Lin, Jenny J.; Egorova, Natalia; Franco, Rebeca; Bickell, Nina A.

Introduction

Vulnerable women, those who are poor, of minority race or have lower educational attainment have historically experienced delays in breast cancer treatment, lower adherence to therapy recommendations, and higher mortality rates.^1–7 Furthermore, although rates of screening have equalized among racial groups, Black women are still more likely to die of breast cancer, even after controlling for breast cancer stage, treatment, comorbidity burden, and insurance type.^8–10 Stage for stage, patients with breast cancer with Medicaid suffer higher mortality than their privately insured counterparts after controlling for age and comorbidities, suggesting that quality of care may vary.¹¹ These disparities in outcomes may be partially due to patient-level factors such as higher burden of comorbidities, lower health literacy, or increased medical mistrust.^12–15 In addition, many of these patients receive care at safety-net hospitals, which predominantly serve poorer and minority communities, care for a larger proportion of Medicaid patients, and often rely on Disproportionate Share Hospital (DSH) payments to cover uncompensated care or Medicaid underpayments. One of the goals of the Affordable Care Act (ACA) of 2010 is to enable access to care and thereby improve patient care and outcomes.¹⁶ Yet, patients cared for in hospitals with higher Medicaid volume may experience treatment disparities due to issues with access to care or system navigation as a result of lack of coordination between medical subspecialties.^6,17

In New York State (NYS), 83% of Medicaid patients with breast cancer get their cancer surgery at hospitals that rely on DSH payments. As part of the mandate under the ACA, DSH reimbursements to these hospitals have decreased with the expectation that these hospitals will make up the revenue loss with higher volumes of patients now insured through the ACA.^18,19 In fact, hospitals that rely on DSH payments have experienced greater financial penalties due to having higher readmission rates,²⁰ thereby straining already stressed resources. Moreover, these hospitals tend to be in more rural areas and serve more vulnerable patients who may have higher burden of comorbidities or more access to care issues.^21–23

As women gain more choice and access after the enactment of the ACA, they will require information about hospitals' quality and outcomes of cancer care. We sought to determine whether there are differences in survival among patients with breast cancer who received primary cancer surgery at NYS hospitals with varying proportions of Medicaid volume. Because operative mortality for breast cancer surgery is negligible, we focus on 90-day survival as a proxy measure for the quality of surgical and postsurgical care received by women undergoing breast cancer surgery.

Methods

This study was conducted using NYS Vital Statistics death records data linked with NYS discharge inpatient and ambulatory surgery databases (SPARCS), which are publicly available and de-identified. We used the National Center for Health Statistic (NCHS) urban–rural classification scheme for counties to identify rural and urban location of hospitals.²⁴ We selected women aged 21 years and older who underwent primary surgery for breast cancer between 2008 and 2013. We chose to look after 2008 because there was a Medicaid expansion in the previous year. We excluded women who were not NYS residents, had other cancers, or had metastatic breast cancer because they would not be eligible for breast cancer surgery. We also excluded two free standing cancer centers because their care delivery structure is different from general hospitals that treat patients with multiple different conditions. For the remaining facilities, we used all Medicaid discharges from NYS hospitals and ambulatory centers (both hospital-based and free standing) and calculated proportion of Medicaid discharges yearly for each facility to create quintiles of facilities based on Medicaid volume. We named these quintiles very low Medicaid, low Medicaid, medium Medicaid, high Medicaid, and very high Medicaid facilities.

Using hospital claims, we ascertained breast cancer surgery (ICD9 procedure codes 85.20-85.23 or 85.41-85.48) to identify the index definitive procedure in women with a diagnostic code for breast cancer (ICD9 174.x, 233.0, and V10.3). For ambulatory surgery, we identified patients who had both breast cancer (ICD9 174.x, 233.0, and V10.3) and breast cancer surgery (CPT codes: 19301-19307). We excluded women with breast cancer and a diagnosis of advanced or secondary cancer (ICD9 197.0-198.7, 198.81, 198.82, 198.89, 199, 209.71-209.74, 511.81, and 789.51).

