child abuse is also mentioned in literature from physicians, dating to medieval times and
earlier (Lewis 2007). The historical view on child abuse has suggested that children were
traumatized in a variety of manners, including freezing, burning, drowning, shaking,
mutilation, and throwing (De Mause 1974).

Ironically, child abuse, a cause of more than 2,000 deaths in the United States
today, is absent in the archaeological record (Lewis 2007; Brogden 1998). In an
examination of more than 5,000 skeletal remains of prehistoric children, Walker and
colleagues (1997) found no examples that were comparable to present day instances of
child abuse. In addition to multiple fractures in different stages of healing, subperiosteal
lesions were examined in modern cases of chronic abuse. These lesions appear localized,
asymmetrically located, and in various stages of healing (Walker et al 1997). While
subperiosteal lesions are undetectable by radiography, an experienced physical
anthropologist should be able to readily identify any healed lesions present (Walker et a1
1997). Such subperiosteal lesions are absent in the archaeological record, indicating that
this pattern of physical abuse did not occur until modern times (Walker et a1 1997). With
the fragmented and sometimes poorly preserved sub-adult remains, it may be difficult to
properly identify and examine the lesions. For instance, if an isolated long bone fragment
were found, the distribution of other lesions on the individual could not be identified.
Additionally, preservation issues may affect the quality of bone available, especially
compared to modern forensic examples. Finally, such a lesion could be mistaken for bone
inflammation or periostitis due to some underlying disease process (Ortner 2003). Other
postulations as to the reasons for the lack of skeletal evidence of physical child abuse

have also been made. Walker suggests that the abuse of children may have been

22

impossible in past societies, as kin-based groups would have led to an intervention
(Walker 2001a). The small sample sizes of bioarchaeology have been criticized for
hiding the evidence of child abuse. Even considering today’s rates of child abuse, one
would need to examine approximately 2,000 infant remains to find just one instance of
child abuse in the skeletal record (Waldron 2000). At the same time, while Battered Child
Syndrome would be expected to be uncommon, there are even more rare pathologies and
disorders that have been found in the archaeological record, indicating that if child abuse
was present, researchers would have inevitably seen at least several instances in the
extensive bioarchaeological examinations that have taken place (Walker 2001b). It is the
inexplicable lack of skeletal evidence for child abuse that has led numerous researchers to
conclude that it did not exist in the past.

With various historical documents referring to the beating of children and to some
extent its acceptance, there seems to be a contradiction between what has been described
and what is actually observed on skeletal remains. Walker (2001b) suggests that this
contradiction may be partially due to the differences in the patterns of child abuse. For
instance, historical records often refer to older children, while the modern definition of
battered child syndrome is typically seen in children under the age of three (Walker
2001b; Kempe et a1 1962). Additionally, if past child abuse was less fatal, children would
survive into adulthood with very little evidence of trauma ever occurring. Walker (2001b)
also postulates, that if Battered Child Syndrome is a recent occurrence, social conditions,
control and support may have degraded and contributed to the increased violence seen

toward children today.

23

The past of child abuse remains questionable, yet the current presence of these
traumatic occurrences emphasizes the importance of examination of pediatric injuries.
With the broad definition of child abuse, including physical harm, sexual assault, neglect,
and physiological abuse, not all abused children will demonstrate skeletal injury
indicating physical abuse (Kleinman 1987a). Furthermore, not all pediatric skeletal
injuries can be indicative of child abuse. Instead, it is essential to consider the reality of
abuse when clinically examining children or when anthropologically examining sub-adult

remains.

Radiography of Pediatric Fractures

With child abuse cases, a complete skeletal survey is performed to assess the
current injuries and possibility of past trauma. Radiographs are considered imperative in
the skeletal survey of all infants demonstrating physical abuse (Merten et al 1983). A
skeletal survey of potentially abused children will typically consist of an anterior view of
the appendicular skeleton and both anterior and lateral views of the axial skeleton
(Kleinman et a1 2004). In the past, a “babygram,” a single image of an infant exposed on
one or two radiographic films, has been considered sufficient documentation, yet more
thorough radiography of the trunk, spine, and skull are necessary when examining
potential abuse victims (Kleinman et a1 1989). Postmortem radiographs are also of use
when examining the extent and sequence of pediatric fractures in unexpected deaths
(McGraw et a1 2001). Radiographs obtained contribute to valuable evidence that may be

used in prosecution of abusers.

24

Radiography has also played a part in distinguishing child abuse from accidental
trauma encountered in infants and young children. In one such study, Cumming analyzed
a sample of 23 infants who suffered fracture from delivery attempting to distinguish birth
trauma from child abuse (1979). The sites of skeletal trauma resulting from delivery
included fiactures to the clavicle, humerus and femur. Additionally, Cumming states that
calcification on radiographs of fiactures at the time of birth would be present between
seven and eleven days after birth, thus enabling an examiner to distinguish between
postnatal accidents from perinatal trauma (1979). The accessibility of pediatric
radiographs and the possibility of extracting evidence from the images, generate a need to
establish more precise radiographic healing stages.

Few studies have addressed the radiographic assessment of fracture healing in
children. The studies that do attempt to correlate specified radiographic signs with a
certain amount of healing. Yeo and Reed (1994) examined one aspect of radiographic
fracture healing, the callus, in a preliminary study of 25 children fi'om birth to age 14 to
determine the influence of sex and age on healing time. Serial radiographs of the femur
were examined and categorized into three stages of callus formation. Stage 1 was
indicated by the first radiographically visible sign of callus. Stage 2 was demonstrated
when complete bridging of the callus was present. Stage 3 was defined by a smooth,
homogeneous density callus with a fracture line still visible. Results indicated no
difference in healing times based on sex and no statistically significant results for age.
The researchers did state that a trend toward increased healing time with patient age was

present at Stage 3 only; with an increased sample size more dramatic results may be

25

expected (Yeo and Reed 1994). Yeo and Reed concluded that further studies and a more
accurate method of assessing fracture healing are needed (1994).

Following Yen and Reed, Islam and colleagues (2000), examined a collection of
radiographic features in a sample of 707 radiographs from 141 children, ages one tol7.
Two hundred five fractures to the radius or ulna were present in the sample, with between
two and eight radiographs per fracture. The researchers examined the effect of sex. age,
displacement, and location on the radiographic signs (sharp/blunt fracture margins,
fracture gap, periosteal reaction, callus density, bridging, and remodeling). Blunting of
fracture margins was noted at one week, though the presence of casts limited this finding.
Periosteal reactions, fracture gap widening, increased bone density, and callus
calcification were seen by two weeks. Partial bridging at the fracture site was seen as
early as three weeks and remodeling was seen as early as four weeks. No radiographic
signs of healing were present prior to five to seven days after the traumatic event. In the
analysis of the radiographic characteristics, the researchers found no statistically
significant difference associated with sex, age, displacement or fracture site (Islam et al
2000). The study, while demonstrating the importance of assessing more than one
radiographic feature was limited in the number of individuals under the age of four. The
small amount of individuals in an age group particularly vulnerable to abuse,
demonstrates the need for larger studies to assess fracture healing in infants and young
children (Prosser et al 2005).

Whether abuse is suspected or not, pediatric bone may be difficult to assess
radiologically. While increasing fracture healing rates, the anatomy and biomechanical

properties of sub-adult bone makes radiographic examination slightly problematic.

26

Children are undergoing endochondral ossification, thus the unfused epiphyses of
children may be variably radiolucent (Ogden 2000). Generally, bones should be
radiographed in two different planes: anterior-posterior and lateral. While one View will
usually depict the fracture, not all fractures are seen on both views (Ogden 2000).
Furthermore, the tendency for pediatric fractures to be incomplete or minimally displaced
increases the difficulty of clinically or radiographically identifying the trauma (Dunbar et
al 1964). Understanding the ossification and fracture patterns of pediatric bone is

essential when radiographically examining skeletal injury.

27

CHAPTER 4: BACKGROUND - INFLUENCE OF FRACTURE LOCATION
Forearm

Fractures of the forearm represent a common injury in both children and adults,
yet the age group most likely to suffer such an injury is individuals between the ages of
five and 14 years (Chung and Spilson 2001). Additionally, forearm fractures are the most
common location of skeletal trauma in childhood, accounting for approximately 25-40%
of all fractures. and 62% of upper limb fractures (Auringer 2002; Jones 2003; Armstrong
et al. 2003; Waters and Mih 2006). Of forearm fractures, the distal aspect of the forearm
is the most likely site of fracture, but the shaft of the forearm is the most likely reason for
orthopedic involvement (Waters and Mih 2006, Mehlman and Wall 2006). The high
frequency of skeletal trauma to the forearm and the extent of orthopedic participation in
these injuries, especially in childhood, make the forearm an ideal location to examine in
the present study.

Fractures to the shaft of the forearm tend to occur when an individual falls onto an
outstretched hand. In children, this ofien results in a torus or buckle fracture; with greater
force, a greenstick or complete fracture may occur (Keogh et al. 2002) Often increased
force results in fracture to both the radius and ulna, though in such circumstances the
fracture to the radius may be more severe (Keogh et al. 2002, Cox and Sonin 2002).
Forearm fractures are often group grouped generally into “both-bone fractures” and
“greenstick fractures” (Mehlman and Wall 2006). When complete both-bone forearm
fractures occur, the injury may be adequately manipulated in a closed reduction.
Greenstick fractures may require complete breakage before the fracture can be properly

reduced (Cox and Sonin 2002). Both-bone fractures. which may demonstrate fracture at

different levels for the radius and ulna, occur when a child falls (Mehlman and Wall
2006). Isolated fractures, on the other hand, tend to occur from direct blows or associated
dislocations (Cox and Sonin 2002: Mehlman and Wall 2006).

Forearm fractures may be diagnosed and monitored radiographically and
clinically. To radiographically examine forearm fractures, an anterior-posterior view and
lateral view are required to confirm the diagnosis (Cox and Sonin 2002). While precise
healing times vary, the radiological union of fracture fragments in the forearms of adults
may be indicated around 8 weeks, whereas in children, forearm fractures heal at a more

rapid rate (Hertel and Rothenfluh 2006: Baitner et al. 2007).

Leg

Nonphyseal fractures of the leg are among the most common in the lower
extremities of children, and are the third most common pediatric long bone injury (Jones
2003; Heinrich and Mooney III). A majority, 70%, of tibial fractures is an isolated
occurrence; the remaining 30% involve the fibula as well (Heinrich and Mooney III).
When more force is involved, both bones are more likely to fracture, whereas. when an
isolated tibial shafi fracture occurs it commonly presents as a spiral fracture. Spiral
fractures in particular have often been cited as evidence of child abuse (Mellick et al
1988). The tibia is one of the most commonly fractured bone in abused children, and
often, unlike what is expected with Battered Child Syndrome, a single fracture may be all
that is present (Loder and Bookout 1991; King et al 1988). Due to the prevalence of tibial
fractures in childhood and their potential association with child abuse, the tibia was

selected as another location to examine in the present study.

