Association of Race and Ethnicity as Risk Factors for OASIS: A Systematic Review

Abstract

Importance

Understanding how race and ethnicity are contextualized as risk factors for obstetric anal sphincter injuries (OASIS) can improve efforts to address maternal health disparities.

Objectives

Our primary aim was to conduct a systematic review and meta-analysis to examine the association between race/ethnicity and OASIS incidence. The secondary aim was to critique how included studies contextualized race/ethnicity as risk factors for OASIS.

Study design

This systematic review analyzed observational studies reporting OASIS incidence by race and ethnicity. The primary outcome was the overall incidence of OASIS, with odds ratios (OR) calculated using a random effects model. We quantified the racial representation in the included studies using a representation quotient (RQ) analysis and applied the Obstetrics & Gynecology’s Equity Rubric for structured critique.

Results

Out of 3,896 studies, 12 were included. The overall OASIS incidence was 4.44% (95% CI, 2.69%−6.60%), with Asian American and Pacific Islander (AAPI) individuals having the highest incidence at 5.69% (95% CI, 4.04%−7.59%). AAPI individuals had higher odds of OASIS compared to White individuals (OR 1.61; 95% CI, 1.12–2.31). Two studies provided disaggregated AAPI data by ethnicity. Native American (RQ 0), Black (RQ 0.15), and Hispanic (RQ 0.20) individuals were underrepresented. Among studies that identified racial disparities, 33% attributed findings to systemic factors, 45% did not address disparities, and 22% suggested biological differences without substantiated evidence.

Conclusions

AAPI individuals had 61% higher odds of experiencing OASIS than White individuals. However, most studies aggregated AAPI data, failed to represent the U.S. population, and overlooked system-level factors contributing to the racial disparity in OASIS incidence.

Introduction

Obstetric anal sphincter injuries (OASIS) refer to lacerations of the perineum and anal sphincter complex that occur during childbirth, affecting approximately 6–7% of vaginal deliveries.¹ OASIS is associated with short and long-term morbidity including pelvic floor disorders, sexual dysfunction, postpartum depression, and anxiety.^2,3,4,5 Consequently, OASIS is monitored as a quality measure for obstetric care in the United States.⁶ Although efforts have been made to identify modifiable risk factors to reduce the incidence of OASIS (e.g., operative vaginal delivery and episiotomy), a recent global systematic review indicated that individuals of Asian descent are at a higher risk for experiencing OASIS when compared to White individuals, including those in the United States.⁷ Despite the persistent nature of this disparity, to our knowledge, there has been minimal progress made toward understanding and addressing OASIS disparity facing the diverse Asian American and Pacific Islander (AAPI) community.

Salient barriers to advancing disparity research for the AAPI population include racial stereotypes (specifically the perpetual foreigners and model minority stereotypes) that inaccurately mischaracterizes this group of over 22 million individuals as a homogenous population.^8,9 AAPI communities represent a diverse array of socioeconomic statuses, ethnic backgrounds, and intergenerational migration histories.¹⁰ Several studies highlighted the importance of data disaggregation in understanding and alleviating health disparities for the diverse AAPI communities.^11,12,13 Incomplete or inaccurate data can limit the capacity of researchers, clinicians, and policymakers to implement targeted interventions for specific groups that might otherwise be overlooked.¹⁴ Therefore, it is crucial for women’s health researchers to assess the current state of OASIS disparity research in order to move towards actions that produce equitable outcomes and quality for all birthing people in the U.S. As a result, the primary objective of this study was to perform a systematic review and meta-analysis to examine the relationship between race, ethnicity, and the incidence of OASIS. The secondary objectives were to examine how investigators contextualized the associations between race and ethnicity and OASIS incidence and assess the racial representation in included studies.

