The principles of prophylaxis against post-surgical infection were established in the laboratory in the early 1960s (Burke 1961). The administration of antibiotic prior to surgery is now widely accepted. The period of administration of prophylaxis has been reduced but the optimal duration remains uncertain. Antibiotic prophylaxis during the operative management of closed fractures has been claimed to reduce infection rates from around five per cent to less than one per cent (Bodoky 1993). As the pathogenesis of post-surgical infection is similar after osteosynthesis of any closed fracture, it has been suggested that combining data from similar prophylactic regimens used during different surgical procedures is quite appropriate (Platt 1991).
Closed hip fractures in the elderly are common, and surgical management is normal in the developed world. The majority of studies of the efficacy of antibiotic prophylaxis in closed fracture fixation have focused on this group of patients. Early randomised trials completed in the 1970s suggested a small but definite prophylactic effect. These trials were individually small, and some used prolonged courses of antibiotic. Many further trials have been reported over the following 25 years. These have addressed a range of issues: duration of administration, route of administration, and antimicrobial spectrum.
A number of descriptive reviews of antibiotic prophylaxis in orthopaedic surgery have been published (Doyon 1989; Norden 1991), in which some attempt has been made to assess methodological quality and include this in the interpretation of results. However, there is sufficient persisting uncertainty about the efficacy, optimal duration, and cost-effectiveness of antibiotic prophylaxis during the surgical treatment of hip and other long bone fractures to justify a systematic review of the evidence from randomised trials.
The objective of this review is to determine whether the prophylactic administration of antibiotics in patients undergoing surgical management of hip or other closed long bone fractures reduces the incidence of wound and other hospital acquired infections.
The following hypotheses are tested:
1. Antibiotic prophylaxis leads to a reduction in the proportion of patients developing a wound infection, either deep or superficial, compared with those given a placebo or no prophylaxis.
2. 'Single dose' antibiotic prophylaxis leads to a significant reduction in the proportion of patients developing a wound infection compared with those given 'longer duration' prophylaxis.
3. There is a significant reduction in the proportion of patients with a post-operative wound infection who receive prophylaxis using broad-spectrum antibiotics when compared with those who receive narrow spectrum agents.
4. Antibiotic prophylaxis leads to a significant reduction in the proportion of patients developing septicaemia, respiratory or urinary tract infection, compared with those given a placebo or no prophylaxis.
5. 'Single dose' antibiotic prophylaxis leads to a significant reduction in the proportion of patients developing septicaemia, respiratory, or urinary tract infection after surgical management of a hip or other long bone fracture compared with those given three or more doses.
6. Oral administration of a prophylactic regimen leads to a significant reduction in the proportion of participants experiencing a wound infection, respiratory or urinary tract infection, or adverse drug effect, compared with those receiving parenteral prophylaxis.
7. There is a significant increase in the proportion of patients with conditions such as gastro-intestinal symptoms or skin reactions in those allocated antibiotic prophylaxis when compared with those receiving placebo or no prophylaxis.
Criteria for considering studies for this review
Types of studies
The predetermined inclusion criteria were broad so as to include any controlled study testing a prophylactic antibiotic in closed fracture surgery.
1. The study must test some method of antibiotic prophylactic intervention aimed at reducing the wound infection rate in closed fracture surgery and compare it against a placebo or alternative intervention group.
2. The study must be a controlled study, randomised or quasi-randomised.
3. The study population must be defined to enable identification of the operative intervention, ideally with relevant subgroups given if more than one.
4. Wound infection must be one of the primary outcome measures.
Types of participants
Any person undergoing surgery for internal fixation or replacement arthroplasty as treatment for a closed fracture of the proximal femur, or any other long bone.
Types of interventions
Any regimen of systemic antibiotic prophylaxis administered at the time of surgery.
Types of outcome measures
1. Wound infection. The reference definition of wound infection for quality assessment was:
Deep wound infection -
A surgical wound infection which occurs within one year, if an implant is in place and infection involves tissues or spaces at or beneath the fascial layer.
Superficial wound infection -
A surgical wound infection which occurs at the incision site within 30 days after surgery and involves the skin subcutaneous tissue, or muscle located above the fascial layer. In the assessment of methodological quality, more weight was given to studies in which wound infection had been confirmed by microbiological analyses.
2. Urinary tract infection
3. Respiratory tract infection
4. Adverse reaction to antibiotic (gastro-intestinal symptoms, skin reactions)
5. Cost-effectiveness outcomes - length of hospital stay
- reoperation due to infection.
