Repetitive events data analysis is quite common in biomedicine. The literature review indicates that most statistical models used for such data are often based on time to the first event or consider events within subject as independent. Also, most applied analysis data taking into account the non-independence of repeated events within subjects are done with continuous risk interval models which may not be relevant for infectious disease such as malaria.

This work aimed to analyze repeated malaria episodes with continuous and discontinuous risk interval models in order to identify the best model estimating malaria risk respective to covariates, in endemic areas.

Discontinuous and continuous intervals models gave similar RRs estimated in all models. This may be due to the short washout period in relative to the long follow-up period of the study. Still, the discontinuous intervals models should be preferred because it is more appropriate in the epidemiological point of view to take into account the time a subject is not at risk to a disease in a given period of time.

Anderson-Gill, PWP-CP and Frailty models estimated RRs with higher magnitude compared to GEE Poisson family model which showed also an overdispersion sign and