Often the indirect forces associated with falls generate a spiral or oblique fracture
in the tibia with no associated fracture of the fibula; high impact, direct injuries are more
likely to involve both bones (Helms and Major 2002). Specifically in children under the
age of four, tibial fractures tend to be the former, often located in the distal and the
middle one third of the bone (Heinrich and Mooney III 2006). Spiral fractures of the tibia
may require open reduction and internal fixation due to the high frequency of delayed
union and re-fracture occurrence (Helms and Major 2002). Toddler’s fractures, a specific
type of lower limb fracture, are an undisplaced spiral fracture of the distal tibia that
occurs in children around the age of two (Dunbar et al 1964). Such injuries are commonly
sustained as a result of one of a child’s numerous falls. After the fall, the child refuses to
bear weight on the leg, with radiography revealing the spiral fracture (Helms and Major
2006)

Fractures of the shaft of the tibia and/or fibula require the entire length of the
bone or bones to be radiographed; this radiography is needed to ensure rotation is
evaluated and to determine if fractures are present on the other bone at a site distant from
the fracture of the first bone (Helms and Major 2002). Often spiral fractures may be seen
in either an anterior-posterior radiograph or a lateral radiograph, but not visualized well
in the other; therefore, two views are required to confirm the fracture (Helms and Major
2002).

Tibial fractures are notorious for their long healing times. In adults, tibial shaft
fractures heal in approximately 16 weeks in adults, but in children the healing occurs at a
faster rate (Helms and Major 2002). The exact healing time of tibial shaft fractures

depends on the severity of the injury, including the displacement of fracture fragments,

30

the degree of comminution, whether the injury is open or closed, and whether other
complications, such as infection. are present (Helms and Major 2002). While typically
fractures will be healed after several months, the fracture may be defined as a delayed
union if the site remains open beyond 20 to 24 weeks. Nonunion is indicated by the
fracture line remaining open without formation of a periosteal callus for 6 months to a
year after the fracture occurrence (Helms and Major 2002). In addition to the decreased
healing time, the distal tibia is known for an increased frequency of nonunions and
delayed unions when compared with other elements throughout the skeletal system

(Helms and Major 2002).

31

CHAPTER 5: QUESTIONS AND HYPOTHESES

The aims of the present study are to determine the applicability of radiographic
assessment to pediatric cases, produce stages of radiographic healing and descriptions,
and provide a timeline for the stages of healing in infants and young children. There has
been little research done on this topic in the past, as most investigations have centered on
known adult radiographs or traumatized non-human long bone specimens. While these
past studies have had a different focus from the present study, the radiographic
characteristics and features of fractures described in the past may be modified and used
when describing the radiographs and developing the stages of fracture healing to be

analyzed in this study.

Specific Aims

The specific aims of this study are as follows:

1. To collect and digitize a set of radiographs from a sample of infants and
young children that represent different stages in the healing process of tibial
and radial fractures for each subject.

2. To develop a series of stages (from the literature) to describe and measure the
typical bone fracture repair process.

3. To evaluate, for each radiograph, the stage of fracture repair.

4. To segment the sample by age group (0-1 years, 2-3 years, 4-5 years) and
evaluate the variation in fracture repair rates. An average number of days of

healing for each stage will be calculated for each group and an analysis of

32

variance (ANOVA) will be used to determine the significance of calculated
means.

5. To evaluate the variation in fi‘acture healing rates among different skeletal
elements (tibia and radius). An average number of days of healing for each
stage will be calculated for each skeletal element and a t-test will be used to
calculate the significance of the calculated means.

6. To summarize the data and present mean healing times for each of the stages
of fracture repair for the three age groups and for each skeletal element

examined.

Hypotheses
Hypotheses were generated to allow for the statistical analysis in Specific Aims 3,
4 and 5. Hypotheses of this study are as follows:
1. Null hypothesis 1: No differences in mean healing time will be seen among
Stages 1-6.
2. Null hypothesis 2: No difference in mean healing time will be seen among
group A (0-1 years), group B (2-3 years), and group C (4-5 years).
a. Null hypothesis 2a: No difference in mean healing time in Stage 1 will
be seen among group A (0-1 years), group B (2-3 years), and group C
(4-5 years).
b. Null hypothesis 2b: No difference in mean healing time in Stage 2 will
be seen among group A (0-1 years), group B (2-3 years), and group C

(4-5 years).

33

Null hypothesis 20: No difference in mean healing time in Stage 3 will
be seen among group A (0-1 years), group B (2-3 years) and group C
(4-5 years).

Null hypothesis 2d: No difference in mean healing time in Stage 4 will
be seen among group A (0-1 years), group B (2-3 years) and group C
(4-5 years).

Null hypothesis 2e: No difference in mean healing time in Stage 5 will
be seen among group A (0-1 years), group B (2-3 years) and group C
(4-5 years).

Null hypothesis 2f: No difference in mean healing time in Stage 6 will
be seen among group A (0-1 years), group B (2—3 years) and group C

(4-5 years).

3. Null hypothesis 3: No difference will be seen in mean healing time between

the tibial shaft and the distal radius.

a.

b.

C.

d.

Null hypothesis 3a: No difference will be seen in mean healing time in
Stage 1 between the tibial shaft and the distal radius.
Null hypothesis 3b: No difference will be seen in mean healing time in
Stage 2 between the tibial shaft and the distal radius.
Null hypothesis 3c: No difference will be seen in mean healing time in
Stage 3 between the tibial shaft and the distal radius.
Null hypothesis 3d: No difference will be seen in mean healing time in

Stage 4 between the tibial shaft and the distal radius.

34

e. Null hypothesis 3e: No difference will be seen in mean healing time in
Stage 5 between the tibial shaft and the distal radius.
f. Null hypothesis 3f: No difference will be seen in mean healing time in

Stage 6 between the tibial shaft and the distal radius.

Expected Results

Through the radiographic analysis of stages of healing and the background
research, the following results are anticipated:

1. The younger the individual, the faster the healing will occur.

2. Forearm fractures will heal faster than lower limb fractures.

3. These differences will be most evident in the beginning of the fracture healing

process.

35

CHAPTER 6: METHODS AND MATERIALS

Materials

For this project, a collection of radiographs obtained from Dr. Edward Sladek (an
orthopedic specialist in Lansing, MI) was used. Radiographs of fractured bones (radius
and tibia) from infants and young children of ages 0-1, 2-3, and 4-5 were included, with
ages taken from the time of initial injury. The individuals in the sample represented male
and female children of known age and time of injury from a mid-Michigan population.

The radiographs examined consisted of sets of images from each individual. If
only one radiograph was available, the individual was not excluded, as there may have
been healing evident with an associated time frame. The number of serial radiographs
ranged from one to 11, with an average of three radiographs per patient. Any individuals
demonstrating co-morbidity or any systemic disorder, such as diabetes or osteogenesis
imperfecta, which may affect the bone healing rate was excluded from the study. The
total sample that was examined consisted of 107 individuals and 294 radiographs. Table 1

summarizes the categorization of the sample.

Table 1 — Summary of Sample

 

 

 

 

 

Age 0-1 Age 2-3 Age 4-5 TOTAL
Forearm 15 15 30 60
Leg 16 16 15 47
TOTAL 31 31 45 107

 

 

 

 

 

36

 

Methods

The radiographs were digitized. Digitization was performed by photographing the
radiographic images on a light board. A tripod was set-up and a Canon Rebel X digital
SLR camera with a 50mm macro lens was directed perpendicular to the radiographs.
Images were saved on an external hard drive and were labeled according to individual
and radiograph date. Multiple radiographs taken on the same date of the same individual
were labeled alphabetically. For instance, the first radiograph taken on July 15, 1996 of
individual 1, would be labeled “1_7.15.96a”.

The sample was catalogued in an Excel 2003 worksheet. Data for each individual
included: reference number, date of birth, date of fracture, date of subsequent
radiographs, fracture location, fiacture diagnosis, age at time of fracture, days of healing
(calculated as the number of days between fracture occurrence and subsequent
radiograph), and other comments. Other comments included information concerning
reductions or surgical interventions.

Once digitized and cataloged, the radiographs were examined for evidence of
fi‘acture healing. Examination was performed by opening each image individually in
Adobe Photoshop CS2. The fracture healing was categorized into six stages, modified
from Hufnagl 2005. The characteristics of each of the stages were as follows:

Stage 1 - No healing: sharp fracture lines, absence of bridging and callus

formation.

Stage 2 — Granulation: beginning of resorption along fracture line, “fluffy” callus

formation, blurring of fi'acture line, absence of a complete mature callus

37

Stage 3 — Callus: mature callus formation around fracture site; callus bulges over
site and demonstrates a radiopaque appearance, fracture line visible but may be
blurred
Stage 4 — Bridging: fracture gap is connected across the fracture site in some, but
not all areas (less than 50%), blurring of the fracture line, callus may still be
present
Stage 5 — Clinical Union: fracture line is significantly blurred; fracture line is
connected in most areas (more than 50%), callus presence minimal
Stage 6 — Completion: No evidence of fracture line, callus presence minimal or
not observable.
Each radiograph was assigned a stage and this information was added to the Excel
worksheet. If there was any uncertainty between two stages, the radiograph was assigned
to the earlier stage. If the radiograph was not clear enough to identify a stage, it was

excluded fi'om the study.

Statistical Analyses

Stages

A database including the days of healing and the stage scored for each radiograph
was created for statistical analysis. All statistics were calculated using the Data Analysis
application in Excel 2003. One-way Analysis of Variance (ANOVA) was performed to
determine if there was a significant difference in the healing time among each of the six
stages. One-way AN OVA is typically used when comparisons of the mean need to be

made among three or more independent groups and one independent variable has been

38

manipulated (Salkind 2007). An F-ratio, comparing the between-group variability and the
within-group variability, is compared to the F-critical value to determine the significance
of the data. In this instance, AN OVA was used to compare the mean days of healing
(dependent variable) among each of the stages (independent variable) to determine
whether the stages would be useful for separating the sample. In addition to AN OVA, the
Bonferroni correction, a post hoc test of multiple comparison, was added to determine
where inequalities were present in the samples (Sirkin 2006). AN OVA was used to see if
there was a statistically significant difference among the stages, and the Bonferroni

correction was used to identify where the difference or differences are occurring.

Age

A database including the days of healing, the stage of healing, and the age of each
individual was created for statistical analysis. All statistics were calculated using the Data
Analysis application in Excel 2003. The statistics were used to determine if there was a
significant difference in the time of healing among the three age groups (0-1 years old, 2-
3 years old, and 4-5 years old) at each of the stages; therefore, enabling the researcher to
determine if differences among the healing times of the age groups are present and at
what point in the healing process, earlier verse later, they occurred. For each of the six
stages, One-way Analysis of Variance (ANOVA) with Bonferroni correction was
performed for the three age groups. In this instance, AN OVA was used to compare the
mean days of healing (dependent variable) in each stage among the three age groups

(independent variable). Additional AN OVA with Bonferroni correction analyses where

39

performed grouping the stages into three groups (Stages 1-2, Stages 3-4, and Stages 5-6)
and into two groups (Stages 1-3 and Stages 4-6).

In addition to the original three age groups, the sample was also divided into two
age groups: individuals from birth to two years old and individuals between three and
five years old. The statistics were used to determine if there was a significant difference
in the time of healing between the two age groups at each of the stages. For each of the
six stages, Welch’s t-test, used for unequal sample size and unequal variance, was
performed for the two age groups. An independent, unpaired t-test is used to examine
whether there is a difference between the means of two independent groups (Salkind
2007). The t-test allowed the mean days of healing (dependent variable) to be compared

between the two age groups (independent variable) for each of the stages.