Search methods for identification of studies
Trials were identified by searches up to the end of August 2000.
i. The Cochrane Central Register of Controlled Trials, The Cochrane Library, Issue 3, 2000 was searched.
ii. MEDLINE (1966-October 2000); OVID web (see Appendix 1)
iii. EMBASE (to October 2000), OVID web (see Appendix 1)
iv. LILACS, Current Contents, Dissertation Abstracts, and Index to UK Theses were searched to the end of August 2000.
v. The bibliographies of identified literature reviews, original articles, and relevant chapters of published books were examined for undetected articles.
vi. Contacts were made in person or by mail with identified trialists and other experts in the field.
No language restriction was applied.
Data collection and analysis
Selection of studies
All identified studies were read. We included all randomised or quasi-randomised controlled trials reporting the results of antibiotic prophylaxis in patients undergoing hip or other long bone fracture surgery with wound infection as one of the primary outcome measures. Reports in which participants were not allocated at enrolment in a randomised or quasi-randomised fashion into treatment or control groups were excluded.
Methodological assessment was undertaken by two raters, using the criteria described in Appendix 2, supplemented by a pre-designed coding manual. Disagreement was resolved by discussion between raters.
Data were independently extracted by two reviewers and adjudicated by a third using a data extraction tool which had undergone prior testing. We approached two groups of trialists for clarification of data relevant to the review.
Using Review Manager, for each study, risk ratios and 95 per cent confidence intervals were calculated for dichotomous outcomes. For comparable groups of trials pooled odds ratios with 95 per cent confidence intervals were derived, and where appropriate absolute risk reduction was calculated.
Description of studies
The search strategy identified 48 trials in which antibiotic prophylaxis had been compared with no treatment, a placebo, or another antibiotic regimen in orthopaedic surgery. Twenty-five of these were excluded either because the participants had sustained an open fracture prior to the administration of antibiotics, because participants had not sustained a fracture, or because no usable data were reported for a fracture fixation subgroup in a wider study. In 23 studies, data were available for patients undergoing closed fracture surgery. We were unable to assess one study (Chiu 1993), which has been reported only in abstract. The remaining 22 studies are described in detail in the 'Characteristics of included studies' table. Some reported more than one comparison.
The comparisons evaluated were:
1. A pre-operative dose and two or more post-operative doses of parenteral (injected) antibiotic compared with a placebo or with no treatment. There were 10 trials in this category, in seven of which the participants underwent hip fracture surgery (Ericson 1973; Boyd 1973; Tengve 1978; Burnett 1980; Hedstrom 1987; Buckley 1990; Bodoky 1993), and in three of which (Bergman 1982; Gatell 1984; Paiement 1994) other closed fracture fixation procedures were carried out.
2. A single preoperative dose of parenteral antibiotic compared with a placebo or no treatment. There were seven trials in this category, in five of which (Buckley 1990; Hjortrup 1990; Luthje 2000; McQueen 1990; Kaukonen 1995) the participants underwent hip fracture surgery, and in two of which a range of closed fracture fixation procedures were carried out (Hughes 1991; Boxma 1996).
3. A single dose of parenteral antibiotic compared with multiple doses of the same agent. There were two trials in this category, in one of which (Buckley 1990) the participants underwent hip fracture surgery, and in the other (Gatell 1987), a range of different closed fracture fixation procedures were carried out.
4. A single dose of parenteral antibiotic using an agent with a long half-life, compared with multiple doses of other agents with shorter half lives. There were three trials in this category, of which one (Garcia 1991) included both participants undergoing hip fracture surgery and others undergoing other closed fracture surgery. In Karachalios 1987 the participants underwent hip fracture surgery, and in Jones 1987 B a range of different closed fracture fixation procedures were carried out.
5. Multiple doses of parenteral antibiotic administered over 24 hours or less, compared with a longer period of administration. Two trials (Nelson 1983; Hedstrom 1987) whose participants underwent hip fracture surgery reported comparisons in this category.
6. Oral administration of antibiotic compared with parenteral administration. There was one trial (Nungu 1995) in this category.
