Objective
We aimed to systematically determine the incidences of wound infection and dehiscence after primary obstetric anal sphincter injury repair.
Data Sources
MEDLINE, Embase, CINAHL, EmCare, the Cochrane Library, and Trip Pro databases were searched from inception to February 2021.
Study Eligibility Criteria
We included observational clinical studies reporting the incidences of wound infection and dehiscence after primary obstetric anal sphincter injury repair. Case series and reports were excluded. Conference articles and observational study abstracts were included if they contained enough information regarding study design and outcome data.
Methods
Data were analyzed as incidence (percentage) with 95% confidence intervals. Moreover, the prediction intervals were calculated to provide a predicted range for the potential incidence of wound complications when applied to an individual study setting. Study quality and risk of bias were assessed using the relevant tool from the Joanna Briggs Institute.
Results
Of 956 studies found, 39 were selected for full-text review. Moreover, 10 studies (n=4767 women) were eligible and included in the meta-analysis. All 10 studies were conducted in high-income countries (Denmark [n=1], the United Kingdom [n=3], and the United States [n=6]). The incidences of wound infection (n=4593 women) and wound dehiscence (n=3866 women) after primary obstetric anal sphincter injury repair ranged between 0.1% to 19.8% and 1.9% to 24.6%, respectively. The overall incidences were 4.4% (95% confidence interval, 0.4–8.4) for wound infection and 6.9% (95% confidence interval, 1.6–12.2) for wound dehiscence. The prediction intervals were wide and suggested that the true incidences of wound infection and dehiscence in future studies could lie between 0.0% to 11.7% and 0.0% to 16.4%, respectively. Overall, 8 studies had a high or unclear risk of bias across ≥1 assessed element. None of the studies used the same set of clinical parameters to define wound infection or dehiscence. Furthermore, microbiological confirmation with wound swabs was never used as a diagnostic measure.
Conclusion
This was a systematic review and meta-analysis of wound infection and dehiscence incidences after primary obstetric anal sphincter injury repair. The incidence estimates from this review will be useful for clinicians when counseling women with obstetric anal sphincter injury and when consenting them for primary surgical repair.
Why was this study conducted?
A paucity of evidence in the literature surrounding the administration of antibiotics immediately after delivery after primary obstetrical anal sphincter repair means that there is a disparity in management globally. It is important that clinicians appreciate the incidence of wound complications after obstetric anal sphincter injury (OASI) so preventative measures can be considered.
Key findings
The incidences of wound infection and wound dehiscence after primary repair of OASIs were 4% and 7%, respectively. Prediction intervals suggested that their true incidence in future studies could lie between 0.0% to 11.7% and 0.0% to 16.4%, respectively.
What does this add to what is known?
Wound infection and dehiscence occur commonly after primary repair of OASIs. The calculated prediction interval estimates could be used to guide sample size and statistical power calculations for future epidemiologic studies.
Introduction
Approximately 3 million women will deliver vaginally in the United States. The 1998–2010 Nationwide Inpatient Sample reported an obstetric anal sphincter injury (OASI) incidence in the United States of 4.4%. In addition, a national survey in the United Kingdom between 2009 and 2010 reported an incidence of 2.9%. However, in primiparous women, the incidence of OASI may be as high as 19% in centers where midline episiotomy is performed and 6.1% where mediolateral episiotomy alone is practiced. , Because of the involvement of the anal sphincter complex and the anatomic location of these tears, which are close to the rectum, organisms from the skin surrounding the perineum and endogenous mucosal surfaces (genitourinary tract and gastrointestinal tract) can contaminate the wound. , This can result in wound infection and breakdown of the primary OASI repair. Perineal wound infection and dehiscence are associated with peripartum morbidity and prolonged postnatal recovery. This can adversely affect women’s quality of life and general well-being because of symptoms, such as pain, dyspareunia, voiding dysfunction, defecatory problems, and body image concerns. , Moreover, this can negatively affect relationships with the newborn, partner, and relatives. Moreover, distress caused by the management of perineal injury and its complications can potentially lead to an increase in the number of women electing for cesarean deliveries in future pregnancies. ,
Because of the risk of wound complications after primary OASI repair, the Royal College of Obstetricians and Gynaecologists (RCOG) recommends prophylactic broad-spectrum antibiotic administration to reduce the risk of these postoperative complications. However, the American College of Obstetricians and Gynecologists (ACOG) does not, because of the lack of high-quality research evidence supporting the postpartum clinical benefit of antibiotic administration. Few studies have identified risk factors for wound complications after primary OASI repair, such as operative vaginal delivery, increasing body mass index, smoking, and fourth-degree tear. , However, no meta-analysis has been conducted to determine the epidemiologic incidence of wound complications after primary OASI repair. Knowing the true incidence of these complications will encourage clinicians to modify the practice to prevent them and will provide further evidence to support postnatal antibiotic administration after OASI.
