Performance of the COVID-19 surveillance system in Honduras 2020: Benford's law analysis

Abstract

COVID-19 is an infectious disease responsible to date for 6, 369, 703 deaths worldwide. To control it, it is essential to adapt and strengthen national health systems and expand surveillance capacities, with the daily report of new cases and deaths from COVID-19 being one of the main tools for understanding and responding to the pandemic.. The work focused on identifying inconsistencies in the data reported on confirmed cases and deaths of COVID-19, by the Honduran health system during 2020. Official data published between March 11 and December 31 were analyzed. 2020, by the Johns Hopkins University coronavirus resource center, the data was divided into four analysis segments, the first corresponded to the first 50 days of the pandemic in the country, the second included data up to June 24, the third until December 1 and the fourth included all the data for 2020. For the analysis, Benford's law was used, a widely used statistical phenomenon that allows detecting anomalous data in the surveillance systems of each country. As of December 31, the Honduran health system confirmed 121, 827 positive cases and 3, 130 deaths from COVID-19 in Honduras. Regarding the detection of anomalous data, the second analysis segment obtained a better result; however, it was observed that the accumulated analysis corresponding to 2020 did not adjust to Benford's Law. In general, surveillance improved performance after the first 50 days of the pandemic; however, the data for 2020 reveal a non-conformity with Benford's law, which is interpreted as a notification performance unsuitable.