Sociodemographic information was obtained from the databases. To evaluate the burden of comorbidities, we used the Charlson comorbidity algorithm.²⁵ The study outcome (dependent variable) was overall (all-cause) 90-day overall survival as determined from NYS Vital Statistics death records. The primary independent variable was Medicaid quintile and confounders that were adjusted for included age, race, comorbidities, overall comorbidity score, insurance, and year of breast cancer surgery. Survival times were calculated from date of initial surgery to the date of death. Subjects still living after 90 days from the surgery were censored.

Statistical Analysis

Continuous variables were reported as mean values with standard deviation and analyzed using t-test. Categorical variables were reported as rates and analyzed using chi-squared test. We calculated hazard ratios using a marginal Cox model with a robust sandwich variance estimator controlling for clustering of patients within the facilities, age, race, insurance, year of surgery, comorbidities, and Medicaid quintiles. All analyses were performed with SAS 9.3 (SAS, Cary, NC) using two-tailed p values. The institutional review board at our institution approved this study.

Results

The total number of hospital facilities in NYS increased from 336 in 2008 to 366 in 2013 (Table 1), although this increase was primarily driven by a greater number of ambulatory surgery sites which increased from 102 to 149 from 2008 to 2013 rather than inpatient surgical sites which decreased in the same time period from 234 to 217. Moreover, the number of facilities that performed breast cancer surgery decreased from 197 to 177 during this same time period, although the annual number of patients with breast cancer increased from 7,582 in 2008 to 7,792 in 2013. Last, the number of patients with breast cancer receiving surgery from the higher quintile Medicaid hospitals (high Medicaid and very high Medicaid sites) increased from 3,179 in 2008 to 4,283 in 2013, and the number of patients in the very highest Medicaid quintile almost doubled from 922 to 1,694 in the same time period.

Table 1

Over the course of 6 years from 2008 to 2013, a total of 1,157 of patients with breast cancer received breast cancer surgery at the lowest Medicaid quintile facilities, 9,538 at low Medicaid facilities, 13,027 at medium Medicaid facilities, 14,673 at high Medicaid facilities, and 7,397 at the highest Medicaid quintile facilities (Table 2). Minorities, younger women, and those with more comorbidities received surgery at facilities with higher Medicaid volume (30.2% Black or Hispanic patients in the highest Medicaid quintile facilities vs. 5.3% in the lowest Medicaid quintile hospitals, p < .001). Women who received breast cancer surgery in the very high Medicaid facilities had higher 90-day mortality (2.1% vs. 0.07%, p < .001) compared with those who received surgery in the very low Medicaid facilities (Figure 1). The 90-day survival difference remained in multivariable analysis after adjusting for age, race, comorbidities, overall comorbidity score, insurance, and year of breast cancer surgery (Table 3). Compared with women who underwent surgery at very low Medicaid facilities, those who had surgery at very high Medicaid facilities had a greater than 2-fold risk of death in the 90 days after primary breast cancer surgery (hazard ratio: 2.73; 95% confidence interval 1.22–6.11).

Table 2-a

Table 2-b

Figure 1

Table 3-a

Table 3-b

Limitations

There are a few limitations to this study. We did not have data about breast cancer stage, but we excluded all patients with metastatic cancer, and the likelihood of death from early-stage breast cancer is low. Moreover, stage should not significantly have an impact on short-term 90-day survival. Similarly, we also did not have data about receipt of chemotherapy or radiation treatment, which can affect survival, but most nonmetastatic breast cancer survivors will survive greater than 90 days after surgery, and the receipt of chemotherapy or radiation treatment should also not have an impact on such short-term survival. We were also unable to obtain data on delay in treatment or poorer treatment adherence, but 90-day postsurgical survival should also not be affected by these factors. Moreover, Pfister et al²⁶ recently showed that it is possible to demonstrate outcome differences by hospital type with administrative data and that the addition of patient-level data (such as cancer stage or date of diagnosis) do not appreciably alter the findings. In addition, although year of surgery was included as a confounding variable, we can not explain variations in 90-day survival by year of surgery. Last, although facilities with the highest Medicaid volume had sicker patients who had a greater number of comorbidities, even after adjustment for type and number of comorbidities, we found a survival difference between women who received breast cancer surgery at the very low Medicaid quintile facilities compared with the very high Medicaid quintile facilities.