Objectives
Given the paucity of evidence, the primary objective of this review was to systematically determine the incidence of wound infection and wound dehiscence after primary OASI repair. Our secondary objective was to determine the prediction intervals associated with wound infection and dehiscence after primary OASI repair, to guide sample size calculations for future studies.
Methods
This systematic review and meta-analysis was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. Moreover, the meta-analysis of Observational Studies in Epidemiology guidelines for reporting meta-analyses of observational studies were observed ( Appendix 1 ). The protocol was developed (CRD 42021239678) and registered on February 26, 2021, with the International Prospective Register of Systematic Reviews.
Search strategy and eligibility criteria
Our primary review question was “What is the incidence of wound infection and wound dehiscence after primary repair of OASIs?” Medline, Embase, CINAHL, EmCare, the Cochrane Library, and Trip Pro databases were searched from inception to February 2021 using the terms “wound infection,” “wound dehiscence,” “wound breakdown,” “obstetric anal sphincter injury,” “third-degree perineal laceration,” “fourth-degree perineal laceration,” and “severe perineal trauma,” including Medical Subject Headings terms, with no restriction on language or year of publication ( Appendix 2 ). The results were exported to Zotero reference management system and deduplicated. Additional references were manually searched from identified studies to find other relevant studies. There was no restriction placed on the date of publication or language. No funding was required to complete this review. However, the Croydon Childbirth Charitable Trust provided an educational grant for the first author.
Inclusion and exclusion criteria
We included randomized control trials (RCTs) and observational studies reporting the incidence of wound infection and dehiscence after primary OASI repair. Case series and case reports were excluded. However, conference articles and abstracts were included if they contained enough information regarding study design and outcome data. Of note, 2 independent reviewers (N.A.O. and K.W.W.) screened the titles and abstracts of all retrieved studies to obtain studies for full-text assessment. Any disagreements surrounding eligibility for full-text assessment were resolved by the senior reviewers or through consensus-based discussion. Authors of included studies were contacted if the full text was unobtainable and if the data reported was published in a manner that was unclear or not extractable. To avoid potential intervention bias, RCTs, which were designed to investigate the effect of an intervention on wound complications after primary OASI repair, were excluded. Full-text articles, which met the inclusion criteria, were assessed by the 2 reviewers (N.A.O. and K.W.W.) independently.
Data extraction
Data were collected and managed using a standardized electronic data extraction form on Microsoft Excel. Data extracted included study characteristics (first author, publication year, study design, setting, and sample size), outcome measures (incidence of wound infection or wound dehiscence) and outcome definition. Of note, 2 investigators (N.A.O. and K.W.W.) performed the data extraction to prevent errors. Any disagreements were resolved through consensus.
Risk of bias of included studies
The methodological quality of the selected studies and risk of bias were assessed at an outcome level by the 2 reviewers (N.A.O. and K.W.W.) independently using the relevant tool from the Joanna Briggs Institute (JBI). Any disagreements surrounding eligibility for full-text assessment were resolved by the senior reviewers or through consensus-based discussion. These results were integrated into the data analysis when interpreting and drawing conclusions from the meta-analysis findings.
