An online nationwide cross-sectional survey study was conducted and reported following the Consensus-Based Checklist for Reporting of Survey Studies (CROSS) [24]. All approximately 4000 practicing pharmacists in the Netherlands [25,26,27] were eligible for this study. Practicing pharmacists were defined as licensed pharmacists who are involved in the provision of pharmaceutical care for individual patients, in any healthcare setting: community pharmacy, outpatient and inpatient hospital pharmacy, or other healthcare settings (e.g., nursing home or general practice).

The Dutch questionnaire was adapted from an English version that had previously been developed. First, a Swedish version was developed at Uppsala University, based on previous literature [8, 16, 17], and piloted among clinical pharmacists in Sweden in 2022 [28]. During a 12-h workshop at the Pharmaceutical Care Network Europe (PCNE) conference in Hillerød, Denmark (8–11 February 2023), the questionnaire was translated into English and adapted. The workshop, moderated by two researchers (TK and SKS), involved seventeen pharmacists and academics from eleven countries (across Europe and Taiwan) who collaborated to refine the English version for use in countries where pharmacists have no or limited prescribing rights. This questionnaire addressed participant demographics, pharmacist prescribing models, settings, preconditions for introduction, and associated benefits and risks. It featured four pharmacist prescribing models inspired by practices in New Zealand (model A), Ontario, Canada (model B), and the UK (models C-D; Fig. 1).

Potential models of pharmacist prescribing rights in the Netherlands with varying difference of independence, as presented in the questionnaire

The questionnaire included 4-point Likert scale questions, ranging from "Disagree" to "Agree," with the additional option of "Don't know/No opinion," and an open text section for comments. The Dutch version was translated from the English using DeepL Translator [29]. Errors were corrected, some questions were adjusted for Dutch language nuances, and country-specific demographic questions were added. The survey was reviewed and revised by three researchers (BK, TK, DP) and then sent to other team members (HFK, AH, MH, LvD, DZ) for further input. Their feedback was incorporated, resulting in the final version of fifteen questions (Supplementary Material, Appendix 1).

The Dutch survey was piloted in October 2023 with three pharmacists from inpatient and outpatient hospital pharmacy settings. This pilot tested for the comprehensibility of the questionnaire content, including the introduction and questions, as well as the survey's formatting and functionality. Changes to participant demographic questions and minor textual adjustments were made; examples of pharmacist prescribing models from New Zealand (model A), Ontario, Canada (model B) and the UK (model C-D) were added before the model-related questions.

Practicing pharmacists were recruited through the Utrecht Pharmacy Practice network for Education and Research (UPPER), the Royal Dutch Pharmacists Association (KNMP), the Dutch Association of Hospital Pharmacists (NVZA), and an informal network of general practice-based pharmacists, ensuring a broad reach across the target group. The UPPER network [30] has an e-mail database of over 2000 members, the KNMP/NVZA network comprises 5470 members, and the network of general practice-based pharmacists includes fifteen members. An e-mail was sent to all UPPER network members and general practice-based pharmacists, inviting them to participate in the study. KNMP and NVZA added a survey link to their newsletter. There is overlap between the UPPER and KNMP/NVZA networks, leading to some pharmacists receiving multiple participation requests. By utilizing these four networks of pharmacists, the aim was to maximize exposure to the survey and increase the likelihood of participation. In all four approaches, potential participants received study information and a link to Qualtrics (Provo, Utah, USA), an online survey tool used for the questionnaire. By submitting the completed questionnaire, participants consented to participate and understood that due to data anonymity, their responses could not be withdrawn once submitted. Detailed participant information was provided in Qualtrics to ensure informed consent before participation. The survey was open for four weeks in October—November 2023. A follow-up email and newsletter were sent two weeks after the initial questionnaire opening. Pharmacists were explicitly asked to complete the questionnaire only once.

Completed questionnaires were collected in Qualtrics and exported to IBM Statistical Package for Social Sciences (SPSS) version 29 (Armonk, New York, USA) for analysis. Participants were excluded if they were not practicing pharmacists. Descriptive statistics were applied to all quantitative results. The open-text comments were analysed, categorized and presented descriptively where applicable or textually summarized. Pre-defined subgroup analyses were conducted by practice setting, frequency of direct patient contact and years of working experience to explore how these characteristics were associated with agreement with different prescribing models. For these subgroup analyses, the 4-point Likert scale was simplified by combining 'Somewhat disagree' with 'Disagree' and 'Somewhat agree' with 'Agree'. Differences were assessed using chi-square tests (statistical difference at p < 0.05), with post hoc analyses to adjust for multiple comparisons.