Discussion

Even after adjusting for multiple patient-level factors, we found that 90-day survival was significantly lower at NYS facilities in the highest Medicaid quintile, suggesting that these hospitals may provide poorer quality care for patients with breast cancer. The enactment of the ACA has improved access to care and has decreased racial disparities in insurance coverage, particularly in states with expanded Medicaid programs.^27–30 However, the challenge of improving patient outcomes through the ACA will persist if quality of care varies at under-resourced, financially strained hospitals that serve a predominantly Medicaid population.

Ly et al³¹ showed that hospitals that are more financially strained who were at the bottom 10% of operating margins had worse processes of care than those in the top 10% of operating margins. Similarly, we found that survival rates after breast cancer surgery were lowest in the highest Medicaid quintile facilities which are likely operating on the lowest margins. It is quite possible that the worse quality of care at financially strained hospitals ensues because of reduced ability to invest in quality improvement efforts or being short-staffed. Moreover, these lower-margin hospitals tend to be in more rural areas and serve more vulnerable patients who may have more comorbidities or access to care issues.^21–23

To further compound these issues, with the implementation of the ACA, safety-net hospitals that rely predominantly on DSH payments are experiencing greater penalties under the Hospital Readmissions Reduction Program, thus exposing them to even greater financial constraints.²⁰ Neuhausen et al. demonstrated that safety-net hospitals in California could face substantial funding gaps of up to $1.5 billion by 2019 because decreases in uncompensated care costs would not match increases in revenue from insured patients under the current ACA implementation plans with DSH payment reductions.³² Given that these hospitals take care of our most vulnerable patients with higher medical complexity, it is crucial that we better understand the factors associated with worse quality of care to develop interventions that can reduce these disparities in care.

One factor that has been shown to be associated with worse outcomes and quality of breast cancer care has been hospital surgical volume.³³ For complex surgeries of relatively uncommon cancers such as pancreatectomy and esophagectomy, the underlying assumption of improved outcomes harkens back to the “practice makes perfect” adage. But breast cancer is a relatively common cancer and with a fairly straightforward surgical procedure and negligible operative mortality. Our finding of significantly higher short-term mortality rates in high Medicaid facilities raises concerns about the quality of perioperative and postoperative care. Hospitals that perform fewer surgeries have been consistently shown to have higher mortality rates and worse quality process indicators for breast cancer.^34–36 A policy approach that has been undertaken by some states to help improve cancer care quality is to restrict state payments to lower-volume hospitals. For example, in 2009, NYS implemented limitations on Medicaid reimbursement to facilities that performed fewer than 30 breast surgery cases a year.³⁷ But even with this limitation, most NYS breast cancer surgical cases covered by Medicaid are performed at safety-net hospitals, and this limitation has performed little to have an impact on sites of breast cancer care. Furthermore, an unintended consequence of such policies is that they may limit access to breast cancer treatment for more vulnerable rural patients with limited access to care. Thus, development of policies to improve overall cancer care must ensure that they do not adversely affect access and places the responsibility of improving surgical outcomes with the hospitals themselves. The challenge for hospital administrators becomes even greater: how can care be improved when patient volume is increasing and hospital reimbursement decreasing? Possible strategies include negotiating with organizations such as the Commission on Cancer or American College of Surgeons' National Surgical Quality Improvement Program³⁸ to obtain markedly reduced fees or have Offices of Medicaid incentivize hospitals' participation in such efforts.

Conclusions

In conclusion, patients with breast cancer who received surgery in hospitals with the highest Medicaid volumes had a greater than 2-fold risk of death in the 90 days after surgery compared with those who received surgery in hospitals with the lowest Medicaid volume. These high Medicaid hospitals that already disproportionately serve poorer and minority patients with breast cancer may be at a greater risk of worsening quality of care with reductions in DSH payments and Hospital Readmission Reduction Program penalties under the ACA mandate because patient volume may increase but reimbursements drop. Further research should be made to assess how the implementation of these programs affect outcomes for the most vulnerable cancer patients who are receiving care at these lower-margin, financially strained hospitals.

Implications

Lawmakers and hospital leadership must ensure that quality of care and access are not negatively affected by these policies, which may contribute to worsening healthcare disparities.

References

Journal for Healthcare Quality is pleased to offer the opportunity to earn continuing education (CE) credit to those who read this article and take the online posttest at www.nahq.org/journal/ce. This continuing education offering, JHQ 276 (41.1 Jan/Feb 2019), will provide 1 hour to those who complete it appropriately.

