Antibiotics For Preventing Infection in Open Limb Fractures

Antibiotics for preventing infection in open limb fractures
(Review)
Gosselin RA, Roberts I, Gillespie WJ
This is a reprint of a Cochrane review, prepared and maintained by The Cochrane Collaboration and published in The Cochrane Library
2007, Issue 2
http://www.thecochranelibrary.com
Antibiotics for preventing infection in open limb fractures (Review) 1

Copyright © 2007 The Cochrane Collaboration. Published by John Wiley & Sons, Ltd
TABLE OF CONTENTS
ABSTRACT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
PLAIN LANGUAGE SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
BACKGROUND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
OBJECTIVES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
CRITERIA FOR CONSIDERING STUDIES FOR THIS REVIEW . . . . . . . . . . . . . . . . . . 2
SEARCH METHODS FOR IDENTIFICATION OF STUDIES . . . . . . . . . . . . . . . . . . . 2
METHODS OF THE REVIEW . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
DESCRIPTION OF STUDIES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
METHODOLOGICAL QUALITY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
DISCUSSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
AUTHORS’ CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
POTENTIAL CONFLICT OF INTEREST . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
ACKNOWLEDGEMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
SOURCES OF SUPPORT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Characteristics of included studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Characteristics of excluded studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
ANALYSES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
Comparison 01. Any antibiotic treatment versus control . . . . . . . . . . . . . . . . . . . . . 10
INDEX TERMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
COVER SHEET . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
GRAPHS AND OTHER TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Analysis 01.01. Comparison 01 Any antibiotic treatment versus control, Outcome 01 Early wound infection . . . 11
Antibiotics for preventing infection in open limb fractures (Review) i

Antibiotics for preventing infection in open limb fractures
(Review)
Gosselin RA, Roberts I, Gillespie WJ
This record should be cited as:

Gosselin RA, Roberts I, Gillespie WJ. Antibiotics for preventing infection in open limb fractures. Cochrane Database of Systematic
Reviews 2004, Issue 1. Art. No.: CD003764. DOI: 10.1002/14651858.CD003764.pub2.
This version first published online: 26 January 2004 in Issue 1, 2004.

Date of most recent substantive amendment: 29 October 2003
ABSTRACT
Background
Wound and bone infections are frequently associated with open fractures of the extremities and may add significantly to the resulting
morbidity. The administration of antibiotics is routinely used in developed countries as an adjunct to a comprehensive management
protocol that also includes irrigation, surgical debridement and stabilisation when indicated, and is thought to reduce the frequency of
infections.
Objectives
To quantify the evidence for the effectiveness of antibiotics in the initial treatment of open fractures of the limbs.
Search strategy
We searched the Cochrane Musculoskeletal Injuries Group specialised register (April 2003), the Cochrane Central Register of Controlled
Trials (The Cochrane Library issue 1, 2003), MEDLINE (1966 to April 2003), EMBASE (1988 to April 2003), LILACS (1992 to
June 2002) and reference lists of articles. Proceedings of meetings of the American Academy of Orthopaedic Surgeons (1980 to 2001),
the Orthopaedic Trauma Association (1990 to 2001) and the Société Internationale de Chirurgie Orthopedique et Traumatologique
(1980 to 2001) were hand searched. We also contacted published researchers in the field.
Selection criteria
Randomised or quasi-randomised controlled trials involving: participants - people of any age with open fractures of the limbs; in-
tervention - antibiotic administered before or at the time of primary treatment of the open fracture compared with placebo or no
antibiotic; outcome measures - early wound infection, chronic drainage, acute or chronic osteomyelitis, delayed unions or non-unions,
amputations and deaths.
Data collection and analysis

Two reviewers independently screened papers for inclusion, assessed trial quality using an eight item scale, and extracted data. Additional
information was sought from three trialists. Pooled data are presented graphically.
Main results
Data from 913 participants in seven studies were analysed. The use of antibiotics had a protective effect against early infection compared
with no antibiotics or placebo (relative risk 0.41 (95% confidence interval (CI) 0.27 to 0.63); absolute risk reduction 0.08 (95% CI
0.04 to 0.12); NNT 13 (95% CI 8 to 25)). There were insufficient data in the included studies to evaluate other outcomes.
Authors’ conclusions
Antibiotics reduce the incidence of early infections in open fractures of the limbs. Further placebo controlled randomised trials are
unlikely to be justified in middle and high income countries. Further research is necessary to the determine the avoidable burden of
morbidity in countries where antibiotics are not used routinely in the management of open fractures.
PLAIN LANGUAGE SUMMARY
Antibiotics are effective in preventing early infection in open fractures of the limbs
Wound and bone infections are common complications following open fractures of the limbs. For more than 20 years, in developed
countries, the use of antibiotics has been a part of a standard management protocol that also includes irrigation, surgical debridement,
and stabilisation when indicated. The review of trials found that antibiotics are effective in decreasing the incidence of wound infections,
as compared to no antibiotics or placebo. No studies on bone infection or long-term morbidity were identified.
BACKGROUND Types of participants

