Patients admitted between January 1, 2018 - February 29, 2020 with one of the following five primary diagnoses of acute myocardial infarction, stroke, heart failure, pneumonia, and chronic obstructive pulmonary disease in alignment with the Centers for Medicare and Medicaid Service's (CMS's) condition-specific mortality measures reporting

Patients admitted between May 1, 2022 and May 31, 2023 without a diagnosis for COVID-19 and with one of the following five primary diagnoses of acute myocardial infarction, stroke, heart failure, pneumonia, and chronic obstructive pulmonary disease in alignment with the Centers for Medicare and Medicaid Service's (CMS's) condition-specific mortality measures reporting.

Patients admitted between March 1, 2020 and April 30, 2022 without a diagnosis for COVID-19 and with one of the following five primary diagnoses of acute myocardial infarction, stroke, heart failure, pneumonia, and chronic obstructive pulmonary disease in alignment with the Centers for Medicare and Medicaid Service's (CMS's) condition-specific mortality measures reporting

Impact of COVID-19 pandemic on hospital care delivery

Our primary outcome will be based on a composite measure of in-hospital mortality or discharge to hospice as determined by discharge status code. The primary outcome will itself be expressed as a modified standardized mortality ration (mSMR) calculated as mean-shrunken number of observed deaths or discharge to hospice divided by the expected number of deaths or discharge to hospice for a center in that post-pandemic month assuming the effects of a typical center in the pre-pandemic era.

PICs are complications developed during the hospital stay which may reflect the performance changes post pandemic due to circumstances observed during the pandemic. List of PICs will be curated as has been done previously using Premier Healthcare Database as reported in Korvink et al. Med Care, 2023. For each of the patient, we will calculate the cumulative number of PICs developed during their inpatient stay, which would be a sum of individual counts of PICs out of the list of 74 total possible PICs. Eventually, observed and expected rate of PICs normalized to the length of stay will be calculated with analysis restricted to within hospitals after adjusting for patient level covariates. Subsequently, risk factor model will include hospital level factors to assess association of these factors to aggregate rates of PICs in the cohort to identify factors associated with better or poor performance post pandemic