OBJECTIVES: Studies of malaria vaccines and chemoprevention often classify a person who has suffered a malaria disease episode and received curative drug treatment as non-susceptible to new malaria infection for two to four weeks, depending on the pharmacokinetics of the curative drugs used. This duration is usually subtracted from the person's follow-up time in the calculation of malaria disease incidence and the effectiveness of the interventions in terms of incidence rate ratio. We aimed to examine the implications of this common practice. METHODS: We examined the concepts of disease incidence and protective effect under different research purposes. We mathematically assess the relationship of the estimates of incidence rate and incidence rate ratio between assessments with and without the subtraction. We used published studies to illustrate. RESULTS: Conceptually, evaluation of disease prevention without the subtraction is analogous to intention-to-treat analysis in clinical trials, whereas the common practice of subtracting is analogous to per-protocol analysis. From a real world research point of view, the common practice leads to over-estimation of disease burden and intervention effectiveness. It does not indicate what will happen in real world when a preventive intervention is implemented. Case studies demonstrate that the over-estimation can be practically significant. CONCLUSIONS: To subtract or not subtract a post-disease duration from follow-up time answer different practical questions, which need be clearly formulated. From a health policy and planning perspective, it is likely more appropriate not to subtract.