People of any age with open fractures of the limbs.
The use of antibiotics in the initial treatment of open fractures of
the extremities is almost universal in all high income countries, Types of intervention
where benefits are assumed to out-weight potential risks. In the Antibiotic administered before or at the time of primary treatment
USA, a 100 per cent compliance level is one of many required cri- of the open fracture compared with placebo or no antibiotic. Trials
teria for hospital re-credentialing by the Joint Committee for Ac- comparing different antibiotics, different antibiotic dosages, route
creditation of Healthcare Organisations (JCAHO). In many lower of administration or differences in timing or duration of admin-
income countries antibiotics may not be used routinely for various istration were excluded.
reasons: costs, lack of knowledge, low levels of suspicion or poor
Types of outcome measures
case-recognition from first-line health care workers, availability,
accessibility, use of traditional or alternative health care. In some of • early wound infection
these countries, the use of antibiotics for open fractures is delayed • chronic drainage
until the patient is seen in the secondary or even the tertiary care • acute or chronic osteomyelitis
centres. This delay often exceeds the accepted “golden period” of • delayed union or non-union
six to eight hours used in high income countries. Since an open • amputation
fracture is by definition contaminated, the use of antibiotics is • death
therapeutic, not prophylactic, and aims at preventing subsequent
infectious problems such as cellulitis, myositis, acute or chronic
SEARCH METHODS FOR
osteomyelitis, infected non-unions, recurrent abscesses, chronic
IDENTIFICATION OF STUDIES
drainage with fistula formation, and their associated impairment
and disability. See: Cochrane Bone, Joint and Muscle Trauma Group methods
A recent systematic review (Gillespie 2003) concluded that antibi- used in reviews.
otic prophylaxis should be offered to those undergoing surgical We searched the Cochrane Musculoskeletal Injuries Group
treatment of closed fractures. To our knowledge no such review specialised register (April 2003), the Cochrane Central Register
has been completed for the use of antibiotics in open fractures. of Controlled Trials (The Cochrane Library issue 1, 2003),
Our purpose is to review the evidence for use of antibiotics in the MEDLINE (1966 to April 2003), EMBASE (1988 to April
treatment of open fractures of the extremities. 2003), LILACS (1992 to June 2002) and reference lists of
articles. Proceedings of meetings of the American Academy of
Orthopaedic Surgeons (1980 to 2001), the Orthopaedic Trauma
OBJECTIVES
Association (1990 to 2001) and the Société Internationale de
Chirurgie Orthopedique et Traumatologique (1980 to 2001)
To quantify the evidence for the effectiveness of antibiotics in the
were hand searched. We also contacted published researchers in
initial treatment of open fractures of the limbs.
the field.
In MEDLINE (OVID WEB) the following search strategy was
CRITERIA FOR CONSIDERING combined with all three stages of the optimal trial search strategy
STUDIES FOR THIS REVIEW (Clarke 2003):
Types of studies 1. Antibiotic Prophylaxis/