Fracture Location

A second database including the days of healing, the stage of healing, and the
fracture location was created for statistical analysis. All statistics were calculated using
the Data Analysis application in Excel 2003. The statistics were used to determine if
there was a significant difference in the time of healing between the two fracture
locations (forearm and leg) at each of the stages; therefore, enabling the researcher to
determine if differences in healing times between the fracture locations exist and at what
point in the healing process, earlier verse later, the differences occurred. For each of the
six stages, Welch’s t-test was performed for the two fracture locations. The t—test allowed
the mean days of healing (dependent variable) to be compared between the two fiacture

location groups (independent variable) for each of the stages. Stages were re-categorized

40

into three groups (Stages 1-2, Stages 3-4 and Stages 5-6) and into two groups (Stages 1-3
and Stages 4-6). T-tests were also used when assessing these alternate stage

combinations.

4]

CHAPTER 7: RESULTS
Radiographic Atlas of Pediatric Fracture Healing
A main objective of this research was to develop a set of images that
demonstrated the radiographic features evident in pediatric fracture healing. Several
individuals will be discussed and illustrated here. Additional examples of each stage are
included in Appendix A (Figures 15-33).
The first example, Individual #10, was a five year old who suffered a forearm

fracture. Figures 6-10 demonstrate the radiographic healing process.

 

Figure 6 — Individual #10: Forearm fracture, Stage 1, 0 days healing

42

 

Figure 7 — Individual #10: Forearm fracture, Stage 1, 3 days healing

 

Figure 8 — Individual #10: Forearm fracture, Stage 2, 17 days healing

43

 

Figure 9 — Individual #10: Forearm fracture, Stage 3, 30 days healing

 

Figure 10 — Individual #10: Forearm fracture, Stage 3, 46 days healing

A second example is Individual #107, a five year old, who suffered a fracture to

the leg. Figures 11-14 demonstrate the radiographic healing process.

 

Figure 11 — Individual #107: Leg fracture, Stage 1, 6 days healing

 

Figure 12 — Individual #107: Leg fracture, Stage 2, 11 days healing

45

 

Figure 13 — Individual #107: Leg fracture, Stage 2, 28 days healing

 

Figure 14 - Individual #107: Leg fracture, Stage 3, 49 days healing

46

Stage Analysis

A statistical aim of this research was to determine the usefulness of categorizing
radiographs by the stages of healing presented. It was expected that the mean healing
times would increase across the stages. This expectation was confirmed; all differences
demonstrated among the mean healing times for each stage indicated increased healing
time with later stages, with Stage 1 showing the shortest mean healing time and Stage 6
indicating the longest mean healing time. The mean healing time in Stage 1 was 3.3 days,
with a standard deviation of 3.4. The mean healing time in Stage 2 was 21.0 with a
standard deviation of 10.5. The mean healing time in Stage 3 was 38.4 days, with a
standard deviation of 13.4. The mean healing time in Stage 4 was 43.9 days, with a
standard deviation of 15.2. The mean healing time in Stage 5 was 65.2 days, with a
standard deviation of 48.2. The mean healing time in Stage 6 was 313.3 days, with a
standard deviation of 235.7. These results are summarized in Table 2.

Table 2 — Stages and Healing Times

 

 

 

 

 

 

 

Mean healing time Range in healing time Standard
STAGE (days) (days) Deviation
Stage 1 3.3 0-14 3.4
Stage 2 21.0 4-50 10.5
Stage 3 38.4 15—75 13.4
Stage 4 43.9 24-93 15.2
Stage 5 65.2 24-156 48.2
Stage 6 313.3 42-750 235.7

 

 

 

 

 

 

47

Due to the clinical purpose of the radiographs, the images analyzed in this study
may have been taken at any point in the stage in which they were categorized. The
impossibility of knowing the exact time in the healing process when each stage is reached
limits the utility of looking at the ranges when assessing healing time. Individuals may
have demonstrated healing times on either side of the mean healing time, but the range of
possible values for each stage may not have been depicted by the sample. In order to
apply such data to other individuals, the standard deviations, indicative of the dispersion
of the data, should be taken into account.

Analysis of Variance (AN OVA) was performed to determine if statistical
differences in the mean healing time were present among the stages of healing (Appendix
B, Table 3). An F value of 100.08 and a p value of less than 0.0001 were obtained. These
results indicate that the differences in the healing time among the stages were statistically
significant at the 0.0001 level and null hypothesis 1 should be rejected. To determine the
extent of the difference in mean healing time, the Bonferroni correction was added to the
ANOVA analysis. At a 95% confidence interval, Stage 1 was statistically significant
from all other stages. Stage 2 was statistically significant from Stages 1, 4, 5 and 6, but
was not statistically significant from Stage 3. Stage 3 was statistically significant from
Stages 1 and 6, but not statistically significant from Stages 2, 4 and 5. Stage 4 was
statistically significant from Stages 1, 2 and 6, but not statistically significant from Stages
3 and 5. Stage 5 was statistically significant from Stages 1, 2 and 6, but not statistically
significant from Stages 3 and 4. Stage 6 was statistically significant from all other stages.
The results indicate statistical differences were demonstrated between some, but not all

stages (Appendix B, Table 3). The variations in mean healing time were most clearly

48

defined at the beginning and end of the healing process, as evident from Stages 1 and 6

demonstrating statistical significance from all other stages.

Age Analysis

Another goal of the research was to determine the influence of the patient’s age
on the rate of the healing process. Three age groups (0-1 years, 2-3 years, and 4-5 years
old) were compared at each of the six stages using AN OVA. It was expected that younger
individuals would have shorter mean healing times than older individuals. This
expectation was confirmed at some, but not all, stages. At Stage 1, an F-value of 3.67 and
p value of 0.0292 were obtained, indicating the mean healing times were statistically
significant among the age groups at the 0.05 level. The Bonferroni correction indicated
that the mean healing time of individuals ages 0-1 year old was statistically significant
fi'om the mean healing time of individuals ages 4-5 years old at the 0.05 level with the
younger individuals having a shorter mean healing time. Individuals ages 0-1 years and
2-3 years and individuals 2-3 years and 4-5 years were not statistically significant from
one another (Appendix B, Table 4). At Stage 2, an F-value of 0.55 and a p value of
0.5776 were obtained, indicating the mean healing times were not statistically significant
among the age groups (Appendix B, Table 5). At Stage 3, an F-value of 0.61 and a p
value of 0.5459 were obtained, indicating the mean healing times were not statistically
significant among the age groups (Appendix B, Table 6). At Stage 4, an F-value of 0.29
and a p value of 0.7537 were obtained, indicating the mean healing times were not
statistically significant among the age groups (Appendix B, Table 7). Due to the small

amount of radiographs demonstrating healing in Stages 5 and 6, the healing times were

49

combined into one group. At Stages 5-6, an F-value of 7.11 and a p value of 0.0082 were
obtained, indicating the mean healing times were statistically significant among the age
groups. The Bonferroni correction indicated that the mean healing time of individuals
ages 0-1 years was statistically significant from the mean healing time of individuals ages
4-5 years at the 0.05 level. Additionally, the mean healing time of individuals ages 2-3
years was statistically significant from the mean healing times of individuals ages 4-5
years at the 0.05 level. In both of these cases, the younger group had a shorter mean
healing time. The mean healing times of individuals 0-1 year and individuals 2-3 years
were not statistically significant form one another (Appendix B, Table 8). These results
indicate age influenced the mean healing time at Stage 1 and Stages 5-6 (Appendix B,
Table 9). Null hypothesis 2 may be rejected as differences were seen in the mean healing
times among the three age groups. Null hypotheses 2a, 2e and 2f may be rejected at the
95% confidence level, whereas null hypotheses 2b, 2c and 2d should be accepted. When
statistically significant results were present, younger individuals demonstrated a shorter
mean healing time than older individuals. Additionally, statistically significant
differences arose at the beginning and end of the healing process.

To further assess the influence of age on the healing process and to expand
sample size, the stages were grouped into sets of two, dividing the healing process into
beginning (Stages 1-2), middle (Stages 3-4) and end (Stages 5-6). Three age groups (0-1
years, 2-3 years, and 4-5 years old) were compared at each of these three stages using
ANOVA. At Stages 1-2, an F -va1ue of 2.18 and p value of 0.1 156 were obtained,
indicating the mean healing times were not statistically significant among the age groups

(Appendix B, Table 10). At Stages 3-4, an F -value of 0.52 and a p value of 0.5974 were

50

obtained, indicating the mean healing times were not statistically significant among the
age groups (Appendix B, Table 11). Stages 5-6 were previously discussed due to small
sample size (Appendix B, Table 8). The results of combining the stages into three groups
indicate age influenced the mean healing time only at Stages 5-6 (Appendix B, Table 12).

Stages were also grouped into a beginning (Stages 1-3) and end (Stages 4-6) of
the healing process. Three age groups (0-1 years, 2-3 years, and 4-5 years old) were
compared at each of these two stages using ANOVA. At Stages 1-3, an F -value of 0.84
and p value of 0.4330 were obtained, indicating the mean healing times were not
statistically significant among the age groups (Appendix, Table 13). At Stages 4-6, an F-
value of 1.62 and a p value of 0.2093 were obtained, indicating the mean healing times
were not statistically significant among the age groups (Appendix B, Table 14). These
results indicate a limited role of age in the fracture healing process (Appendix B, Table
15).

Due to the small sample size of later stages, age was also assessed by grouping
the sample into two age groups instead of the original three age groups. Individuals age
0-2 years old were compared with individuals age 3—5 years old through a t-test at each of
the six stages. At Stage 1, a t-value of -2.35 and a t-critical value of 1.66 were obtained,
indicating the mean healing times were statistically significant between the age groups at
the 0.01 level. Individuals ages 0-2 years had a shorter mean healing time than older
individuals (Appendix B, Table 16). At Stage 2, a t-value of 0.128 and a t-critical value
of 1.67 were obtained, indicating the mean healing times were not statistically significant
between the age groups (Appendix B, Table 17). At Stage 3, a t-value of -0.40 and a t-

critical value of 1.69 were obtained, indicating the mean healing times were not

51

statistically significant between the age groups (Appendix B, Table 18). At Stage 4, a t-
value of 0.27 and a t-critical value of 1.77 were obtained, indicating the mean healing
times were not statistically significant between the age groups (Appendix B, Table 19).
At Stage 5, a t-value of -5.97 and a t-critical value of 2.92 were obtained, indicating the
mean healing time of 0-2 year olds was shorter than the mean healing time of 3-5 year
olds. This finding was statistically significant at the 0.01 level (Appendix B, Table 20).
At Stage 6, a t-value of -2.53 and a t-critical value of 2.13 were obtained, indicating the
mean healing time of 0-2 year olds was shorter than the mean healing time of 3-5 year
olds. This finding was statistically significant at the 0.05 level (Appendix B, Table 21).
These results indicate age influenced the mean healing time at Stage 1, Stage 5 and Stage
6 (Appendix B, Table 22). Again, statistically significant differences were present at the
beginning and end of the healing process.

As previously assessed with the three age groups, the healing stages were also
divided into a beginning (Stages 1-2), middle (Stages 3-4) and end (Stages 5-6).
Individuals age 0-2 years old were compared with individuals age 3-5 years old through a
t-test at each of these three stages. At Stages 1-2, a t-value of -1.67 and a t-critical value
of 1.66 were obtained, indicating the mean healing times were statistically significant
between the age groups at the 0.05 level. Individuals ages 0-2 years had a shorter mean
healing time than older individuals (Appendix B, Table 23). At Stages 3-4, a t-value of -
0.14 and a t-critical value of 1.68 were obtained, indicating the mean healing times were
not statistically significant between the age groups (Appendix B, Table 24). At Stages 5-
6, a t-value of -2.63 and a t-critical value of 1.89 were obtained, indicating the mean

healing time of 0-2 year olds was shorter than the mean healing time of 3-5 year olds.

