Excess Hospital Death During the First Year of the COVID-19 Pandemic, Bishkek, Kyrgyzstan, 2020

Health Information Systems (Including Surveillance)

Background: COVID-19 was first detected in Kyrgyz Republic in March 2020. The largest COVID-19 surge occurred in June–July when reported COVID-19 deaths also peaked. Reported deaths do not capture all deaths resulting from the COVID-19 pandemic. Estimating excess deaths (ED) by hospital helps identify most affected facilities and periods during pandemic to improve healthcare management.
Methods: To assess changes in hospital mortality, we used monthly death data from 20 hospitals in Bishkek. We estimated ED, calculated in-hospital 2020 death rate per 10,000 admission and compared them to the expected level (EL) of deaths using the upper bound of the 95% confidence interval of historical average over the 5-years (2015–2019) preceding the pandemic (threshold). ED percent is expressed as ED count divided by expected deaths threshold.
Results: Of 2,593 registered deaths (in-hospital death rate was 149 per 10,000), the estimated number of ED was 1,015; 73% greater than expected. Elevated mortality was observed in 17 (85%) of the 20 hospitals with ED ranging from 5 to 1,026% above expected. Of the 20 hospitals, 14 (70%) were repurposed for COVID-19 patient intake because of insufficient hospital beds. In March–May, estimated ED was 172 (64% above EL), in June–August (peak period of COVID-19 transmission) ED increased to 554 (177% above EL) and in September–December ED decreased to 262 (42% above EL).
Conclusions: Our results confirm that hospitals in Bishkek experienced greater than expected mortality during the COVID-19 pandemic. Ongoing surveillance is recommended at hospitals to support rapid response when surges in patient load and increased deaths are observed. Greater community mitigation and increased uptake of vaccines, especially among persons at greater risk of hospitalization and death, would reduce the burden on healthcare systems during surges.

Please email us if you have any corrections.

If this abstract has been converted into a full article, please email us the link. We would love to help promote your work.