A comparative study between the first and second waves of COVID-19 in the Balearic Islands using machine learning on electronic health records
Khajuria, T.; Pou Goyanes, J. A.; Mirasso C.; Vicente. R.
Submitted , (2021)
Successive waves of COVID-19 are having a devastating effect on societies and health services all over the world. Aspects such as the possibility of early diagnostic and treatment, social measures or viral mutations have changed between the first and second waves of COVID-19. Here, we evaluate whether clinical markers of patients admitted to hospital with COVID-19 in an insular region have also changed during the two waves. In particular, we develop a stratified analysis together with machine learning methods to assess whether the set of clinical markers is discriminative of the wave in which a patient was admitted. This multivariate analysis takes into account non-linear combinations of clinical markers in order to reveal consistent changes between the waves. The analysis indicates that only weak changes between the waves occur once differences in the number and types of clinical tests conducted are discarded. The methodology developed can be used to reveal undetected changes in either the overall pattern of clinical response or clinical markers in upcoming epidemic waves