Improving local prevalence estimates of SARS-CoV-2 infections using a causal debiasing framework

Area of Science:

• Epidemiology
• Biostatistics
• Public Health

Background:

• Current SARS-CoV-2 surveillance relies on symptomatic testing, leading to biased prevalence estimates.
• Targeted testing schemes are unrepresentative of the general population, affecting public health policy decisions.
• Accurate real-time monitoring of infectious disease spread is crucial for effective pandemic response.

Purpose of the Study:

• To develop a causal framework for debiased, fine-scale spatiotemporal estimation of SARS-CoV-2 prevalence and transmission.
• To integrate data from targeted testing with a randomized surveillance study (REACT) for improved accuracy.
• To provide more reliable estimates of the effective reproduction number (R_t) to guide public health interventions.

Main Methods:

• Developed a probabilistic model incorporating a bias parameter to adjust for differential testing probabilities.
• Combined observed test counts from targeted schemes with data from the UK's REACT randomized surveillance study.
• Validated the causal framework on held-out data over a 7-month period.

Main Results:

• The framework successfully produced debiased, fine-scale spatiotemporal estimates of SARS-CoV-2 prevalence and R_t.
• Local R_t estimates accurately predicted 1- and 2-week ahead changes in case numbers.
• Observed increases in prevalence and R_t correlated with the spread of Alpha and Delta variants.

Conclusions:

• Randomized surveys significantly enhance targeted testing for more accurate infectious disease monitoring.
• The developed causal framework offers improved statistical accuracy for tracking SARS-CoV-2.
• This approach is valuable for monitoring emerging and ongoing infectious disease outbreaks.