Surgical Care for Women with Endometrial Cancer in Florida

Introduction

Minimally invasive surgery (MIS) was designated as a quality measure for endometrial cancer by the Society of Gynecologic Oncology and the American College of Surgeons Commission on Cancer in 2014.^1–4 According to data in the National Inpatient Sample database, in 2012–2013, 52.4% of women with endometrial cancer were treated with open surgery versus 47.6% treated with MIS. ² In addition, significant disparities were reported in the races of patients receiving minimally invasive approaches to hysterectomy as well as disparities in the volumes, locations, and ownership characteristics of the organizations providing these surgeries. ² Furthermore, complications and cost per case were higher in total abdominal hysterectomies, compared to minimally invasive approaches. ² Hysterectomies performed in the outpatient setting are not included in the National Inpatient Sample, which limits the generalizability of conclusions regarding hysterectomies performed in the inpatient setting.

Subsequent studies using the SEER-Medicare database and the National Surgical Quality Improvement Project database showed significant increases in minimally invasive hysterectomy for endometrial cancer. Using the SEER-Medicare database, Wright et al. reported that minimally invasive hysterectomy increased from 9.3% in 2006 to 61.7% in 2011 with robotic-assisted procedures accounting for 62.3% of the minimally invasive cases. ⁵ Casarin et al.'s analysis from the American College of Surgeons' National Surgical Quality Improvement Project database showed that MIS increased from 24.2% in 2008 to 71.4% in 2014, with substantial decreases in 30-day morbidity, readmissions, and reoperations for women treated for endometrial cancer. ⁶

Although studies using national datasets have shown pertinent disparities in hysterectomy routes, it is not known if such disparities persist in populations with high adoption of minimally invasive hysterectomy. Florida, a state located in the southeastern United States, has high utilization of MIS techniques. Using contemporary data available for characterizing inpatient and outpatient hysterectomies in Florida, the current study's primary objective was to determine if racial, social, and/or geographic disparities in routes of hysterectomy for endometrial cancer still exist in this state with its high utilization of MIS techniques. Secondary objectives included assessments of complications and associated costs of abdominal hysterectomy versus minimally invasive hysterectomy.

Materials and Methods

Results

There were 6513 patients identified who underwent hysterectomy for the treatment of endometrial cancer between June 2014 and December 2016. During this time period, 5300 (81.4%) patients underwent MIS while 1213 (18.6%) patients underwent abdominal hysterectomy (Table 1). Age was the only variable that was not statistically different between patients who underwent abdominal surgery versus MIS. There were 210 (3.2%) surgeries performed by physicians who only performed 1 hysterectomy for endometrial cancer during the time period. Hospitals providing hysterectomies included 74.5% teaching, 92.4% large-sized, and 66.5% high-volume hospitals. Patients who underwent abdominal hysterectomy were more likely to be obese (41.1% versus 28.7% for MIS; p < 0.001), treated for metastatic cancer (23.2% versus 7.6% for MIS; p < 0.001), and experience complications (17.2% versus 2.6% for MIS; p < 0.001). Patients undergoing minimally invasive hysterectomy were more likely to have lymphadenectomy performed (70.2 % versus 38.8% for abdominal surgery; p < 0.001).

Table 2 shows factors associated with patients undergoing MIS versus abdominal hysterectomy with ORs and 95% CIs. Patients who had concurrent lymphadenectomy (OR: 3.33; 95% CI: 2.88–3.85), were located in Central Florida (OR: 1.71; 95% CI: 1.17–2.48), or the Florida Peninsula (OR: 1.73; 95% CI: 1.17–2.56), and in high-volume (OR: 1.69; 95% CI: 1.30–2.20) or medium-volume (OR: 3.11; 95% CI: 2.37–4.08) hospitals were significantly more likely to receive MIS than abdominal surgery. African Americans (OR: 0.41: 95% CI: 0.34–0.50), other non-white race patients (OR: 0.64; 95% CI: 0.49–0.84), patients on Medicaid insurance (OR: 0.42; 95% CI: 0.30–0.59), or those who received care at a teaching hospital (OR: 0.82; 95% CI: 0.68–0.98) or government hospital (OR: 0.50; 95% CI: 0.38–0.65) were less likely to undergo MIS.

