Modeling Covid-19 dynamics for real-time estimates and projections: an application to Albanian data

Erida Gjini

The copyright holder for this preprint . https://doi.org/10.1101/2020.03.20.20038141 doi: medRxiv preprint true symptomatic cases, then the effective symptomatic proportion the model would 'see' would be lower, and slightly different model parameters would result. An R 0 so high, of the order of 5, estimated by the early dynamics of confirmed cases in Albania, may be related to superspreaders having dominated the early transmission events in the epidemic in this country, or related to big public gatherings and festivities at national level (political meetings, International Women's day, etc.) just preceding the first confirmed case. Other studies have also shown that mean R 0 ranges from 2.24 (95%CI: 1.96-2.55) to 3.58 (95%CI: 2.89-4.39) (15) , and is very sensitive to the reporting rate. If the reporting rate is increased by 8-fold, the estimate is expected to approach the more widely agreed values in the literature (2; 3) . So far the model does not take into account reporting rate, but it is likely that Covid-19 cases in Albania may be under-reported, due to the low number of overall tests performed, compared to the total population size.