Clinical diagnosis of COVID-19: a prompt, feasible, and sensitive diagnostic tool for COVID-19 based on a 1,757-patient cohort (The AndroCoV Clinical Scoring for COVID-19 diagnosis).

Abstract

Importance: In the COVID-19 pandemic, a limiting barrier for more successful approaches to COVID-19 is the lack of appropriate timing for its diagnosis, during the viral replication stage, when antiviral approaches could demonstrate efficacy, precluding progression to severe stages. Three major reasons that hamper the diagnosis earlier in the disease are the unspecific and mild symptoms in the first stage, the cost- and time-limitations of the rtPCR-SARS-CoV-2, and the insufficient sensitivity of this test as desired for screening purposes during the pandemic. More sensitive and earlier methods of COVID-19 detection should be considered as key for breakthrough changes in the disease course and response to specific therapeutic strategies. Our objective was to propose a clinical scoring for the diagnosis of COVID-19 (The AndroCoV Clinical Scoring for COVID-19 Diagnosis) that has been validated in a large population sample, aiming to encourage the management of patients with high pre-clinical likelihood of presenting COVID-19, at least during the pandemics, independent of a rtPCR-SARS-COV-2 test. Materials and methods: This is a compounded retrospective and prospective analysis of clinical data prospectively collected from the Pre-AndroCoV and AndroCov Trials that resulted in a clinical scoring for COVID-19 diagnosis based on likelihood of presenting COVID-19 according to the number of symptoms, presence of anosmia, and known positive household contact, in a variety of combinations of scoring criteria, aiming to the detect scorings that provided the highest pre-test probability and accuracy. Sensitivity, specificity, positive predictive value, negative predictive value, positive likelihood ratio, and accuracy were calculated for subjects screened in two different periods and altogether, for females, males, and both, in a total of nine different scenarios, for combinations between one, two, or three or more symptoms, or presence of anosmia in subjects without known positive household contacts, and no symptoms, one, two, or three or more symptoms, or presence of anosmia or ageusia in subjects with known positive household contacts. Results: 1,757 patients were screened for COVID-19. Among the multiple combinations, requiring two or more symptoms with or without anosmia or ageusia for subjects without known contact and one or more symptoms with or without anosmia or ageusia with known positive contacts presented the highest accuracy (80.4%), and higher pretest probability and accuracy than virtually all rtPCR-SARS-CoV-2 commercially available kit tests. Conclusion: The AndroCoV Clinical Scoring for COVID-19 Diagnosis was demonstrated to be a feasible, quick, inexpensive and sensitive diagnostic tool for clinical diagnosis of COVID-19. A clinical diagnosis of COVID-19 should avoid delays and missed diagnosis, and reduce costs, and should therefore be recommended as a first-line option for COVID-19 diagnosis for public health policies, at least while SARS-CoV-2 is the prevailing circulating virus.