Core biostatistical competencies and attitudes towards biostatistics among health researchers: a Delphi study

The importance of biostatistics is well established, as evidenced by its integration in health-related training programs. However, consensus on core biostatistical competencies required across research profiles remains lacking, and attitudes towards biostatistics insufficiently explored. This study aimed to establish expert consensus on essential biostatistical knowledge and competencies for health researchers and identify key attitudes towards biostatistics. A Delphi study involved a multidisciplinary panel of 25 experts affiliated or collaborating with French-Speaking Belgian universities, including biostatisticians, health researchers, and psychometricians. A study-specific questionnaire was developed from literature and existing questionnaires encompassing biostatistical knowledge, competencies, and attitudes towards biostatistics. Across three rounds, panellists rated each item on a 7-point Likert scale and provided qualitative feedback. Consensus was defined as median score ≥ 6 and an interquartile range ≤ 1. Items were iteratively refined, added, or removed between rounds. Response stability between the first two rounds was assessed using the Wilcoxon signed-rank test. Strong consensus was achieved for competencies related to fundamental biostatistical knowledge, interpretation, communication, and critical appraisal with a total of 64 out 71 items (90.1%) reaching consensus across three rounds. In contrast, items concerning advanced statistical methods and regression models reached lower agreement, likely reflecting heterogeneity across researcher profiles. Consensus on attitudes towards biostatistics was harder to achieve, with only 26 of 70 items (37.1%) reaching consensus. While perceived value and interest were more consistently endorsed, items related to anxiety and discouragement showed greater variability. Differences between expert groups were observed, particularly regarding the expected degree of autonomy in biostatistical practice and perceptions of statistical complexity. This study proposes a consensus framework defining the fundamental biostatistical competencies required of health researchers in a French-speaking academic context in Belgium. This context-specific framework could support the development of assessment tools, training programs, curricula, and lifelong learning strategies. The findings further suggest that attitudes, although less consistently characterized, may influence how knowledge is engaged with in practice. Future research could evaluate researchers according to these competencies and attitudes to identify training needs and guide targeted educational interventions. Further validation is warranted before extending its application to other settings.