OBJECTIVES: Testing is widely accepted as critical to fighting the COVID-19 pandemic. To understand the dynamics of tests conducted could help to assess a country response to the pandemic. In that sense, our aim was to verify if there is a relationship between pandemic dynamics and molecular tests conducted in Colombia.

METHODS: We retrieved publicly available data from The Colombian National Institute of Health from March 2 to December 31, 2020, on testing and outcomes related to COVID-19. Pearson correlation coefficients were calculated between molecular tests conducted per 100000 persons and death or incidence rate per 100000 people of every department of Colombia. To assess if there a cyclic relationship between the daily number of molecular tests and daily COVID-19 cases in Colombia, we executed a cointegration analysis and evaluated the hypothesis with an augmented Dickey-Fuller (ADF) test. A critical value of -3.42 for rejecting the null hypothesis at 5% was used.

RESULTS: There was a positive correlation between molecular tests conducted and COVID-19 incidence and death rate (r = 0.79, p < 0.01 and r = 0.64, p < 0.01, respectively). The cointegration (ADF) test revealed a statistically significant and closely time-dependent stochastic structure between daily COVID-19 cases and number of molecular tests (ADF, -3.50; p < 0.01).

CONCLUSIONS: In Colombia, the molecular tests conducted are reactive to COVID-19 incidence and, in contrast to other scenarios, the molecular testing increase does not reduce COVID-19 incidence or mortality.