Sie sind auf Seite 1von 6

This is an enhanced PDF from The Journal of Bone and Joint Surgery The PDF of the article you

requested follows this cover page.

A study of the clinical incidence of infection in the use of banked allograft bone
WW Tomford, RJ Starkweather and MH Goldman J Bone Joint Surg Am. 1981;63:244-248.

This information is current as of September 27, 2007 Reprints and Permissions Click here to order reprints or request permission to use material from this article, or locate the article citation on jbjs.org and click on the [Reprints and Permissions] link. The Journal of Bone and Joint Surgery 20 Pickering Street, Needham, MA 02492-3157 www.jbjs.org

Publisher Information

Copyrtghl

981 by

The

Journal

of Bone

and

Joint

Surgery.

Incorporated

A Study

of the Clinical of Banked


M.D4, Naval

Incidence Allograft
AND Medic-al M. H.

of Infection Bone*
M.D., Bethesda, BETHESDA, Maryland MARYLAND Institute,

in the Use
BY W. W. TOMFORD, Frotn the M.D.t, Clinical R. J. Investigation

STARKWEATHER, Department.

GOLDMAN,

Research

ABSTRACT:

in grafting
567

To determine the incidence of infection procedures utilizing banked allograft bone, were sent to collaborating surgeons Tissue Bank freeze-dried allograft bone 1973 to October 1976. Three hundred were sufficiently completed Twenty-one patients were to re-

Materials

and

Methods are procured for tissue colunder sterile

questionnaires who used Navy from October

United States Navy Tissue Bank tissues from donors who meet the standard criteria lection reported previously . Bones collected conditions are cleaned ground bone, longitudinal and ters of all soft sections

and three questionnaires be included in the study. ported twelve ered

tissue. For crushed or are cut from the shaft

as showing evidence were considered minor major according to the course. In eleven were positive cultures
ten there were

of infection, of which and nine were consideffect on the patients of the twenty-one paas proof of infection;
not. Analysis of the

then processed into pieces two to four cubic milhimein size. Cortical plates or strips are cut into fifteen-

postoperative tients there


in the

remaining

proved

infections

showed

that

the

allograft

was

prob-

centimeter lengths. Long bones are left whole or are divided at mid-shaft. All tissue is stored in liquid-nitrogen freezers (- 196 degrees Celsius) to await results from aerobic and anaerobic touch cultures made of all deposits during procurement and processing. If cultures are negative, the tissue is freeze-dried and
10 per

ably not primarily responsible in most of the patients. Based on the data obtained in this study, the incidence of infection with the use of banked allogenous bone appears ported bone. Orthopaedic steadily increasing Tissue Bank, which surgeons lesions7 has noted that benign cystic unions3, allografts demand for allograft over the past several has attempted applications to fracture bone years. this has The been Navy from nonof to compare favorably with infection for orthopaedic procedures utilizing rates reautogenous

to less stored at

than room

ten micrometers temperature


of all tissues

of vapor evacuated
As

pressure bottles.
many as days.

in
pro-

cent

of cultures 90 The per

cessed
twenty-one

in the Tissue cultures grow

Bank either

are positive

after
cent

being

held
acnes

for
posior cent

Approximately

of the 10 per

tive

Corynebacterium remaining

Staphylococcus

epidermidis.

to meet

demand,

grow either Staphylococcus that has

non-hemolytic aureus (less culture

Streptococcus, than 2 per cent).

or, rarely, Any tissue and of the

have broadened malunions and

a positive

is considered

is irradiated.

In this

study,

fewer

than

contaminated 5 per cent

large segmental for tumor surgery6.

defects, and whole-bone Long-term evaluation

these grafts has been an important part of the Tissue Banks role in the medical community, and studies of the physical and chemical suitability of the tissues have been initiated intermittently24. However, no determination of the incidence of infection following the use of processed and stored allograft bone has been performed. For this reason, a retrospective study of the results of grafting procedures tuted,
*

allografts required irradiation. These allografts were cultured after irradiation and all cultures were negative. Questionnaires were sent to 567 collaborating surgeons who used Navy Tissue Bank bone during the period
October 1973 to October 1976. The questionnaire con-

sisted of a history requested remarks

of the patients regarding any

grafting procedure complications that

and en-

sued that possibly were related questionnaires mailed, 313 (55

to infection. Of the 567 per cent) were returned,

using Navy Tissue and the findings are


This work was supported

Bank allograft reported here.