Risk of bias in included studies
The quality of most studies as reported was poor to moderate. The score for each attribute is recorded for each trial in the 'Characteristics of included studies' table. Only four of the 22 published trials were adequately powered to test the research hypothesis with confidence. In seven trials the description of the randomisation process indicated that prior concealment of allocation was satisfactory, in 12 it was unclear, and in three assignment was inadequately concealed. Placebos were used appropriately in 12 trials, and were not employed where their use would have been appropriate in 10 trials. In trials with a short (inadequate) follow up, analysis by intention to treat was usually possible, but in those in which follow up would have been likely to identify late onset post-operative infection, losses were larger. The trialists' reported definitions of outcome measures varied. In most reports, the definition of wound infection was clinical and did not require microbiological confirmation, although it was often sought. Blinding of the outcome assessor was rarely mentioned, although it could be assumed in those studies which were fully placebo controlled.
Effects of interventions
Results in each comparison category are shown in the analyses. We accepted that there was variation between studies in the reported definition of the main outcome measures, but considered that these definitions were clinically sufficiently consistent to permit pooling. Also, although the regimens of antibiotic prophylaxis within each category also varied in respect of the agent and the details of timing, we found that all reported trials employed agents likely to be widely effective at the time of the study against Staphylococcus aureus, the principal organism implicated in post-operative wound infection.
1. A pre-operative dose and two or more post-operative doses of parenteral (injected) antibiotic compared with a placebo or with no treatment.
Data from 1896 participants from 10 trials (11 datasets) were pooled. Regimens in this category significantly reduced the incidence of deep wound infection (risk ratio (RR) 0.36, 95% confidence intervals (CI) 0.21, 0.65), and of superficial wound infection (RR 0.48, 95% CI 0.28 to 0.81), and reduced the incidence of infection of the urinary tract (RR 0.66, 95% CI 0.43 to 1.00). There was no significant reduction in the rate of respiratory infection (RR 0.81, 95% CI 0.41 to 1.63). Adverse effects were rarely reported but appeared more common in participants given antibiotics (RR 1.83, 95% CI 0.96 to 3.50). The absolute risk of deep wound infection in the control patients was 4.3% (0.043), and the risk difference -0.03, (95% CI -0.04 to -0.01).
2. A single preoperative dose of parenteral antibiotic compared with a placebo or no treatment.
Data were pooled from seven trials, including one large multi-centre trial of good quality (Boxma 1996). Regimens in this category reduced the incidence of deep wound infection (RR 0.40, 95% CI 0.24 to 0.67), superficial wound infection (RR 0.69, 95% CI 0.50 to 0.95), urinary tract infection (RR 0.63, 95% CI 0.53 to 0.76), and respiratory infection (RR 0.46, 95% CI 0.33 to 0.65). The absolute risk of deep wound infection in the control patients was 3% (0.03), and the absolute risk reduction -0.02 (95% CI -0.03 to -0.01).
3. A single dose of short-acting parenteral antibiotic compared with multiple doses of the same agent.
Data were pooled from two trials, one of which (Gatell 1987) dominates the analysis by virtue of its size. This trial is of moderate quality. The analysis indicates that a single dose, in the circumstances described in Gatell 1987 was less effective (with marginal statistical significance) in preventing deep wound infection (risk ratio 7.89, 95% CI 1.01 to 61.98), superficial wound infection (RR 4.82, 95% CI 1.08 to 21.61) and urinary tract infection (RR 1.81 to 95% CI 1.01 to 3.23) after surgery for closed fracture than a multiple dose regimen (see 'Discussion').
4. A single dose of parenteral antibiotic using an agent with a long half-life, compared with multiple doses of other agents with shorter half life.
Data were pooled from three trials in this category. One of these (Garcia 1991) was a large trial of moderate quality, but despite its size, analysis of the pooled data failed to show a significant difference between the two types of regimen for the outcomes of deep wound infection (RR 0.57, 95% CI 0.20 to 1.64), superficial wound infection (RR 1.01, 95% CI 0.35 to 2.93), urinary tract infection (RR 0.69, 95% CI 0.37 to 1.32), or respiratory infection (RR 0.31, 95% CI 0.04 to 2.48).
5. Multiple doses of parenteral antibiotic administered over 24 hours or less, compared with a longer period of administration.
Data were pooled from the two small trials in this category. There was no evidence of difference between the two types of regimen for the outcomes of deep (RR 1.10, 95% CI 0.22 to 5.34) or superficial wound infection (RR 0.57, 95% CI 0.17 to 1.93).
