Third-degree and fourth-degree vaginal lacerations—tears that occur during childbirth involving the anal sphincter and the rectal mucosa, respectively—are a complication of vaginal deliveries that are relatively common when operative delivery is performed.1–3 Sequelae of third-degree and fourth-degree lacerations include persistent anal sphincter defects resulting in fecal urgency and incontinence.4,5 Besides forceps and vacuum delivery, other established risk factors include nulliparity, second-stage labor abnormalities, midline episiotomy, and large birth weight.1,6,7
Given the long-term morbidity associated with third-degree and fourth-degree lacerations, prevention of these complications is an important clinical goal. The Agency for Healthcare Research and Quality has proposed third-degree and fourth-degree laceration rates in operative and nonoperative deliveries as patient safety indicators.8,9 Experts in patient safety have advocated using third-degree and fourth-degree laceration as a quality measure indicator.10 The Joint Commission included third-degree or fourth-degree laceration in its 2002 Pregnancy and Related Conditions Core Measure set along with inpatient neonatal mortality and vaginal birth after cesarean delivery rates, and the measure was subsequently endorsed by the National Quality Forum in 2003.11 However, concerns have arisen regarding the preventability of third-degree and fourth-degree lacerations and their validity in measuring quality. Endorsement was subsequently withdrawn by the National Quality Forum citing concerns around unreliable data (health care providers being encouraged to not code partial third-degree lacerations), a majority of risk factors not being amenable to prevention, and no interval change in laceration rates after 2003, when laceration rates were adopted as a care quality measure.11
Despite the potential importance of third-degree and fourth-degree laceration as a quality measure, little is known about how hospital-level and patient-level demographic, medical, and obstetric factors affect incidence of this complication. The purposes of this study were to 1) better characterize factors associated with use of third-degree and fourth-degree laceration and 2) describe hospital-level variability for these delivery complications.
MATERIALS AND METHODS
Data from the Nationwide Inpatient Sample from the Agency for Healthcare Research and Quality were used for this analysis. The Nationwide Inpatient Sample, the largest publicly available all-payer inpatient care database in the United States, contains a random sample of approximately 20% of discharges from all hospitals within the United States. The sampling frame from the Nationwide Inpatient Sample includes nonfederal, general, and specialty-specific hospitals throughout the United States. Sampled hospitals include both academic and community facilities. Institutional review board exemption was obtained from the Columbia University institutional review board. Given that the data are fully deidentified and have been reviewed by the Columbia University institutional review board, subsequent analyses by our group have been deemed exempt through 2015.
Women 18 years of age or older and younger than 50 years of age who underwent vaginal delivery between 1998 and 2010 were identified using an enhanced methodology to capture delivery hospitalizations based on billing and procedure codes.12 Women with third-degree and fourth-degree lacerations were identified by International Classification of Diseases, 9th Revision codes 664.2x and 664.3x, respectively. Patient demographics analyzed included age (younger than 20, 20–24, 25–29, 30–34, and 35 years of age or older), year of delivery, race (white, black, Hispanic, and other), and insurance status (Medicare, Medicaid, commercial, self-pay, and unknown). Comorbidity was estimated using an index specifically designed to assess risk of death, acute organ injury, or both in obstetric patients.13 This index was developed by evaluating the association of obstetric, demographic, and medical risk factors in pregnant patients with outcomes such as acute renal failure, liver failure, heart failure, disseminated intravascular coagulation, sepsis, and shock among other conditions; increased comorbidity score is associated with higher risk of any of these outcomes and death. Hospital characteristics included location (urban, rural, unknown; n=1,868, 1,227, 9, respectively), teaching status (teaching, nonteaching, unknown; n=623, 2,307, 174, respectively), geographic region (Midwest, Northeast, South, West; n=960, 494, 1,018, 632, respectively), and hospital size based on bed number (fewer than 400 beds, 400–600 beds, more than 600 beds, unknown; n=822, 983, 1,290, 9, respectively). Major obstetric, medical, and fetal diagnoses were included as covariates. Risk factors for third-degree and fourth-degree laceration were identified by a literature review and an iterative process and included in the analysis. These include labor abnormalities, shoulder dystocia, fetal distress and heart rate abnormalities, operative vaginal delivery, fetopelvic disproportion and suspected macrosomia, labor induction, postdates pregnancy, episiotomy, regional anesthesia, gestational and pregestational diabetes, hypertensive diseases of pregnancy, obesity, and other conditions (Table 1).