Methods

We conducted a systematic review of the literature in accordance with PRISMA (Preferred Reporting Items for Systematic reviews and Meta-Analyses) guidelines (Supplemental Content 1). In partnership with a librarian scientist (JR), we searched PubMed (MEDLINE), Embase (Elsevier), Web of Science (Clarivate), and the Cochrane Database of Controlled Trials (Wiley) from August 1^st 2012 (when American College of Obstetricians and Gynecologists endorsed the classification of OASIS) to November 1st 2022 using search terms relevant to OASIS as well as U.S. racial and ethnic groups (Supplemental Content 2).¹⁵ As this review utilized publicly available data, the study was deemed not to involve human subjects and received exempt status from the Institutional Review Board at Tufts Medical Center.

Four authors (TM, SR, EH, MC) performed the initial screening and independently reviewed abstracts and full texts to identify eligible studies for inclusion. Conflicts between two reviewers were resolved by a third author. We included peer-reviewed observational studies reporting the incidence of OASIS for individuals from all U.S. racial and ethnic groups. The outcome of interest was third- or fourth-degree OASIS, which include disruption of the internal and/or external anal sphincter muscle with or without anal mucosal disruption.¹⁵ We included all vaginal deliveries with or without instrumentation (e.g., forceps or vacuum). We excluded conference abstracts or non-peer reviewed articles as well as studies that aggregated multiple racial groups together. Of the included studies, 2 authors (TM, OB) extracted study information, including year of publication, study population geographic location using U.S. census regions, study design, sample size, overall incidence of OASIS, race and ethnicity groups included, method used to specify race and ethnicity as well as OASIS diagnosis.¹⁶ The numbers of OASIS events, numbers of all birthing individuals, unadjusted and adjusted odds ratios (ORs), and 95% confidence intervals (CIs) from each study were also included.

In accordance with the MOOSE (Meta-analysis Of Observational Studies in Epidemiology) guidelines, a random-effect meta-analysis was performed when there were two or more included studies in the analysis. The primary outcome for this systematic review was assessing the OASIS incidence across U.S. racial and ethnic groups. We also performed subgroup analyses by years of publication and U.S. census regions. Both inconsistency index (I²) and χ² test P values were used to quantify heterogeneity. An I² value of greater than 75% and 1-sided P<.10 were considered significant.¹⁷ Additionally, we performed post hoc sensitivity analyses by removing studies categorized as high-risk for bias and/or with potentially overlapping data sources. The meta-analysis and forest plots were generated using the metafor package in R version 4.5.1.

We determined each racial group’s representation in the included studies relative to their respective representations in the U.S. population via a representation quotient (RQ) analysis. RQ is calculated as the ratio of the racial and ethnic group’s population in the included study combined across all U.S. regions then divided by U.S. census data on each specific group, with underrepresentation defined as RQ<1.^18,19 Finally, two authors (TM, OB) used questions from sections of the Obstetrics & Gynecology’s Equity Rubric to perform a structured critique of how the included studies used race and ethnicity as risk factors for OASIS in their investigation and interpretation.²⁰ The Equity Rubric was created by a committee within the journal of Obstetrics & Gynecology, which included experts in obstetrics-gynecology, urogynecology, public health, epidemiology, maternal fetal medicine, midwifery, critical race theory, doula services, gynecologic oncology, and family planning.²⁰ The Equity Rubric was developed as a tool to help researchers to systematically center health equity as they conceptualize, design, analyze, interpret, and evaluate research in obstetrics and gynecology.²⁰ Two authors (TM, OB) also independently assessed the risk of bias using a validated tool specific for prevalence studies.²¹ All disagreements were resolved by discussion.

Results

We screened 3,897 nonduplicated records and identified 12 studies meeting the inclusion criteria (Figure 1 for study selection process flow and Supplemental Content 3 for study characteristics).^{22,23,24,25,26,27,28,29,30,31,32} Among the 12 studies, 11 were retrospective cohort studies, while one study was a prospective cohort study. The studies included data from the years 2012 to 2022, comprising 5 national U.S. samples, with the remainder being focused on a specific region: 3 from the Western U.S., 2 from the Midwest, and 2 from the Southern U.S. All studies involved White individuals; 9 studies included Asian American and Pacific Islander individuals, 8 studies included Black individuals, 6 studies included Hispanic individuals, 1 study included Native American individuals, and 7 studies reported “Other” group with limited definitions. Of the 9 studies that included AAPI individuals, 2 studies provided disaggregated ethnicity data, which encompassed Chinese, Filipino, Japanese, and Pacific Islander in both studies, as well as Asian Indian, Vietnamese, and Korean ethnicities in 1 out of the 2 studies. We were unable to perform meta-analysis for Native American and disaggregated AAPI ethnic groups due to their limited data available. The risk of bias was assessed for each study, revealing that 4 studies were categorized as low risk, 6 as moderate risk, and 2 as high risk (Supplemental Content 3).