6. Oral administration of antibiotic compared with parenteral administration.
Only one small trial (Nungu 1995) has evaluated this comparison for the outcomes of deep and superficial wound infection, and urinary infection. No significant difference between the routes was demonstrated (deep infection: RR 0.29, 95% CI 0.01 to 7.07; superficial infection: RR 0.17, 95% CI 0.02 to 1.47; urinary infection: RR 1.10, 95% CI 0.62 to 1.94). This trial was underpowered to identify any difference between oral and parenteral routes of administration.
Antibiotic prophylaxis has been widely used in fracture surgery since the 1970s. The 22 trials reported span over quarter of a century. In the early studies, penicillins effective against gram positive cocci were used. As resistant organisms appeared, subsequent generations of penicillins, and then cephalosporins have been trialed in prophylaxis, both against placebo or no treatment, and against other agents. As cephalosporins have a wider antimicrobial spectrum than the penicillins used in the early studies, their efficacy in reducing the occurrence of infections of the urinary and respiratory tracts, which were secondary outcome measures in many of the included trials, may be superior.
Over time, shorter durations of prophylaxis have been used. For effective prophylaxis, the minimum inhibitory concentration (MIC) of the antibiotic in the tissues must be exceeded for at least the period from incision to wound closure (Burke 1961). In practice, this initially meant using regimens with several consecutive doses, and the pooled data support their effectiveness. The availability of agents with a long elimination half life has allowed single dose prophylaxis reliably to meet Burke's prerequisite. Amongst published trials the large multi-centre Netherlands trial (Boxma 1996), which used an antibiotic which provided concentrations exceeding MIC for 12-24 hours, dominates the analysis in this category. This trial supported the hypothesis that single dose intravenous prophylaxis is effective in reducing the incidence of deep wound infection, superficial wound infection, and urinary and respiratory tract infections. The effect sizes are similar to those of from multi-dose prophylaxis.
Despite the lack of power in many of the individual studies, pooling of the accumulated evidence supports the hypothesis of effectiveness of antibiotic prophylaxis. Boxma 1996 included an economic evaluation based on the effectiveness data which indicates a cost saving of a little under 500US dollars per patient given prophylaxis. This may be conservative as the evaluation does not appear to take into account the costs of urinary and respiratory infections prevented. Albers and colleagues (Albers 1994) have calculated that antibiotic prophylaxis for closed fracture surgery is cost effective if the absolute risk reduction is 0.25% or more; the pooled estimates for absolute risk reduction for single dose or multiple dose prophylaxis in this review exceed that. Therefore, without taking into account the effects of wide adoption of prophylactic regimens on the development of antibiotic resistance, antibiotic prophylaxis appears cost-effective.
Direct comparisons of multiple and single dose prophylaxis have also been conducted by Gatell 1987 and Buckley 1990. Buckley 1990 was underpowered to distinguish between the two regimens tested. Gatell 1987, a large single centre study, concluded that single dose prophylaxis was inferior , but the statistical significance is marginal. Also, the tissue MIC may not have been reliably exceeded throughout the procedure for all participants in Gatell 1987, in which an agent (cefamandole 2g) with a short half life was administered 30 minutes prior to the planned onset of surgery.
Implications for practice
1. Antibiotic prophylaxis for closed fracture surgery is an effective intervention. Application of cost-modelling to the effectiveness data has indicated that it also appears to be cost-effective, in the absence of any data on the impact on the development of antibiotic resistance.
Implications for research
1. Further placebo controlled trials to evaluate the effectiveness of antibiotic prophylaxis in closed fracture surgery would be unlikely to be ethical.
The authors acknowledge the help of the following who assisted with searching, retrieval of studies, and methodological assessment: Dr Helen Handoll, Mrs Lesley Gillespie, Dr Chris Hoffman. We would also like to thank the following for useful comments from the initial editorial review in 1998: Dr Helen Handoll, Dr Rajan Madhok, Prof Gordon Murray and Dr Antony Berendt.
Data and analyses
- Top of page
- Authors' conclusions
- Data and analyses
- What's new
- Contributions of authors
- Declarations of interest
- Sources of support
- Index terms
Appendix 1. Search strategies
MEDLINE (OVID Web)
1 randomized controlled trial.pt.
2 controlled clinical trial.pt.
3 random allocation/
4 double blind method/
5 single blind method/
6 exp Cross-Over Studies/
8 ((clinical or controlled or comparative or placebo or prospective$ or randomi#ed) adj3 (trial or study)).tw.
9 (random$ adj7 (allocat$ or allot$ or assign$ or basis$ or divid$ or order$)).tw.