by the Naval Medical

bone

was

insti-

Research

and

Dc-

with 303 of the 3 13 completed sufficiently to be used to compile data. No case was included that did not have at least an eighteen-month follow-up. The diagnoses, procedures performed, and types of grafts used in the 303 patients are listed in Tables I, II, and Results
Twenty-one (6.9 per course;
THE

velopment Command, Work Unit No. M0095-PN.00l.0003. The opinions or assertions contained herein are the private ones of the authors and are not to be construed as official or reflecting the views of the U.S. Navy Department or the naval service at large. 1 Department of Orthopaedic Surgery. Massachusetts General Hospital, Boston, Massachusetts 02 1 14. Wenatchee Valley Clinic, Wenatchee, Washington 98801. Division of Transplant Surgery, University of Virginia Medical Center, Richmond, Virginia 23298.

III,

respectively.

cent) the

of

the

303 282
AND

completed

questionnaires
tients postoperative

reported

evidence

of
remaining
OF BONE

infection

in the
reported

pano

244

JOURNAL

JOINT

SURGERY

CLINICAL

INCIDENCE

OF

INFECTION

IN

THE

USE

OF

BANKED

ALLOGRAFT

BONE

245
noted No none in three pato a in this intraveof the in in nine first ospa-

evidence infection
the patients

of this complication. Of necessity, evidence of was determined by the reporting surgeon because
could not be personally evaluated by us. Cases

heal tients

primarily. (Cases produced antibiotic because The group course

Wound

drainage 9), but each antibiotics. drainage, none and infections

was

2, 4, and
chronic therapy, of

of these

responded procedure required removal resulted V), the

short-term group nous graft

of oral

reported
healing

as
such

infection
as

included
and

complications
delayed

of wound
the

wounddrainage

erythema

epithehialization responding

necessitated none (Table from had patients

(with
alone.

or

without the

wound twenty-one
basis

drainage),
received

and
from

infection,

From groups

information on the

teomyelitis. into
on tients, of major most of was different in several these aspects group. (Cases

surgeons,
two

patients
of the effect

could
of the

be

divided

complication

the postoperative

course.

Cases

classified

as showing

eviproducthose

Diagnostically,

tumors

dence of minor infection ing little or no effect classified required These analyzed
The

involved complications on the patients recovery; of major a chronic

14 through 20). The


teogenic (giant-cell three fracture

17) or traumatic tumors included


sarcomas) tumors and of the two had

wounds (Cases two malignant


potentially end been proximal of the

18, 19, and lesions (oslesions tibia). on at The least

as showing reoperation two groups separately.


minor

evidence or involved are listed

infection either complication. V and were

malignant operated

in Tables
(Table IV),
I

IV and
in twelve

non-unions

infections

patients,

twice previously, teomyehitis) at

and all some time

three prior
II

had been infected (osto implantation of the

TABLE DIAGNOSES

TABLE
PROCEDURES

Lesion

No. of Patients

Procedure Curettage and pack Onlay grafts Whole or partial bone

No. 229 49
replacement

Cysts Unicameral Aneurysmal Simple Mandibular Rheumatoid


Tumors

I 17 19 6 14 1

25

TABLE ALLOGRAFTS Type

III

No.

Malignant Benign Fibrous Fracture Spine fusion Mandibular Limb-length Congenital Realignment Osteonecrosis Osteopetrosis Iliac defect Sterile abscess reconstruction discrepancy pseudarthrosis osteotomy dysplasia

l6 48 27 20 18 4 4 3 2 I 1 1 1

Crushed Ground
Cortical

cortical cancellous

strips Ilium strips Combination


Whole bone

183 15 26 17 37

25

allograft. bone cyst

The (Case

two 13)

remaining and

patients

had

a unicameral of the

a congenital

pseudarthrosis

tibia
operated

(Case grafted
months Grafts

21). once
prior used

The with

unicameral
.

bone bone,

cyst

had

not

been
had

on previously

The use second

congenital of the group

pseudarthrosis

been
nine

autogenous

unsuccessfully,
more corti-

to the in the

allograft. included

occurred chiefly in benign cysts or fusions. There seven benign bone cysts, two tumors (chondrosarcoma giant-cell tumor), one non-union after a lengthening tibia, one spine fusion for congenital scohiosis, maxillary osteotomy for maxillary reconstruction.

were and

cal bone

than

in the

first

group.

