Diagnostic accuracy and predictive value of clinical symptoms for the diagnosis of mild COVID-19

Abstract

Objective To assess the diagnostic accuracy of clinical symptoms and their predictive values in patients with suspected mild COVID 19 and to identify target groups for self isolation and outpatient treatment without additional testing in the primary health care system. Methods We conducted an open label prospective study in both male and female patients aged 18 to 72 years with suspected mild COVID 19 who were sequentially enrolled in the study. The clinical diagnosis was performed in accordance with the WHO recommendations based on the acute onset of such symptoms as olfactory dysfunction, hyperthermia, myalgia, nasal congestion, nasal discharge, cough, rhinolalia, sore throat, without pneumonia or hypoxia in persons in contact with a confirmed case of COVID 19. The physician assessed clinical symptoms using a 4-point scale. The patient self-assessed clinical symptoms using a ten point visual analogue scale (VAS). All enrolled patients underwent laboratory testing to confirm the diagnosis of COVID 19. Results Of the 120 patients underwent testing, the diagnosis of mild COVID 19 was confirmed in 96 patients and ruled out in 24 patients. When assessing symptoms by a physician according to the correlation analysis, hyperthermia, myalgia, nasal congestion and rhinolalia have a positive predictive value with a significance level of more than 0.6. When self-assessing symptoms by a patient, fever, myalgia and nasal congestion have a diagnostic accuracy with a significance level of more than 0.5. Nasal discharge, cough and sore throat have negative predictive values. Discussion The presence of these symptoms in patients with an acute onset of the disease can help to make a clinical diagnosis of coronavirus disease and identify target groups for self-isolation and outpatient treatment without additional testing. Highly suspect asymptomatic patients are not considered as those who have possible mild COVID 19 infection. Keywords: diagnostic accuracy, predictive values, COVID 19, symptom