Figure 1.

Flowchart of reviewed and included studies.

Data from the 12 studies included, which encompassed over 39 million vaginal deliveries, estimated the overall pooled incidence of OASIS to be 4.44% (95% CI, 2.69%−6.60%; I²=100%; P<.001). The highest incidence of OASIS was 5.69% (95% CI, 4.04%−7.59%; I²=100%; P<.001) observed among AAPI individuals, followed by the category labeled as ‘Others’ at 4.61% (95% CI, 3.80%−5.49%; I²=80.3%; P<.001). The pooled OASIS incidence for White individuals was 4.55% (95% CI, 2.40%−7.33%; I²=100%; P<.001), for Black individuals 2.49% (95% CI, 0.74%−5.05%; I²=98.3%; P<.001), and for Hispanic individuals 2.25% (95% CI, 1.65%−2.94%; I²=98.1%; P<.001) (Figure 2). No associations were observed based on geographic or temporal variations across all racial groups. In comparison to White individuals, AAPI individuals had higher odds of OASIS (OR 1.61; 95% CI, 1.12–2.31), whereas Black and Hispanic individuals showed lower odds (OR 0.47; 95% CI, 0.38–0.57 and OR 0.63; 95% CI, 0.63–0.64, respectively). Our findings remained consistent after conducting sensitivity analyses that excluded studies with potentially overlapping populations^22,24 and the two studies identified as having a high risk of bias^26,30.

Figure 2.

Forest plot of obstetric anal sphincter injuries incidences.

Due to the limited availability of disaggregated AAPI data, a meta-analysis for each AAPI ethnic group could not be performed. The two studies that included disaggregated AAPI ethnic OASIS data revealed significant variability in incidence rates across all reported ethnicities. For instance, among the 4 ethnicities reported by both studies, Wagner et al. reported OASIS rates of 1.9% for Chinese individuals, 1.6% for Filipinas, 1.5% for Japanese individuals, and 0.9% for Pacific Islanders based on a national database.²³ In contrast, de Silva et al. provided regional data from the Western U.S., indicating OASIS rates of 4.3% for Chinese individuals, 7.6% for Filipinas, 6.8% for Japanese individuals, and 9.1% for Pacific Islanders.³³ In Wagner et al.’s national sample encompassing 7 AAPI ethnicities, Asian Indian individuals experienced the highest OASIS rate at 2.8%, while de Silva et al.’s regional sample indicated that Pacific Islanders had the highest OASIS rate among the reported AAPI ethnic groups.^23,33

Our representation quotient (RQ) analyses and structured critiques using the Obstetrics & Gynecology’s Equity Rubric demonstrated gaps in racial and ethnic representation in the studies as well as interpretation of racial differences in the study observations. We observed a considerable underrepresentation of Native American (RQ 0), Black (RQ 0.15), and Hispanic (RQ 0.20) populations in the included studies. Conversely, there was an overrepresentation of OASIS data for White (RQ 1.27) and AAPI (RQ 1.50) groups. Although AAPI group was overrepresented in the included studies, there exists a lack of disaggregated data to indicate adequate representation across the diverse AAPI ethnicities.

The structured critique of the interpretation of racial differences in the findings across the 12 included studies using the Equity Rubric, none of the studies explicitly stated how their research could enhance health outcomes for the marginalized racial populations included. Among the 9 studies that recognized racial disparities in OASIS incidence, 33% attributed their findings to systemic or non-patient factors (e.g., clinician practice differences²⁵, challenges in reporting accurate racial and ethnic data,²⁸ and social determinants of health²⁹), 45% did not address the disparity, and 22% suggested biological differences without providing relevant evidence (e.g. connective tissue differences^22,26). Of the 3 studies that did not find racial disparities in OASIS incidence for AAPI individuals, 2 studies attributed their findings to their smaller sample sizes while one study aggregated AAPI individuals as “Other” without a clear rationale for this decision.

