Racial Disparities in Surgical Outcomes Among Women with Endometrial Cancer.

This retrospective study used data extracted from the multi-institutional NSQIP database for all analyses 11 . The NSQIP is an ongoing quality improvement initiative that included over 700 US hospitals by 2020. 12 . Data in the NSQIP, including preoperative and 30-day postoperative patient-level information, was abstracted by a trained surgical clinical reviewer from electronic health records and medical charts according to standardized definitions by applying a systematic sampling protocol. This sampling protocol identifies the first 35 consecutive surgical cases meeting inclusion criteria within an 8-day cycle to reduce selection bias. The inter-rater reliability calculated for NSQIP data showed an overall 2% disagreement rate among participating hospitals 11 . This study was approved by the Institutional Review Board of Nationwide Children’s Hospital. We identified women older than 18 years who underwent a hysterectomy for EC treatment between 2014 and 2020 for study inclusion. This was a single-sex study as only women develop EC. Women with EC who were treated with hysterectomy were identified by Current Procedural Terminology (CPT) codes (see supplemental materials ) in conjunction with postoperative diagnosis codes 179.0, 182.0, 182.1, 182.8 (International Classification of Diseases 9th Revision [ICD-9]) and C54.0, C54.1, C54.2, C54.3, C54.8, C54.9 (10 th Revision [ICD-10]). A total of 29,290 women with EC who were treated with hysterectomy were identified. Of these, 4,971 were excluded from further analysis due to missing information on race/ethnicity; 1,541 women classified as “other,” including Asians, American Indians, and Native Hawaiians, were also excluded from analyses due to group heterogeneity and relatively small sample sizes. After applying these exclusion criteria, 22,778 women with EC treated with hysterectomy remained and were included in the analytical sample. Race/ethnicity was designated as the primary exposure, categorized as follows: Non-Hispanic White (hereafter referred to as White), Non-Hispanic Black (hereafter referred to as Black), and Latina. We defined postoperative complications as previously described 13 – 15 and included (1) wound complications (deep organ space surgical site infection, deep surgical site infection, and surgical wound dehiscence); (2) pulmonary complications (pneumonia and pulmonary embolism); (3) renal complications (acute renal failure, progressive renal insufficiency, and urinary tract infection); (4) neurologic complications (coma lasting > 24 hours and stroke); and (5) cardiovascular complications (deep venous thrombosis, myocardial infarction, and cardiac arrest). SAEs included a composite of the following: sepsis (yes, no), septic shock (yes, no), bleeding requiring transfusion (yes, no), mechanical ventilation lasting > 48 hours (yes, no), unplanned reintubation (yes, no), and unplanned reoperation (yes, no) 16 , 17 . All variables are available and coded in the NSQIP database. Length of hospital stay was calculated as the difference (in days) between the date of surgery and the date of discharge. We assessed the following covariates: age at diagnosis (<60 years, ≥60 years), body mass index (BMI; <29, 30–39, ≥40), American Society for Anesthesiologists (ASA) classification (<3, ≥3), diabetes mellitus (yes, no), smoking within one year prior to surgery (yes, no), preoperative dialysis (yes, no), functional dependency (yes, no), congestive heart failure (yes, no), history of chronic obstructive pulmonary disease (COPD; yes, no), hypertension requiring medication (yes, no), steroid use for a chronic condition (yes, no), preoperative weight loss (yes, no), operation time (mean and interquartile range), surgical approach (laparoscopy, abdominal hysterectomy, vaginal hysterectomy), and lymph node assessment (yes, no). CPT codes for each surgical approach are presented in the supplemental materials . We estimated the risk-adjusted odds ratios (ORs) for postoperative complications and SAEs with their respective 95% confidence intervals (CIs) using multivariable logistic regression after adjusting for covariates with standardized differences greater than 10% between racial/ethnic groups, including age, diabetes mellitus, smoking within one year of surgery, hypertension requiring medication, surgical approach, and lymph node assessment. We used Poisson regression with robust standard errors for parameter estimates to compare lengths of hospital stay across racial/ethnic groups, presented as incidence rate ratios (IRRs) and 95% CIs. All analyses were performed using Stata/IC 15 software (StataCorp LP, College Station, Texas). A p-value < 0.05 was considered significant.

