Association between SARS-CoV-2 in wastewater and COVID-19 hospitalizations in three countries, 2022-2024.

Importance: As fewer jurisdictions report national COVID-19 hospitalization rates and testing of mild and asymptomatic cases is reduced, there is a need to better understand the relationship between COVID-19 hospitalizations and alternative measures of COVID-19 circulation, such as wastewater surveillance. Objective: We described the association between levels of SARS-CoV-2 in wastewater and COVID-19 hospitalization rates at the national level during and after the pandemic and explored whether wastewater virus level can predict COVID-19 hospitalization rates. Design setting and participants: Retrospective analysis of public health reports of national wastewater surveillance and COVID-19 hospitalizations from Denmark, the Netherlands, and the United States from 2022 to 2024. Main outcomes and measures: For each country, we calculated Pearson correlation coefficients and hospitalization multipliers (defined as the hospitalization rate for a given scaled wastewater virus level), for the overall study period and by Omicron subvariant predominance. Additionally, we developed linear regression models using scaled wastewater virus levels to predict concurrent and projected (1-4 weeks) COVID-19 hospitalization rates. Results: There was a strong correlation between national SARS-CoV-2 wastewater virus levels and weekly new COVID-19 hospitalization rates per million (0.86 [95% CI, 0.82 to 0.90], 0.80 [95% CI, 0.72 to 0.85], and 0.89 [95% CI, 0.85 to 0.92] in Denmark, the Netherlands, and the United States, respectively). Correlations were consistently strong across all subvariant predominance periods in all three countries (range, 0.72 to 1.0). Results from linear regression models showed that hospitalization rate lagged wastewater SARS-CoV-2 levels by approximately 1 week. Linear regression models best predicted hospitalizations 1 week into the future (range of mean absolute percentage error, MAPE, 11.2 to 22.6%) with decreasing prediction accuracy within the range of 2-4 weeks (range of MAPE, 32.5 to 62.3% at 4 weeks). The median hospitalization multiplier (defined as ratio of weekly hospitalization rate to scaled wastewater SARS-CoV-2 level) were 859.3 (IQR, 621.7 to 1210.4), 178.3 (IQR, 133.7 to 243.7), and 245.9 (IQR, 184.0 to 293.7) for Denmark, the Netherlands, and the United States, respectively, during the study period. Regression models identified significant reduction in hospitalizations for a given wastewater virus level over time in all three countries. Conclusions and Relevance: SARS-CoV-2 wastewater virus levels were strongly correlated with COVID-19 hospitalization rates in the upcoming week. Wastewater to hospitalization ratios can be leveraged to enhance public health decision-making and resource allocation.