Stage | Recommendation | Rationale |
---|---|---|
Sample and survey | Conduct systematic sampling and oversample participants with major potential risk factors for long-term symptoms | Relate study population to a well-defined source population and increase statistical power to conduct hypothesis test |
Collect information on pre-COVID symptoms and conditions | Allow to distinguish long-term symptoms and pre-COVID symptoms or the population’s baseline level | |
Conduct multiple follow-up activities to examine changes in long-term symptoms over time | Establish temporality and compare the rate of symptom development between comparison groups | |
Design and analysis | Apply appropriate analytical methods, including confounding adjustment for major demographic and socioeconomic factors, pre-existing conditions, and comorbidities | Control for major factors related to long-term symptoms |
Use causal knowledge and graphs to guide covariate adjustments and provide a rationale for a priori selection of potential confounders | Reduce confounding and decrease the risk of including variables that could increase bias | |
Present information on number of cases and population at risk by COVID severity status and other important risk factors, and results of both crude and adjusted models | Allow for a close examination on main study results and the uncertainty that may result from small numbers |