Episiotomy is a common obstetric procedure, estimated to be performed in 25% of vaginal deliveries in the United States in 2004.1 Restrictive use of episiotomy has been recommended given the risks of the procedure and unclear benefits of routine use. In 2006, the American Congress of Obstetricians and Gynecologists recommended against routine episiotomy,2 and, in 2008, the National Quality Forum recognized limiting routine episiotomy as an important measure of quality and patient safety, noting increased risks of pain, laceration, and anal incontinence with the procedure.3
The ideal rate of episiotomy is unknown. Decreasing use of the procedure was documented in the 1990s4; however, whether rates have continued to decrease after evidence-based recommendations is unknown. Also, little is known regarding how hospital and patient factors are related to rates of performing episiotomy or the variation of rates among hospitals. We used a large insurance database to investigate contemporary use of episiotomy.
The Perspective database was used to identify women who underwent a vaginal delivery from 2006-2012. The all-payer database includes more than 500 hospitals (approximately 15% of hospitalizations nationally, including deliveries). Data undergo a rigorous quality-control process. Delivery hospitalizations and use of episiotomy were captured using International Classification of Diseases, Ninth Revision, codes. Deliveries with shoulder dystocia, fetal distress, and fetal heart rate abnormalities were excluded given that these may be accepted indications for episiotomy. The analysis was deemed exempt by the Columbia University institutional review board.
Multivariable mixed-effects log-linear models, including clinical and demographic covariates and a random intercept for the procedural hospital (which accounts for hospital-level clustering), were developed. We further evaluated between-hospital variation by calculating hospital-specific rates of episiotomy using the generalized linear mixed model after making adjustments for patient, obstetric, and hospital characteristics and by assessing the top and bottom 10% of hospitals. We estimated hospital-specific rates of episiotomy using empirical Bayes predictions5 for each hospital with patient matching selection criteria. A 2-sided P < .05 was considered significant. All statistical analyses were performed with SAS version 9.4 (SAS Institute Inc).
Within the cohort of 2 261 070 women who were hospitalized for a vaginal delivery in 510 hospitals, 325 193 underwent episiotomy (14.4%; 95% CI, 14.4%-14.5%). There was a temporal decline in rates of episiotomy between 2006 (17.3%; 95% CI, 17.2%-17.4%) and 2012 (11.6%; 95% CI, 11.5%-11.7%) (P < .001).
Several demographic characteristics were associated with receipt of episiotomy: 15.7% (95% CI, 15.6%-15.8%) of white women vs 7.9% (95% CI, 7.8%-8.0%) of black women (P < .001) and 17.2% (95% CI, 17.1%-17.3%) with commercial insurance vs 11.2% (95% CI, 11.1%-11.3%) with Medicaid insurance (P < .001). Hospital factors (rural location and teaching status) were associated with less use. These factors remained statistically significant in a multivariable model (Table).
Within the multivariable model (Figure), adjusted hospital rates of episiotomy use demonstrated significant variation. Among the 10% of hospitals that used the procedure most frequently, the mean adjusted hospital episiotomy rate was 34.1% (95% CI, 32.0%-36.3%) vs 2.5% (95% CI, 2.2%-2.8%) in the 10% of hospitals that used the procedure least frequently.
Our findings demonstrate that use of episiotomy declined over time, possibly reflecting adoption of recommendations. The analysis demonstrated substantial between-hospital variation in episiotomy not accounted for by demographic, obstetric, and hospital characteristics. In addition, demographic and hospital factors, such as insurance, race, rurality, and hospital teaching status, were associated with the episiotomy rate. These observations suggest nonmedical factors are related to use of episiotomy.
Even though our study benefits from the inclusion of a large sample, we recognize important limitations. Claims data are used primarily for billing purposes, and so there is the possibility episiotomy use was not captured in a small number of patients. Second, hospitals included in the analysis may not be representative of all hospitals in the United States. Third, some demographic variables such as nulliparity may be undercoded. Fourth, confounding by unmeasured medical risk factors is possible. Further strategies are needed to enhance adherence to evidence-based recommendations.
Corresponding Author: Alexander M. Friedman, MD, Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, Columbia University College of Physicians and Surgeons, 622 E 168th St, New York, NY 10032 (email@example.com).
Author Contributions: Dr Friedman had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Friedman, Ananth, D’Alton, Wright.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Friedman, Prendergast, Wright.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Friedman, Ananth, Prendergast, Wright.
Obtained funding: Wright.
Administrative, technical, or material support: D’Alton, Wright.
Study supervision: D’Alton, Wright.
Conflict of Interest Disclosures: The authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported.
JL. Episiotomy in the United States: has anything changed? Am J Obstet Gynecol
. 2009;200(5):573.e1-573.e7.PubMedGoogle ScholarCrossref
American College of Obstetricians-Gynecologists. ACOG Practice Bulletin: episiotomy: clinical management guidelines for obstetrician-gynecologists: Number 71, April 2006. Obstet Gynecol
. 2006;107(4):957-962.PubMedGoogle Scholar
EK. New perinatal quality measures from the National Quality Forum, the Joint Commission and the Leapfrog Group. Curr Opin Obstet Gynecol
. 2009;21(6):532-540.PubMedGoogle ScholarCrossref
SJ. Is there a benefit to episiotomy at spontaneous vaginal delivery? a natural experiment. Am J Obstet Gynecol
. 1996;175(4 pt 1):897-901.PubMedGoogle ScholarCrossref
D. Approximate Bayesian inference in conditionally independent hierarchical models (parametric empirical Bayes models). J Am Stat Assoc
. 1989;84(407): 717-726.Google ScholarCrossref