Table 1 shows the clinical characteristics of the sample, both overall and according to race. Patients older than 65 years represented 46% of the sample, and the median BMI was 35 kg/m 2 (interquartile range: 29–42 kg/m 2 ). In addition, 58.3% of patients had an ASA classification ≥3 (n = 13,267), and 57% had hypertension requiring medication. Fewer than 25% of patients had diabetes mellitus, and 7.9% had a history of smoking within a year of surgery. Laparoscopic hysterectomy was the most common approach, applied in 85.2% of cases, whereas 13.6% were treated with abdominal hysterectomy. A higher percentage of Black women were diagnosed with either diabetes mellitus (29.9% vs. 21.4%) or hypertension requiring medication (73.9% vs. 55.5%) compared with White women. Finally, 26.3% of Black women underwent abdominal hysterectomy compared with 11.8% of White women. Latina women tended to be younger (≥65 years: 26.6% vs. 48.3%) and more likely to have diabetes mellitus (33.7% vs. 21.4%) than White women. Finally, a greater percentage of White women (86.9%) underwent laparoscopic hysterectomy to treat EC compared with other racial/ethnic groups (72.7% of Black women and 82.9% of Latina women). Of the 22,708 patients undergoing an EC-related hysterectomy, 711 (3.1%) developed postoperative complications within 30 days following surgery ( Figure 1 ). Both Black (adjusted OR: 1.62, 95% CI: 1.05–2.48) and Latina women (adjusted OR: 1.79, 95% CI: 1.04–3.09) had higher risks of postoperative complications compared with White women. The 30-day incidence of SAEs was 5.0% (n = 1,146) overall. Compared with White women, Black women (adjusted OR: 1.55, 95% CI: 1.13–2.15) but not Latina women (adjusted OR: 1.13, 95% CI: 0.73–1.76) were more likely to develop SAEs. Finally, Black women were likely to have a longer length of hospital stay (adjusted IRR: 1.18, 95% CI: 1.07–1.30; Figure 2 ) compared with White women, but no difference in length of hospital say was observed for Latina women compared with White women ( Figure 2 ). We further explored various components of composite postoperative complications and SAEs ( Figure 3 ). Among postoperative complications, wound complications (2.9% vs. 1.7%), pulmonary complications (1.5% vs. 0.8%), and renal complications (2.9% vs.1.7%) occurred at higher rates among Black women than among White women, whereas only wound complications occurred more frequently in Latina women than in White women (2.8% vs. 1.7%). Among SAEs, bleeding and septic complications occurred at increased frequencies among Black and Latina women compared with White women.

