A unified framework for diagnostic test development and evaluation during outbreaks of emerging infections

Abstract

Evaluating diagnostic test accuracy during epidemics is difficult due to an urgent need for test availability, changing disease prevalence and pathogen characteristics, and constantly evolving testing aims and applications. Based on lessons learned during the SARS-CoV-2 pandemic, we introduce a framework for rapid diagnostic test development, evaluation, and validation during outbreaks of emerging infections. The framework is based on the feedback loop between test accuracy evaluation, modelling studies for public health decision-making, and impact of public health interventions. We suggest that building on this feedback loop can help future diagnostic test evaluation platforms better address the requirements of both patient care and public health.

Fig. 1. Model-based demonstration of effect of delays in implementing public health interventions on number of individuals needing treatment during an epidemic.

Time course of the number of individuals requiring intensive care therapy if measures are taken at different time points to reduce the effective reproduction number below 1 (reproduction number (R) at the start set to R = 2, reduced to R = 0.9 after 7, 14, 21, 28 days). Assumptions of the Susceptible-Exposed-Infectious-Recovered (SEIR) model: pre-infection time: 3 days, infectious time: 7 days, proportion of individuals requiring intensive care unit out of all infectious patients: 2%, length of stay in intensive care unit: 15 days, population size: 80.000.000, number of susceptible individuals at the start: 79.950.000, number of exposed individuals at start: 40.000, number of infectious individuals at the start: 10.000, number of immune individuals at the start: 0, number of individuals requiring intensive care at the start: 0.

Fig. 2. Example model forecasts of number of individuals needing treatment during an epidemic, under varying assumptions about antibody test specificity during parameterisation.

Forecast for number of individuals requiring intensive care therapy if measures as described for Fig. 1 are taken after 28 days, with the model parametrised using different assumptions for antibody test specificity. SEIR model with Infectious compartment split into ‘Detected’ and ‘Undetected’ compartments. Parametrisation as described in Fig. 1, but with 2% of detected infectious cases requiring intensive care. Proportion of detection obtained from data from the ‘Heinsberg’ seroprevalence study, corrected for antibody test sensitivity of 0.9, and specificity of 0.9, 0.93, 0.96, 0.99, 1.

Fig. 3. Feedback loop at core of proposed framework.

Schematic representation of the feedback triangle (in green) between diagnostic test accuracy results, the parametrization of modelling studies and their results, the consequences for decision analysis, and the test strategy chosen based on these decisions.