Core CPHQ Examination Content Area

IV. [Domain - Quality Review and Accountability.]

Breast Cancer: Does type of hospital where you get surgery affect survival?

Learning Objectives:

After reading this article and taking this test, the learner should be able to:

1. Recognize that women with breast cancer who undergo surgery at hospitals with highest Medicaid volume have higher 90-day mortality rates.
2. Identify factors that may impact hospitals’ quality of care.
3. Question the impact of the Affordable Care Act on improving patient outcomes at under-resourced hospitals.

• 1. Women who receive breast cancer surgery at hospitals with higher Medicaid volume are more likely to have the following characteristics:
• a. Older, minority race, fewer comorbidities
• b. Older, minority race, more comorbidities
• c. Younger, minority race, fewer comorbidities
• d. Younger, minority race, more comorbidities
• 2. The 90-day mortality hazard ratio for women who received breast cancer surgery at hospitals with the highest Medicaid volume compared to those who underwent surgery at hospitals with the lowest Medicaid volume was:
• a. 2.1
• b. 2.5
• c. 2.7
• d. 3.2
• 3. Disparities in breast cancer care for women with Medicaid may be due to the following:
• a. Lower health literacy
• b. More comorbidities
• c. More medical mistrust
• d. All of the above
• 4. Patients' 90-day mortality rates after breast cancer surgery may be affected by the following:
• a. Adherence to treatment recommendations
• b. Delay in receipt of surgery
• c. Peri- or post-operative care
• d. Receipt of chemotherapy or radiation therapy
• 5. Part of the mandate under the Affordable Care Act (ACA) includes:
• a. Decreased Disproportionate Share Hospital (DSH) reimbursements
• b. Increased Disproportionate Share Hospital (DSH) reimbursements
• c. No change in Disproportionate Share Hospital (DSH) reimbursements
• d. None of the above
• 6. Worse quality of surgical care at hospitals may be due to the following:
• a. Higher surgical volume
• b. Increased financial strain
• c. Increased staffing
• d. None of the above
• 7. In New York State, the percentage of breast cancer patients with Medicaid who receive care at hospitals that rely on Disproportionate Share Hospital (DSH) reimbursements is:
• a. 2%
• b. 30%
• c. 57%
• d. 83%
• 8. With enactment of the Affordable Care Act (ACA), safety-net hospitals in California that rely on Disproportionate Share Hospital (DSH) reimbursements may
• a. Face substantial funding gaps
• b. Increase patient volume
• c. Lose insured patients to other hospitals
• d. Receive more funding
• 9. Hospitals that rely on Disproportionate Share Hospital (DSH) reimbursements may
• a. Experience financial hardship with the enactment of the Affordable Care Act (ACA)
• b. Have higher readmission rates
• c. Both A and B
• d. None of the above
• 10. An unintended consequence of policies to limit reimbursements to hospitals with low surgical volume may be:
• a. Decreased post-operative complications
• b. Decreased treatment access for vulnerable patients with limited access to care
• c. Increased volume of insured patients at safety-net hospitals
• d. None of the above

Authors' Biographies

Jenny J. Lin, MD, MPH, is an associate professor in the Division of General Internal Medicine, Icahn School of Medicine at Mount Sinai in New York, NY. She is a primary care physician and cancer researcher investigating comorbidity management in cancer survivors. For this project, she was responsible for the conduct and interpretation of study findings, manuscript writing, and revising.

Natalia Egorova, PhD, MPH, is an associate professor in the Department of Population Health Science and Policy, Icahn School of Medicine at Mount Sinai in New York, NY. She is expert in analysis of large healthcare survey and discharge databases. For this project, she was responsible for the data analysis, manuscript writing, and revising.

Rebeca Franco, MPH, is a project manager in the Department of Population Health Science and Policy, Icahn School of Medicine at Mount Sinai in New York, NY. She is responsible for managing several multisite studies. For this project, she participated in the execution of the study protocol and data analysis and was involved with manuscript development.

Nina A. Bickell, MD, MPH, is a professor in the Department of Population Health Science and Policy, Icahn School of Medicine at Mount Sinai in New York, NY. She is a health services researcher investigating cancer disparities. For this project, she was responsible for the design, conduct, and interpretation of study findings and manuscript writing and revising.