Discussion

While acknowledging the high heterogeneity of the included studies, AAPI individuals had 61% increased odds of experiencing OASIS than their White counterparts during vaginal deliveries in the U.S. While the finding of AAPI individuals experiencing higher incidence of OASIS is not new, we found most of the existing OASIS literature aggregated the diverse AAPI communities into a single category, failed to represent the diverse racial and ethnic demographics of the U.S. population, and missed the opportunity to consider factors beyond patient characteristics that could be contributing to the disparity in OASIS incidence within AAPI communities.

We found that reporting disaggregated OASIS data for AAPI individuals was infrequent and is therefore a missed opportunity to consider factors beyond patient characteristics that could be contributing to the OASIS disparity in the AAPI communities. The AAPI communities are highly diverse, encompassing over 50 ethnic groups and more than 100 distinct languages.³³ The aggregated category of Asian Americans includes various groups such as Asian Indian, Chinese, Filipino, Japanese, Korean, Vietnamese, and other Asian ethnicities. Meanwhile, the aggregated category for Pacific Islanders comprises Native Hawaiian, Guamanian or Chamorro, Samoan, and other Pacific Islanders.³⁴ According to the National Bureau of Economic Research, the AAPI population is one of the most economically divided racial groups in the U.S.³⁵ When a lower socioeconomic status is associated with poorer health outcomes, relying on aggregated data regarding health and health-related social needs can reinforce the model minority myth, which falsely characterizes AAPI individuals as uniformly successful racial group while obscuring the unique challenges faced by many individuals within the communities.³⁶ In the limited disaggregated data included in this study, certain ethnicities were noted to have higher OASIS incidence rates, such as Asian Indian in Wagner et al.’s national dataset and Pacific Islanders as reported by de Silva et al. from a single center in Honolulu, Hawai’i.^23,33 The discrepancies identified in these two studies pointed to the need for additional disaggregated data in order to better understand if regional or geographic differences in OASIS disparity exists. This further emphasizes the necessity for disaggregated AAPI data regarding OASIS incidence in the U.S. in order to improve clinicians’ understanding and their ability to address this disparity.

We also found that the current OASIS literature contains incidence data that fails to accurately reflect the U.S. population. We observed that Black, Hispanic, and Native American birthing people were not adequately represented in the OASIS studies we reviewed. This finding raises concerns about the validity of the conclusions that suggest Black and Hispanic individuals are less likely to experience OASIS compared to their White counterparts, and whether race is genuinely an independent and reliable factor linked to OASIS incidence. Understanding racial group representation in OASIS data is needed to evaluate the generalizability of the studies’ findings and to identify gaps for more inclusive reporting strategies in OASIS research. Underrepresented data may produce misinformation on maternal outcomes, which run of the risk of impairing efforts to reduce OASIS disparities and improve maternal health.³⁷ We advocate for inclusion of all birthing individuals from diverse racial and ethnic backgrounds in maternal health research in order to better inform clinical practice, improve patient care, and promote equitable maternal outcomes in the U.S.

While most of the studies reported increased OASIS incidence for the AAPI community, many failed to address this disparity or proposed biological racial differences without substantiating evidence. Race is a sociopolitical construct; and while it does have biological implications, it does not imply inherent biological differences.³⁸ Researchers must ensure that the rationale for including race should not be proposed to explain genetics or a biological determinant of disease.³⁹ One approach to achieve this is by employing conceptual models that illustrate how race and racism may influence the outcomes of interest.⁴⁰ Consumers of OASIS research should question whether the conclusions derived from the study could inadvertently reinforce harmful racial stereotypes among AAPI women. Examples of such include the contradictory portrayal of AAPI women as both hypervisible due to their perceived exoticization and simultaneously invisible in contexts where their struggles are frequently ignored.^41,42,43 The repeatedly naming of Asian race as an independent and non-modifiable risk factor for OASIS implies this disparity is immutable, potentially resulting in inaction and the continuation of inequities.⁴⁴ Therefore, it is imperative that future OASIS research uses race and ethnicity variables to provide thoughtful reporting of data and interpretation of results with attention to equity.