In this large cohort of racially and ethnically diverse women with EC treated with hysterectomy, Black and Latina women experienced more postoperative complications than White women, and Black women also experienced more SAEs and longer lengths of hospital stay than White women. These racial/ethnic disparities in surgical outcomes among women with EC treated with hysterectomy persisted despite adjusting for baseline differences in comorbid conditions, surgical approaches, and lymph node assessment, which are characteristics that vary by race and impact surgical outcomes. To our knowledge, few studies have examined racial disparities in surgical outcomes among women with EC treated with hysterectomy. The results of the present study suggest that minority women diagnosed with EC experience more short-term morbidity following hysterectomy than their White counterparts. Because hysterectomy is the standard of care for EC treatment and management, the identification of racial/ethnic disparities in post-hysterectomy outcomes may allow for the development of updated, evidence-based recommendations for the surgical management of EC and subsequent postoperative care. Although racial/ethnic disparities in surgical outcomes among women who undergo EC-related hysterectomies have not been extensively explored, our findings agree with the general body of evidence, which informed the 2003 Unequal Treatment report published by the Institute of Medicine (IOM) 18 , and other studies that have been published since the release of the seminal IOM report 19 . Evidence suggests that Black Americans, as compared with White Americans, have significantly higher morbidity following various surgical procedures. For example, Black individuals undergoing bariatric surgery, nonobstetric or obstetric procedures, or thyroidectomy experience higher rates of postoperative complications than their White counterparts 20 – 23 . Significant racial disparities in the occurrence of postoperative complications have also been observed following surgical oncological procedures. Using the Nationwide Inpatient Sample (NIS), Sukumar and colleagues 24 examined data from over 300,000 patients who underwent one of eight major surgical oncological procedures and found that Black and Hispanic patients were 24% and 5% more likely to experience postoperative complications, respectively, than White patients. Further, in analyses stratified by surgical procedure, Black women who underwent hysterectomy were 43% more likely to experience postoperative complications than White patients, whereas no differences in postoperative complications were noted between Latina and White patients who underwent hysterectomy. In the NIS analysis, gastrointestinal, genitourinary, and vascular complications occurred with increased frequency following hysterectomy among Black patients compared with White patients. By contrast, the NSQIP sample used in the present study showed that pulmonary and wound complications occurred at a higher frequency following hysterectomy among Black compared with White patients. In contrast to the present study findings, two previous NSQIP analyses found no differences in postoperative complications between Black and White patients undergoing EC-related hysterectomy using multivariable-adjusted analyses 9 , 10 . In the first report, Madhi et al. analyzed NSQIP data including 2,899 White and 349 Black women who were diagnosed with EC and underwent hysterectomy between 2005 and 2011 and reported that 21% of Black women and 12% of White women experienced postoperative complications within 30 days following hysterectomy 9 . However, a multivariable-adjusted analysis stratified by surgical approach (laparotomy vs. laparoscopy) found no racial differences in postoperative complications. In the second NSQIP analysis, which included 13,720 White and 1,553 Black women who were diagnosed with EC and treated with hysterectomy between 2010 and 2015, Lee et al. reported that 22.5% of Black women and 13.6% of White women experienced postoperative complications. However, when these analyses were adjusted by surgical approach, age, BMI, preoperative lab values, and operative time, the association between race and postoperative complications was attenuated. Differences between the findings of the present study and those of previous reports may be due to differences in sample sizes. Our sample included the largest sample size to date, representing NSQIP data from 22,708 women with EC treated with hysterectomy, and included Latina women, making our sample the largest and most comprehensive sample studied for racial/ethnic disparities in surgical outcomes. This analysis adds to the sparse and conflicting body of knowledge regarding short-term morbidity disparities among women with EC who undergo hysterectomy. SAEs represent an important subset of postoperative complications reflecting the overall surgical performance 17 , 25 . Our findings also show that Black women are more likely than White women to experience SAEs. To our knowledge, this is the first study to specifically examine racial disparities in SAEs among women with EC who undergo hysterectomy; thus, our findings present a novel addition to the literature. We also observed that Black women who underwent EC-related hysterectomies experienced 18% longer hospital stays than White women, which agrees with a previous report 26 . Longer lengths of hospital stay have important economic ramifications, increasing the total procedural costs. Longer lengths of hospital stay among Black women who undergo EC-related hysterectomy may be due to comorbidities among this group of