Logistic regression analysis to predict prognosis in COVID-19 patients at tertiary care center in Rewa, Madhya Pradesh, India

Abstract

Introduction SARS-CoV-2 was declared as a global pandemic by World Health Organization (WHO) owing to its high infectivity and pathogenicity that has been quickly entangling the world since its first reported outbreak in China in December, 2019. The main objectives of the present work were to quantify the influence of age, duration of hospitalization, duration of presenting symptoms and comorbidities on the probability of mortality from COVID-19 disease. Methods This was a cohort study involving adult in-patients (30th April to 31st December’20) with laboratory-confirmed (RT-PCR) COVID-19 at Sanjay Gandhi and Gandhi Memorial Hospital, Rewa, Madhya Pradesh, India, whose data was retrospectively extracted from hospital records. These files were selected randomly attributing to 25% of the total number of in-patients fulfilling the inclusion and exclusion criteria. Out of the 131 cases, 26 cases were deceased while 105 cases were recovered. A binary logistic regression analysis was employed to quantify various factors responsible for mortality among COVID-19 patients. Results The result showed that age & duration of presenting symptoms had a positively significant influence on COVID- 19 mortality. Based on the binary logistic regression model, there is an increased risk of COVID-19 mortality of 1.028 per year of age & 1.213 times per day of presenting symptoms. Further analysis also showed that COVID-19 patients who was hospitalized since 10 days, was approximately 8 times less likely to die as compared to a COVID-19 patient who was hospitalized since 1 day. This is explained by the presence of co morbidities among older aged patients as reported in previous literature. Discussion Among several factors contributing to mortality; in particular, older age and duration of presenting symptoms was associated with higher odds of in-hospital death, while duration of hospitalization with lower odds of in hospital death. The current study confirmed that increased age was associated with death in patients with COVID-19.

Keywords

• COVID-19, Logistic Regression, Prognosis, Mortality, Risk Factors