There are established strategies in the literature to guide clinicians and researchers on how to thoughtfully and purposefully incorporate race and ethnicity in research to avoid reinforcing disparities.⁴¹ Researchers must clearly justify the inclusion of race, ensuring it is understood as a social construct shaped by racism and structural factors, not linked to biological determinism.⁴⁰ Data collection should prioritize self-identification, disaggregate racial categories, and use inclusive, non-stigmatizing terminology based on study participants’ input or leveraging the expertise of researchers from other fields who have interrogated this science for decades.^45,46 Adopting a race-conscious, multidisciplinary approach, such as the Public Health Critical Race Praxis, is essential for achieving methodological rigor and addressing systemic racism while promoting equity.^47,48,49 Researchers should also critically evaluate whether their findings may perpetuate inequality or bias by engaging diverse study participants throughout the research process to ensure their lived experiences are appropriately assessed and accurately reflected, such as through the use of community-based participatory research models.⁵⁰ Finally, researchers must interpret results thoughtfully, consider the role of structural racism and intersecting societal factors in health disparities, and ensure that studies contribute to reducing inequities rather than merely naming them.^47,51

This research has its limitations. One limitation is the high heterogeneity present, as not all studies reported uniform and comprehensive details of each vaginal delivery episode. The high heterogeneity level impacts the certainty of our estimates. Additionally, while our literature review was conducted systematically using strict criteria, it is possible that we overlooked pertinent publications that included secondary findings related to OASIS incidence. Nevertheless, our study was designed to assess articles that primarily concentrate on OASIS research within the framework of race and ethnicity, rather than on peripheral analyses. Another limitation could stem from the potential for abstraction errors; however, the first and senior authors evaluated the precision of all data abstraction through multiple quality control checks.

In conclusion, we found that AAPI individuals were 61% more likely to experience OASIS compared to their White counterparts during vaginal deliveries. Moreover, our review revealed that disaggregated data for AAPI individuals were scarce, and numerous studies failed to address the disproportionate rates of OASIS among AAPI individuals or focused solely on patient characteristics that do not lead to actionable next steps or solutions. Lastly, we observed a significant lack of data representing other racially marginalized groups and their OASIS rates, which may hinder our confidence in accurately capturing the true incidence of OASIS within the broader U.S. population. We strongly encourage future research to adopt a multilevel approach to understand and mitigate the OASIS disparities faced by the diverse AAPI communities, while also striving for data equity to report disaggregated data that reflects the diverse demographics of the U.S. birthing population. By prioritizing equity, we can improve our understanding of the multilevel factors contributing to OASIS disparities and enhance the likelihood of achieving improved maternal pelvic floor outcomes in the U.S.

Why this matters?:

Obstetric anal sphincter injuries (OASIS), also referred to as third- and fourth-degree perineal lacerations, are associated with significant morbidity, which can result in pelvic floor disorders, mental health issues, and sexual dysfunction. Research indicates that individuals of Asian American and Pacific Islander (AAPI) descent consistently face a heightened risk for OASIS in the United States. Despite the persistent nature of this finding, to our knowledge, there has been limited advancement in understanding and addressing the higher prevalence of OASIS within the diverse AAPI community. This study found that AAPI individuals had 61% increased odds of experiencing OASIS during vaginal deliveries compared to White individuals. Additionally, we found limited disaggregated data for AAPI individuals thus inaccurately portraying this group as homogenous in the literature. We also found underrepresented reporting of OASIS outcomes for other racially minoritized groups, particularly among Native American, Black, and Hispanic populations. These limitations hinder our understanding of racial and ethnic disparities in OASIS and signal an opportunity for an equity-driven approach in order to understand and address the drivers contributing to the differential burden of OASIS in AAPI communities.