Case study: Is there any impact of insurance type on disposition outcomes in cancer patients on comfort care and hospital bed availability?

Abstract

e23242 Background: This retrospective study evaluates if insurance type affects disposition outcomes and demographics of cancer patients on comfort care. Our aim is to review if there are any potential disparities related to insurance status and its impact on hospital bed availability. Methods: We analyzed data from 518 comfort care patients with cancer treated at Upstate New York Hospitals (2022-2024). Patients were grouped by insurance type (Medicare, Medicaid, Private Insurance (PI)) and analyzed for factors like sex, age, race, cancer type, Case Mix Index (CMI), ICU involvement. Difficult-to-discharge (D-t-d) patients were identified based on medical needs, financial issues, and social determinants of health. Time from comfort care transition to discharge was assessed using propensity score matching (PSM) and linear regression. ED boarders served as a proxy for hospital bed availability, with correlation analysis using robust standard errors to find the link between comfort care volume and ED boarders. Results: Most patients had Medicare insurance (360), followed by Medicaid (88) and PI (70). Medicare patients were older with lower CMI, more balanced sex distribution, and more frequently discharged to skilled nursing facilities (SNF). Medicaid patients were mainly males, younger, had higher CMI, more ICU stays, and more of them discharged to SNF. PI patients were younger, had higher CMI, fewer D-t-d cases, more ICU stays, and were more often discharged home. White patients predominated in all groups, while African Americans were more in Medicaid. Most patients in all groups passed away in the hospital. Unmatched data showed that Medicare patients had significantly longer hospital stays than PI (median stay: Medicare 2.1 days, Medicaid 1.8, PI 1.3; p = 0.031). Matched data based on disposition, D.t.p, ICU stay, race, sex, age, CMI revealed no differences among groups. Patients with ICU stay before comfort care, linked to shorter hospital stays (β = -2.54, p < 0.001), were excluded as they passed away promptly. Re-analysis of hospital stays after comfort care, excluding ICU stays from the matched data, showed a significant decrease in PI compared to Medicare (β = -1.12, p = 0.046), with no differences in other comparisons. Linear regression showed that for each additional comfort care patient per month, ED boarders increased by 10.64 patients (p = 2.4x10-5), stressing the effect on bed availability. Conclusions: Despite higher CMI, fewer D-t-d PI groups discharged home and more expeditiously than the Medicare group. Based on the results, better screening tools based on insurance type and individuals can be created to aid in finding better solutions for effective discharge. Our analysis suggests the direct impact of the delayed disposition of comfort care patients on hospital bed availability, as one of the factors, causing a potential increase in associated healthcare costs.