A multiple logistic model for prediction of urinary tract infections in an urban community: A public health perspective

Abstract

Urinary tract infection (UTI) is one of the most common infectious diseases globally. A lot of clinical research has been done on UTI patients, but a questionnaire-based study on UTI is scarce. A cross-sectional study was conducted on outpatients with a high suspicion of uncomplicated UTI in Hayes Memorial Mission Hospital at Prayagraj (Eastern part of Northern India) to find out the frequency of symptoms and predisposing factors and their relationship towards the prediction of UTI. Logistic regression analysis showed a significant association between UTI and some of the variables. Also, the factors responsible for the occurrence of UTI are “gender”, “how many times you urinate from morning till night”, “a sudden desire to urinate, which is difficult to hold”, “weakness of urinary stream”, “splitting or spraying of the urinary stream” and “fever”. A statistical model (multiple logistic model) has been also established for the prediction of UTIs with an accuracy of 82.2%. It is also observed that the prevalence rate (odds ratio) of UTI in females is 2.38 times that of males. The study created a screening questionnaire for patients suspected of having UTI. A multiple logistic model has been established for the prediction of UTI which can be instrumental for clinicians from a public health perspective in the management of Urinary Tract Infections in this era of escalating AMR.