Network Analysis of US Hospital Financial and Operating Characteristics Before, During and After the COVID-19 Pandemic, 2011–2023

(Background) The COVID-19 pandemic and the federal response changed hospital revenue, coverage and relief simultaneously. Studies have described changes in the levels of hospital financial measures, but not whether the relationships among them changed. (Objective) To compare networks of hospital financial and operating characteristics before, during and after the pandemic. (Methods) Medicare cost reports for US short-term acute care and critical access hospitals (2011–2023; reports of 150–366 days) provided one observation per hospital per period for 13 characteristics. Correlation and partial correlation networks were estimated for 2011–2019 (4,554 hospitals), 2020–2021 (4,177) and 2022–2023 (4,141). Global differences were tested by permutation and edge differences by Fisher z tests with false discovery rate control. (Results) Overall connectivity was unchanged during the pandemic (mean absolute correlation 0.155 versus 0.151; p = 0.272) and weaker afterwards (0.139; p = 0.002 versus before). Nevertheless 38 of 78 edges changed between the pre-pandemic and pandemic periods after correction. Charity care and the Medicaid share decoupled (+0.155 to -0.039), bad debt and the Medicaid share decoupled (+0.098 to -0.040), and days cash and outside income became tied to the Medicare share (+0.108 to +0.259 and +0.104 to +0.251). Operating margin lost centrality (node strength 2.06 to 1.53) while the Medicare share gained it. (Conclusions) The pandemic left the overall structure of hospital finances intact while rewiring its components. Aggregate measures of connectivity can conceal substantial reorganization.