Introduction: A better understanding of the determinants involved in general practitioners' (GPs) decision-making processes when it comes to prescribing statins as primary prevention in patients with multimorbidity could provide insights for improving implementation of primary prevention guidelines.

Methods: We conducted a qualitative study using a deductive framework-based and inductive analysis of GPs' semi-structured interviews verbatim, from which expertise profiles of prescribers were also drawn. The analytical framework was built from a pragmatic synthesis of the evidence-based medicine, Modelling using Typified Objects (MOT) model of clinical reasoning processes, Theoretical Domains Framework, and shared decision-making frameworks.

Results: Fifteen GPs were interviewed between June 2019 and January 2020. Diabetes seemed to represent a specific motivation for deciding about statin prescription for primary prevention purposes; and in situations of multimorbidity, GPs differentiated between cardiovascular and non-cardiovascular multimorbidity. Expert prescribers seemed to have integrated the utilisation of cardiovascular risk calculation scores throughout their practice, whereas non-expert prescribers considered them difficult to interpret and preferred using more of a "rule of thumb" process. One interviewee used the risk calculation score as a support for discussing statin prescription with the patient.

Conclusion: Our results shed light on the reasons why statins remain under-prescribed for primary prevention and why non-diabetic multimorbid patients have even lower odds of being prescribed a statin. They call for a change in the use of risk assessment scores, by placing them as decision aids, to support and improve personalised shared decision-making discussions as an efficient approach to improve the implementation of recommendations about statins for primary prevention.