Machine Learning Analysis of National Vaccination Schedules and Rates of Compliance Reveals Correlation with COVID19 Mortality

Abstract

We questioned whether previous vaccination histories, as reflected by the national vaccination schedules and compliance rates, might correlate with mortality rates from COVID-19. Using a machine learning algorithm, we aligned the WHO data on national vaccinations schedules and compliance rates with national mortality rates, as published daily by official health authorities and posted on the WHO website. Vaccination schedule and the national rates of compliance were significantly associated with national mortality rates. The five most influential vaccines were Rubella, inactivated Polio, DTP (Diphtheria, Tetanus, Pertussis) and the Pneumococcal vaccine. These vaccinations are more likely to be associated with younger generations whereas mortality is reported mainly in older people. Our findings suggest that further studies exploring the possible mechanism(s) underlying this association with vaccination may provide important information in dealing with an ongoing pandemic. Supporting our conclusions, a recent computational study predicted an epitope similarity between the viruses causing COVID_19, measles, and rubella, the latter two covered by the triple vaccine MMR identified in our studies. This suggests that childhood vaccinations contribute to the differential clinical presentation of COVID_19 in young and older populations. Some recent publication also hypothesized that there would be a non-specific protective effect for the MMR vaccine.