SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. References

Objective To test the hypothesis that intravenous antibiotics given intra-operatively reduce the failure rate of vesico-vaginal fistula repair.

Design A single blind, randomised controlled trial.

Setting A district hospital in Benin, West Africa.

Population Seventy-nine women undergoing repair of an obstetric vesico-vaginal fistula by a single surgeon at Hopital Evangelique; two women had repeat operations.

Methods Participants in the treatment group (n = 41) received ampicillin 500 mg intra-operatively. Controls (n = 40) received no prophylactic antibiotics.

Main outcome measures Failure of fistula closure and objective incontinence (a positive pad test) at hospital discharge. Secondary outcomes included febrile morbidity, other antibiotic use and urinary infection.

Results Antibiotic prophylaxis did not reduce the odds of failed repair (OR 2.1 95% CI 0.75–6.1) or of objective incontinence (OR 1.9; 95% CI 0.72–5.1). The women in the antibiotic prophylaxis group received less post-operative antibiotics and had less urinary infections at day 10.

Conclusions Prophylactic antibiotics should not be used in vesico-vaginal fistulae repair in the developing world outside randomised controlled trials.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. References

Prophylactic antibiotics reduce many types of surgical wound infection, such as that after caesarean delivery1 or hysterectomy2,3. If they also reduced vesico-vaginal fistulae repair wound infections, this might be translated into more cures. Antibiotic prophylaxis pre-, intra- and post-operatively) is used routinely in the Addis Ababa Fistula Hospital, but in developing countries antibiotics are expensive. Many patients pay for antibiotics, but they should not be used without good reason; at least one experienced active fistula surgeon does not use them4. To our knowledge, this is the first trial of intra-operative prophylactic antibiotics in vesico-vaginal fistulae repair to be reported. Informed consent was not obtained because the hospital board felt that this was not appropriate in the local culture, and approved the trial without it. The hospital did not have a research ethics committee

METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. References

Seventy-nine consecutive patients with obstetric fistula repair operations in a small mission hospital in Benin were studied. Before the operation the women were randomly allocated to two groups, using sealed opaque envelopes. The intervention group received ampicillin 500 mg intravenously at the start of the operation, while the control group underwent repair in the usual way without antibiotic prophylaxis. There is no universally accepted classification of fistulae, therefore these were graded according to the system of Waldijk4. All operations were performed by the first author (A. J. T.) who was trained in gynaecology in the UK. Before the trial he had performed ten fistulae repairs, but after the 26th randomised case he received further training at Addis Ababa Fistula Hospital.

Outcome measures

The urethral catheter was usually removed at the earliest on day 10, dependent on the grade of the fistula. On the following day, a pad test was performed by applying a pre-weighed dry pad to the perineum. The patient walked for 20 minutes, and then the pad was re-weighed. A weight gain of > 10 g was classed as objective incontinence. The pad test was carried out by a nurse who did not know the trial group allocation. All patients were also examined by the first author who classified the operative result as healed or not healed. When making this assessment he was aware of the treatment allocation. In the latter part of the trial some patients underwent examination of catheter specimens of urine for urinary tract infections on day 3 and day 10. Patients who had > 10 white blood cells per high power field were deemed to have infection (cultures were not performed at the hospital). No long term follow up was performed because of transport difficulties and limited patient finance.

Statistics

The significance of differences in the number of days of pyrexia, or of antibiotic use was measured using non parametric tests (Mann-Whitney). Rates of adverse outcomes in both groups are given as odds ratios and 95% confidence intervals. All analyses were by intention-to-treat unless otherwise stated.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. References

Seventy-nine women underwent 8 1 operations (4 1 treatment, 40 control). Two patients were randomised a second time when their original operation had failed. Their characteristics at entry and the type of surgery are listed in Table 1, the size being the maximum diameter of the fistula. Three patients who did not undergo surgery were discovered to have ureteric fistulas after randomisation. Although one later had a successful ureteric reimplantation, all three were classified as failures for analysis (all were classified as grade 3). The details of the fistula grades are shown in Table 2 and the outcomes in Table 3. Rates of nonhealing and objective incontinence were higher in the group who received antibiotics although the confidence intervals around the odds ratio included one. Less post-operative antibiotics were administered in the group allocated prophylaxis and rates of urinary infection were lower in this group. Operative success rates before the first author received fistula repair training at Addis Ababa (33%, 5/15 women in antibiotics group; 54% 6/11 controls) rose after training to 69% (18/26 women in antibiotics group) and 83% (24/29 controls).

Table 1.  Characteristics of two groups of patients. Values are given as mean (range) on n.
 TreatmentControl
Parity2.5 (1–10)2.9 (0–14)
Duration (months)34 (2–168)43 (4–268)
Maximum dimension of fistula (cm)1.2 (1–3)1.4 (1–3)
Previous fistula surgery1410
Operation
 Vaginal route3735
 Abdominal route21
 Combined approach12
 No operation12
Operated after training in Addis Abbaba2629
TOTAL4140
Table 2.  Details of grades of fistula. CM = closing mechanism.
GradeTreatmentControl
1. Not involving the (CM)1918
2Aa. Involving CM without urethral involvement and defect not circumferential1212
2Ab. (As 2Aa) plus defect circumferential02
2Ba. Urethral involvement and defect not circumferential73
2Bb. (As 2Ba) plus defect circumferential02
3. Other (eg, ureterovaginal)33
Table 3.  Outcome for patients in the two groups. Values are given as n (%) on mean [range].
 Treatment (n = 41)Control (n = 40)(OR 95% CI)
  1. *P < 0.08.