Table 3 details factors associated with patients experiencing major perioperative complications. Patients who had abdominal hysterectomy (OR: 6.31; 95% CI: 4.81–8.28), metastatic cancer (OR: 2.61; 95% CI: 1.98–3.45), or more comorbidities (OR: 4.56; 95% CI: 1.81–11.49 for >5 comorbidities), or who had received care at not-for-profit hospital ownership (OR: 2.56; 95% CI: 1.68–3.91), or medium-volume (OR: 1.80; 95% CI: 1.07–3.02) or high-volume (OR: 1.75; 95% CI: 1.04–2.93) hospitals were more likely to experience complications. Patients receiving care in Central Florida (OR: 0.42; 95% CI: 0.23–0.77), or West Florida (OR: 0.55; 95% CI: 0.32–0.96), or at a teaching hospital (OR: 0.46; 95% CI: 0.35–0.61) were less likely to experience complications.

Table 4 presents the costs of the procedures and their associations with complications. Costs were significantly increased for abdominal surgery versus MIS (mean difference: $2,035.20; p < 0.001) and for abdominal procedures with complications versus MIS with complications (mean difference: $4,541.50; p < 0.001). Costs were similar for non-complicated abdominal surgery and MIS.

Discussion

Evidence published in the last 2 years suggests that, while the overall rate of hysterectomies had decreased from 2000 to 2014, a large proportion of hysterectomies have been performed in outpatient settings.^18,19 Thus, inpatient-only data might not capture changing treatment patterns for gynecologic conditions accurately, ¹⁹ including endometrial cancer. In addition, it was not known if disparities persist in hysterectomy routes in populations with high adoption of minimally invasive hysterectomy. The present study was performed to address this critical gap by examining inpatient and outpatient hysterectomies in a state with a high utilization of MIS techniques to determine if disparities persisted in the routes, complications, and costs of hysterectomy for endometrial cancer. The study showed that certain patient populations, including African Americans, Medicaid-insured, and obese individuals, were more likely to undergo abdominal hysterectomy than MIS. In contrast, high- and medium-volume hospitals, those with nonteaching status, and those with nongovernment hospital ownership were associated with greater use of MIS. Of note, abdominal hysterectomy was more expensive and associated with increased complication rates. Together, these findings have important implications for clinicians, payers, and policy makers.

Encouragingly, Florida has a high adoption of MIS for endometrial cancer, compared to previous national studies,^2,20 with a large proportion of minimally invasive hysterectomy noted (81.4%). The inclusion of outpatient cases likely explains why the proportion of hysterectomies performed in a minimally invasive manner was so much higher than in previous studies, which utilized inpatient-only data.^2,20 The dissemination of minimally invasive technologies, including laparoscopy—especially robotics—has resulted in a shift from inpatient hysterectomies toward outpatient hysterectomies. ²¹ MIS appears to lead to shorter lengths of stay and fewer complications even among high-risk populations, such as obese and morbidly obese women.^22–25 This supports MIS appropriateness for outpatient hysterectomies and likely explains its increasing use in outpatient settings, given that early ambulation and discharge is a desired goal in these settings.

It has also been suggested that the adoption of MIS in outpatient settings might improve access for disadvantaged patient groups. ²¹ Although all patient groups have benefited from the increased prevalence of MIS in Florida, significant disparities still persist. Racial minorities, such as African American patients and those of other non-white races were less likely to undergo minimally invasive hysterectomy than white patients. This finding is consistent with previous research that has reported racial disparities in patients undergoing minimally invasive hysterectomy among women receiving inpatient treatment for endometrial cancer in the United States.^2,20

In addition, Medicaid insurance, government-owned hospitals, and teaching hospitals were associated with increased use of abdominal hysterectomy, which calls for considerations of systematic aspects contributing to disparities, such as bias against poorer patients, reduced access to care among minority patients, and uneven distribution of hospital resources or technical expertise. For example, teaching and government hospitals are likely to have more African American patients whose payers are CMS-based, and these women often have less access to care and thus present with more advanced disease and with more comorbidities (e.g., obesity). Given how tangled these relationships are, providing more minimally invasive procedures in teaching and government hospitals, without efforts to improve access to care among African American women with endometrial cancer, could prove to be unhelpful.

