A systematic overview of prospective cohort studies of cardiovascular disease in sub-Saharan Africa: reply to Bovet et al., and Gao et al.

AFRICA 469 letter to the Editor a systematic overview of prospective cohort studies of cardiovascular disease in sub-Saharan africa: reply to Bovet et al., and Gao et al. dear Sir Two groups of investigators have recently provided evidence supporting the need for elaborated longitudinal studies to inform successful health service and policy solutions to the growing problem of chronic and cardiovascular disease in sub-Saharan Africa (SSA). In one of those studies, published in the Cardiovascular Journal of Africa, 1 our group reached such a conclusion on the basis of a systematic review of relevant existing cohort studies conducted in SSA, published and indexed 1 The feedback received from colleagues both from Africa and beyond testifies to the interest and also the expectations of the scientific community at large for longitudinal studies on chronic diseases in SSA. We are particularly grateful to Drs Bovet and Shamlaye, 4 and Drs Gao and Yuan, 5 who through two letters published in the Cardiovascular Journal of Africa, have made a significant contribution to the debate. Drs Bovet and Shamlaye 4 provided evidence suggesting that our review missed some relevant studies fulfilling our entry criteria and published in leading medical journals. They further suggested that we omitted some SSA countries from our search. We did acknowledge in the limitations sections of our article that for a number of reasons, there was still a possibility that our search did not capture all relevant studies. Therefore, we welcome the contribution of Bovet and Shamlaye and call for an ongoing register of African cohort studies, possibly in the columns of the Cardiovascular Journal of Africa along the lines of the cohort profiles in the International Journal of Epidemiology. 6 However, of the eight studies listed by the two colleagues, at least four do not fulfil the eligibility criteria of our review, including a study from Mauritius published one year after the completion of our review, 7 a study with a follow-up duration shorter than six months, 8 one in which none of the predictors of interest was assessed at baseline, 9 and one cross-sectional study with no follow-up component. 10 It would have been more appropriate to repeat the systematic search using our strategy, or any other judged appropriate by the authors, and quantify the gap, if any, between our study and what should have been optimal. Furthermore, unlike the authors' suggestion, we made no restriction by country …