52

This finding was statistically significant at the 0.05 level (Appendix B, Table 25). These
results indicate age influenced the mean healing time at the beginning (Stages 1-2) and
end (Stages 5-6) of the healing process (Appendix B, Table 26).

The original six stages were also split into two groups: beginning (Stages 1-3) and
end (Stages 4-6) of the healing process. Individuals age 0-2 years old were compared
with individuals age 3-5 years old through a t-test at each of these stages. At Stages 1-3, a
t-value of -1.41 and a t-critical value of 1.65 were obtained, indicating the mean healing
times were not statistically significant between age groups (Appendix B, Table 27). At
Stages 4-6, a t-value of -l .68 and a t-critical value of 1.69 were obtained, indicating the
mean healing times were not statistically significant between age groups (Appendix B,
Table 28). These results indicate that age doe not influence the healing rate throughout

the healing process (Appendix B, Table 28).

Fracture Location Analysis

A final statistical goal of the research was to determine the influence of the
fracture location on the rate of the healing process. Two fracture locations, forearm and
leg, were compared at each of the six stages through t-tests. It was expected that forearm
fractures would have a shorted mean healing time when compared to leg fractures. At
Stage 1, a t-value of 1.10 and a t-critical value of 1.66 were obtained, indicating the mean
healing times were not statistically significant between the fracture locations (Appendix
B, Table 30). At Stage 2, a t-value of -3.95 and a t-critical value of 1.67 were obtained,
indicating the mean healing time of forearm fractures was shorter than the mean healing

time for leg fractures. This finding was statistically significant at the 0.01 level

53

(Appendix B, Table 31). At Stage 3, a t-value of -l.72 and a t-critical value of 1.68 were
obtained, indicating the mean healing time of forearm fractures was shorter than the mean
healing time for leg fractures. This finding was statistically significant at the 0.05 level
(Appendix B, Table 32). At Stage 4, a t-value of 0.17 and a t-critical value of 1.81 were
obtained, indicating the mean healing times were not statistically significant between
fracture locations (Appendix B, Table 33). At Stage 5, a t-value of -0.42 and a t-critical
value of 2.35 were obtained, indicating the mean healing times were not statistically
significant between fracture locations (Appendix B, Table 34). At Stage 6, a t-value of
0.09 and a t-critical value of 2.35 were obtained, indicating the mean healing times were
not statistically significant between fracture locations (Appendix B, Table 35). These
results indicate fracture location influenced the mean healing time at Stages 2 and 3
(Appendix B, Table 36). Null hypothesis 3 may be rejected due to the differences
indicated between the mean healing times of forearm fracture and leg fractures. Null
hypotheses 3b should be rejected at the 0.01 level, and null hypothesis 3c should be
rejected at the 0.05 level. Null hypotheses 3a, 3d, 3e and 3f should be accepted.
Additionally, when significant differences were found, forearm fiactures healed at an
increased rate compared to leg fractures.

The stages were combined into a beginning (Stages 1-2), middle (Stages 3-4) and
end (Stages 5-6) of the healing process. The two fracture locations were compared at
each of these three stages through t-tests. At Stages 1-2, a t-value of -2.59 and a t-critical
value of 1.66 were obtained, indicating the mean healing times were shorter for
individuals with forearm fractures than individuals with leg fractures. This finding was

statistically significant at the 0.01 level (Appendix B, Table 37). At Stages 3-4, a t-value

54

of -l .04 and a t-critical value of 1.67 were obtained, indicating the mean healing times
were not statistically significant between fracture locations (Appendix B, Table 38). At
Stage 5-6, a t-value of -0.09 and a t-critical value of 1.80 were obtained, indicating the
mean healing times were not statistically significant between fracture locations
(Appendix B, Table 39). These results indicate fiacture location influenced the mean
healing time at the beginning (Stages 1-2) of the healing process (Appendix B, Table 40).
Additionally, when significant differences were found, forearm fractures healed at an
increased rate compared to leg fractures.

The original six stages were also split into two groups: beginning (Stages 1-3) and
end (Stages 4-6) of the healing process. The two fracture locations were compared at
each of these stages. At Stages 1-3, a t-value of —1.59 and a t-critical value of 1.65 were
obtained, indicating the mean healing times were not statistically significant between
fracture locations (Appendix B, Table 41). At Stages 4—6, a t-value of -0.20 and a t-
critical value of 1.71 were obtained, indicating the mean healing times were not
statistically significant between fracture locations (Appendix B, Table 42). These results
indicate fracture location does not influence the mean healing rate during the fracture

repair process (Appendix B, Table 43).

55

CHAPTER 8: DISCUSSION

The six stages examined in this study demonstrate a timeline of fracture healing in
children. The images of each of these stages present radiographic examples of the
appearance of fracture lines, fiacture bridging, and fracture callus formation throughout
the healing process. Future radiographs may be compared to these images and the stage
descriptions to obtain an estimate of healing time in young children and infants from birth
to five years old.

Each of the six stages demonstrated an increased healing time when compared to
the previous stages and overall the six stages were significantly different. When
individually comparing stages, mean healing times were significantly different for most
stages. Certain stages (Stages 2 and 3, Stages 3 and 4, Stages 3 and 5, and Stages 4 andS)
did not demonstrate statistical significance from one another. It was expected that all
stages would show differences from one another; therefore, the stages lacking statistical
significance may have a commonality that prevented identifying each as a unique
radiographic stage. These stages were all close together in the healing process and
demonstrated an overlap in the ranges of healing times. The reasons the individual stages
lacked significance could include natural variation in the fracture healing once a fracture
healing has begun, thus indicating the importance of the initial fracture response when
examining fracture healing. Altemately, the limited sample size, especially in later stages,
may have prevented differences from being accurately depicted. A third reason for the
lack of significance is that healing in the latter aspect of one stage would be difficult to
differentiate from healing in the earlier part of the subsequent stages. With experience the

researcher could identify, for instance, a late Stage 2. This may suggest that the stages

56

themselves were too broad, and it would be necessary to further narrow the stages based
on additional radiographic features. Future research, with a larger sample size, could
confirm or refute this possibility. Finally, the lack of significance may be due to other
contributing factors influencing the healing process. Nonetheless, radiographic features
present in the six stages should be considered when analyzing radiographic images, but
an examiner should be cautious when employing a healing time estimate. Such discretion
is particularly imperative when differentiating the middle stages of fracture healing and
when an examiner has limited experience analyzing radiographs.

The patient’s age did influence the fracture healing process and all statistically
significant differences demonstrated increased mean healing time with increased age.
When the sample was divided into three age groups, this finding was only significant at
Stages 1 and 5-6. The difference indicated between the age groups, suggests that there
may be a certain point in the development process where fracture healing may slow,
possibility around the age of three. To further examine this result, age groups should be
expanded and older children and adolescents included. The lack of statistical significance
in the middle of the healing process may be a reflection of contributing factors or due to
the original lack of significance between certain stages, as previously discussed.

When the stages were grouped together into beginning, middle and end,
statistically significant differences in the mean healing time were only present at Stages
5-6. By combing the stages into only two groups, no statistically significant differences
were present. The decreased significance of the results when the stages were combined
suggests the uniqueness of each fracture healing stage and furthers the hypothesis that an

increased number of stages may lead to increased statistical significance. If the stages

57

were not unique categories, the regrouping of the stages may have increased the
significance; instead the significance decreased.

When the sample was divided into two age groups, age also influenced the
healing process. This finding suggested that age influences healing at Stages 1, 5 and 6,
though statistical significance was greater at the beginning of the healing process. Stages
2, 3 and 4 remained unaffected by age. While age differences in healing were expected at
all stages, these stages are where the majority of healing in occurring. While increased
age may prevent Stage 2 from beginning, as evident by the increased healing times in
older individuals in Stage 1, age may not play an active role in the osseous reaction
characteristic of Stage 2, the callus formation of Stage 3 or the bridging of the callus in
Stage 4. Another possibility is a certain factor, unrelated to age, may be delaying or
speeding up the osseous reaction that occurs in these stages. If there was a contributing
factor present largely at Stages 2, 3 or 4, but not necessary at other stages, it could
prevent the influence of age being demonstrated at this time in the healing process.
Alternatively, this finding may be a reflection of the original significance of the stages.

When healing stages were combined, the significance of the results declined.
With beginning (Stages 1-2), middle (Stages 3-4) and end (Stages 5-6) stages, the results
were still significant at the beginning and end of the healing process, but at lower
confidence levels. When the stages were divided into two groups, confidence levels were
lowered even more though still suggesting age may be a factor in the healing times. Once
again, this finding confirmed the uniqueness of the stages, as more confidence was
expressed in the statistically significant differences found when analyzing the original six

stages.

58

To completely understand the influence of age in the fracture healing process, it is
also necessary to examine the rates evident in adult fracture healing. Hufnagl (2005) used
similar stages to assess a sample that ranged in age from two to 93 years. Using
comparable stages, the mean healing times were as follows: Stage 1: 0.22 days; Stage 2:
22.37 days; Stage 3: 79.42 days; Stage 4: 116.96 days; Stage 5: 124.20 days; and Stage 6:
260.86. Hufiragl’s sample of 62 individuals included three equal age groups (2-10, 11—45,
and 45 and older) with results indicating statistically significant differences among the
age groups. It can be assumed that the mean healing times given would have been skewed
towards individuals over the age of five. Comparing such healing times to the present
study demonstrates some differences. Stage 1 and 2 demonstrated healing times within a
couple days of each other. Stages 3, 4 and 5 were shortened by approximately one half in
the present study of young children, whereas Stage 6 demonstrated a longer mean healing
time. Stages 3, 4 and 5 may be demonstrating the stages in fiacture repair where age is
most influential when comparing adults to children. When looking more precisely at
children in the present study, differences were found earlier in the healing process.
Additionally, while Stage 6 would demonstrate a deviation fiom expectation, in both
studies Stage 6 was the most limited in number of individuals. In this regard, Stage 6
warrants further examination when comparing across samples and methodologies.
Additionally, more samples of all ages should be examined in a similar manner to obtain
accurate comparisons across certain groups.

Fracture location also influenced the healing time in the sample examined. This
finding was evident at Stages 2: Granulation and Stage 3: Mature Callus. When location

did affect the healing time, forearm fi’actures healed faster then leg fractures. While it was

59

expected such a finding would be consistent across the stages, there are several reasons
why the mean healing time in Stages 2 and 3 may have been influenced by the fiacture
location. Stages 2 and 3 are the first stages where osseous reaction in occurring. Fracture
location may be a particularly important factor for the initiation of fracture healing, with
diminished impact after healing has begun. Additionally, the other stages may be more
likely to encounter other contributing effectors. For instance, another important factor
cited as impacting healing time is mobility. During the beginning stages of fiacture
repair, mobility may be limited due to discomfort or internal or external stabilization. In
later stages, mobility would increase and could become more influential in the fracture
healing rate. Furthermore, the age of an individual may be an indirect factor. While in
adults there may be significant differences in the mobility of upper and lower limbs when
injury is sustained, these variations in movement may not be as pronounced in young
children and infants. Additionally, in adults the vascularity of bone is decreased when
compared to children (Odgen 2000). This difference, while obviously differentiating
children from adults, may also influence the effects of fracture location on healing time.
For instance, the tibia is known for its long healing times; a feature that may be linked to
the diminished vascularity of the bone. With children’s increased vascularity, the tibia
may heal significantly faster in children compared to adults, but when comparing
children with children the expected difference may be diminished. At the same time,
fracture location did impact healing time in these children, demonstrating that fracture
location may impact healing rates regardless of patient age. It is important to realize the
connection between fracture location and age, as fracture location may demonstrate an

increased impact at a certain point in the development process.

