Procedure Prioritization During a Nationwide Ban on Non-Urgent Healthcare: A Quasi-Experimental Retrospective Study of Hospital Data in Switzerland

Abstract

Background: During the COVID-19 lockdown in spring 2020, Switzerland restricted non-urgent healthcare services to safeguard capacity. While prioritization of care was supposed to be driven by medical urgency, demographic factors or economic incentives might have influenced the hospitals’ resource allocation decisions. Objectives: This study investigates potential determinants of procedure prioritization in hospitalized patients during the lockdown period. Design: Quasi-experimental retrospective study of hospital data in Switzerland. Methods: We analyzed 496 456 adult patients with known insurance status and a recorded procedure, admitted for cardiovascular, orthopedic/musculoskeletal or oncological reasons from January 2017 (3 years before the COVID-19 outbreak) to mid-April 2020 (in the first year of the COVID-19 pandemic), to obtain admission rate ratios (ARRs, “lockdown” admission rates divided by “normal” rates) from negative binomial regression analysis of fortnightly admissions for frequent procedure-diagnosis combinations. Quade and Wilcoxon signed-rank tests compared ARRs between sex×age, insurance and comorbidity strata. Results: Admission rates showed significant reductions for 29 of 53 procedure-diagnosis combinations. Reductions varied strongly by emergency, with largest decreases in orthopedic procedures for arthrosis (osteoarthritis) and non-arthritic joint disorders, and the smallest in cerebral imaging for stroke patients and surgical procedures for malignant neoplasms. The only difference in ARRs between strata was a stronger decrease in admission rates for cardiovascular combinations for patients with private versus basic health insurance. Conclusion: While medical procedures were affected to varying degrees by the ban on non-urgent healthcare during the COVID-19 lockdown, we found no robust evidence that factors other than medical urgency influenced healthcare prioritization.