Each

of the

tumor

resec-

of the and one The

tions necessitated the use of a large piece of cortical bone and in three (Cases 14, 15, and 17) this type of graft was used exclusively. The operations for fracture non-union (Cases
pseudarthrosis

types of bone graft used included crushed or ground bone (Cases 1 through 8), a combination of crushed or cancelbus bone with cortical bone (Cases 9, 1 1 , and 12), and cortical bone alone (Case 10). Procedures in this group consisted (Cases
bloc

18,

19, bone

and
(Case

20) (Case

as

well
utilized

as

for grafted

the

congenital
bone. The

2 1) also

cortical

unicameral cortical
The

cyst

13) was bone.


in

with

crushed
group inre-

and more
of the

corticocancellous
performed

mainly

of curettage

and

packing

of the

lesion volved
Three

procedures

the

second

1 through 8). One procedure (Case 9) utilized en excision of a large portion of bone, and the remaining
(Cases 10, 1 1 , and

extensive
tumor

surgery
resections

than
(Cases

in
14,

the
15,

first
and

group.
17)

three

12) utilized resulted one


1981

onlay

grafts

for group

fuof not

sion. No serious patients


VOL. 63-A,

complications IV).
2, FEBRUARY

in the first (Case

quired extensive excision of both soft tissue and bone. The fracture non-unions (Cases 18, 19, and 20) required wide exposure because of previous infection. The tibial pseudarthrosis (Case 2 1 ) required extensive dissection to

(Table
NO.

Only

incision

4) did

246

W.

W.

TOMFORD,

R.

J.

STARKWEATHER,
TABLE IV
OF MINOR

AND

M.

H.

GOLDMAN

PATIEN

TS

WITH

EVIDENCE

INFECTION

Case
I

Diagnosis Mandibular Non.ossifying cyst

Procedure
Curettage Curettage and and pack pack Ground

Allograft
cancellous cortical

Complication
Localized erythema Drainage

Cultured Yes Yes No Yes No No


No No

Result Pos. Pos.


-

Organism Not listed Not listed


-

Treatment
10

Result Wound
healed primarily

wound
at 4th day

days of oral antibiot.

Crushed

10 days of oral antibiot.


-

fibroma 3
4 5 6
7

Wound healed primarily Wound healed


primarily

Unicameral bone cyst


Mandibular Simple cyst Mandibular Unicameral bone cyst cyst

Curettage
Curettage Curettage

and pack
and pack and pack

Crushed
Ground Crushed

cortical
cancellous cortical

Localized
erythema Drainage Localized erythema

wound at 5th day wound

Pos.
-

Not listed
-

Wound
10

incised and drained; days of oral antibiot.


-

Wound healed secondarily


Wound healed primarily Wound Wound Wound healed healed healed healed healed healed

Curettage
Curettage

and pack
and pack

Ground
Crushed

cancellous
cortical cortical

Localized wound erythema


Localized wound wound

primarily
-

bone cyst
8 Giant-cell

erythema
Curettage

primarily
-

tumor

and pack

Crushed Femoral
Cortical

Localized

erythema 9
10
II

primarily at 4th day


wound

Chondrosarcoma
Non-union of limb-

Partial bone replacement


Onlay graft

condyle
strip

Drainage Localized

Yes
No

Pos.
-

Enterobacter species
-

10 days of oral
-

antibiot.

Wound Wound Wound

primarily primarily

lengthening Congen. Crouzons scoliosis syndrome Onlay Onlay graft graft Rib matchsticks, crushed conical Rib segment. ilium strip

erythema Localized wound erythema Localized


erythema

No Yes

primarily Pos. Not listed 10 days of oral antibiot. Wound healed


primarily

12

wound

allow

placement

of

the

nine-centimeter

fibular-shaft

ailograft. All drainage character reporting and were

tures were a culture


Although

negative until was positive


the graft resorbed

six months postoperatively for Staphylococcus


and there was no

when
aureus. evidence of

nine

patients

in

the

second

group

had

wound

postoperatively, and each had of the drainage was not described surgeons. Of nine cultures, three

a culture. The by any of the (Cases 14, 17, six cultures in three of

healing antibiotic

of the pseudarthrosis, therapy. Discussion

the

infection

cleared

with

19) showed no growth. positive. The organisms

The remaining were not named

Twenty-one

patients

with

evidence

of infection

out

of

these (Cases 16, 18, and 20), but the other three grew either Staphylococcus epidermidis (Case 13) or Staphylococcus auerus (Case 21), or both (Case 15). Antibiotics were used in the treatment of five patients (Cases 15, 16, 18, 20, and 21). Complications dence of major ensued in all patients showing eviinfection that were not noted in patients