Surprisingly, there are quite a few cases of hysterectomy performed by surgeons who only operated on 1 endometrial cancer case per year (3.2% overall). Further analysis is required to determine why such a surprisingly high proportion of hysterectomies are performed by surgeons who only perform surgery on 1 case of endometrial cancer per year. It could be that these cases were handled by surgeons who had a gynecologist–oncologist or a general surgeon as a backup. It is also possible that these endometrial cancers were not diagnosed preoperatively, as a surprisingly high rate of unsuspected uterine malignancies (1.44%) has been reported among women undergoing hysterectomies, particularly among patients who were age 55 or older.^26,27 This latter possibility should be considered by the Centers for Disease Control and Prevention, which might soon be reviewing the preoperative evaluation of women scheduled for hysterectomies or myomectomies to treat uterine sarcomas. ²⁸ Additional research is needed to understand factors contributing to the high proportion of hysterectomies performed by surgeons who only operate on 1 case of endometrial cancer per year, and to assess the impact this could be having on clinical outcomes for patients with endometrial cancer.

When considering complications and associated costs, the open approach to hysterectomy appears to be more expensive,^29,30 and this is more so when the procedure is accompanied by complications.^2,20 In this regard, the current study showed that patients undergoing abdominal hysterectomy were six times more likely to experience complications. It also appears that patients who underwent abdominal hysterectomy in the current study were sicker, which was evidenced by a greater number of comorbidities and a greater proportion of metastatic disease. This has serious ramifications for clinicians and other parties involved in clinical decision-making regarding the route of hysterectomy. While the higher rate of complications following abdominal hysterectomy might be due to the invasiveness of the open approach, this could be compounded further by the tendency for patients with metastatic disease and greater comorbidities to be treated with abdominal hysterectomy.

Of note, the risk of complications appeared to be independently increased by the presence of metastatic disease and comorbidities. Overall, these findings indicate that the occurrence of complications are more likely to be affected by clinical characteristics, such as route of hysterectomy and patient disease burden, rather than demographic characteristics such as race, age, or insurance status. This also has important implications for payers and hospital administrators, as the difference in costs between abdominal and minimally invasive hysterectomy was significantly increased by the presence of complications. Thus, the costs of hysterectomy for endometrial cancer could be significantly reduced if stricter criteria were applied regarding the appropriateness of hysterectomy route, if needed at all, considering the patient's clinical state and risk for complications.

The current study had important limitations. The findings might not be generalizable to other populations with significantly different demographics. However, the population of Florida is geographically and socioeconomically diverse, which makes it appropriate for studying disparities associated with provision and utilization of health services, including cancer treatment. In addition, the administrative and retrospective nature of the data limited the ability to assess certain aspects of the care-delivery process, and disease progression. This nature of the data also limited the ability to discover more granular detail about the patients, procedures, and disease states. These limits constrained the ability to account accurately for all possible factors that might have affected the choice of hysterectomy route. For example, the current authors were unable to define patients who had had multiple hospitalizations or outpatient visits over time.

In addition, there are other anatomical or historical factors that could influence surgeon choice for route of hysterectomy, such as uterine size or past surgical history, that were not captured in this administrative data, although these confounders were unlikely to account fully for the differences reported here. In addition, the analysis of complications associated with the route of hysterectomy was limited to complications encountered during hospital stays or initial outpatient visits, as patient identifiers were not available to track patient revisits or readmissions.

Conclusions

Although Florida has a high adoption of MIS for endometrial cancer, disparities persist regarding surgical route chosen for certain patients. Consistent with national studies, African American, Medicaid-insured, obese, and sicker patients are less likely to undergo minimally invasive hysterectomy. Additional differences in hysterectomy routes were noted by geographic regions as well as hospital volumes, teaching status, and ownership. It is important for individual practices to review policies and hospital procedures to determine if there are systematic aspects contributing to these disparities. Regions with low adoption of MIS should be scrutinized closely to determine if surgeon expertise, availability of minimally invasive technologies, physician training, or unmeasured patient characteristics are reducing the odds of patients receiving minimally invasive hysterectomy. Efforts by state-level entities should also focus on improving access to care and minimally invasive hysterectomy for African American women with endometrial cancer. These efforts should include getting minimally invasive technologies to areas with low utilization of MIS or creating a better referral system for patients who would otherwise go to low-volume hospitals.