10 ((singl$ or doubl$ or trebl$ or tripl$) adj7 (blind$ or mask$)).tw.
11 (cross?over$ or (cross adj1 over$)).tw.
12 ((allocat$ or allot$ or assign$ or divid$) adj3 (condition$ or experiment$ or intervention$ or treatment$ or therap$ or control$ or group$)).tw.
15 exp Orthopedic Procedures/
16 exp Orthopedic Equipment/
17 exp Orthopedics/
18 exp Fractures/
19 exp Fracture Healing/
20 exp Amputation, Traumatic/
21 exp Arm Injuries/
22 exp Athletic Injuries/
23 exp Dislocations/
24 exp Hand Injuries/
25 exp Leg Injuries/
26 exp Soft Tissue Injuries/
27 exp Contusions/
28 exp Contusions/
29 exp "Sprains and Strains"/
30 exp Tendon Injuries/
31 exp Arthroscopy/
33 (hip$ or femur$ or femoral$ or trochant$ or pertrochant$ or intertrochant$ or subtrochant$ or intracapsular$ or extracapsular$ or condyl$ or shaft$ or diaphys$).tw.
34 (pelvi$ or acetabul$ or knee$ or patell$ or tibia$ or fibula$ or pillon or malleol$ or ankle$ or calcan$ or tarsal$ or talus$ or talar$ or cuneiform$ or navicular$ or cuboid$ or metatars$ or phalang$).tw.
35 (leg$1 or (lower adj limb$) or arm$1 or (upper adj limb$) or shoulder$ or humer$ or radi$ or ulna$ or scapula$).tw.
36 (elbow$ or wrist$ or supracondyl$ or epicondyl$ or metacarpal$ or hand$ or finger$ or trapez$ or hamate$ or capitate$ or scaphoid$ or lunate$ or triquetral$ or pisiform$).tw.
37 ((styloid$ or coronoid$) adj2 process$).tw.
38 (ligament$ or tendon$ or menisc$).tw.
40 (fracture$ or injur$ or wound$ or dislocat$ or sublux$ or rupture$ or tear$ or torn).tw.
42 (fracture$ adj3 (open or closed)).tw.
45 14 and 44
46 Animal/ not Human/
47 45 not 46
48 exp Antibiotic Prophylaxis/
49 Surgical Wound Infection/ or Postoperative Complications/ or Wound Infection/
50 (antibiotic$ or antimicrob$).tw.
EMBASE (OVID Web)
1. exp Randomized Controlled trial/
2. exp Double Blind Procedure/
3. exp Single Blind Procedure/
4. exp Crossover Procedure/
6. ((clinical or controlled or comparative or placebo or prospective$ or randomi#ed) adj3 (trial or study)).tw.
7. (random$ adj7 (allocat$ or allot$ or assign$ or basis$ or divid$ or order$)).tw.
8. ((singl$ or doubl$ or trebl$ or tripl$) adj7 (blind$ or mask$)).tw.
9. (cross?over$ or (cross adj1 over$)).tw.
10. ((allocat$ or allot$ or assign$ or divid$) adj3 (condition$ or experiment$ or intervention$ or treatment$ or therap$ or control$ or group$)).tw.
13. Animal/ not Human/
14. 12 not 13
15. exp Orthopedic Surgery/
16. exp Orthopedics/
17. exp Orthopedic Equipment/
18. exp Fracture Healing/
19. exp Avulsion Injury/
20. exp Contusion/
21. exp Limb Injury/
22. exp Musculoskeletal Injury/
23. exp Pelvis Injury/
24. exp Traumatic Amputation/
25. exp Soft Tissue Injury/
26. exp Sport Injury/
28. (hip$ or femur$ or femoral$ or trochant$ or pertrochant$ or intertrochant$ or subtrochant$ or intracapsular$ or extracapsular$ or condyl$ or shaft$ or diaphys$).tw.
29. (pelvi$ or acetabul$ or knee$ or patell$ or tibia$ or fibula$ or pillon or malleol$ or ankle$ or calcan$ or tarsal$ or talus$ or talar$ or cuneiform$ or navicular$ or cuboid$ or metatars$ or phalang$).tw.
30. (leg$1 or (lower adj limb$) or arm$1 or (upper adj limb$) or shoulder$ or humer$ or radi$ or ulna$ or scapula$).tw.