303

surgical

procedures

employing

banked

bone

ahlografts

is a 6.9 per cent incidence


the incidence banked bone tion and
grafting

of infection.

No other

studies
the

of

of clinical infection following have been reported. Cruse reported in 372 patients and a 3.9 per
site in 986 patients

use of an infec-

rate of 2.7 per cent excision of a lesion,


in the donor
.

after cent

bone biopsy incidence of


who had auto-

infection

showing evidence of minor infection. The unicameral bone cyst (Case 13) drained for six months and then cleared without antibiotic therapy. In each of the patients with tumors , drainage developed during the postoperative course. One patient (Case 14) had drainage for six months and union was delayed. One patient (Case 17) began to have drainage at one separated at two weeks, weeks. The graft was week postoperatively; the wound and the graft was removed at three removed in another patient with a (Case 15) after a before one, one at The for ancul-

The
with

6.9

per cent
report,

infection
implies

rate
that the

in our study,
risk

compared
with

Cruses

of infection

the use of allograft autograft


ported
-

bone However,

is approximately many
surgeons

double complications

that

with re-

bone.
by the

of the

responding

as evidence

of infection

erythema,

delayed

epithelialization,

and These

drainage phenomena tissue without

are not necessarily proof of infection. occur in the presence of devascularized

tumor (Case 16) after one week, and in another drainage developed eight months postoperatively, prolonged course of irradiation and chemotherapy and after surgery. Of the three allograft (Case 18) was removed four months third fracture several tibiotics. (Case (Case 20), non-union fracture at two

non-unions, weeks and

both secondary to reinfection. (Case 19) drained minimally the infection cleared without with a congenital pseudarthrosis shortly after surgery, but

bacterial contamination. Therefore, the incidence based on all reported infections may not accurately reflect the true risk of infection. The evidence in each case must be exammed to determine if an infection proved by bacterial contamination occurred, and if the allograft can be implicated as the source of the contamination.
Eleven patients had a positive culture: five with evi-

months and then In the patient 2 1), drainage began

dence major

of a minor infection and six with evidence of infection. For four of the five with positive cultures

in the former

group

(Cases

I , 2, 4, and
OF BONE

12) the organism


AND JOINT SURGERY

THE JOURNAL

CLINICAL

INCIDENCE

OF

INFECTION

IN
TABLE

THE
V
OF

USE

OF

BANKED

ALLOGRAFT

BONE

247

PATIENTS

WITH

EvID ENCE

AJOR

INFEcrIoN

Case

Diagnosis

Procedure

Allograft

Complication Drainage

Cultured Yes

Result Pos.

Organism

Treatment Dry dressing, no antibiot. Drained

Result for 6 mos.,

13
14

Unicameral bone cyst


Giant-cell tumor

Curettage
Whole

and pack
bone

Ilium strip, crushed conical


Prox. end including

at 4th day at 5th day

Staph.
epidermidis

then cleared Dry dressing, no antibiot. Drained


then

of tibia
10 cm of femur 12 cm

Drainage

Yes

Neg.

No growth

for 6 mos.,
cleared;

of shaft
15 Osteogenic sarcoma Whole bone Distal end including Drainage at 8 mos. Yes Pos.

delayed Staph. aureDry dressing. long-term oral antibiot. us. Staph.

union

Drainage continued at 18 mos.;

of shaft 16 17
18

epidermidis
Drainage at 4th day Yes

delayed Graft removed at 1 wk.; IV/oral antibiot. Graft


Graft

union

Giant-cell Osteogenic
sarcoma

tumor

Curettage Whole
Onlay

and pack bone

Cortical

plate, crushed conical shaft

Pos. Neg.
Pos.

