Investigation into drug dose practices and proportional reasoning in mathematics among medical students.

Abstract

Multilingual preclinical South African medical students, receiving a developing quantitative literacy program, were studied. Their preparedness for drug dosage practices, aspects they found difficult, the underlying reasons, and factors or interventio ns that helped them were investigated A mixed methods action -research-type study involved three cycles, each testing a different consenting student cohort. In Cycle 1 dosage calculation ability, when using different concentration formats, was assessed progressively over two years. Individual teaching observations explored struggling students' experience. In Cycle II, the focus broadened to include retrieving dosage information from guidelines and preparing a dose in a syringe. In Cycle ill, assessment in eluded drug selection, requiring interpretation of statistics to compare treatments. In all cycles, the percentage of successful students was calculated, error frequencies were determined, and associations were sought. Observations were coded for themes of student difficulties, supportive strategies and improvement, responses were mapped to stage progress toward proportional reasoning and assessments were analysed for linguist ic and mathematical difficulty. In Cycle I (n = 364), competence rose from 23% to 66% after extended exposure to training and assessment, peer learning and individual tuition. Observed tuition themes produced curriculum change. In Cycle II, 239 students were randomised to four groups. The two groups given standard numerical information fared best (46% success with the addition of equipment and 43% without). Where information was extracted from guidelines, 25% were successful when provided with equipment and only 10% without. Groups with equipment calculated fewer implausible doses and u sed fewer incorrect units. In Cycle ill, after training, only 26% of 231 students understood relative risk, but a mere 6% understood the concepts of absolute risk and number needed to treat. Sixty students (26%) made a rational drug choice. Proportional re asoning ability was not associated with interpreting risk statistics successfully or making a rational treatment selection. In all cycles, English as home language predicted success, although in Cycle I, this effect was lost after peer tuition. To prepare medical students for the complexities of drug dose practices, early training and repeated assessment should include extraction of information from guidelines, use of the relevant equipment as well as training in interpreting risk statistics for rational drug selection.