Clinical, demographical characteristics and hospitalisation of 3,010 patients with COVID-19 in Friuli Venezia Giulia Region (Northern Italy). A multivariate, population-based, statistical analysis

OBJECTIVES: to describe the clinical and demographical characteristics of COVID-19 infected people in the Friuli Venezia Giulia Region (FVG, Northern Italy). DESIGN: retrospective cohort study with an individual level record linkage procedure of different administrative databases. SETTING AND PARTICIPANTS: the cohort included 3,010 patients residing in FVG who tested positive for COVID-19 between 1 March and 15 May 2020, 2020. Regional hospital admissions and deaths without hospital admissions up to June 1st, 2020 were analysed. Determinants of the probability of a highly severe illness were investigated in terms of hospitalisations or death without hospital admission. MAIN OUTCOME MEASURES: COVID-19 patients were identified from regional epidemiological data warehouse. Demographical and clinical variables such as gender, age, patient’s comorbidities, vaccinations, ARBs/sartans prescriptions, and geographical residence variables were collected by linking different databases. Descriptive analyses were performed. Logistic multivariate regressions were used to estimate the probability of hospitalisation or death, whichever came first. Model coefficients and odds ratios (OR) were reported. RESULTS: COVID-19 population in FVG had a mean age of 60 years and 59% were females. The study found that 37% had hypertension while patients with cardiologic diseases, diabetes, and cancer were around 15%; 22% of the cases were residing in retirement homes. Approximately 30% received flu or pneumococcal vaccination and a similar proportion of patients had at least one prescription of ARBs /sartans in the previous 6 months. Statistical models showed a higher probability of a worst course of disease for males, elderly, highly complicated (in terms of resource use) subjects, in the presence of cardiologic diseases, diabetes, and pneumococcal vaccination. People living in retirement homes had a lower probability of hospitalisation/death without hospital admission. The cohort was divided into two groups: COVID-19 patients infected in the territory and infected in retirement homes. Among COVID-19 patients infected in the territory, the probability of hospitalisation/death was higher for males, for older individuals, and for those with comorbidities. Diabetes resulted to be a risk factor (OR 1.79; 95%CI 1.23- 2.62), as well as pneumococcal vaccination (OR 1.64; 95%CI: 1.18-2.29), which is a likely proxy of fragile patients with pulmonary disease. The flu vaccination showed a potential protective effect with a 40% lower probability of hospitalisation or death (OR 0.62; 95%CI 0.44-0.85). Among the retirement homes cohort group, a higher probability of a bad course of disease emerged for males and for more complex patients. CONCLUSIONS: the greatest risk of hospitalisation/death as a measure of more severe illness was confirmed for males, elderly, and for individuals with comorbidities. Flu vaccination seemed to have had a protective effect while pneumococcal vaccination likely identified a group of high-risk patients to be actively monitored. For patients infected in the territory, different hospitalisation strategies were implemented by the regional health districts.