Combinatorial (bio-geo-temporal) and non-combinatorial analysis of the Covid-19 dissemination that affected Georgia (the country) in 2021
Bibliographic Data
Introduction To extract more information from the same data and support decision-making, we explored whether bio-geo-temporal (BGT) variables could distinguish municipalities where BGT variables differed and (if prioritized in subsequent interventions) could lead to cost-effective decision-making. This study was conducted investigating COVID-19 outbreaks reported in Georgia in 2021. Methods Using both commercial and non-commercial proprietary software packages, we explored the test positivity rate or TP % (percentage of test-positive results among tested individuals), the date and the georeferenced location of municipalities where tests were conducted, as well as the associated population and road density (municipal road length / municipal area). Analyses included spatial (Getis-Ord and Moran’s I) and non-spatial statistical tests. Results While the TP% was not linearly related with any one variable, combinations of BGT variables displayed distinct data patterns that separated two groups of municipalities (one composed of just two municipalities). The two-municipality group showed a statistically significantly greater median road density than the remaining municipalities. The same two municipalities also possessed a greater ability to detect asymptomatic cases: they expressed 20.3 times larger TP%/km 2 but occupied a territory 18.3 times smaller than that of other municipalities. Combinatorial analyses that included testing emphasis (tests conducted per thousand inhabitants) provided novel ways to monitor and/or adjust testing policy at local levels. In contrast, analysis of the same adjusted TP% with three spatial statistical approaches identified much larger territories to be intervened. Discussion Findings supported the view that methods exploring dynamic combinations of BGT relationships may identify highly connected municipalities (those likely to behave as network nodes during disease dissemination processes) when tested as TP%/km 2 . If selectively intervened, such nodes may lead to earlier cessation of epidemics. To explore the reproducibility of this proof-of-concept, further studies are recommended
Geographic information system · Georeference · Outbreak · Population · Population density · Spatial analysis · Statistical analysis · COVID-19 epidemiological studies · Data-Driven Disease Surveillance · Zoonotic diseases and public health
| Highly cited | No |
|---|