Cost-effectiveness of wastewater-based environmental surveillance for SARS-CoV-2 in Blantyre, Malawi and Kathmandu, Nepal: A model-based study

Abstract

Wastewater-based environmental surveillance (ES) has been demonstrated to provide an early warning signal to predict variant-driven waves of pathogens such as severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Our study evaluated the potential cost-effectiveness of ES for SARS-CoV-2 compared with clinical testing alone. We used the Covasim agent-based model of COVID-19 to simulate disease transmission for hypothetical populations in Blantyre, Malawi, and Kathmandu, Nepal. We simulated the introduction of a new immune-escaping variant over 6 months and estimated health outcomes (cases, deaths, and disability-adjusted life years [DALYs]) and economic impact when using ES to trigger a moderate proactive behavioral intervention (e.g., increased use of masks, social distancing) by policymakers versus no ES and hence a delayed reactive intervention. Costs considered included for ES, clinical testing, treatment, and productivity loss for the entire population due to implementation of the behavioral intervention. We calculated the incremental cost-effectiveness ratios and compared these with local willingness-to-pay thresholds: $61 for Malawi and $249 for Nepal. We performed sensitivity analyses to evaluate the impact of key assumptions on the results. Costs are reported in 2022 US dollars. We estimate that if ES were implemented, approximately 600 DALYs would be averted in Blantyre and approximately 300 DALYs averted in Kathmandu, over the six-month period. Considering health system costs, ES was cost-effective in Blantyre and cost-saving in Kathmandu. Cost-effectiveness of ES was highest in settings with low clinical surveillance, high disease severity, and high intervention effectiveness. However, from the societal perspective, ES may not be cost-effective depending on the magnitude of population-wide productivity losses associated with the proactive behavioral intervention and the cost-effectiveness threshold. SARS-CoV-2 ES has the potential to be a cost-saving or cost-effective tool from the health system perspective when linked to an effective public health response. From the societal perspective, however, the length of the behavioral intervention and its consequences for productivity losses of the entire population may make ES not cost-effective. Implementing ES for multiple pathogens may improve its cost-effectiveness.

Fig 1. Examples of simulation time series of the default scenario with ES (green) and without ES (purple), for 10 different random seeds.

Note that Blantyre and Kathmandu have different y-axis scales.

Fig 2. Relationship between the number of additional days in the simulation that the intervention is active when ES is used (x-axis) and the number of DALYs averted (y-axis).

For example, a simulation for Malawi would lie on the line of best fit if the intervention started 10 days earlier due to ES, and this simulation had 175 fewer DALYs compared with the same simulation without ES. Note that Blantyre and Kathmandu have different axis scales. Abbreviations: DALY, disability-adjusted life year; ES, environmental surveillance.

Fig 3. Total DALYs incurred due to COVID (left), and percentage of DALYs averted by ES (right).

Abbreviations: DALY, disability-adjusted life year; ES, environmental surveillance; YLD, years of life with disability; YLL, years of life lost.

Fig 4. Estimated costs due to COVID, the surveillance programs, and productivity loss.

Note the factor of 10 difference in scales between Blantyre and Kathmandu, which is largely due to Kathmandu’s roughly three-times-larger population size and roughly three-times-higher per capita GDP than Blantyre. Note: These data are also listed in Table 3. Abbreviation: ES, environmental surveillance.

Fig 5. Sensitivity analysis of model parameters.

Bars show standard errors of the mean. Values for the base case parameters values are listed in Table 1. Note that Blantyre and Kathmandu have different x-axis scales. Abbreviations: DALY, disability-adjusted life year; ES, environmental surveillance.