Evaluating the impact of local tracing partnerships on the performance\n of contact tracing for COVID-19 in England

Le résumé fourni par la source

Assessing the impact of an intervention using time-series observational data\non multiple units and outcomes is a frequent problem in many fields of\nscientific research. In this paper, we present a novel method to estimate\nintervention effects in such a setting by generalising existing approaches\nbased on the factor analysis model and developing a Bayesian algorithm for\ninference. Our method is one of the few that can simultaneously: deal with\noutcomes of mixed type (continuous, binomial, count); increase efficiency in\nthe estimates of the causal effects by jointly modelling multiple outcomes\naffected by the intervention; easily provide uncertainty quantification for all\ncausal estimands of interest. We use the proposed approach to evaluate the\nimpact that local tracing partnerships (LTP) had on the effectiveness of\nEngland's Test and Trace (TT) programme for COVID-19. Our analyses suggest\nthat, overall, LTPs had a small positive impact on TT. However, there is\nconsiderable heterogeneity in the estimates of the causal effects over units\nand time.\n

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.