60

Several aspects of the present study warrant further examination. The limited
sample size of radiographs in later stages of healing was problematic. Radiographs are
most commonly used and examined at the beginning of the fracture healing process.
Once a patient feels better, he or she may not return for subsequent radiographs (Ogden
2000). Additionally, once significant healing is shown, it may not be necessary to
continually radiograph the fracture site. Overall, the need for radiographs is determined
on a clinical basis which may not be conducive to research interests. The original intent
of the radiographs should be acknowledged as a limitation of the study.

Stage 6 may have led to false confidence in the statistical differences between
Stage 5 and 6. Any radiograph examined after healing had taken place would have been
categorized as a Stage 6. In other words, any radiograph, whether taken days or years
later, after Stage 5 would be a Stage 6. There is no limit to the end of Stage 6, so the data
would be more likely to indicate statistically significant differences if radiographs were
taken after an individual had previously reached Stage 6. A remedy for this dilemma in
future research may be to limit the days healing for radiographs examined, but
acknowledge that Stage 6 would continue indefinitely.

Radiograph quality and the presence of external stabilization (casts), especially
around the beginning of the healing process limited the ability to accurately categorize
the sample into stages. If external stabilization or radiograph quality limited the ability to
see a stage it was excluded, but if the researcher was able to identify whether it was either
of two stages, the radiograph was classified as the earlier of the stages. For instance, what
was identified as a late Stage 2/early Stage 3 would have been classified as a Stage 2.

Therefore, an early Stage 3, obscured by casting material, may have been incorrectly

61

identified as a Stage 2. In future research, a sample devoid of external stabilization would
be ideal, though such a sample would be difficult to gather due to the clinical need for
fracture stabilization. Additionally, in a real world situation, a fracture in the beginning of
the healing process would likely be casted. The limitation of external stabilization in
radiographs should be acknowledged as a limitation of this research.

Another potential concern for the present research is the possibility of other
contributing direct or indirect factors influencing the rate of the fracture healing process.
Of particular concern for the present research would be the type of fiacture encountered.
While not a goal of this research, the type of fi'acture should be considered when
assessing healing time in future research. When examining the ranges of the stages in the
present study, the fractures with healing times at the beginning of the range tended to be
incomplete, particularly torus, fractures. An incomplete fracture would be less traumatic
than a complete fracture, and it is possible this factor was influencing the mean healing
times examined when comparing ages and fracture location. In a sample of children, it is
particularly important to determine the influence of incomplete fractures as they are a
more common occurrence in this age group. Finally, while unusual instances of fracture
repair were limited by the exclusion of co-morbidity, there may have been other factors
or medical conditions which the researcher was not aware of that may have influenced
healing rates. Great significance should not be placed on one individual demonstrating a

delayed or increased healing rate.

62

CHAPTER 9: CONCLUSION

Fracture healing can demonstrate a unique set of radiographic features that can be
monitored throughout the healing process. The fracture repair process can be influenced
by a number of factors. The goal of the present study was twofold. First, the research
aimed to identify a set of radiographic stages of healing for infants and young children. A
set of six radiographic stages of fracture healing was developed based on key
radiographic features and past studies. A second goal was to determine how fi'acture
location and patient age may influence the healing times of these radiographic stages. A
sample of 107 individuals and 294 radiographs were examined to meet these goals. The
sample consisted of forearm and leg fractures in children from birth to age five. The
sample was separated into three age groups: birth to one year old, two to three years old,
and four to five years old. Each radiograph was then placed into one of the six
radiographic stages of fracture healing. Unpaired t-tests were used to compare fracture
location at each of the six stages and Analysis of Variance (ANOVA) was used to
compare patient age at each of the stages. These tests identified how fracture location and
patient age influenced the mean healing time at each radio graphic stage of fracture
healing.

It was expected that the stages used would demonstrate increased healing time
with later stages. While this finding was confirmed, stages in the middle of the healing
process overlapped in their ranges and had limited statistical significance. The stages
were usefiil in categorizing the sample and demonstrate the utility of assessing
radiographic features in the fi'acture healing process. The limited significance of certain

stages from one another in the middle of the healing process suggest caution should be

63

used when estimating healing times. Additionally, future research should examine
whether the radiographic stage of fracture healing may be refined into a larger number of
more specific stages.

Through the analysis, the present study identified patient age as a factor
influencing the healing rate of fractures in children. The differences identified in the
mean healing time among the age groups suggests that fracture healing occurs at an
accelerated rate in younger individuals. While this finding was not acknowledged at all
stages of healing, the beginning and end of the fracture repair process were identified as
key instances where age may be involved. The middle of the fracture healing process, a
time when a large amount of osseous repair is occurring, may not be influenced by age in
children. Instead age may play more of a role in delaying the start of the healing, as
evident by the influence of age at Stage 1. Age may also demonstrate an increased
influence on the fracture healing process, and perhaps later in the process, when children
are compared with adults.

Fracture location was also identified as a factor involved in the rate of the repair
of skeletal trauma. The variations in mean healing time between forearm and leg fractures
indicated that forearm fractures heal at an accelerated rate. This finding was identified in
Stages 2 and 3 of the healing process, and may indicate an increased role in the initiation
and osseous reaction of fracture healing. Contributing factors, such as age and fiacture
site mobility, may influence the effects of fracture location on the rate of the healing
process.

Additional analysis of both patient age and fracture location by combining the

stages of healing also has implications for the radiographic stages of fracture healing.

While statistically significant differences were still demonstrated when healing stages
were grouped together, the level of confidence was reduced with combination,
substantiating the uniqueness of each of the stages. In firture studies, the stages should not
be combined, instead further separation should be encouraged.

This thesis provides statistical evidence of patient age and fracture location
impacting the fracture healing process. It also provides a method to assess pediatric
fractures through radiographic analysis and estimate a time since injury. While this thesis
supports the use of radiographic stages of fracture healing in young children and infants,
more research is needed to confirm and reassess the implications of age and fracture
location. Age and fracture location themselves may influence other factors in the healing
process. For instance, consider other factors that may accompany age. In children, the
types of skeletal trauma will differ with an increased prevalence of incomplete fiactures
(Lewis 2007). The healing rates between complete and incomplete fractures would likely
differ. Additionally, the connection between fracture location and age is important to
understand, as certain age groups are more likely to suffer fractures in specific locations
(Jones 1994). The present study attempts to address age and fiacture location in a broad
sense, but if other factors in fracture healing can be identified a more precise timeline
could be developed for radiographically assessing fractures.

The use of radiographic fracture healing stages has implications for the field of
forensic anthropology. Skeletal remains may be assessed through visual examination, but
may also be radiographed to elicit more information on antemortem trauma. Increasingly,
it is becoming important to approach trauma, especially in children, with a multifaceted,

integrated methodology. To facilitate a complete understanding and examination of

65

childhood trauma, forensic anthropology must incorporate clinical literature from
pediatric biology and radiography. The in-depth inquiries required to assess trauma in
children result in a demand to incorporate research methods to gain as much information
as possible. While challenges loom and create a daunting future for the research in
childhood trauma, the knowledge to be gained through extensive examination is

priceless.

66

APPENDIX A

67

Radiographic Atlas of Pediatric Fracture Healing
Stage 1 — No healing: sharp fracture lines, absence of bridging and callus

formation.

 

Figure 15 — Individual #8: Age 1, Forearm fracture, 0 days healing

 

Figure 16 — Individual #17: Age 4, Forearm fracture, 5 days healing

68

 

Figure 17 - Individual #92: Age 1, Leg fracture, 1 day healing

 

Figure 18 — Individual #109: Age 5, Leg fracture, 0 days healing

69

Stage 2 - Granulation: beginning of resorption along fracture line, “fluffy” callus

formation, blurring of fracture line, absence of complete mature callus.

 

Figure 19 — Individual #14: Age 4, Forearm fracture, 27 days healing

 

Figure 20 — Individual #74: Age 1, Forearm fracture, 7 days healing

70

 

Figure 21 — Individual #67: Age 2, Leg fracture, 54 days healing

 

Figure 22 — Individual #75: Age 5, Leg fracture, 39 days healing

7l

Stage 3 — Callus: mature callus formation around fracture site; callus bulges over
site and demonstrates a radiopaque appearance, fracture line visible but may be

blurred.

 

Figure 23 — Individual #47: Age 5, Forearm fracture, 50 days healing

 

Figure 24 — Individual #86: Age 1, Forearm fracture, 75 days healing

72

 

Figure 25 — Individual #98: Age 3, Leg fracture, 42 days healing

 

Figure 26 - Individual #100: Age 5, Leg fracture, 40 days healing

73

Stage 4 - Bridging: fracture gap is connected across the fracture site in some, but not all

areas (less than 50%), blurring of the fracture line, callus may still be present.

 

Figure 27 — Individual #12: Age 3, Forearm fracture, 51 days healing

 

Figure 28 — Individual #35: Age 5, Forearm fracture, 40 days healing

74

 

Figure 29 — Individual #26: Age 5, Leg fracture, 63 days healing

75

Stage 5 — Clinical Union: fracture line is significantly blurred; fracture line is

connected in most areas (more than 50%), callus presence minimal.

 

Figure 30 — Individual #7: Age 0, Forearm fracture, 49 days healing

 

Figure 31 —- Individual #49: Age 3, Forearm fracture, 111 days healing

76

Stage 6 — Completion: No evidence of fracture line, callus presence minimal or

not observable.