The primary outcomes of the analysis were third-degree and fourth-degree lacerations. The associations between third-degree and fourth-degree laceration and clinical and demographic variables were compared using χ2 tests. To account for the influence of clinical and demographic factors on third-degree and fourth-degree laceration, we developed mixed-effects log-linear regression models. In addition to clinical and demographic characteristics, the models included a hospital-specific random intercept for the facility to account for hospital-level clustering. This analytic approach allows for determination of the effect that an individual center may have on the outcome variable. For example, two otherwise similar hospitals may have a different rate of a particular outcome because of practice patterns within the hospital. Thus, the models allowed for evaluation of fixed effects that would be expected to have a constant relationship with the outcome (eg, race, maternal age) and random effects in which the relationship between the variable and outcome may vary (the effect of delivery in an individual hospital).
We calculated between-hospital variation to evaluate the variation in third-degree and fourth-degree laceration across hospitals after accounting for the influence of individual patient, obstetric, and hospital characteristics.14,15 We used a stepwise approach starting with a regression model, which contained no covariates and included only the hospital-specific random intercept term. We then developed a fully adjusted model incorporating all observed patient, obstetric, and hospital characteristics. We further evaluated between-hospital variation by calculating hospital-specific rates of third-degree and fourth-degree laceration using the generalized linear mixed model following adjustments for patient, obstetric, and hospital characteristics. We estimated hospital-specific rates of third-degree and fourth-degree laceration using Empirical Bayes predictions. It has been shown that Empirical Bayes predictions tend to bring hospital rates toward the mean. Estimates are based on sample size so bias is limited for hospitals with a small denominator.14,16 We then plotted these rates by ranking each hospital by its predicted probability of third-degree and fourth-degree laceration. We calculated overall probability of third-degree and fourth-degree laceration with 95% confidence intervals (CIs) averaged across all hospitals. Finally, we performed a sensitivity analysis by removing all operative deliveries from our model and repeating the analysis with third-degree or fourth-degree laceration as the primary outcome. The generalized linear mixed models were implemented using the GLIMMIX procedure. All analyses were performed with SAS 9.3.
Of 7,096,056 women who underwent a vaginal delivery in 3,104 hospitals and were included in the cohort, 3.3% (n=232,762) experienced a third-degree laceration and 1.1% (n=76,347) experienced a fourth-degree laceration. The rate of lacerations declined during the study period (1998–2010), from 4.0% to 2.5% for third-degree and 1.7–0.6% for fourth-degree lacerations. Associations between third-degree and fourth-degree laceration and patient demographic, obstetric, and medical factors are displayed in Tables 1 and 2. Major determinants of third-degree laceration included forceps with (18.3%) and without (19.3%) episiotomy, vacuum delivery with (12.5%) and without (8.1%) episiotomy, shoulder dystocia (7.6%), fetal distress (7.2%), and long labor (12.3%). These risk factors were also associated with higher rates of fourth-degree laceration. In comparison, demographic variables and episiotomy without operative delivery were not major predictors of third-degree and fourth-degree vaginal laceration. Comorbid obstetric risk was not associated with clinically significant increased probability of third-degree and fourth-degree vaginal laceration. Conditions such as pregestational and gestational diabetes, maternal cardiac disease, obesity, and hypertensive diseases of pregnancy were associated with similar, decreased, or marginally increased rates of lacerations.