2. exp Antibiotics/
Randomised and quasi-randomised (e.g. date of birth, alternation) 3. (antibiotic$ or antimicrob$).tw.
controlled trials. 4. or/1-3
5. Infection/ 1=not mentioned
6. exp Wound Infection/ 2=states numbers and reasons for withdrawal, but analysis
7. Sepsis/ unmodified
8. infect$.tw. 3=primary analysis based on all cases as randomised
9. or/5-8
C. Assessment of outcome. Were assessors of outcome blinded to
10. and/4,9
treatment status?
11. Fractures, Open/
1=not done or not mentioned
12. exp Fractures/
2=moderate chance of unblinding of assessors
13. (open or compound).tw.
3=action taken to blind assessors, or outcomes such that bias is
14. and/12-13
unlikely
15. (infect$ adj3 (bone$ or fracture$)).tw
16. or/11,14,15 D. Were treatment and control groups comparable at entry?
17. and/10,16 1=large potential for confounding or not discussed
2=confounding small; mentioned but not adjusted for
No language restriction was applied.
3=unconfounded; good comparability of groups or confounding
adjusted for
METHODS OF THE REVIEW E. Was a placebo treatment assigned as part of the randomisation?
1=No
Two reviewers independently screened each record for eligibility 3=Yes
by initially examining titles, abstracts and key words. Reports
F. Were exclusion criteria clearly defined?
identified by either reviewer were retrieved. We searched the
1=not defined
reference lists of relevant trials and reviews in which the largest
2=poorly defined
number of eligible trials had been published. We contacted the
3=well defined
authors of eligible trials and reviews to ask about unpublished
trials. Reports of potentially eligible trials were obtained and two G. Was the method of assessment of wound infection stated?
reviewers independently assessed each one for eligibility. Two 1=not stated
reviewers independently extracted data from included reports 2=clinical decision, or definite criteria without a microbiological
using a standard pro forma. Data were extracted on the type of diagnosis
participants, the type of antibiotic used, the number randomised 3=definite criteria including a microbiological diagnosis
to intervention or control groups, the quality of allocation
concealment, and the outcome measures stated in the protocol. H. Was the method and duration of surveillance stated?
We wrote to the authors of reports if relevant information was 1=not stated, or not active
missing. Trials in which we could not confirm that random or 2=active, but less than three months
quasi-random allocation had been used to allocate participants 3=active, and at least one year
were excluded. For each intervention, we estimated the pooled relative risk in
Quality assessment a fixed-effects model. We calculated 95 per cent confidence
intervals for each outcome. Selection bias was assessed using
Methodological quality for each study was independently assessed Egger’s weighted regression method and Begg’s rank correlation
by two reviewers using a schedule derived from the generic test and funnel plot. Heterogeneity was assessed using a Chi-square
evaluation tool developed by the Cochrane Musculoskeletal test at a five per cent significance level. A stratified analysis based
Injuries Group. Disagreements were resolved by consensus. The on the timing of antibiotic administration was not possible as the
scores were designed to give readers a general impression of trial necessary data were generally not reported.
quality and were not used in any quantitative manner.
A. Was the assigned treatment adequately concealed prior to

