Estimates of COVID-19 hospitalization and ICU burden from existing severe acute respiratory infections surveillance

By Shanet Susan AlexReviewed by Danielle Ellis, B.Sc.Feb 15 2022

In a recent study posted to the medRxiv* preprint server, researchers estimated the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) hospitalization and intensive care unit (ICU) treatment during the first three waves of the coronavirus disease 2019 (COVID-19) pandemic from the hospital sentinel based on the 10th revision of International Classification of Diseases (ICD) for severe acute respiratory infections (ICOSARI) in Germany. Further, the team compared these estimations with COVID-19 notification data.

Following the emergence of COVID-19, nations worldwide adopted existing surveillance systems such as syndromic surveillance systems and mandatory notification systems to establish SARS-CoV-2 surveillance. Further, in the light of the repeated emergence of mutated SARS-CoV-2 variants and waning natural infection- and vaccine-induced immunity, the ongoing severe SARS-CoV-2 cases surveillance is vital to evaluate disease seriousness. However, the actual burden of COVID-19 hospitalization and ICU treatment is difficult to assess since accurate data on hospitalized cases is not available easily or easy to analyze.

*Important notice: medRxiv publishes preliminary scientific reports that are not peer-reviewed and, therefore, should not be regarded as conclusive, guide clinical practice/health-related behavior, or treated as established information.

In Germany, the mandatory COVID-19 reporting system obtains a broad range of epidemiological data from laboratory-confirmed COVID-19 cases, including data about COVID-19-related hospital admission, ICU, and mortality. Since continuous monitoring by the local public health authorities is required to obtain data on COVID-19-associated hospitalization, these data are mostly incomplete. Hence, there is an immediate requirement for comprehensive data on hospitalized SARS-CoV-2 patients to validate COVID-19 notification data on severe cases and provide a robust estimate of SARS-CoV-2 hospitalization and ICU burden. 

About the study

In the present study, the researchers estimated the weekly SARI incidence using data from ICOSARI and validated the same retrospectively against full population data. The team evaluated the age-specific incidence of COVID-19 hospitalization and ICU requirement during the initial three waves of the SARS-CoV-2 pandemic in Germany based on the data of SARI cases. Further, the researchers compared the hospitalization and ICU burden of COVID-19 from notification data with the estimations derived from sentinel data and demonstrated potential underreporting of COVID-19 hospitalization in the notification data.

Findings

The results indicate that the estimations of SARI incidence from hospital sentinel data were in line with the full population data regarding timing and levels of peaks and trends. Although the estimated COVID-SARI incidence corresponded very well with the notified incidence of COVID-19 hospitalization in the first COVID-19 pandemic wave, it was far higher during the second and third pandemic waves. 

The fraction of unspecified hospitalization status of SARS-CoV-2 cases within the notified COVID-19 cases was a lot higher in the second and third waves than the initial COVID-19 wave. The median proportion of SARS-CoV-2 patients with unknown hospitalization status was 12%, 23%, and 18% during the initial, second, and third COVID-19 waves, respectively, in the notification data. 

The estimated ICU incidence in the COVID-SARI was substantially higher than the results derived from notification data in all three SARS-CoV-2 pandemic waves. Similarly, the proportion of ICU admissions of hospitalized COVID-19 cases demonstrated a significant difference between the hospital sentinel and notification data. ICU admissions among SARS-CoV-2 hospitalizations was 40%, 34%, and 35% for the COVID-SARI cases in the sentinel, and 15%, 7.4%, and 11% in the notification data during the first, second, and third COVID-19 pandemic waves, respectively.

In the hospital sentinel and notification data, the incidence of COVID-19 hospitalization and ICU admission was very low for those below 15 years in all three waves of the SARS-CoV-2 pandemic. For the 35 years or older age groups, the COVID-19 hospitalization and ICU admission incidence from the two data sources matched very well in the first COVID-19 wave and less in the second and third waves. In both data sources, hospitalization and ICU incidence of the age group 35 to 59 years were lower in the second COVID-19 wave than in the third wave.

According to both data sources, the 80 years or older age groups had the highest incidences of hospital admission and the ICU requirements in the first and second COVID-19 waves. Further, while 80 years or older age groups had a slightly higher incidence of COVID-19 hospitalization than the 60-70 years age groups, the latter age groups had a higher ICU admission incidence than the 80 years or older age groups in the third SARS-CoV-2 wave. 

Conclusions

The study findings show that using data from the syndromic surveillance system, ICOSARI provides valid and crucial information for evaluating SARS-CoV-2 hospitalization and ICU burden, particularly in periods with high COVID-19 cases, where notification data might be incomplete.

The study suggests the significance of SARI surveillance to estimate disease severity, inform the introduction or relaxation of restrictions, and monitor the transmission of diseases. The method of assessing the incidence of hospitalization and ICU admission from the sentinel data is significant in evaluating the transition of a pandemic to an epidemic. Moreover, it can be generalized in evaluating other diseases such as the respiratory syncytial virus (RSV) and seasonal influenza infections. Further, using an existing system has several advantages like known potential biases, stability and catchment population, and baseline information from previous years.

*Important notice: medRxiv publishes preliminary scientific reports that are not peer-reviewed and, therefore, should not be regarded as conclusive, guide clinical practice/health-related behavior, or treated as established information.