In multivariable models of factors associated with third-degree and fourth-degree lacerations, vacuum and forceps delivery retained significance. For fourth-degree lacerations, risk ratios (RRs) for forceps with and without episiotomy were 10.6 (95% CI 10.3–10.8) and 8.8 (95% CI 8.5–9.1), respectively, and vacuum delivery with and without episiotomy were RR 7.4 (95% CI 7.3–7.6) and RR 3.3 (95% CI 3.2–3.4), respectively (Table 3). Risk of third-degree lacerations was similarly increased with vacuum and forceps deliveries. Other factors associated with increased risk of fourth-degree laceration included severe preeclampsia or eclampsia (RR 2.5, 95% CI 1.8–3.6), shoulder dystocia (RR 2.7, 95% CI 2.6–2.8), episiotomy without operative delivery (RR 1.7, 95% CI 1.6–1.7), and commercial insurance (RR 1.2, 95% CI 1.1–1.4). These factors also were associated with risk of third-degree lacerations. Black race was associated with decreased risk of both third-degree (RR 0.6, 95% CI 0.6–0.6) and fourth-degree (RR 0.7, 95% CI 0.7–0.7) lacerations.
The overall risk-adjusted means for all hospitals in the model (n=3,070) was 1.1% for fourth-degree lacerations and 3.0% for third-degree lacerations. Between-hospital variation in third-degree and fourth-degree lacerations is displayed graphically in Figures 1 and 2, respectively. Each hospital was ranked in ascending order on the X axis by the mean probability of laceration at that hospital over the study period. Variation in the model was limited: 9.4% of hospitals had a 95% CI above an adjusted third-degree laceration rate of 4.0%, and 5.6% of hospitals had a 95% CI below an adjusted third-degree laceration rate of 2.0%. The remaining 85.0% of hospitals had a third-degree laceration rate 95% CI that fell between 2.0% and 4.0%. Similarly, 93.3% of hospitals had a 95% CI for adjusted rate of fourth-degree laceration that fell between 0.5% and 1.5%. 6.1% of hospitals had a 95% CI that was above 1.5%, and 0.6% of hospitals had a 95% CI below 0.5%. In an adjusted model excluding operative deliveries from the analysis (Fig. 3), variation in combined third-degree and fourth-degree laceration rates was found to be limited: 88.8% of hospitals had a 95% CI for combined third-degree and fourth-degree laceration rate that fell between 2.0% and 4.0%.
In the setting of a continuing search for evidence-based quality metrics in obstetrics,17 our findings suggest that third-degree and fourth-degree lacerations may be of limited value clinically. Care quality metrics should be measurable, valid, clinically meaningful, and amenable to improvement with application of evidence-based practices.18–20 The large majority of hospitals in our analysis had adjusted laceration rates that were statistically indistinguishable when including 95% CIs, precluding meaningful comparisons between different institutions. Although care quality in regard to preventing lacerations may have varied between centers, these differences did not result in statistically observable variation. Furthermore, apart from the concerns raised by the National Quality Forum regarding data collection validity,3 our study found that the main determinants of laceration risk were nonmodifiable obstetric complications and operative delivery. Although operative delivery rates—and particularly use of forceps—have declined recently21 amidst an increasing national cesarean delivery rate,22 it is unclear that indicated operative delivery should be considered an adverse quality event, especially if a patient is well counseled about the risks of operative delivery and the downstream health risks of the alternative (cesarean delivery) if expedited delivery is required. Use of third-degree and fourth-degree lacerations as an adverse quality event could further encourage health care provider use of cesarean delivery. Furthermore, episiotomy, the one modifiable clinical decision that may affect third-degree and fourth-degree laceration rates for which restricted use is advocated,23,24 was not found to be a major determinant of risk in this analysis and, regardless, can be measured independently.