DESCRIPTION OF STUDIES
allocation?
1=states random, but no description, or quasi-randomisation
The search strategy identified eleven trials in which the use of any
(Category C)
antibiotics for open fractures of the limbs was compared to placebo
2=small but real chance of disclosure of assignment (Category B)
or no antibiotics. Four studies were excluded either because there
3=method did not allow disclosure of assignment (Category A)
was no randomisation (two), patients were randomised after an
B. Were the outcomes of patients who withdrew described and initial dose of antibiotics (one), or it was impossible to disaggre-
included in the analysis (intention to treat)? gate data on open fractures and soft tissue injuries only (one). The
remaining seven studies are described in detail in the “Character- confidence interval (CI) 0.27 to 0.63). The absolute risk of wound
istics of Included Studies” table. infection was 13.4 per cent (49/366) in the controls (no antibiotic)
and 5.5 per cent (30/547) in those receiving any antibiotics, giving
There were two randomised double-blinded trials (Bergman 1982;
an absolute risk reduction 0.08 (95% CI 0.04 to 0.12); NNT 13
Braun 1987), three quasi-randomised trials (Rojczyk 1983; Dickey
(95% CI 8 to 25), i.e. 13 patients would need to be treated with
1989; Suprock 1990), and two trials where the randomisation
antibiotics to prevent one wound infection.
methods were unclear (Patzakis 1974; Sloan 1987).
Some studies were limited to open fractures of fingers (Sloan 1987; When the analysis was stratified according to whether controls re-
Suprock 1990), some excluded hand and finger fractures (Rojczyk ceived placebo or not, the following results were found: the placebo
1983; Braun 1987), and the others included all fractures of the group had an absolute risk of wound infection of 24.3 per cent
extremities. Four studies included only patients in the ill-defined (18/74) for the controls and 5.8 per cent (6/103) for those receiv-
’adult age group’ (Bergman 1982; Braun 1987; Sloan 1987; Dickey ing antibiotics (risk difference 18.5 per cent, RR 0.24 ), while the
1989). Two studies included patients of all ages (Patzakis 1974; no-placebo group had an absolute risk of 10.6 per cent (31/292)
Rojczyk 1983) and one study did not mention age (Suprock 1990). for the controls and 5.4 per cent (24/444) for those receiving an-
tibiotics (risk difference 5.2%, RR 0.51).
Only two studies used a placebo (Bergman 1982; Braun 1987).
Many different antibiotic regimens were used: Penicillin/Strep-
tomycin and Cephalothin (Patzakis 1974), Penicillin and Di- DISCUSSION
cloxacillin (Bergman 1982), Cloxacillin (Braun 1987), Cefazolin
(Rojczyk 1983; Dickey 1989), first generation Cephalosporin, Di- The use of antibiotics has been part of the standard care of open
cloxacillin or Erythromycin (Suprock 1990), or Cephradine (Sloan fractures of the extremities since the mid 1970s. This is in great
1987). There were also differences in duration and route of ad- part due to the seminal article by Patzakis et al (Patzakis 1974),
ministration of treatment. which demonstrated a significant reduction in the infection rate in
those receiving Cephalothin, compared to those receiving nothing
or a combination of Penicillin and Streptomycin. Considering that
METHODOLOGICAL QUALITY
open fractures are by definition contaminated, this made empirical
sense. Over the next 16 years, only six more studies meeting this re-
The quality of the included studies varied from good to poor,
view’s inclusion criteria were added to form the evidence on which
with scores ranging between 11 and 20 out of a maximum score
this therapeutic rationale is based. All authors agree that antibio-
of 24, the mean being 14.2. Concealment was judged adequate
therapy is an adjunct, not a replacement for a comprehensive open
in only two studies (Bergman 1982; Braun 1987). The trialists’
fracture management protocol that includes early lavage, surgical
reported definitions of outcome varied, as did length of follow-up
debridement, fracture stabilisation when appropriate, and bone
periods, but wound infection required microbiological confirma-
grafting and wound coverage if necessary. In developed countries,
tion in five studies (Patzakis 1974; Bergman 1982; Rojczyk 1983;
antibiotics are now almost universally used in the management of
Braun 1987; Sloan 1987). Of the outcomes we initially wanted
open fractures, although a number of different regimens are still
to look at, the data permitted only the analysis of wound infec-
advocated. The wide variety of antibiotic regimens in the seven
tion. Most follow-up periods were too short to assess osteomyeli-
studies reflects the changes seen over time in the bacterial flora
tis, chronic drainage or infected non-unions, and there were no
involved, antimicrobial resistance and pharmacological advances.
reported amputations for infection or infection-related deaths.
Although it did not meet the inclusion criteria for this review,
the study by Miller (Miller 1986) showing similar results between
RESULTS the group receiving only one dose of antibiotics and the group
receiving a full course in type 1 and type 2 open fractures, may
Although there was no uniform definition of the main outcome raise some interesting hypotheses, particularly when applied to
measure, we considered that those used in each study were clini- low-resource environments - is a short (single dose) course indeed
cally sufficiently consistent to permit pooling. The antibiotic reg- as effective as a standard multiple dose regimen? If so, how do
imens varied significantly in terms of agents used and duration, different antibiotics compare from a cost-effectiveness standpoint?
but we found each regimen likely to be effective at the time of
the study against the bacterial flora usually associated with open POTENTIAL WEAKNESSES
fractures.
Despite the low quality and relative lack of power of most studies,
The data from 913 participants in the seven studies were pooled, the pooled data supports the effectiveness of antibiotics in decreas-
and it was found that antibiotics significantly reduced the inci- ing the incidence of wound infection following open fractures of
dence of wound infection, with a relative risk (RR) of 0.41, (95% the limbs, at least in the short term. The precision of the results

