Martine Kruijtbosch

129 De t e rmi nan t s t ha t i n f l uence mo ra l r eason i ng pe r spec t i ves Data analysis The determinants are gender, age, type of pharmacy, job profile and type of drug shortage (Appendix (Table 7) of Chapter 3.2). Along with these determinants, pharmacist MRP in itself was used as a determinant for possible variation in likelihood of intended actions to address the shortages. SPSS 25 was used for all data analysis. Differences in determinants between the pharmacists with different dominant MRPs were first tested with a univariate and then multivariate binary logistic model. Possible associations between pharmacists with different dominant MRPs and the likelihood of the nine intended action options were tested with a chi-square test. Ethics approval The Institutional Review Board Utrecht of the division Pharmacoepidemiology & Clinical Pharmacology at Utrecht University formally approved the research. RESULTS Two hundred sixty-seven respondents (94% community pharmacists and 6% early career pharmacists practising in community pharmacy) completed the survey. Of these respondents, 63% were women, 39% worked in chain pharmacies and 79% were managing pharmacists (primarily responsible pharmacists) at their pharmacies, and the median age was 42 years (IQR = 32–52 years). References of these characteristics to national data are included in Appendix (Table 7) of Chapter 3.2. Nine pharmacists did not provide information about their job profiles. The data from these nine respondents were excluded from the determinant analysis. For the remaining 258 pharmacists, dominant MRPs were calculated for 3 scenarios*258 respondents = 774 scenarios (Table 2). For testing possible associations between pharmacists with different dominant MRPs and the likelihood of the nine intended action options, the data of all 267 respondents were included (Table 3). Table 1 shows the number of identified pharmacists with dominant PE-MRPs, RR-MRPs and BO-MRPs and unclear MRPs in the Osteoporosis, the Contraceptive and the Parkinson’s drug shortage scenarios and in all these scenarios together. Because the number of identified pharmacists with dominant RR- or BO-MRPs is too small, no further determinant analysis could be performed within these groups of pharmacists. Therefore, for the determinant analysis, the group of identified pharmacists with dominant PE-MRPs was compared with the total group of pharmacists who did not have dominant PE-MRPs (i.e. pharmacists with dominant RR-MRPs, dominant BO-MRPs and unclear MRPs).