Data synthesis
Stata (version 15.1; StataCorp, College Station, TX) and Meta-Essentials (version 1.5) were used to analyze the data. For each outcome (wound infection or wound dehiscence), the incidence rates (percentage) with 95% confidence intervals (CIs) were calculated. Moreover, the prediction interval was calculated to provide a predicted range for the potential incidence of wound complications when applied to an individual study setting. Heterogeneity in the effect estimates of the individual studies was calculated using the I 2 statistic. An I 2 of >50% was considered significant heterogeneity. The results were pooled, and a meta-analysis was performed if each outcome was represented in at least 2 studies. The random-effects model (DerSimonian and Laird) was used if heterogeneity was significant ( I 2 >50%). In addition, Tau2 values were reviewed to assess the effect size variance among studies.
Results
Study selection
Overall, 956 articles were initially identified. After removal of duplicates and screening of study titles and abstracts, 39 articles were selected for full-text review. Of note, 10 studies were eligible and included in the meta-analysis ( Figure 1 ). A full list of excluded studies is provided in Appendix 3 .
Study characteristics
Table 1 describes the methodological characteristics of the included studies in further detail and their reported incidence of wound infection and wound dehiscence. The mean sample size was 476 women, and the studies were published between 1970 and 2020. The design of these studies included 4 prospective observational studies, , 5 retrospective observational studies, , and 1 reporting on incidence. Among the included studies, all were conducted in high-income countries, with 1 completed in Denmark, 3 in the United Kingdom, , , and 6 in the United States. , , ,
| Author, y | Country | Study design | Sample size (n) | Outcome | Outcome definition | Outcome follow-up | Wound infection incidence (%) | Wound dehiscence incidence (%) |
|---|---|---|---|---|---|---|---|---|
| Ajibade et al, 2013 | United Kingdom | Study reporting incidence | 171 | Wound dehiscence | Unclear | Unclear | NA | 2.3 |
| Goldaber et al, 1993 | United States | Retrospective cohort | 390 | 1. Wound infection 2. Wound dehiscence | 1. Presence of purulence and cellulitis at the repair site was usually accompanied by a temperature of 38.0°C. 2. Complete or partial separation of the layers of the repair site. Vaginal mucosal or superficial perineal skin separation was not considered dehiscence. | Unclear | 3.6 | 4.6 |
| Gommesen et al, 2019 | Denmark | Prospective cohort | 200 | 1. Wound infection 2. Wound dehiscence | 1. Presence of purulent discharge or a wound abscess according to the CDC definition for episiotomy site infection 2. Gap of >0.5 cm between wound edges | 11–21 d | 3.0 | 12.5 |
| Groves et al, 2007 | United States | Retrospective case-control | 62 | 1. Wound infection 2. Wound dehiscence | Unclear | 5 wk | 4.8 | 11.3 |
| Harris et al, 1970 | United States | Prospective cohort | 870 | 1. Wound infection | Unclear | 6 wk | 0.1 | NA |
| Johnson et al, 2012 | United Kingdom | Prospective cohort | 29 | 1. Wound infection 2. Wound dehiscence | 1. Presence of ≥2 of the following markers: perineal pain, wound dehiscence, and/or purulent vaginal discharge 2. Unclear | 21 d | 10.3 | 3.4 |
| Kaltreider et al, 1948 | United States | Retrospective cohort | 710 | 1. Wound infection 2. Wound dehiscence | 1. Unclear 2. Skin separation to complete breakdown of incision | Unclear | 2.5 | 6.3 |
| Lewicky-Gaupp et al, 2015 | United States | Prospective cohort | 268 | 1. Wound infection 2. Wound dehiscence | 1. Of note, ≥3 of the following on examination: heat, erythema, edema, or purulent discharge 2. Wound breakdown of at least 1 cm | 1 wk | 19.8 | 24.6 |
| Stock et al, 2013 | United States | Retrospective cohort | 909 | 1. Wound infection 2. Wound dehiscence | Unclear | Unclear | 4.3 | 1.9 |
| Wan et al, 2020 | United Kingdom | Retrospective cohort | 1147 | 1. Wound infection 2. Wound dehiscence | 1. Perineal tenderness, erythema, exudate, odor, and edema 2. Gaping of the perineal wound >0.5 cm with or without pyrexia | 3 mo | 2.7 | 3.3 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