of this systematic review may be decreased by the relatively small ment of open fractures of the limbs. Appropriate antibiotics re-
number of included studies, and the relatively small number of flecting the local infectious agents and in line with sensitivity/re-
total participants (913). This is unlikely to change in the future sistance patterns should be used.
as it is doubtful that further trials using placebo or no antibiotics
Implications for research
will ever be done.
1. Further placebo controlled trials to evaluate the effectiveness of
The validity of the study may be weakened by the fact that only two
antibiotics for open fractures of the limbs are unwarranted.
studies were judged to have adequate concealment. The finding
that the stratified analysis appears to show a greater effect in the 2. For those populations for which antibiotics are not universally
blinded than the non-blinded studies may be slightly reassuring. used in the management of open fractures, reliable epidemiologi-
Another potential weakness comes from the relatively short periods cal data on the incidence and prevalence of open fractures, preva-
of follow-up in most studies, although it is impossible to estimate lence of antibiotic utilisation, and incidence and prevalence of
in which direction this would bias the results. On the other hand, infectious complications would allow quantifying the avoidable
very few patients were lost to follow-up, and all data analyses burden associated with this problem.
appear to have been done according to intent-to-treat.
3. Both in these low-resource settings, where management of open
Another potential weakness is selection bias, but we feel confident fractures is inconsistent, and in settings where antibiotic adminis-
that the search strategy was thorough and comprehensive enough tration is routine, there may be a justification for a study compar-
to minimise this possibility. To our knowledge, there is no other ing the benefits of single versus multiple doses of antibiotics, and
systematic review or meta-analysis with which to compare our re- which antibiotic by which route.
sults. The Eastern Association for the Surgery of Trauma (EAST
2000) has published a recent literature review leading to “practice 4. An economic analysis would then allow the most cost-effective
management guidelines for prophylactic antibiotic use in open intervention regimen to be determined, and to value the costs and
fractures” which is in agreement with the findings of this review. benefits associated with burden reduction.
Gillespie 2003 showed evidence for effectiveness of antibiopro-
phylaxis in the surgical management of closed fractures, which is
also concordant with our results. POTENTIAL CONFLICT OF
INTEREST
This review presents evidence for the effectiveness of antibiotics in
decreasing the rate of early infection in open fractures of the limbs. None known.
This intervention is not utilised as much in developing countries
as it is in developed countries. The extent to which results from
studies carried out in developed countries, where antibiotics are ACKNOWLEDGEMENTS
part of a comprehensive open fracture management protocol, are
applicable to developing countries is debatable. There are indeed The authors would like to acknowledge the following for their help
no studies from developing countries comparing antibiotics as in the preparation of this review: Ms Leeann Morton, Mrs Lesley
the main, or even the only initial treatment of open fractures Gillespie and Ms Frances Bunn. We would also like to thank the
of the limbs, to placebo or no treatment. Nevertheless, it seems following for their useful comments during the editorial review
reasonable to assume that, at least to some extent, the findings of process: Professor John Stothard, Professor Marc Swiontkowski,
this review would apply to low-resource settings. If good data on Professor Rajan Madhok, Mr Peter Herbison and Dr Janet Wale.
open fracture prevalence and incidence were available for a given
population, it could be possible to estimate the morbidity burden
that might be avoidable. An economic analysis could then evaluate
SOURCES OF SUPPORT
the costs and benefits.
External sources of support

AUTHORS’ CONCLUSIONS
• No sources of support supplied
Implications for practice
Internal sources of support
The use of antibiotics is an effective intervention in the manage- • No sources of support supplied

