Workflow from Scientific Research

Reliable assessments of NPI effectiveness require consistent outcome data with meta-information, and carefully collected raw intervention data allowing application of different coding. Data should be analysed considering both variation between populations and over time. By use of mechanistic modelling, the effectiveness of NPIs should be related to relative changes in transmission. Modelling approaches should be tested and compared with rigorous model validation and sensitivity analyses. Effectiveness should be assessed for individual NPIs, accounting for uncertainty and variation between populations. In the figure, sources of uncertainty are marked with dashed outline. NPI=non-pharmaceutical intervention. +=and.

DOI