Not listed No growth


Not listed

10 days
;

of

Wound healed secondarily


Wound healed healed
secondarily

Mid-femoral

Drainage
wound

at 7th day;
separation

Yes
Yes

removed
removed

at 3 wks.
at 2 wks.;

no antibiot. 10 days of

Fracture

graft

non-union 19 Fracture
non-union

Fibular shaft, crushed cortical

Drainage

at 5th day

Wound

oral antibiot. Drainage at 10th day Yes Neg. No growth Dry dressing, no antibiot.

secondarily Spot drained


several

Onlay

graft

Conical

plate

for

months,

then cleared 20
21

Fracture non-union
Congen.

Onlay Onlay

graft graft

Conical Fibular

plate shaft

Drainage Drainage

at 4 mos. at 7th day

Yes Yes

Pos. Pos.

Not listed Staph. aureus

Graft removed
antibiot.

at 4 mos. antibiot.

6 wks.

of IV

Wound
Aseptic

healed secondarily
drainage

tibial pseudarthrosis

10 days of oral

for 6 mos. , then cultured Staph.


aureus

was not listed; each antibiotics. For the listed


well with major not and as

of these four patients was treated fifth patient (Case 9) the organism required
For the 18, the was (Cases ten wound patients (Cases 3, erythema and three used. 13, from 16, of

with was as
of was three a

the
the

tissue.
tissue,

In addition,
the Navy surgeon eleven culture patients (Case culture tissue from irradiated. by may the the the The final have glass culture

as a final
Tissue Bank the with 13) used was in that contaminating touch been storage (or even preoperatively

check
recommends

on the
immediately postoperative to show

sterility
that the prior

of
col-

and
positive

the

patient
cultures (Cases in each treatment group

incision
three group 20), the For the of with

and
the

drainage
six patients evidence organism was removed

laborating to implantation. Of tures, organism Culture to was been or the the removal not shipment missed tissue

allograft positive reported and patient

antibiotic infection

treatment.

cula similar prior tissue have Bank, time results of of

the graft

only

one on

listed, antibiotic

but in this

postoperatively. was negative and the may Tissue at the The was

remaining

of the

patients

15, and 21),


did not have 8,

the organism
positive 10, drainage. and cul1 1) No

Navy

Tissue culture

Bank, at the

was
tures. were

listed.
The noted remaining patients to have Seven

organism

5 through
without

contaminated container. if one

from

culture was grown and no antibiotics were used in any of these patients. All wounds healed primarily. The three remaining patients (Cases 14, 17, and 19) had negative cultures. none Each were drainage an organism. of treated these wih had drainage antibiotics. from In the Cases wound, 14 and but 19, never sev-

preoperative

performed)

were
with tion had

not
positive Of the who

reported
cultures. five

for
patients

any

of the

ten
evidence

remaining
of minor (Cases no incisions evidence

patients
infechealed of os-

showing culture, The there group evidence and

had

a positive drainage.

three surgical was (Cases for

2, 4, and 9)

drainage
The grew

continued
fluid

for
was

several
cultured 17,

months
the wound

and

then
but separated

cleared.

postoperative however,

periodically

primarily,

In Case from

teomyelitis
other erythema two

at any
patients was listed

time

in the postoperative

course.
1 and infection. 12),

For

the

eral
not

days
cultured.

postoperatively
Material

and healed use


culture

the
the

graft
wound

was
was

removed
negative

but
on

in this as the
,

wound In each

culture, tions and

and
A positive

the

wound the
wound

without
does not

further
provide

comphicaevidence

without

of antibiotics.

of these superficial
taminated Ofthe of The but two after major

five patients infection


graft, patients infection

the reported information suggested a that was unlikely to be due to a conthis than had cannot Case positive be proved. evidence two (Cases 13) showing cultures,

although (other who

of the source implantation

of the infection. Factors apart from allograft such as extensive surgical dissection, exinstruments or to bacterial conthe role of the
wound infection is

cessive operative time, or contaminated personnel may produce or contribute tamination. One method of evaluating
-

18 and

20)

were

known
been thatthe

to have

had

a previous
for the infection

infection.
reinfection, recurred. months
15)

graft may have it is more likely patients, the surgical infections

responsible original occurred In one

In

allograft

in the

event

of a postoperative

only

at

several (Case

to culture cessing
are
VOL.

the
just
NO.

allograft
prior

prior

to implantation. Tissue Bank,


and

In the touch
vacuum-packing

pro-

procedure.

of them

an

of tissues

at the Navy

cultures

grown
63-A,

to freeze-drying
1981

infected draining sinus developed after eight chemotherapy. The graft cannot be eliminated

months of as a source

2. FEBRUARY

248

W.

