Importance: Although the hospital at which a patient is treated is a known source of variation in mortality after inpatient surgery, far less is known about how the neighborhoods from which patients come may also contribute.
Objective: To compare postoperative mortality among Medicare beneficiaries based on the level of neighborhood deprivation where they live and hospital quality where they received care.
Design, setting, and participants: This cross-sectional study examined outcomes among Medicare beneficiaries undergoing 1 of 5 common surgical procedures (colon resection, coronary artery bypass, cholecystectomy, appendectomy, or incisional hernia repair) between 2014 and 2018. Hospital quality was assigned using the Centers for Medicare & Medicaid Services Star Rating. Each beneficiary's neighborhood was identified at the census tract level and sorted in quintiles based on its Area Deprivation Index score, a composite measure of neighborhood quality, including education, employment, and housing quality. A risk matrix across hospital quality and neighborhood deprivation was created to determine the relative contribution of each to mortality after surgery. Data were analyzed from June 1 to December 31, 2021.
Exposures: Hospital quality and neighborhood deprivation.
Main outcomes and measures: The main outcome was risk-adjusted 30-day mortality after surgery using a multivariable logistic regression model taking into account patient factors and procedure type.
Results: A total of 1 898 829 Medicare beneficiaries (mean [SD] age, 74.8 [7.0] years; 961 216 [50.6%] male beneficiaries; 28 432 [1.5%] Asian, 145 160 [77%] Black, and 1 622 304 [86.5%] White beneficiaries) were included in analyses. Patients from all neighborhood deprivation group quintiles sought care at hospitals across hospital quality levels. For example, 9.1% of patients from the highest deprivation neighborhoods went to a hospital in the highest star rating of quality and 4.2% of patients from the lowest deprivation neighborhoods went to a hospital in the lowest star rating of quality. Thirty-day risk-adjusted mortality varied across high- and low-quality hospitals (4.3% vs 7.2%; adjusted odds ratio [aOR], 1.78; 95% CI, 1.66-1.92) and across the least and most deprived neighborhoods (4.5% vs 6.8%; aOR, 1.58; 95% CI, 1.53-1.64). When combined, comparing patients from the least deprived neighborhoods going to high-quality hospitals vs patients from the most deprived neighborhoods going to low-quality hospitals, the variation increased further (3.8% vs 8.1%; aOR, 2.20; 95% CI, 1.96-2.46).
Conclusions and relevance: These findings suggest that characteristics of a patient's neighborhood and the hospital where they received treatment were both associated with risk of death after commonly performed inpatient surgical procedures. The associations of these factors on mortality may be additive. Efforts and investments to address variation in postoperative mortality should include both hospital quality improvement as well as addressing drivers of neighborhood deprivation.