REFERENCES
References to studies included in this review References to studies excluded from this review
Almanza 1999
Bergman 1982 {published data only}
Almanza AJ, Reyes AG, Diaz RR. Antibiotic treatment in open frac-
Bergman BR. Antibiotic prophylaxis in open and closed fractures: a
tures [Tratamiento antimicrobiano en las fracturas expuestas]. Revisa
controlled clinical trial. Acta Orthopaedica Scandinavica 1982;53(1):
Mexicana Ortopedica Traumatologia 1999;13(5):470–1.
57–62.
Cutler 1944
Braun 1987 {published data only} Cutler EC, Morton PC, Sandusky WR. Observations on the prophy-
Braun R, Enzler MA, Rittmann WW. A double-blind clinical trial lactic use of penicillin in the wounds of aerial warfare. British Journal
of prophylactic cloxacillin in open fractures. Journal of Orthopaedic of Surgery 1944;32:207–11.
Trauma 1987;1(1):12–17.
Miller 1986
Dickey 1989 {published data only} Miller SD, Bray RC, Hughes GNF. Antibiotics in open fractures:
Dickey RL, Barnes BC, Kearns RJ, Tullos HS. Efficacy of antibi- a prospective randomised, double-blind study of wound infection
otics in low-velocity gunshot fractures. Journal of Orthopaedic Trauma [abstract]. Journal of Bone and Joint Surgery. British Volume 1986;68
1989;3(1):6–10. (5):850.
Peacock 1988
Patzakis 1974 {published data only}
Peacock KC, Hanna DP, Kirkpatrick K, Breidenbach WC, Lister GD,
∗
Patzakis MJ, Harvey JP, Ivler D. The role of antibiotics in the
Firrell J. Efficacy of perioperative cefamandole with postoperative
management of open fractures. Journal of Bone and Joint Surgery.
cephalexin in the primary outpatient treatment of open wounds of
American Volume 1974;56(3):532–41.
the hand. Journal of Hand Surgery 1988;13(6):960–4.
Patzakis MJ, Ivler D. Antibiotic and bacteriologic considerations in
Additional references
open fractures. Southern Medical Journal 1977;70 Suppl 1:46–8.
Clarke 2003
Rojczyk 1983 {published data only} Clarke M, Oxman AD, editors. MEDLINE highly sensitive search
Rojczyk M. Antibiotic doses in open fractures. Hefte zur Unfall- strategy for b.1) SilverPlatter-MEDLINE, b.2) OVID-MEDLINE,
heilkunde 1980;148:861–2. and b.3) PubMed. Cochrane Reviewer’s Handbook 4.2.0 [updated
March 2003]; Appendix 5B. In: The Cochrane Library [database on
∗
Rojczyk M. Treatment results in open fractures, aspects of antibiotic disk and CDROM]. The Cochrane Collaboration. Chichester, UK:
therapy [Behandlungsergebnisse bei offenen frakturen, aspekte der John Wiley & Sons; 2003, issue 4.
antibiotikatherapie]. Hefte zur Unfallheilkunde 1983;162:33–8. EAST 2000
Luchette FA, Bone LB, Born CT, DeLong WG, Hoff WS, Mullins
Rojczyk M, Malottke R. The effect of antibiotic prophylaxis in the
D, et al. EAST practice management guidelines work group: Practice
treatment of open fractures. Hefte zur Unfallheilkunde 1979;138:
management guidelines for prophylactic antibiotic use in open frac-
355–7.
tures. Eastern Association for the Surgery of Trauma, http://www.
Sloan 1987 {published data only} east.org/tgp/openfrac.pdf accessed 01/03/02.
Sloan JP, Dove AF, Maheson M, Cope AN, Welsh KR. Antibiotics in Gillespie 2003
open fractures of the distal phalanx?. Journal of Hand Surgery. British Gillespie WJ, Walenkamp G. Antibiotic prophylaxis for surgery for
Volume 1987;12(1):123–4. proximal femoral and other closed long bone fractures (Cochrane
Review). In: The Cochrane Library, 1, 2003. Chichester, UK: John
Suprock 1990 {published data only}
Wiley & Sons.
Suprock MD, Hood JM, Lubahn JD. Role of antibiotics in open
fractures of the finger. Journal of Hand Surgery. American Volume
1990;15(5):761–4. ∗
Indicates the major publication for the study
TABLES
Characteristics of included studies
Study Bergman 1982

Methods RCT
Double-blinded
Characteristics of included studies (Continued )
Placebo
Cultures for diagnosis
Quality Score: 20
Participants 90 over 30 months:
60 cases treated with dicloxacillin or benzyl penicillin
30 controls with placebo
Interventions 1. Intervention: 2g dicloxacillin: first dose given pre-operatively with infusions repeated every six hours for
two days.
2. Intervention: 3 million units benzyl penicillin: first dose given pre-operatively with infusions repeated
every six hours for two days.
3. Control: 100 ml saline: in the same dosing regime as for the intervention groups.
Outcomes Early wound infection: a wound was considered infected when signs of inflammation were present, i.e.
suprafascial drainage and a positive bacterial culture.
Deep infection was defined as a subfascial process going down to the bone or osteosynthesis material.
Superficial thrombophlebitis was defined as a palpable fibrotic vessel or visible inflammation along the course
of the vessel.
Notes F/U until wound healed or drainage
Allocation concealment A – Adequate
Study Braun 1987

Methods RCT
Double-blinded
Placebo
Quality Score: 19
Participants 100 consecutive open fractures- time?
13 excluded (hand fractures, skull fractures, colon injury, previous use of antibiotics)
87 participants:
43 cases treated with cloxacillin
44 controls treated with placebo
Interventions 1. Intervention: During the first four days, four 1g doses of cloxacillin were administered intravenously.
Thereafter, four 1.5g doses of cloxacillin were given orally for six days.
2. Control: Indistinguishable placebo given in the same dosing regimen as for the intervention group.
Outcomes Early wound infection: Wound swabs were taken at weekly intervals from the base of the wound and the
surrounding skin.
Notes F/U up to 10 months
Cloxacillin group:
- 2 urticaria
- 3 Gastro-intestinal symptoms
- 1 phlebitis
Placebo group:
- 7 phlebitis
Allocation concealment A – Adequate
Study Dickey 1989