31. (elbow$ or wrist$ or supracondyl$ or epicondyl$ or metacarpal$ or hand$ or finger$ or trapez$ or hamate$ or capitate$ or scaphoid$ or lunate$ or triquetral$ or pisiform$).tw.
32. ((styloid$ or coronoid$) adj2 process$).tw.
33. (ligament$ or tendon$ or menisc$).tw.
35. (fracture$ or injur$ or wound$ or dislocat$ or sublux$ or rupture$ or tear$ or torn).tw.
37. (fracture$ adj3 (open or closed)).tw.
41. antibiotic prophylaxis/ or infection prevention/
42. (antibiotic$ or antimicrobial$).tw.
44. hospital infection/ or infection complication/ or postoperative infection/ or surgical infection/ or wound infection/ or "bone and joint infections"/
Appendix 2. Quality assessment criteria
A. Was the assigned treatment adequately concealed prior to allocation?
1=states random, but no description, or quasi-randomisation (Category C)
2=small but real chance of disclosure of assignment (Category B)
3=method did not allow disclosure of assignment (Category A)
B. Were the outcomes of patients who withdrew described and included in the analysis (intention to treat)?
2=states numbers and reasons for withdrawal, but analysis unmodified
3=primary analysis based on all cases as randomised
C. Assessment of outcome. Were assessors of outcome blinded to treatment status?
1=not done or not mentioned
2=moderate chance of unblinding of assessors
3=action taken to blind assessors, or outcomes such that bias is unlikely
D. Comparability of treatment and control groups at entry
1=large potential for confounding or not discussed
2=confounding small; mentioned but not adjusted for
3=unconfounded; good comparability of groups or confounding adjusted for
E. Was a placebo treatment assigned as part of the randomisation?
F. Were exclusion criteria clearly defined?
G. Method of assessment of wound infection.
2= clinical decision, or definite criteria without a microbiological diagnosis
3=definite criteria including a microbiological diagnosis
H. Surveillance for wound infection
1=not stated, or not active
2=active, but less than three months
3=active, and at least one year
Comment sent 12 January 1999
I very much enjoyed reading the new Cochrane review of antibiotic usage in fracture prosthetic surgery. This is full of fascinating information, and is a good marriage of high-quality statistical and clinical input. I have made our Orthopaedic surgeons aware of the review's findings, and have proposed a change to our agreed prophylactic protocols as a result.
The review usefully examined the effects of long- and short-halflife agents, but another common point of debate is the choice of narrow-spectrum (mainly isoxazolyl penicillin) versus broad-spectrum (e.g. cefuroxime, or isoxazolyl penicillin plus gentamicin) prophylaxis. There is a suggestion in the classic Lidwell/MRC study reports that broad spectrum prophylaxis may reduce infection rates (by reducing deep Gram-negative infections), but others have been concerned about costs, side effects (principally Clost. difficile diarrhoea) and induction of resistance associated with broader-spectrum agents. Did the group consider this approach? Could it be examined in the future?
Many thanks for your interest in our review. We did not identify any trial which compared narrow and broad spectrum agents in the closed fracture population. There are in joint replacement (Pollard 1979, Vainionpaa 1988), and in open fracture management (Patzakis 1977); these studies were in our excluded trials table. If we have missed any that you know about, I would be most interested to hear of them. In Nelson 1983 (a study which we included comparing antibiotic versus placebo), the choice of antibiotic was at the discretion of the treating surgeon, but the trial did not compare these options. There seems to be evidence of a beneficial effect in limiting urinary tract infection when a cephalosporin was compared with a placebo (see the analyses in our review).
Comment sent from:
Dr Mark Farrington, Cambridge, UK
Prof William Gillespie, Dunedin, New Zealand
Dr Helen Handoll, Edinburgh, UK
Dr Rajan Madhok, Hull, UK (criticism editor)
Last assessed as up-to-date: 25 November 2000.
Protocol first published: Issue 1, 1995
Review first published: Issue 4, 1998
Contributions of authors
Both reviewers (WJ Gillespie and G Walenkamp) appraised and extracted data from all included papers. WJ Gillespie was respnsible for drafting and for entering modifications to the text. WJ Gillespie is the guarantor of the review.
Declarations of interest
Sources of support
- HealthCare Otago Endowment Trust, Dunedin, New Zealand.
- Chief Scientist Office, Department of Health, The Scottish Office (Original Review), UK.
- Health Research Council of New Zealand (Update), New Zealand.
Medical Subject Headings (MeSH)
MeSH check words
* Indicates the major publication for the study