COVID-19 Hospitalisation in Portugal, the first year: Results from hospital discharge data

Using Portuguese diagnosis-related groups (DRGs) data we analysed the dynamics of the COVID-19 hospitalisation in Portugal, with the ultimate goal of estimating parameters to be used in COVID-19 mathematical modelling.

After processing the data, corresponding to the period between March 2020 and the end of March 2021, an exploratory analysis was conducted, in which the length of stay (LoS) in hospital care, the time until death (TuD), the number of admissions and mortality count were estimated and described with respect to the age and gender of the patients, but also by region of admission and evolution over time.

The median and mean of LoS was estimated to be 8 and 12.5 days (IQR:4-15) for non-ICU patients and for ICU patients as 18 and 24.3 days (IQR:11-30). The percentage of patients that, during their hospitalisation, required ICU care was determined to be 10.9%. In-hospital mortality of non-ICU patients (22.6%) and of ICU patients (32.8%) were also calculated. In a visual exploratory analysis we observed that they changed with time, age group, gender and region.

Various univariable probability distributions were fitted to the LoS and TuD data, using maximum likelihood estimation (MLE) but also maximum goodness-of-fit estimation (MGE) to obtain the distribution parameters.

AMS Subject Classification

62-07,62E07,62F10,62P10