Positivity rate: an indicator for the spread of COVID-19

Objective This paper investigates the relationship between the positivity rate and numbers of deaths and intensive care unit (ICU) patients. In addition, it explores the use of the positivity rate as an indicator for the spread of COVID-19. Methods We used COVID-19 datasets for eight countries - Canada, USA, UK, Italy, Belgium, Ireland, Colombia and South Africa - and considered two correlation cases. The first case considers the correlation of the number of confirmed cases with each of deaths and ICU patients. The second case considers the correlation of the positivity rate with each of deaths and ICU patients. When obtaining the correlation, we considered different lagging periods between the date of confirming a case and the date of its ICU admittance or death. We compared the obtained correlation coefficient values for each of the two considered cases to explore whether the positivity rate is a better indicator for the spread of the disease than confirmed cases. For each of the eight considered countries, we obtained the daily reproduction number using each of confirmed cases and positivity rate. The two obtained sets of reproduction number values for each country were statistically compared to investigate whether they are significantly different. Results When considering the daily positivity rate instead of the daily number of confirmed cases, the maximum correlation with deaths is increased by 349.9% for the USA (the country with the highest increase) and 4.5% for the UK (the country with the lowest increase), with an average increase of 60.8% considering the eight countries. Considering the daily positivity rate instead of the daily number of confirmed cases caused the maximum correlation with the number of ICU patients to be increased by 74.7% for the USA (the country with the highest increase) and 2.2% for the UK (the country with the lowest increase), with an average increase of 25% over the considered countries. The results for the daily reproduction number obtained using the positivity rate are statistically different from those obtained using daily confirmed cases. Conclusion The results indicate that positivity rate is a better indicator for the spread of the disease than the number of confirmed cases. Therefore, it is highly advised to use measures based on the positivity rate when indicating the spread of the disease and considering responses accordingly because these measures consider the daily number of tests and confirmed cases.