Methods CCT quasi-randomised (alternate days)
No blinding
No placebo
No cultures
Quality Score: 11
46 cases with treatment
50 controls without treatment
Results reported on 32 cases treated with cefazolin, 35 without treatment
Interventions 1. Intervention: One day of intravenous cefazolin, three 1g doses given eight hourly.
2. Control: No antibiotics
Outcomes Early wound infection: any wound complication including prolonged drainage, erythema, or any physical
findings such as cellulitis, localised fluctuance or drainage.
Notes F/U until bony union
30% loss to F/U
low-velocity gunshot wounds only
case-definition?
Allocation concealment C – Inadequate
Study Patzakis 1974

Methods RCT
No blinding
No placebo
Quality Score: 12
Participants 310 consecutive patients, over 12 months
212 cases treated with penicillin and streptomycin, or cephalotin
Interventions 1. Intervention: Adults received 10 million units benzyl penicillin per day in a continuous infusion and 0.5g
streptomycin intramuscularly every twelve hours. Children received 100,000 units of penicillin per kilogram
of body weight per day in a continuous infusion and 7.5mg of streptomycin per kilogram of body weight
every twelve hours intramuscularly.
2. Intervention: Both adults and children received cephalothin, 100mg per kilogram of body weight per day
in divided dosage intravenously every six hours.
3. Control: No antibiotics.
Outcomes Early wound infection: clinical signs and symptoms of wound infection present such as fever, erythema,
tenderness, and wound drainage with either a positive gram stain or a positive culture. Either of the latter
two had to be present before the wound was classified as infected.
Notes Length of F/U not specified.
If gunshot wounds excluded (none got an infection) infections in cases: 11 in 176
Infections in controls: 11 in 79
Study Rojczyk 1983

No blinding
No placebo
Quality Score: 12
111 cases treated with cefazolin
Interventions 1. Intervention: Cefazolin 1g intravenously every six hours for five days.
2. Control: no antibiotics
Outcomes Early wound infection
Notes Hand and feet open fractures excluded.
80% F/U at 1 year
Allocation concealment D – Not used
Study Sloan 1987

Methods RCT
No blinding
No placebo
Quality Score: 12
Participants Initially 85, but stopped after 40 because of 30% infection rate in placebo group
Study continued with remaining three antibiotics groups- time?
40 participants:
30 cases treated with one of three regimen of cephradine
10 controls treated with placebo
Interventions 1. Intervention: Cephradine 500mg orally every six hours for five days
2. Intervention: Cephradine 1g intravenously followed by 500mg orally every six hours for five days
3. Intervention: Cephradine 1g intravenously followed by one dose of 1g orally
Control: Placebo
Outcomes Early wound infection: Swabs were taken if there was any evidence of infection (erythema, pus or exudate).
Notes Only fingers
F/U 5 days
8 lost to F/U
Study Suprock 1990

No blinding
No placebo
No cultures
Quality Score: 13
45 cases treated with cephalosporin, dicloxacillin or erythromycin
Interventions 1. Intervention: either a first generation cephalosporin, dicloxacillin or erythromycin for three days - dose
and route of administration were not reported.
2. Control: no antibiotics.
Outcomes Early wound infection: clinical signs of infection - full details not reported.
Notes Only fingers
F/U up to one year
No antibiotic dosage reported.
Cultures done only if clinical suspicion.
F/U: follow-up

Characteristics of excluded studies
Study Reason for exclusion

Almanza 1999 CCT. Non randomised comparison.
Cutler 1944 CCT. Non randomised comparison. Fractures are not seperated from soft tissue injuries only.
Miller 1986 RCT. All participants received an initial dose of antibiotics, then were randomised to more or no more antibiotics.
Peacock 1988 RCT. Fractures are not separated from soft tissue injuries only.
ANALYSES
Comparison 01. Any antibiotic treatment versus control
No. of No. of
Outcome title studies participants Statistical method Effect size
01 Early wound infection 7 913 Relative Risk (Fixed) 95% CI 0.41 [0.27, 0.63]
INDEX TERMS
Medical Subject Headings (MeSH)
∗ Antibiotic Prophylaxis; Fractures, Open [∗ complications]; Randomized Controlled Trials; Wound Infection [∗ prevention & control]
MeSH check words
Humans
COVER SHEET
Title Antibiotics for preventing infection in open limb fractures
Authors Gosselin RA, Roberts I, Gillespie WJ
Contribution of author(s) All reviewers contributed to the protocol development and the editing of the review. One
reviewer (RG) executed the search strategy. Two reviewers (RG and IR) independently
assessed trial quality and extracted data. Richard Gosselin is the guarantor of the review.
Issue protocol first published 2002/3
Review first published 2004/1
Date of most recent amendment 23 August 2005
Date of most recent 29 October 2003
SUBSTANTIVE amendment
What’s New Information not supplied by author
Date new studies sought but 08 April 2003
none found
Date new studies found but not Information not supplied by author
yet included/excluded
Date new studies found and Information not supplied by author