 

Figure 32 — Individual #40: Age 5, Forearm fiacture, 418 days healing

 

Figure 33 - Individual #1 15: Age 2, Leg fracture, 188 days healing

77

APPENDIX B

78

Table 3 — Analysis of Variance with Bonferroni Correction for Six Stages

79

l-way AN OVA
11 294
Groups n Mean SE Pooled SE SD
STAGE 1 99 3.3 0.34 3.65 3.4
STAGE2 91 21.0 1.11 3.81 10.5
STAGE 3 58 38.4 1.76 4.77 13.4
STAGE 4 30 43.9 2.78 6.63 15.2
STAGE 5 9 65.2 16.07 12.11 48.2
STAGE 6 7 313.3 89.09 13.73 235.7
Mean
Source of variation Sum squares DF square F statistic p
Groups 6604644 5 1320929 100.08 <0.0001
Residual 3801112 288 1319.8
Total 10405756 293
Bonferroni
Contrast Difference 95% Cl
STAGE 1 V STAGE 2 -17.6 -333 to .20 (significant)
STAGE 1 V STAGE 3 -351 -529 to -173 (significant)
STAGE 1 V STAGE 4 -405 -63.0 to -18.1 (significant)
STAGE 1 V STAGE 5 -61.9 -99.3 to -24.5 (significant)
STAGE 1 V STAGE 6 -3100 -3520 to -267.9 (significant)
STAGE 2 v STAGE 3 -175 -355 to 0.6
STAGE 2 V STAGE 4 -229 -455 to 03 (significant)
STAGE 2 V STAGE 5 -443 -81.8 to -6.7 (significant)
STAGE 2 V STAGE 6 -292.3 -334.5 to 2501 (significant)
STAGE 3 v STAGE 4 -5.4 -29.6 to 18.7
STAGE 3 v STAGE 5 -26.8 -65.3 to 11.7
STAGE 3 V STAGE 6 —274.9 -317.9 to -231.8 (significant)
STAGE 4 v STAGE 5 -214 -62.2 to 19.5
STAGE 4 V STAGE 6 -269.4 -314.6 to -2243 (significant)
STAGE 5 V STAGE 6 -248.1 -3023 to -1939 (significant)

Table 4 — Analysis of Variance with Bonferroni Correction for Three Age Groups with

Six Stages: Stage 1

 

 

 

 

l-way ANOVA
n | 99
Pooled
STAGE 1 11 Mean SE SE SD
0-1 30 2.4 0.54 0.61 3.0
2-3 24 2.7 0.57 0.68 2.8
4-5 45 4.3 0.57 0.50 3.8
Source of Sum Mean F
variation squares DF square statistic J)
STAGE 1 81.7 2 40.9 3.67 0.0292
Residual 1069.9 96 1 l .1
Total 1151.7 98
Bonferroni
Contrast I Difference 95% CI
0-1 v 2-3 -0.3 -2.5 to 1.9
0-1 v 4-5 -1.9 -3.9 to -0.0 (significant)
2-3 v 4-5 -1.6 -3.7 to 0.4

Table 5 - Analysis of Variance for Three Age Groups with Six Stages: Stage 2

 

 

 

 

 

1-way ANOVA
n l 91
Pooled
STAGE 2 n Mean SE SE SD
0-1 17 19.9 2.61 2.57 10.8
2-3 25 22.8 1.91 2.12 9.6
4-5 49 20.4 1.58 1.51 11.0
Source of Sum Mean F
variation squares DF square statistic p
STAGE 2 124.1 2 62.1 0.55 0.5776
Residual 9890.8 88 112.4
Total 10014.9 90
Bonferroni
Contrast | Difference 95% CI
0-1v 2-3 -3.0 -11.1 to 5.2
0-1 v 4-5 -0.5 -7.8 to 6.8
2-3 v 4-5 2.5 -3.9 to 8.8

80

Table 6 - Analysis of Variance with Bonferroni Correction for Three Age Groups with

Six Stages: Stage 3

 

 

 

l-way ANOVA
n l 58
Pooled
STAGE 3 11 Mean SE SE SD
0-1 16 36.4 3.89 3.38 15.6
2-3 14 36.6 4.14 3.62 15.5
4-5 28 40.5 2.10 2.56 11.1
Source of Sum Mean F
variation squares DF square statistic p
STAGE 3 224.1 2 112.1 0.61 0.5459
Residual 10070.1 55 183.1
Total 10294.2 57
Bonferroni
Contrast Difference 95% Cl
0-1 v 2-3 -0.2 -12.4 to 12.0
0-1 v 4-5 -4.0 -14.5 to 6.4
2-3 v 4-5 -3.8 -14.8 to 7.1

 

81

Table 7 — Analysis of Variance with Bonferroni Correction for Three Age Groups with

Six Stages: Stage 4

 

 

 

 

l-way AN OVA
11 I 30
Pooled
STAGE 4 It Mean SE SE SD
0-1 5 47.4 11.51 6.97 25.7
2—3 10 41.2 3.56 4.93 11.3
4-5 15 44.5 3.63 4.03 14.1
Source of Sum Mean F
variation squares DF square statistic p
STAGE 4 138.9 2 69.5 0.29 0.7537
Residual 6564.5 27 243.1
Total 6703.5 29
Bonferroni
Contrast Difference 95% CI
0-1 V 2-3 6.2 -15.6 to 28.0
0-1 v 4-5 2.9 -l7.6 to 23.5
2-3 v 4-5 -3.3 -19.5 to 13.0

82

Table 8 - Analysis of Variance with Bonferroni Correction for Three Age Groups with

Six Stages: Stage 5-6

 

 

 

 

 

l-way ANOVA
n l 16
Pooled
STAGE 5-6 11 Mean SE SE SD
0-1 5 64.8 28.37 66.09 63.4
2-3 6 93.0 28.16 60.33 69.0
4-5 5 379.6 110.46 66.09 247.0
Source of Sum Mean F
variation squares DF square statistic p
STAGE 5-6 3103450 2 1551725 7.11 0.0082
Residual 2838980 13 21838.3
Total 5942430 15
Bonferroni
Contrast Difference 95% CI
0-1 v 2-3 -28.2 -273.9 to 217.5
0-1 v 4-5 -314.8 -571.4 to -58.2 (significant)
2-3 v 4-5 -286.6 -532.3 to -40.9 (significant)

Table 9 — Summary of Analysis of Variance for Three Age Groups with Six Stages and

 

 

 

 

 

 

 

 

 

 

Mean Healing Times
Mean Healing Time (days)
Statistically Age 0-1 Age 2-3 Age 4-5
Significant
Stage 1 0.05 2.4 2.7 4.3
Stage 2 Not significant 19.9 22.8 20.4
Stage 3 Not significant 36.4 36.6 40.5
Stage 4 Not significant 47.4 41.2 44.5
Stages 5-6 0.01 64.8 93.0 379.6

 

 

 

 

83

 

Table 10 — Analysis of Variance with Bonferroni Correction for Three Age Groups with

3 Stages: Stages 1-2

 

 

 

1-way ANOVA
n | 190
Pooled
STAGES 1-2 n Mean SE SE SD
0-1 47 8.7 1.59 1.70 10.9
2-3 49 13.0 1.77 1.66 12.4
4-5 94 12.7 1.20 1.20 11.6
Source of Sum Mean F
variation squares DF square statistic J)
STAGES 1-2 591.5 2 295.7 2.18 0.1156
Residual 25335.8 187 135.5
Total 25927.3 189
Bonferroni
Contrast Difference 95% CI
0-1 v 2-3 -4.3 -10.0 to 1.5
0-1 v 4-5 -4.0 -9.0 to 1.0
2-3 v 4-5 0.3 -4.7 to 5.2

 

84

Table 11 — Analysis of Variance with Bonferroni Correction for Three Age Groups with

3 Stages: Stages 3-4

 

 

 

1-way ANOVA
n I 88
Pooled
STAGES 3-4 n Mean SE SE SD
0-1 21 39.0 4.01 3.12 18.4
2-3 24 38.5 2.82 2.92 13.8
4-5 43 41.9 1.86 2.18 12.2
Source of Sum Mean F
variation squares DF square statistic p
STAGES 3-4 211.8 2 105.9 0.52 0.5974
Residual 17370.1 85 204.4
Total 17581.9 87
Bonferroni _
Contrast Difference 95% CI
0-1 v 2-3 0.5 -9.9 to 10.9
0-1 v 4-5 —2.8 -12.1 to 6.5
2-3 v 4-5 -3.3 -12.2 to 5.6

 

Table 12 — Summary of Analysis of Variance for Three Age Groups with Three Stages

 

 

 

 

 

 

 

 

and Mean Healing Times
Mean Healing Time (days)
Statistically Age 0-1 Age 2-3 Age 4-5
Significant
Stages 1-2 Not significant 8.7 13.0 12.7
Stage 3-4 Not Sigrificant 39.0 38.5 41.9
Stages 5-6 0.01 64.8 93.0 379.6

 

 

 

85

 

Table 13 — Analysis of Variance with Bonferroni Correction for Three Age Groups with

Two Stages: Stages 1-3

 

 

l-way ANOVA
n l 248
Pooled
STAGES 1-3 n Mean SE SE SD
0-1 63 15.7 2.16 2.09 17.2
2-3 63 18.2 2.06 2.09 16.4
4-5 122 19.1 1.48 1.50 16.4
Source of Sum Mean F
variation squares DF square statistic p
STAGES 1-3 461.7 2 230.8 0.84 0.4330

Residual 67336.3 245 274.8
Total 67798.0 247

 

 

Bonferroni
Contrast Difference 95% CI
0-1 v 2-3 -2.5 -9.6 to 4.6
0-1 v 4-5 -3.3 -9.5 to 2.9
2-3 v 4-5 -0.8 -7.0 to 5.4

86

Table 14 — Analysis of Variance with Bonferroni Correction for Three Age Groups with

Two Stages: Stages 4-6

 

 

 

1-way ANOVA
n I 46
Pooled
STAGES 4-6 n Mean SE SE SD
0-1 10 56.1 14.72 40.99 46.6
2-3 16 60.6 12.07 32.41 48.3
4-5 20 128.3 41.93 28.98 187.5
Source of Sum Mean F
variation squares DF square statistic p
STAGES 4-6 54517.5 2 27258.8 1.62 0.2093
Residual 7224604 43 16801.4
Total 776977 .9 45
Bonferroni
Contrast Difference 95% CI
to
0'1 V 2'3 -4.5 -134.7 125.6
0-1 v 4-5 -72.2 -197.2 to 52.9
2-3 v 4-5 -67 .6 -175.9 to 40.7

 

Table 15 — Summary of Analysis of Variance for Three Age Groups with Two Stages and

 

 

 

 

 

Mean Healing Times
Mean Healing Time (da S)
Statistically Age 0-1 Age 2-3 Age 4-5
Significant
Stages 1-3 Not significant 15.7 18.2 19.1
Stage 4-6 Not significant 56.1 60.6 128.3

 

 

 

 

 

87

 

Table 16 — T-test for Two Age Groups with Six Stages: Stage 1

 

0-2 3-5

 

Mean

Variance

Observations

Hypothesized Mean Difference
Df

t Stat

P(T<=t) one-tail

t Critical one-tail

P(T<=t) two-tail

t Critical two-tail

2.428571429 3.98245614
8.787456446 13.08897243
42 57
0
96
-2.345619971
0.010526594
1.660882845
0.021053188
1.984985829

 

Table 17 - T-test for Two Age Groups with Six Stages: Stage 2

 

0—2 3-5

 

Mean

Variance

Observations

Hypothesized Mean Difference
Df

t Stat

P(T<=t) one-tail

t Critical one-tail

P(T<=t) two-tail

t Critical two-tail

21 .17857143 20.87301587
109.1891534 113.9513569
28 63
0
53
0.127889339
0.449360289
1.6741 15993
0.898720577
2.005745046

 

Table 18 — T-test for Two Age Groups with Six Stages: Stage 3

 

0-2 3-5

 

Mean

Variance

Observations

Hypothesized Mean Difference
Df

t Stat

P(T<=t) one-tail

t Critical one-tail

P(T<=t) two-tail

t Critical two-tail

37.4 38.97368421
229.9368421 159.269559
20 38
0
33
-0.397312082
0.346847799
1.692360456
0.693695598
2.03451691

 

88

Table 19 — T-test for Two Age Groups with Six Stages: Stage 4

 

 

0—2 3-5
Mean 45.1 43.25
Variance 376.5444444 173.25
Observations 10 20
Hypothesized Mean Difference 0
Df 13
t Stat 0.27183275
P(T<=t) one-tail 0.3950099
t Critical one-tail 1.770931704
P(T<=t) two-tail 0.7900198
t Critical two-tail 2.16036824

 

Table 20 — T-test for Two Age Groups with Six Stages: Stage 5

 

 

0-2 3-5
Mean 34.66666667 126.3333333
Variance 94.26666667 660.3333333
Observations 6 3
Hypothesized Mean Difference 0
Df 2
t Stat -5.969240478
P(T<=t) one-tail 0.01 346801 1
t Critical one-tail 2.91998731
P(T<=t) two-tail 0.026936023
t Critical two-tail 4.302655725

 

Table 21 — T-test for Two Age Groups with Six Stages: Stage 6

 

 

0-2 3-5
Mean 135.6666667 446.5
Variance 6610.333333 51499.66667
Observations 3 4
Hypothesized Mean Difference 0
Df 4
t Stat -2.531340198
P(T<=t) one-tail 0.032286269
t Critical one-tail 2.131846486
P(T<=t) two-tail 0.064572538
t Critical two-tail 2.776450856

 

89

Table 22 — Summary of T-tests for Two Age Groups with Six Stages and Mean Healing

Times

2.
21

37.
45.1
34.

