Predicting NCOVID-19 Probability Factor with Severity Index

ABSTRACT

Hospitals worldwide are struggling to cope up with patient’s admission issues related with the increasing number of COVID-19 patients’ cases mainly driven by Delta variant, as severely ill nCOVID patients are found waiting for hospital beds, which are occupied by non-critical COVID patients. To make the situation worse, people who are partially or fully vaccinated against COVID-19 are also getting re-infected. Due to the absence of prior knowledge of an index of severity for COVID-19 patients, hospitals, with limited number of ventilators and medical equipment, fail to admit patients on any priority basis. With multiple tests kit available in market till now, there is none with an instantaneous index for severity prediction for COVID. This research develops a free and user-friendly algorithm titled “SAHAYATA 1427” (renamed herein Sahayata) which predicts a factor for a patient having the probability of disease nCOVID-19 termed as “probability factor” of COVID-19 for each patient. Concurrently, the algorithm also provides an index for severity by which the patient is affected by nCOVID, termed as “severity index.” The input data is both demographic and patient provided. The severity index is determined using artificial intelligence. Using a logistic regression model with data set of existing COVID patients, Sahayata predicts the probability factor for an nCOVID-19 patient with an accuracy, precision and recall of 88.17%, 100% and 87.3%, respectively. Results indicate that it can be used effectively both at hospitals by trained medical personnel and at home by the general population. Sahayata helps the COVID-19 patients living in rural communities with smaller patients care facilities with limited equipment by providing a way for efficient treatment care. © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.