Prior analyses have demonstrated significant between-hospital variation in the uptake of evidence-based recommendations in obstetrics such as routine postcesarean venous thromboembolism prophylaxis and perioperative cesarean antibiotic administration.25,26 Our findings demonstrated minimal between-hospital variation in third-degree and fourth-degree vaginal lacerations. Between-hospital variation is variation in a practice that cannot be explained by measured factors including measurable hospital characteristics. In essence, between-hospital variability reflects hospital practice patterns, which have been shown to drive rates of other interventions.14,15 The fact that laceration rates did not vary significantly based on institutional practice patterns and were closely associated with operative delivery rates suggests reduction of this complication may not be an actionable target for quality improvement initiatives focused on hospital-level protocols. Furthermore, our study shows a further temporal decline in third-degree and fourth-degree laceration demonstrated by other analyses27 in part resulting from a decrease in operative deliveries; given the decreasing incidence of lacerations, it may be challenging to make adequately statistically powered comparisons between institutions. Nevertheless, that a few centers at either end of the spectrum demonstrated particularly high and low adjusted rates suggests further investigation of strategies to reduce lacerations is warranted.
Although our analysis uses a large data set that includes patients from a cohort of hospitals representative of national practice patterns, we recognize important limitations. First, because claims data are used primarily for billing purposes, there is the possibility laceration was misclassified in some patients and that diagnoses not directly related to billing such as those related to labor were undercoded. Second, individual physician characteristics that may be an important determinant of outcome could not be captured by this analysis. Third, because administrative data do not allow for direct determination of patient attributes such as body mass index and other important information, it is likely that the prevalences of some clinical risk factors are underestimated. Fourth, statistically significant values are highly likely across comparison groups and in the majority of cases, clinical significance is better represented by the effect size rather than statistical significance alone. Fifth, our study concludes in 2010; in subsequent years, there may have been further decreases in laceration rates, a trend that our study was not able to assess. Sixth, clinical practice factors such as perineal massage28 were not able to be accounted for in our analysis. Seventh, for some important demographic variables such as race, significant numbers of patients had missing data. Our findings regarding decreased risk of lacerations among black women require further validation. Finally, we are unable to evaluate whether third-degree and fourth-degree lacerations resulted in increased risk of long-term sequelae because the Nationwide Inpatient Sample lacks longitudinal follow-up.
In conclusion, our study found minimal variation of third-degree and fourth-degree lacerations across centers and based on demographic characteristics. Statistically significant variation is a necessary precondition for assessing the effect of hospital-based practice patterns. Our data suggest that observable reduction of third-degree and fourth-degree lacerations on a hospital level as a result of specific evidence-based recommendations or quality improvement initiatives that utilize pay-for-performance strategies, educational initiatives, public reporting of data, or all of these29,30 may not be feasible. Further care quality and clinical research is needed to better characterize optimal strategies to reduce the incidence of this complication including the factors that led to the temporal decrease observed in this analysis.
1. Combs CA, Robertson PA, Laros RK Jr. Risk factors for third-degree and fourth-degree perineal lacerations in forceps and vacuum deliveries. Am J Obstet Gynecol 1990;163:100–4.
2. Hudelist G, Gelle'n J, Singer C, Ruecklinger E, Czerwenka K, Kandolf O, et al.. Factors predicting severe perineal trauma during childbirth: role of forceps delivery routinely combined with mediolateral episiotomy. Am J Obstet Gynecol 2005;192:875–81.
3. Christianson LM, Bovbjerg VE, McDavitt EC, Hullfish KL. Risk factors for perineal injury during delivery. Am J Obstet Gynecol 2003;189:255–60.
4. Sultan AH, Kamm MA, Hudson CN, Bartram CI. Third degree obstetric anal sphincter tears: risk factors and outcome of primary repair. BMJ 1994;308:887–91.
5. Evers EC, Blomquist JL, McDermott KC, Handa VL. Obstetrical anal sphincter laceration and anal incontinence 5–10 years after childbirth. Am J Obstet Gynecol 2012;207:425.e1–6.
6. Helwig JT, Thorp JM Jr, Bowes WA Jr. Does midline episiotomy increase the risk of third- and fourth-degree lacerations in operative vaginal deliveries? Obstet Gynecol 1993;82:276–9.
7. Handa VL, Danielsen BH, Gilbert WM. Obstetric anal sphincter lacerations. Obstet Gynecol 2001;98:225–30.
10. Mann S, Pratt S, Gluck P, Nielsen P, Risser D, Greenberg P, et al.. Assessing quality obstetrical care: development of standardized measures. Jt Comm J Qual Patient Saf 2006;32:497–505.
