Association between RT-PCR Ct values and COVID-19 new daily cases: a multicenter cross-sectional study.

ABSTRACT

Proactive prediction of the epidemiologic dynamics of viral diseases and outbreaks of the type of COVID-19 has remained a difficult pursuit for scientists, public health researchers, and policymakers. It is unclear whether RT-PCR Cycle Threshold (Ct) values of COVID-19 - or any other virus - as indicator of viral load, could represent a possible predictor for underlying epidemiologic changes on a population level. The study objective is thus to investigate whether population-wide changes in SARS-CoV-2 RT-PCR Ct values over time are associated with the daily fraction of positive COVID-19 tests. In addition, this study analyses the factors that could influence RT-PCR Ct values. A retrospective cross-sectional study was conducted on 63,879 patients from May 4, 2020 to September 30, 2020, in all COVID-19 facilities in the Kingdom of Bahrain. Data collected included number of tests and newly diagnosed cases, as well as Ct values, age, sex nationality, and symptomatic status. Ct values were found to be negatively and very weakly correlated with the fraction of daily positive tests in the population r = -0.06 (CI 95% -0.06; -0.05; p=0.001). The R-squared for the regression model (adjusting for age and number of daily tests) showed an accuracy of 45.3%. Ct Values showed an association with nationality (p=0.012). After the stratification, the association between Ct values and the fraction of daily positive cases was only maintained for the female sex and Bahraini-nationality. Symptomatic presentation was significantly associated with lower Ct values (higher viral loads). Ct values do not show any correlation with age (p=0.333) or sex (p=0.522). We report one of the first and largest studies to investigate the epidemiologic associations of Ct values with COVID-19. Although changes in Ct values showed a moderate association with daily cases, our results indicate that it may not be as predictive within a simple model. More population studies and models from global cohorts are necessary.