 

135.

Table 23 — T—test for Two Age Groups with Three Stages: Stages 1-2

 

 

0-2 3-5
Mean 9.928571429 12.85
Variance 133.5455487 137.2714286
Observations 70 120
Hypothesized Mean Difference 0
Df 146
t Stat -1.672333075
P(T<=t) one-tail 0.048300164
t Critical one-tail 1.655357664
P(T<=t) two-tail 0.096600327

t Critical two-tail

1.976345629

Table 24 - T-test for Two Age Groups with Three Stages: Stages 3-4

 

0-2 3-5

 

Mean

Variance

Observations

Hypothesized Mean Difference
Df

t Stat

P(T<=t) one-tail

t Critical one-tail

P(T<=t) two-tail

t Critical two-tail

39.96666667 40.448275 86
281.1367816 165.3393829
30 58
0
47
-0.137760693
0.445509027
1.677926775
0.891018055
2.01 1738616

 

90

Table 25 — T-test for Two Age Groups with Three Stages: Stages 5-6

 

 

0-2 3-5
Mean 68.33333333 309.2857143
Variance 4261.75 55257.57143
Observations 9 7
Hypothesized Mean Difference 0
Df 7
t Stat -2.634111956
P(T<=t) one-tail 0.016855294
t Critical one-tail 1.894577508
P(T<=t) two-tail 0.033710589
t Critical two-tail 2.36462256

 

Table 26 — Summary of T-tests for Two Age Groups with Three Stages and Mean

Healing Times

Mean H ' Time da
tatisti Si ' 0-2 3-5

1-2 0.0 9.
3-4 Not 8' ° 40.
5-6 0.0 68.

 

Table 27 - T-test for Two Age Groups with Two Stages: Stages 1-3

 

 

0—2 3-5
Mean 16.03333333 19.13291139
Variance 284.5269663 267 .033 1 775
Observations 90 1 5 8
Hypothesized Mean Difference 0
Df 1 80
t Stat -1.407229149
P(T<=t) one-tail 0.080541927
t Critical one-tail 1.653363597
P(T<fi) two-tail 0.161083855
t Critical two-tail 1.9732306]

 

91

Table 28 — T-test for Two Age Groups with Two Stages: Stages 4-6

 

 

0-2 3-5
Mean 56.10526316 112.2222222
Variance 2224.432749 26993.02564
Observations 19 27
Hypothesized Mean Difference 0
Df 32
t Stat -1.679202614
P(T<#) one-tail 0.05142628
t Critical one-tail 1.693888407
P(T<=t) two-tail 0.102852561
t Critical two-tail 2.036931619

 

Table 29 — Summary of T-tests for Two Age Groups with Two Stages and Mean Healing
Times

Mean H ' Time
tatisti Si ' cant 0-2 3-5

1-3 Not Si ' 16.
4-6 Not Si ' 56.1

 

Table 30 — T—test for Fracture Location with Six Stages: Stage 1

 

 

forearm leg
Mean 3.596774194 2.864864865
Variance 13.81835008 8.231231231
Observations 62 37
Hypothesized Mean Difference 0
Df 91
t Stat 1.096756703
P(T<=t) one-tail 0.137820812
1 Critical one-tail 1.661771876
P(T<=t) two—tail 0.275641624
t Critical two-tail 1.986377356

 

92

Table 31 — T-test for Fracture Location with Six Stages: Stage 2

 

 

forearm leg
Mean 17.32692308 25.82051282
Variance 68.38122172 129.4669366
Observations 52 39
Hypothesized Mean Difference 0
Df 66
t Stat -3.94530882
P(T<=t) one-tail 9.78476E-05
t Critical one-tail 1.668270215
P(T<=t) two-tail 0.000195695
t Critical two-tail 1.996563697

 

Table 32 — T-test for Fracture Location with Six Stages: Stage 3

 

 

forearm leg
Mean 36.53846154 42.31578947
Variance 205.4129555 114.5614035
Observations 39 19
Hypothesized Mean Difference 0
Df 46
tStat -1.7189141
P(T<=t) one-tail 0.046176968
t Critical one-tail 1.678658919
P(T<=t) two-tail 0.0923 53935
t Critical two-tail 2.012893674

 

Table 33 — T-test for Fracture Location with Six Stages: Stage 4

 

 

forearm leg
Mean 44.13043478 43
Variance 238.5731225 241 .3333333
Observations 23 7
Hypothesized Mean Difference 0
Df 10
t Stat 0.168798835
P(T<=t) one-tail 0.434660161
t Critical one-tail 1.812461505
P(T<=t) two-tail 0.869320322
t Critical two-tail 2.228139238

 

93

Table 34 — T-test for Fracture Location with Six Stages: Stage 5

 

 

forearm leg
Mean 59.16666667 77.33333333
Variance 1716.566667 4677.333333
Observations 6 3
Hypothesized Mean Difference 0
Df 3
t Stat -0.42291479
P(T<=t) one-tail 0.350421 109
t Critical one-tail 2.353363016
P(T<=t) two-tail 0.700842217
t Critical two-tail 3.182449291

 

Table 35 — T-test for Fracture Location with Six Stages: Stage 6

 

 

forearm leg
Mean 3 17.4 303
Variance 76650.8 26450
Observations 5 2
Hypothesized Mean Difference 0
Df , 3
t Stat 0.085215813
P(T<=t) one-tail 0.46872914
t Critical one-tail 2.3 5336301 6
P(T<=t) two-tail 0.93745 8281
t Critical two-tail 3.182449291

 

Table 36 — Summary of T-tests for Fracture Location with Six Stages and Mean Healing

Times

tatistically

Not 8'
0.01
0.
Not Si
Not S'
Not S'

 

94

Table 37 — T-test for Fracture Location with Three Stages: Stages 1-2

 

 

 

 

 

 

 

 

forearm leg
Mean 9.859649123 14.64473684
Variance 85.5022512 202.9521053
Observations 1 14 76
Hypothesized Mean Difference 0
Df 117
tStat 2587311844
P(T<=t) one-tail 0.005448454
t Critical one-tail 1.657981556
P(T<=t) two-tail 0.010896909
t Critical two-tail 1.98044745
Table 38 — T-test for Fracture Location with Three Stages: Stages 3-4
forearm leg
Mean 39.35483871 42.5
Variance 227.6753041 140.5
Observations 62 26
Hypothesized Mean Difference 0
Df 59
t Stat -1.043986774
P(T<=t) one-tail 0.150375123
t Critical one-tail 1.671091923
P(T<=t) two-tail 0.300750247
t Critical two-tail 2.000997483
Table 39 -— T-test for Fracture Location with Three Stages: Stages 5-6
forearm leg
Mean 176.5454545 167.6
Variance 49705.27273 24228.8
Observations 11 5
Hypothesized Mean Difference 0
Df 11
t Stat 0.092440427
P(T<fi) one-tail 0.4640051
t Critical one-tail 1.795883691
P(T<fi) two-tail 0.928010201
t Critical two-tail 2.200986273

 

95

Table 40 — Summary of T-tests for Fracture Location with Three Stages and Mean

 

 

 

 

 

 

 

 

 

 

Healing Times
Mean Healin Time (days)
Statistically Forearm Leg
Significant
Stage 1-2 0.01 9.9 14.6
Stage 3-4 Not Significant 39.4 42.5
Stage 5-6 Not Significant 176.5 167.

 

Table 41 — T-test for Fracture Location with Two Stages: Stagesl-3

 

 

forearm leg
Mean 16.66013072 20.17895
Variance 250.9889921 307.6804
Observations 153 95
Hypothesized Mean Difference 0
Df 184
t Stat -1.593025337
P(T<=t) one-tail 0.056435453
t Critical one-tail 1.653177151
P(T<=t) two-tail 0.112870906
t Critical two-tail 1.972939572

 

Table 42 — T-test for Fracture Location with Two Stages: Stages 4-6

 

 

forearm leg
Mean 86.97058824 94.91667
Variance 19174.9385 13058.63
Observations 34 12
Hypothesized Mean Difference 0
Df 23
t Stat -0.195489365
P(T<fi) one-tail 0.423363309
t Critical one-tail 1.713870006
P(T<=t) two-tail 0.846726619
t Critical two-tail 2.068654794

 

96

 

Table 43 — Summary of T-tests for Fracture Location with Two Stages and Mean Healing

 

 

 

 

 

 

 

 

Times
Mean Healin Time (days)
Statistically Forearm Leg
Significant
Stages 1-3 Not Significant 16.7 20.2
Stages 4-6 Not Significant 87.0 94.9

 

97

BIBLIOGRAPHY

98

BIBLIOGRAPHY

Armstrong PF, VE Joughin, HM Clarke, RB Willis (2003) Fractures of the Forearm,
Wrist and Hand. In Skeletal Trauma in Children. Green NE and MF Swiontkowski (eds.)
Philadelphia, PA: WB Saunders Company.

Auringer ST (2002) Special Considerations in Children. In Radiology of Skeletal Trauma.
Rogers LF (ed.). Vol. 1, 3rd ed. New York: Churchill Livingstone.

Baitner AC, A Perry, FD Lalonde, TP Bastrom, J Pawelek, and PO Newton (2007). The
healing forearm fracture: A matched comparison of forearm refracture. Journal of
Pediatric Orthopaedics 27(7):743-747.

Blokhais TJ, de Bruine JH, Bramer JA, den Boer FC, Bakker FC, Patka P, Haarimon HJ,
Manolui RA (2001). The Reliability of Plain Radiography in Experimental Fracture
Healing. Skeletal Radiology 30 (3): 151-156.

Brogden, BG (1998). Forensic Radiology. Boca Raton: CRC Press.

Caffrey J (1946) Multiple fiactures in the long bones of infants suffering from chronic
subdural hematoma. American Journal of Roentgenology 56:163-173.

Chung KC and SV Spilson (2001) The frequency and epidemiology of hand and forearm
fi'actures in the United States. The Journal of Hand Surgery 26(5): 908-915.

Cox TD and A Sonin (2002) The Elbow and Forearm. In Radiology of Skeletal Trauma.
Rogers LF (ed.). Vol. 2, 3rd ed. New York: Churchill Livingstone.

Cumming WA (1979) Neonatal Skeletal Fractures. Birth Trauma or Child Abuse?
Journal of the Canadian Association of Radiologists 30: 30—33.

De Mause L (1974) The History of Childhood. New York: Psychohistory Press.

De Palma L, Greco F, Specchia N, Rizzi L. (1991) Evaluation of Fracture Healing with
Computerized Analysis of Radiographic Images. Radiological Medicine 82 (1-2): 44-47.

