gms | German Medical Science

Background and status of (inter)national research: Coronavirus (COVID-19) affected Italy early and severely (1). To limit viral transmission, the Italian government implemented a series movement restriction measures starting the week of March 8th, 2020. While lock-down measures effective in limiting the spread of COVID-19 (2), increasing studies report decreased hospitalization rates for acute myocardial infarction (AMI), stroke and heart failure (3–6).

Question and objective: We quantify changes in hospitalization rates and average length of stay in urgent care and specifically for AMI, stroke and heart failure.

Method or hypothesis: We use a controlled interrupted time series design - the treatment group defined as year 2020, and the control group as the average of corresponding weeks in previous years (2015 to 2019). Our analytical sample included inpatient hospitalizations (COVID-19 excluded) in public hospitals in Tuscany. Our study period was divided into 3 – “pre-lockdown” (Jan 4–Feb 22), “phase-in period” (Feb 22–Mar 9), and “post-lockdown” (Mar 10–July 31). We used segmented linear regression with an autoregressive error model to measure the immediate effect (level change) and overall mean (trend change) of lockdown.

Results: Our study included 234,928 inpatient hospitalizations in 2020 (2,146,939 total), representing 52.5% female, 50.1% junior high school educated, and 93.5% Italian. Inpatient planned care and urgent care both experienced significant decreases as compared to counterfactual (-39 [-51, -26]; -134 [-178, -89], respectively), with a larger percent decrease observed in planned care (-71%). We observed a large percent decrease in hospitalization rates compared to counterfactual for AMI (-30%; -1.32 [-1.98, -0.66]), stroke (-24%; -1.51 [-2.56,-0.44]), and heart failure (-47%; -8.7 [-11.1,-6.3]). We observe a larger percent decrease for those over 65 compared to under 65 for AMI and stroke. Average length of stay did not change significantly.

Discussion: We found large significant decrease in inpatient hospitalization, which is concerning due to the potential healthcare and economic implications of delayed care of time sensitive procedures. This may be explained by patient’s fear of contracting the virus if medical attention was sought, further aggravated through social norms due to stay-at-home orders. Additionally, reorganization of healthcare resources away from elective care may explain the large percent decrease we observe in planned care.

Practical implications: Our findings can inform policy makers on healthcare resource allocation to account for increases in longer-term impacts of delayed care in acute illnesses such as stroke, AMI and heart failure.

Appeal for practice (science and/or care) in one sentence: Stay at home orders should be judiciously balanced with messaging around the importance of timely diagnosis and treatment of acute illnesses.