Trust and the impact of state interventions on healthcare utilization during the COVID-19 pandemic in Germany: An instrumental variables approach

BACKGROUND: Non-pharmaceutical interventions were enacted to minimize COVID-19-related mortality and morbidity during the pandemic. OBJECTIVE: To examine the causal effect of state interventions, implemented to limit COVID-19 mortality and morbidity, on non-COVID-19 hospital utilization, and whether this effect varied by the level of trust in the healthcare system. METHODS: The causal effect of state interventions on non-COVID-19 hospital utilization is estimated using an instrumental variables framework, exploiting the variation in the timing of municipal and state elections, and by level of trust in the healthcare system. Main data source is all hospital admissions from 2016-2021 in Germany. All non-respiratory hospital admissions, admissions for emergency conditions, elective surgeries, and chronic conditions are examined. RESULTS: On average, non-respiratory hospital admissions declined by 19.2 admissions per 10,000 population per four-weeks (-12.5%, p-value<0.001) throughout the pandemic while four-weeks of contact restrictions caused a decline of 2.7 non-respiratory hospital admissions per 10,000 population (-1.8%, p-value:0.05). Among emergency admissions, contact restrictions led to declines only in acute myocardial infarction admissions (-0.3 per 10,000 population, -12.6%, p-value<0.001). Effects varied markedly by trust in the healthcare system. Counties with low trust experienced the largest declines in non-COVID-19 hospital utilization; during a four-week contact restriction, non-respiratory hospital utilization rate was 45.8, 24.6, and 8.1 per 10,000 population lower than the pre-pandemic levels in low, medium, and high trust counties corresponding to declines of 30%, 16%, and 5% (p-values:<0.001), respectively. CONCLUSIONS: Our findings highlight the importance of patient trust in the healthcare system, even within a universal healthcare setting.