Dunbar J W, HF Owen, MB Nogrady (1964) Obscure Tibial Fractures of infants — The
toddler’s fiacture. Journal of the Canadian Association of Radiologists. 15:136-144.

Fitzgerald ER (1977) Postmortem Transition in the Dynamic Mechanical Properties of
Bone. Medical Physics 4 (1): 49-53.

Frost (1989) The Biology of Fracture Healing: An Overview for Clinicians. Part 1.
Clinical Orthopaedics and Related Research 248: 283-293.

99

Glencross B and P Stuart-Macadam (2000) Childhood Trauma in the Archaeological
Record. International Journal of Osteoarchaeology 10:198-209.

Grigoryan M, Lynch JA, Fierlinger AL, Guermazi A, Fan Bo, MacLean DB, MacLean A,
and Genant HK (2003). Quantitative and Qualitative Assessment of Closed Fracture

Healing Using Computed Tomography and Conventional Radiography. Academic
Radiology 10(11): 1267-1273.

Gwinn JL, Levin W, and Petersen HG (1961) Roentgenographic manifestations of
unsuspected trauma in infancy. Journal of the American Medical Association 1761926-
929.

Heinrich SD and JF Mooney III (2006) Fractures of the Shafts of the Tibia and Fibula. In
Rockwood and Wilkins’ Fractures in Children 6th ed., Beaty JH and JR Kasser (eds.)
Philadelphia, PA: Lippincott Williams & Wilkins.

Helms CA and NM Major (2002) The Knee and Shafts of the Tibia and Fibula. In
Radiology of Skeletal Trauma. Rogers LF (ed.). Vol. 2, 3rd ed. New York: Churchill
Livingstone.

Hendrix RW (2002) Fracture Healing. In Radiology of Skeletal Trauma. Rogers LF (ed.).
Vol. 1, 3rd ed. New York: Churchill Livingstone.

Hertel R and DA Rothenfluh (2006) Fractures of the Shafts of the radius and ulna. In
Bucholz RW, JD Heckman, and CM Court-Brown (eds.) Rockwood & Green’s Fractures
in Adults. Philadelphia, PA: Lippincott Williams & Williams.

Heybeli N, Yesilda A, Oyar O, Gulsoy UK, Tekinsoy MA, and Mumcu EF (2002).
Diagonistic Ultrasound Treatment Increase the Bone Fracture Healing Rate in an
Internally Fixed Rat Femoral Osteotomy Model. Journal of Ultrasound in Medicine 21:
1357-1363.

Hobbs CJ (1 989) Fractures. British Medical Journal 298: 101 5-1 01 8.

Hufnagl (2005) An Investigation of Time Since Injury: A Radiographic Study of Fracture
Healing. MA Thesis. Louisiana State University.

Islam 0, D Soboleski, S Symons, LK Davidson, MA Ashworth, and P Babyn (2000)
Development and Duration of Radiographic Signs of Bone Healing in Children.
American Journal of Roentgenology 175: 75-78.

Jones E (2003) in Green NE and Swiontkowski MF (eds.) Fractures in Children. 3rd ed.
Philadelphia: WB Saunders Company.

100

Jones E (1994) In Green NE and Swiontkowski MF (eds.) Fractures in Children.
Philadelphia: WB Saunders Company. Pp.1-14.

Kempe RS, Silverman FN, Steele BF, Droegemueller W, and Silver HK (1962) The
Battered Child Syndrome. American Journal of Medical Science 181 ( 1): 17-24.

Keogh C, D Bergin, and S Eustace (2002) The Wrist. In Radiology of Skeletal Trauma.
Rogers LF (ed.). Vol. 2, 3rd ed. New York: Churchill Livingstone.

Kerley ER (1976) Forensic Anthropology and Crimes Involving Children. Journal of
Forensic Sciences 21 (2): 333-339.

Kerley ER (1978) The Identification of Battered-Infant Skeletons. Journal of Forensic
Sciences 23 (1): 163-168.

King J, D Diefendorf, J Apthorp, VF Negrete and M Carlson (1988) Analysis of 429
fractures in 189 battered children. Journal of Pediatric Orthopaedics 8:585-5 89.

Kleinman PL, PK Kleinman, and JA Savageau (2004) Suspected infant abuse:
Radiographic Skeletal survey practices in pediatric health care facilities. Radiology 233:
477-485.

Kleinman PL, BDBlackbourne, SC Marks, A Karellas, and PL Belanger (1989)
Radiologic Contributions to the Investigation and Prosecution of Cases of Fatal Infant
Abuse. The New England Journal of Medicine 320 (8): 507-511.

Kleinman PK (1987a) Introduction. in Kleinman PK (ed.) (1987) Diagnostic Imaging of
Child Abuse. Baltimore: Williams and Wilkins.

Kleinman PK (1987b) Skeletal TraumazGeneral Considerations in Kleinman PK (ed.)
(1987) Diagnostic Imaging of Child Abuse. Baltimore: Williams and Wilkins.

Kline A, Balian G Harwitz S (2005). BMP-14 Deficiency Inhibits Long Bone Fracture
Healing: A Biochemical, Histologic, and Radiographic Assessment. Journal of
Orthopaedic Trauma 19 (9): 629-634.

Langley-Hobbs S (2003). Biology and Radiological Assessment of Fracture Healing. In
Practice 25: 26-35.

Lewis ME (2007) The Bioarchaeology of Children: Perspectives fiom Biological and
Forensic Anthropology. Cambridge, UK: Cambridge University Press.

Loder RT and C Bookout (1991) Fracture Patterns in Battered Children. Jouranl of
Orthopaedic Trauma. 5(4):428-433.

101

Lovejoy CO and KG Heiple (1981) “The analysis of fractures in skeletal populations
with an exmaple fi'om the Libben Site, Ottowa County, Ohio.” American Journal of
Physical Anthropology 55:529-541.

Maples WR (1986) Trauma Analysis by the Forensic Anthropologist. In Forensic
Osteology: Advances in the Identification of Human Remains, Reiches KJ (ed.)
Springfield: Charles C Thomas.

McGraw EP, JE Pless, DJ Pennington, and SJ White (2001) Postmortem radiography

after unexpected death in neonates, infants, and children: Should imaging be routine?
America] Journal of Roetgenology 1 78: 1 5 1 7-1 521 .

Mehlman CT and EJ Wall (2006) Injuries to the Shafts of the Radius and Ulna. In
Rockwood and Wilkins’ Fractures in Children 6th ed., Beaty JH and JR Kasser (eds.)
Philadelphia, PA: Lippincott Williams & Wilkins.

Mellick LB, KB Reesor, D Demers, KA Reinker (1988). Tibial fiactures of young
children. Pediatric Emergency Care 4(2): 97-101.

Merten DF, MA Radkowski, and JC Leonidas (1983) The abused child: A radiological
reappraisal. Radiology 146: 377-381.

O’Connor JF and J C Cohen (1987) Dating Fractures. in Kleinman PK (ed.) Diagnostic
Imaging of Child Abuse. Baltimore: Williams and Wilkins.

Ogden J A (2000) Skeletal Injury in the Child. 31rd ed. New York: Springer-Verlag, Inc.

Ogden JA (1984) The Uniqueness of Growing Bone. In Fractures in Children.
Rockwood, CA Jr., KE King (eds.) Vol. 3, J .B. Philadelphia: Lippincott Co.

Ogden JA (1982) Anatomy and physiology of skeletal development. In: Ogden JA
(ed.) Skeletal Injury in the Child. Philadelphia: Lea & Febiger.

Ogden JA (1963) Injury to growth mechanisms of the immature skeleton. Skeletal
Radiology 6:23 7-253.

Ortner DJ (2003) Trauma. In Identification of Pathological Conditions in Human Skeletal
Remains. 2nd ed. Ortner DJ (ed.). San Diego: Academic Press.

Paton DF (1992) Fractures and Orthopaedics. Edinburgh: Churchill Livingstone.
Perry MA (2005) “Redefining Childhood through Bioarchaeology: Toward an

Archaeological and Biological Understanding of Children in Antiquity.” Archaeological
Papers of the American Anthropological Association 15(1): 89-1 1 1.

102

Prosser I, S Maguire, SK Harrison, M Mann, JR Sibert, AM Kemp, and Welsh Child
Protection Systematic Review Group (2005) How old is this fracture? Radiologic dating

of fractures in children: A systematic review. American Journal of Roetgenology 184:
1282-1286.

Rang M (1983) Injuries of the epiphysis, growth plate and perichondrial ring. In
Rang M (ed.) Children 's Fractures. Philadelphia: J .B. Lippincott.

Risselada M, Kramer M, de Rooster H, Taeymans O, Verleyen P, and van Bree H (2005).
Ultrasonographic and Radiographic Assessment of Uncomplicated Secondary Fracture
Healing of Long Bones in Dogs and Cats. Veterinary Surgery 34 (2): 99-107.

Salkind NJ (ed.) (2007). Encyclopedia of Measurement and Statistics. Vol. 1. Thousand
Oaks, CA: Sage Reference, p32-35.

Sauer NJ (1998) The Timing of Injuries and Manner of Death: Distinguishing Among
Antemortem, Perimortem, and Postmortem Trauma. In Forensic Osteology: Advances in
the Identification of Human Remains. Reiches KJ (ed.) 2"d ed., Springfield: Charles C
Thomas.

Sirkin RM (2006) Statistics for the Social Sciences. 3rd ed. Thousand Oaks, CA: Sage
Publications, Inc.

Skak SV and TT Jensen (1988) Femoral Shaft Fractures in 265 Children: Log-normal
correlation with age of speed of healing. Acta Orthpedica Scandinavia 59(6): 704-707.

Sledzik PS and Kelley MA. Initial Osseus Remodeling Following Trauma. Manuscript.

Tiedeman JJ, Lippiello L, Connolly JF, and Strates BS (1990). Quantitative
Roentgenographic Denitometry for Assesssing Fracture Healing. Clinical Orthopaedics
and Related Research 253: 279-286.

Toal RL and Mitchell SK (2002) Fracture Healing and Complications. In Textbook of
Vetrinary Diagnostic Radiology. Thrall DE (ed.) 4th ed., Philedelphia: WB Saunders Co.

Waldron T (2000) “Hidden or overlooked? Where are the disadvantaged in the skeletal
record?” In J Hurbert (ed.) Madness, Disability and Social Exclusion. London:
Routledge.

Walker PL (2001a) “A Bioarchaeolgical Perspective on the History of Violence.” Annual
Reviews of Anthropology 30:573-596.

Walker PL (2001b) “Is the Battered-Child Syndrome a Modern Phenomenon?” In press:
Proceedings of the 10‘h European Meeting of the Paleopathology Association.

103

Walker PL, Cook DC, and Lambert PM (1997) Skeletal Evidence for Child Abuse: A
Physical Anthropological Perspective. Journal of Forensic Sciences 42 (2): 196-207.

Waters PM and AD Mih (2006) Fractures of the Distal Radius and Ulna. In Rockwood
and Wilkins’ Fractures in Children 6“1 ed., Beaty JH and JR Kasser (eds.) Philadelphia,
PA: Lippincott Williams & Wilkins.

Yeo LI and MH Reed (1994) Staging of Healing of Femoral Fractures in Children.
Journal of the Canadian Association of Radiologists 45(1): 16-19.

104

MICHIGAN STATE UNIVERSITY LIBRARIES

Ill IIIIII III

3 1293 3062 5804