  2. **P < 0.01

Fistula not healed17 (41.5)10(25)2.1(0.75–6.1)
Patient wet25181.9[0.72–5.1]
Days pyrexial0.8 [0–6]0.75[0–6]* 
No. of days with antibiotic treatment5 [0–151] (median days 5)8[0–16]** (median days 9) 
Urinary infection
CSU day 3 infected8/14 (57)9/15 (60)0.89 (0.16–5.0)
CSU day 10 infected6/15 (40)18/20 (90)0.07 (0–01–0.55)

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. References

We have failed to show any important benefit from the use of prophylactic antibiotics during vesico-vaginal fistulae repair, and the trend was towards nonsignificantly higher operative failure and more objective incontinence in the intervention group. The only apparent benefit was a reduction in the use of other antibiotics and urinary infection. This latter benefit should be interpreted with caution because it is based on incomplete analysis of cases, it was not the primary trial hypothesis, and prophylactic antibiotics have not generally been shown to reduce post-operative urinary infections in other trials of their use.

Ampicillin was the antibiotic chosen because it is inexpensive and easily available. Theoretically, it could be suggested that antibiotic protection against anaerobes would be more effective. A meta-analysis has shown that broad spectrum penicillins are as effective as cephalosporins prophylaxis at caesarean section’. In a review by Duff and Park5 ampicillin as a single agent was one of a number of antibiotics that reduced pelvic infection following vaginal hysterectomy. There are a number of reasons to be cautious in interpreting this trial. Firstly, it was relatively small and the confidence intervals for the odds of failed repair include 0.75. Some surgeons might regard even this reduced failure rate as of clinical importance, and if so it would be reasonable to undertake further trials. Secondly, the study was not blinded. Nevertheless, we do not believe that the surgeon's awareness of the group allocation affected the classification of operative failure, which was typically clear cut. Assessment of objective incontinence is more susceptible to bias, but was performed by a person who was blind to the treatment allocation. The result of the pad test may have been affected by the presence of urinary tract infections causing temporary incontinence; however, as infections were commoner in the control group, if this had any effect it would be in favour of the treatment group.

The overall success rate in this trial, if the three cases not operated upon are excluded, was 66%. Although results are very dependent on the type of fistula included6, this is lower than some surgeons achieve, and may reflect the inexperience of the operator in the early part of the series. Nevertheless, it is well within the range of reported success rates which range from 35%–100%6,7. In the settings of developing countries case series often show slightly higher success rates of 70%–856,9,10. We believe that these results can be generalised to the typical surgeon performing fistula repair in a developing country. In the developed world the aetiology of vesico-vaginal fistulae is different, with most being secondary to surgery (71 %) or malignancy (18%)9 and our results may not be generalised to such settings.

In conclusion, we have shown no important benefit from the use of prophylactic antibiotics, and a trend towards worse results. Since antibiotics are expensive (particularly parenteral preparations) and may have adverse effects, we recommend that they only be used for vesico-vaginal fistulae prophylaxis in randomised controlled trials. Such trials are justified as the confidence intervals for our estimate of the effect is wide enough to include a clinically worthwhile benefit.

Acknowledgements

The authors would like to thank Ms A. J. Farrin for statistical advice.

References

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. References
  • 1
    Smaill F. Prophylactic antibiotics in caesarean section (all trials) [revised 3 Aug 1994]. In: KeirseMJNC, RenfrewMJ, NeilsonJP, CrowtherC, editors. Pregnancy and Childbirth Module. In: The Cochrane Pregnancy and Childbirth Database [database on disk and CDROM]. The Cochrane Collaboration; Issue 2. Oxford : Update Software, 1995. [Available from London: BMJ Publishing Group.].
  • 2
    Hemsall DL, Molly C, Heard RA, Nobles BJ, Bawdon RE, Hemsall PG. Single dose prophylaxis for vaginal and abdominal hysterectomy. Am J Obstet Gynecol 1987; 157: 498501.
  • 3
    Region A, Ravera M, Cocozza E, Andreata M, Mukasa F. Randomised study of antibiotic prophylaxis for general and gynaecological surgery from a single centre in rural Africa. Br J Surg 1996; 83: 356359.
  • 4
    Waaldijk K. Step by Step Surgery of Vesico Vaginal Fistulas. Edinburgh : Campion Press Ltd, 1994. 11.
  • 5
    Duff P, Park RC. Antibiotic prophylaxis in vaginal hysterectomy: a review. Obstet Gynecol 1980; 55(Suppl): 193202.
  • 6
    Elkins TE. Surgery for the obstetric vesicovaginal fistula. Am J Obstet Gynecol 1994; 170: 11081120.
  • 7
    Iloabachie GC. Two-stage repair of giant vesico-vaginal fistula. Int J Gynecol Obstet 1989; 28: 2731.
  • 8
    Raz S. Female Urology. Philadelphia : WB Saunders, 1983: 372377.
  • 9
    Kelly J. Vesico-vaginal and Recto-vaginal fistulae. J Roy Soc Med 1992; 85:257258.
  • 10
    Arrowsmith SD. Genitourinary reconstruction in obstetrical fistulas. J Urol 1994; 152: 403406.
  • 11
    Thornton JG. Should vesico-vaginal fistulae be treated only by specialists Trop Doct 1986; 16: 7879.