Analyzing and Predicting the Impact of COVID-19 on Online Pharmaceuticals Sectors and Pathological Services in India

The novel coronavirus disease 2019 (COVID-19) is a worldwide pandemic that has impacted the healthcare industry. While retail pharmaceutical and medical stores were classified as vital services, online pharmacies emerged as one of the most significant benefactors of India’s pandemic-induced lockdown, as consumers decided to buy medications online to lessen the risk of contamination. The wide-ranging consequences of the COVID-19 outbreak have also prompted healthcare providers to embrace new methods and techniques that allow for at-home lab testing. We conducted a research study and analysis to determine if online pharmacies and at-home lab tests would continue to develop rapidly and become increasingly digital post-COVID pandemic considering that it has been fuelled by the current pandemic crisis and changing customer needs and requirements. The research was carried out considering the factors such as "Gender, Age-group, how often they use online pharmacies, rating their online pharmacy experiences, how often they purchased medicines online before and during COVID, have they used at-home lab testing services, and rating the quality, discounts, safety, genuineness and time taken for at-home lab testing". Based on these factors, in this paper, we have built two prediction models that use supervised learning which would predict how likely a person will utilize online pharmacies and at-home lab tests post COVID 19 pandemic. It consists of 5 probabilities i.e. very likely to use it, likely use it, moderately use it, less likely to use it, and rarely use it. The various classification algorithms employed in the study were logistic regression, decision trees, random forest, support vector machines, and Gradient booster classification algorithm The model’s results are based on the previous data that was collected by asking around 250 individuals. Accuracies obtained for predicting how likely a person will use online pharmacies post COVID are as follows, by gradient booster model the accuracy is ~85%, the accuracy obtained by decision tree classifier is ~71%, by logistic regression model the accuracy is ~85% and accuracy obtained by random forest model is ~78%. Accuracies obtained for predicting how likely a person will use at-home lab tests post COVID are as follows, by decision tree classifier accuracy is ~78%, the accuracy obtained by logistic regression is ~50% and the accuracy obtained by support vector machines is ~64%.