PURPOSE: Episiotomy is one of the most commonly performed procedures among women of childbearing age in the United States. In 2005, a major systematic review conducted by Hartmann and colleagues recommended against routine use of episiotomy and was widely covered in the media. We assessed the impact of the Hartman et al. study on episiotomy trend. METHODS: Based on 100% hospital discharge data from eight states in 2003-2008, we used interrupted time series regression models to estimate the impact of the Hartman et al. review on episiotomy rates. We used mixed-effects regression models to assess whether interhospital variation was reduced over time. RESULTS: After controlling for underlying trend, episiotomy rates dropped by 1.4 percentage points after Hartman et al. publication (p < 0.01 for spontaneous delivery; p < 0.1 for operative delivery). The publication has smaller effect on government hospitals as compared to private hospitals. Mixed effects models estimated negative correlation between cross-time and cross-hospital variations in episiotomy rates, indicating reduced cross-hospital variation over time. CONCLUSIONS: Our results suggested that there has been a gradual decline in episiotomy rates over the period 2003-2008, and that synthesis of evidence showing harms from routine episiotomy had limited impact on practice patterns in the case of episiotomy. The experience of episiotomy illustrates the challenge of using comparative effectiveness and evidenced-based medicine to reduce use of unnecessary procedures.
PURPOSE: Episiotomy is one of the most commonly performed procedures among women of childbearing age in the United States. In 2005, a major systematic review conducted by Hartmann and colleagues recommended against routine use of episiotomy and was widely covered in the media. We assessed the impact of the Hartman et al. study on episiotomy trend. METHODS: Based on 100% hospital discharge data from eight states in 2003-2008, we used interrupted time series regression models to estimate the impact of the Hartman et al. review on episiotomy rates. We used mixed-effects regression models to assess whether interhospital variation was reduced over time. RESULTS: After controlling for underlying trend, episiotomy rates dropped by 1.4 percentage points after Hartman et al. publication (p < 0.01 for spontaneous delivery; p < 0.1 for operative delivery). The publication has smaller effect on government hospitals as compared to private hospitals. Mixed effects models estimated negative correlation between cross-time and cross-hospital variations in episiotomy rates, indicating reduced cross-hospital variation over time. CONCLUSIONS: Our results suggested that there has been a gradual decline in episiotomy rates over the period 2003-2008, and that synthesis of evidence showing harms from routine episiotomy had limited impact on practice patterns in the case of episiotomy. The experience of episiotomy illustrates the challenge of using comparative effectiveness and evidenced-based medicine to reduce use of unnecessary procedures.