RESUMO
Background. Only a proportion of screened potential participants were actually randomized while conducting a phase 1 study of a humanized rabies monoclonal antibody. We aimed to assess the challenges in defining who is a normal volunteer and the issues that affect volunteer recruitment and thus accrual. Methods. One hundred and fifty-six volunteers were screened and 74 (47.4%) were randomized in a phase 1 study. Data on all participants screened for the study were analysed and reasons for their non-randomization were classified. Results. The reasons for volunteers not being randomized were: (i) deranged laboratory parameters (n=62); (ii) nonlaboratory causes (n=4); and (iii) withdrawal of consent (n=16). A large proportion of screen failures were due to low haemoglobin levels, which led to the protocol being amended midway during the study. An informal interview of those who declined consent showed that they had only wanted to get themselves investigated thoroughly or were interested in getting their HIV status evaluated. Conclusions. Our study shows that <50% participants screened for a phase 1 study in a developing country actually get randomized. The main reason for non-randomization is abnormal laboratory tests. This may help investigators and sponsors to plan protocols better, define normal ranges with acceptable variations based on their own populations a priori and have more pragmatic accrual targets.