W.

TOMFORD,

R.

J.

STARKWEATHER,

AND

M.

H.

GOLDMAN

of infection in that patient, but the suppressive effect of chemotherapy on the immune system should be considered as a factor in the etiology of the infected drainage. The second patient (Case 2 1) had aseptic drainage from the surgical site for six months before the cultures became positive. Although impossible to prove, this suggests that the drainage became infected from an exogenous source rather than from the graft. In the remaining patient in this group (Case 16), the allograft was removed at one week, the patient was treated with antibiotics, and the wound healed satisfactorily. No report was available on a culture of the Of allograft before or after surgery. all the cases analyzed, the graft appears to be a strong candidate for the source of infection in only one patient (Case 13). In one additional patient (Case 16) the graft was a possible source, although reports of cultures before surgery were not available. Information provided for the other nine patients not have been a primary analysis cannot be interpreted suggested factor that the allograft in the infection. the innocence may This of

viously infected in the development

fractures, and of infection

these factors are important in any patient undergoing the implantation measures would patients of infection of a large seem to be reported varies as ac-

surgery. When combined with piece of dead bone, preventive particularly In this having

important. analysis of the twenty-one the incidence

an infection,

cording to which cases are included only patients in whom the ahlograft fore less

and which are not. If cultured positively be-

and after the procedure are included, the incidence is than I per cent. If only patients with proved post-

operative infection are included, the incidence is 3.6 per cent. If all patients who were reported to have an infection are included, the incidence is 6.9 per cent. Perhaps the best interpretation of the data is that the incidence does not appear to be greater than 7 per cent and probably is lower. The advantages creased operative and complications with a second eration. The of the use of a banked allograft time and avoidance of postoperative such operative individual as bleeding site surgeon or infection still are together worthy with
-

depain

associated of considthe patient

as proving

the grafts, but it does emphasize the necessity for careful surgical techniques and prophylactic measures when employing an ahlograft. The use of preoperative antibiotics, for example, which was not examined in this study, might be a valuable adjunct. Many of these procedures involve extensive dissection for tumors, or d#{233}bridement with pre-

must decide if the risk of using able. We hope that this study easier.
NoTE: The authors wish to thank Mr. Vernon

allograft bone is acceptwill make that decision

Gambttl

for his valuable

asststance

in the

study.

References
I. 2.
3. 4. 5. Soc., 3: 21 1 . 1978. lncidence of Wound Infection on the Surgical Services. Surg. Clin. North America. 55: 1269-1275. 1975. FRIEDLAENDER. G. E.; STRONG, D. M.; and SELL, K. W.: Studies on the Antigenicity ofBone. I. Freeze-Dried and Deep-Frozen Bone Allografts in Rabbits. J. Bone and Joint Surg. , 58-A: 854-858, Sept. 1976. MCMASTER, P. E.. and HOHL, MASON: Tibiofibular Cross-Peg Grafting. J. Bone and Joint Surg., 57-A: 720-721, July 1975. MANKIN. H. J.; FOGELSON, R. S.; THRASHER. A. Z.; and JAFFER. FAROOQ: Massive Resection and Allograft Transplantation in the Treatment of Malignant Bone Tumors. New England J. Med., 294: 1247-1255, 1976. SPENCE. K. F.. JR.; SELL. K. W.; and BROWN, R. H.: Treatment of Unicameral Bone Cyst with Freeze-Dried Cancellous Bone Allograft. In Proceedings of The American Academy of Orthopaedic Surgeons. J. Bone and Joint Surg.. 50-A: 841-842. June 1968. A. Cortical Bone.
CRUSE.

BRIGHT. BRIGHT.

R. R.

W.: W.,

Decision Making and BURSTEIN.

in Tissue Procurement. Transpl. H.: Mechanical Properties of

Proc., Preserved

S (Supplement

1): 173-179, 1976. Trans. Orthop. Res.

P. J.

E.:

6. 7.

THE

JOURNAL

OF BONE

AND

JOINT

SURGERY