women, as baseline characteristics revealed higher rates of obesity, diabetes mellitus, hypertension requiring medication, and smoking within one year prior to surgery among Black women compared with White women. Moreover, operative time was slightly longer for Black women than White women, which might indicate that Black women presented with more surgically complex disease, requiring a longer recovery time and longer length of hospital stay. Of note, certain uterine features, including the presence and size of fibroids, and nonuterine pelvic features including endometriosis, ovarian cysts, and adhesions are associated with greater surgical complexity 27 and some of these features are more common among Black women 28 . Limitations of the present study included the inability to identify novel predictors of racial disparities among women with EC treated with hysterectomy. Specifically, data on tumor characteristics, such as histological subtype, and information on surgeon and facility characteristics were unavailable for analysis. Differences in the volume of cases handled by facilities and surgeons have been associated with differences in the incidence of postoperative complications and surgical mortality following certain cancer-related surgeries 29 , 30 . Importantly, racial disparities in access to higher volume facilities and surgeons exist among women undergoing hysterectomy 31 . Beyond health care metrics, other social determinants of health may also affect the relationship between race/ethnicity and surgical outcomes among women with EC treated with hysterectomy; however, these factors could not be addressed in the present study using the available NSQIP data. For example, differences in the ability to access high-volume facilities and surgeons for treatment management, differences in living conditions, and differences in the ability to access sufficient work leave for proper recovery following surgery may exist across racial/ethnic groups, and these differences may contribute to the occurrence of racial disparities in short-term morbidity. The strengths of our NSQIP analysis include the large sample size, the inclusion of Latina women, and the use of validated ICD codes to identify postoperative complications and SAEs. In conclusion, we observed significant racial disparities in surgical outcomes among women with EC who underwent hysterectomy between 2014 and 2020. This analysis expands the paradigm of EC disparities to include the surgical experience as an area affected by racial disparities. Implications of these findings also call for improvements to peri-and postoperative care, patient monitoring and patient counseling. Future work exploring the direct and indirect effects of social determinants of health on surgical outcomes remains warranted.

With an estimated 65,000 new diagnoses in the United States (US), endometrial cancer (EC) is the most common gynecologic cancer in women 1 . Well-documented Black-White disparities in EC incidence and mortality exist. EC incidence rates accounting for higher hysterectomy prevalence among Black women are 99.2 per 100,000 Black women compared to 88.4 per 100,000 White women 2 . Additionally, mortality rates for Black women with EC are approximately two-fold greater than those for White women [age-standardized mortality per 100,000: 8.9 (Black) vs. 4.5 (White)] 1 . To date, most research has focused on individual-level differences, such as aggressive tumor characteristics observed in Black women ( e.g ., histology and stage at presentation) and the low rate of guideline-concordant adjuvant treatment among Black women, as the main drivers of existing disparities, with limited focus on structural-level inequities 3 . Additionally, few studies have addressed racial disparities in EC incidence or outcome among non-Black minority populations, particularly Latinas; however, compared with White women with EC, Latina women with EC have been reported to display tumor characteristics linked to worse prognosis and higher cancer-specific mortality 4 . Although racial disparities in EC incidence, presentation, and mortality are known to exist, whether these disparities extend to the surgical experience, particularly the occurrence of postoperative complications, serious adverse events (SAEs), and length of hospital stay following EC-related hysterectomy, remains unknown. Previous studies have focused on surgical characteristics ( e.g ., route of surgery and operative time), age, obesity, and lymph node status as predictors of surgical outcomes rather than racial/ethnic disparities 5 – 8 . Two studies have examined associations between race and postoperative outcomes among women with EC treated with hysterectomy, neither of which observed differences in postoperative complication rates between Black and White women, and no analyses were performed in either study to examine racial/ethnic disparities in SAEs or length of stay 9 , 10 . Further, Latina women were not included in these studies. To address these gaps, the present study examined associations between race/ethnicity and a range of surgical outcomes among women with EC treated with hysterectomy. Here, we more fully characterize racial/ethnic disparities in short-term morbidity among women with EC undergoing hysterectomy and expand on previous studies by examining more recent cases (spanning 2014–2020) and including Latina women in our analyses. Data indicating that minority races may be at higher risks for postoperative complications following EC-related hysterectomy may allow for improved peri-operative care, monitoring and counseling to improve outcomes among these patients.