11. National Quality Forum. National voluntary consensus standards for perinatal care 2008: a consensus report. Washington, DC: National Quality Forum; 2009.
12. Kuklina EV, Whiteman MK, Hillis SD, Jamieson DJ, Meikle SF, Posner SF, et al.. An enhanced method for identifying obstetric deliveries: implications for estimating maternal morbidity. Matern Child Health J 2008;12:469–77.
13. Bateman BT, Mhyre JM, Hernandez-Diaz S, Huybrechts KF, Fischer MA, Creanga AA, et al.. Development of a comorbidity index for use in obstetric patients. Obstet Gynecol 2013;122:957–65.
14. Haymart MR, Banerjee M, Stewart AK, Koenig RJ, Birkmeyer JD, Griggs JJ. Use of radioactive iodine for thyroid cancer. JAMA 2011;306:721–8.
15. Perera HK, Ananth CV, Richards CA, Neugut AI, Lewin SN, Lu YS, et al.. Variation in ovarian conservation in women undergoing hysterectomy for benign indications. Obstet Gynecol 2013;121:717–26.
16. Birkmeyer NJ, Dimick JB, Share D, Hawasli A, English WJ, Genaw J, et al.. Hospital complication rates with bariatric surgery in Michigan. JAMA 2010;304:435–42.
17. Pettker CM. Quality in obstetrics: the search continues. Am J Obstet Gynecol 2014;211:85–6.
19. Meyer GS, Nelson EC, Pryor DB, James B, Swensen SJ, Kaplan GS, et al.. More quality measures versus measuring what matters: a call for balance and parsimony. BMJ Qual Saf 2012;21:964–8.
20. Cassel CK, Conway PH, Delbanco SF, Jha AK, Saunders RS, Lee TH. Getting more performance from performance measurement. N Engl J Med 2014;371:2145–7.
21. Fong A, Wu E, Pan D, Chung JH, Ogunyemi DA. Temporal trends and morbidities of vacuum, forceps, and combined use of both. J Matern Fetal Neonatal Med 2014;27:1886–91.
22. Osterman MJ, Martin JA. Changes in cesarean delivery rates by gestational age: United States, 1996–2011. NCHS Data Brief 2013;1–8.
23. Werner RM, Bradlow ET. Relationship between Medicare's hospital compare performance measures and mortality rates. JAMA 2006;296:2694–702.
24. Carroli G, Mignini L. Episiotomy for vaginal birth. The Cochrane Database of Systematic Reviews 2009, Issue 1. Art. No.: CD000081. DOI: 10.1002/14651858.CD000081.pub2.
25. Brubaker SG, Friedman AM, Cleary KL, Prendergast E, D'Alton ME, Ananth CV, et al.. Patterns of use and predictors of receipt of antibiotics in women undergoing cesarean delivery. Obstet Gynecol 2014;124:338–44.
26. Friedman AM, Ananth CV, Lu YS, D'Alton ME, Wright JD. Underuse of postcesarean thromboembolism prophylaxis. Obstet Gynecol 2013;122:1197–204.
27. Frankman EA, Wang L, Bunker CH, Lowder JL. Episiotomy in the United States: has anything changed? Am J Obstet Gynecol 2009;200:573.e1–7.
28. Shipman MK, Boniface DR, Tefft ME, McCloghry F. Antenatal perineal massage and subsequent perineal outcomes: a randomised controlled trial. Br J Obstet Gynaecol 1997;104:787–91.
29. Council of the Institute of Medicine. America's health in transition: protecting and improving quality. Washington, DC: The National Academies Press; 1994.
© 2015 by The American College of Obstetricians and Gynecologists. Published by Wolters Kluwer Health, Inc. All rights reserved.
30. Petersen LA, Woodard LD, Urech T, Daw C, Sookanan S. Does pay-for-performance improve the quality of health care? Ann Intern Med 2006;145:265–72.