included/excluded
Date authors’ conclusions Information not supplied by author
section amended
Contact address Dr Richard Gosselin
PO Box 1983
El Granada
California
94018
USA
E-mail: [email protected]
Fax: +001 650 712 8928
DOI 10.1002/14651858.CD003764.pub2
Cochrane Library number CD003764
Editorial group Cochrane Bone, Joint and Muscle Trauma Group
Editorial group code HM-MUSKINJ
GRAPHS AND OTHER TABLES
Analysis 01.01. Comparison 01 Any antibiotic treatment versus control, Outcome 01 Early wound infection
Review: Antibiotics for preventing infection in open limb fractures
Comparison: 01 Any antibiotic treatment versus control
Outcome: 01 Early wound infection
Study Antibiotic None Relative Risk (Fixed) Weight Relative Risk (Fixed)
n/N n/N 95% CI (%) 95% CI
Bergman 1982 4/60 6/30 13.7 0.33 [ 0.10, 1.09 ]
Braun 1987 2/43 12/44 20.3 0.17 [ 0.04, 0.72 ]
Dickey 1989 1/46 1/50 1.6 1.09 [ 0.07, 16.88 ]
Patzakis 1974 11/212 11/98 25.8 0.46 [ 0.21, 1.03 ]
Rojczyk 1983 8/111 12/88 22.9 0.53 [ 0.23, 1.24 ]
Sloan 1987 0/30 3/10 8.9 0.05 [ 0.00, 0.91 ]
Suprock 1990 4/45 4/46 6.8 1.02 [ 0.27, 3.84 ]
Total (95% CI) 547 366 100.0 0.41 [ 0.27, 0.63 ]

Total events: 30 (Antibiotic), 49 (None)
Test for heterogeneity chi-square=6.30 df=6 p=0.39 I² =4.8%
Test for overall effect z=4.02 p=0.00006
0.001 0.01 0.1 1 10 100 1000

Favours treatment Favours control


Antibiotics For Preventing Infection in Open Limb Fractures

Uploaded by

Copyright:

Available Formats

Antibiotics For Preventing Infection in Open Limb Fractures

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Antibiotics For Preventing Infection in Open Limb Fractures

Uploaded by

Copyright:

Available Formats

Antibiotics for preventing infection in open limb fractures

Gosselin RA, Roberts I, Gillespie WJ

Antibiotics for preventing infection in open limb fractures (Review) 1

Antibiotics for preventing infection in open limb fractures (Review) i

Gosselin RA, Roberts I, Gillespie WJ

This record should be cited as:

This version first published online: 26 January 2004 in Issue 1, 2004.

Data collection and analysis

BACKGROUND Types of participants

Types of studies 1. Antibiotic Prophylaxis/

A. Was the assigned treatment adequately concealed prior to

Antibiotics for preventing infection in open limb fractures (Review) 4

External sources of support

Antibiotics for preventing infection in open limb fractures (Review) 5

Characteristics of included studies

Study Bergman 1982

Study Braun 1987

Study Dickey 1989

Study Patzakis 1974

Study Rojczyk 1983

Study Sloan 1987

Study Suprock 1990

Antibiotics for preventing infection in open limb fractures (Review) 9

Study Reason for exclusion

Comparison 01. Any antibiotic treatment versus control

Date new studies found and Information not supplied by author

GRAPHS AND OTHER TABLES

Bergman 1982 4/60 6/30 13.7 0.33 [ 0.10, 1.09 ]

Braun 1987 2/43 12/44 20.3 0.17 [ 0.04, 0.72 ]

Dickey 1989 1/46 1/50 1.6 1.09 [ 0.07, 16.88 ]

Patzakis 1974 11/212 11/98 25.8 0.46 [ 0.21, 1.03 ]

Rojczyk 1983 8/111 12/88 22.9 0.53 [ 0.23, 1.24 ]

Sloan 1987 0/30 3/10 8.9 0.05 [ 0.00, 0.91 ]

Suprock 1990 4/45 4/46 6.8 1.02 [ 0.27, 3.84 ]

Total (95% CI) 547 366 100.0 0.41 [ 0.27, 0.63 ]

0.001 0.01 0.1 1 10 100 1000

Antibiotics for preventing infection in open limb fractures (Review) 11

You might also like