Self-perceived learning outcomes of academic detailing discussing rational therapy with proton pump inhibitors among general practitioners in Norway Authors: Jan Schjøtt, Bettina Maria Riedel, Ketil Arne Espnes Published: 2026 ABSTRACT AbstractIntroductionThe increasing prevalence of long-term use of proton pump inhibitors (PPIs) could reflect inappropriate prescribing. Deprescribing PPIs may lead to rebound hyperacidity and subsequent reinitiation of therapy. Education on rational PPI therapy is relevant for general practitioners (GPs).MethodsWe evaluated a Norwegian academic detailing (AD) campaign aimed at promoting rational therapy with PPIs among GPs. Electronic questionnaires were distributed before and after AD visits conducted between August 2024 and July 2025. Respondents rated their agreement with five statements assessing specific knowledge, and two statements assessing general knowledge and skills, related to PPI therapy. A seven-point Likert scale (1 = strongly disagree, 7 = strongly agree) was used to measure self-perceived learning outcomes.ResultsOf 1,018 GPs visited, 138 (14%) completed both questionnaires. The median positive change on the Likert scale was two points for both specific knowledge and general knowledge and skills. Seventy-four percent of GPs reported a 2–4 point improvement in specific knowledge, while 64% reported a similar improvement in general knowledge and skills. Three additional statements regarding the impact of the AD visit were included in the post-visit questionnaire. The statement ‘The AD visit answered all my questions regarding PPI therapy’ received a median rating of 7. ‘The AD visit has changed my practice with PPI’ received a median rating of 6, and ‘The AD visit has confirmed my practice with PPI’ received a median rating of 5. No significant differences were observed between Norwegian health regions in the ratings of statements related to rational prescribing of PPIs or the AD visit.ConclusionsAD addressing rational therapy with PPIs was associated with a positive self-perceived learning outcome among GPs who completed both questionnaires.Keywords: Academic detailing, general practitioner, learning outcomes, proton pump inhibitors, rational therapy Introduction Proton pump inhibitors (PPIs) are among the most widely prescribed medications worldwide and have significantly improved the treatment of acid-related disorders, gastrointestinal bleeding, hypersecretory conditions and other related diseases [1–5]. However, concerns have been raised regarding the increasing prevalence of long-term PPI use (> 4–8 weeks), the potential for inappropriate prescribing practices, and the risk of adverse effects [5–9]. One notable adverse effect of prolonged PPI therapy is rebound acid hypersecretion (RAHS), which may occur following discontinuation due to a compensatory increase in gastric acid production [10,11]. RAHS may present with symptoms such as heartburn and dyspepsia, often prompting patients to resume PPI therapy, assuming these symptoms indicate recurrence of the original condition. Nevertheless, evidence suggests that more than 50% of patients can discontinue or reduce their PPI dosage without compromising symptom control [10]. Recent studies have raised safety concerns associated with the long-term use of PPIs, particularly in relation to risks of gastric cancer and bone fractures [12,13]. These findings underscore the need for educational outreach to promote rational therapy with PPIs. The Norwegian Academic Detailing Program (KUPP) is educational outreach to general practitioners (GPs). The program is a collaboration between the Regional Medicines Information and Pharmacovigilance Centres (RELIS) and clinical pharmacology units at the four regional university hospitals. KUPP campaigns typically target inappropriate prescribing practices. The current campaign was initiated in response to a marked increase in PPI prescribing in Norway over the past two decades. Between 2004 and 2022, the proportion of the population receiving PPI prescriptions rose from 4% to 12%. Additionally, following its market introduction in autumn 2010, prescriptions for Vimovo© (naproxen/esomeprazole) increased steadily, reaching 4% of the population, approximately 220,000 individuals, by 2022 [14]. This study measured Norwegian GPs’ self-perceived learning outcomes from AD sessions focused on rational therapy with PPIs. Methods Study population General practitioners (GPs) in Norway who scheduled and received AD visits between August 2024 and July 2025 were eligible for inclusion in the study. Intervention AD sessions were conducted by detailers from RELIS and the clinical pharmacology units. Each session lasted approximately 20 min and was delivered either in-person or virtually. Virtual sessions were conducted via conferencing platforms such as Whereby (Video Communication Service AS, parent company Videonor AS©) or Microsoft Teams©. In-person AD sessions involved discussion of a printed campaign brochure, while virtual sessions utilized a Microsoft PowerPoint presentation based on the same brochure. The material emphasised the importance of avoiding unnecessary PPI prescriptions, the risk of RAHS following PPI withdrawal, the use of a PPI trial period as a diagnostic tool, the risk of adverse events with long-term PPI use, and appropriate deprescribing strategies. It featured tables outlining recommended PPIs for reflux treatment, appropriate dosing, and maximum daily doses. Non-prescription alternatives such as antacids and histamine H2-receptor antagonists were suggested. Flow-charts facilitated structured decision-making, guiding the GP through therapeutic pathways based on prevailing ­clinical factors. Outcome measures Two electronic questionnaires were developed using Corporater Surveyor (Corporater AS©). The pre-AD questionnaires (pre-AD) were distributed to GPs via email after booking before the visit, while the post-AD questionnaires (post-AD) were distributed after the AD visit. Seven statements assessing knowledge of PPI therapy were rated using a seven-point Likert scale (1 = strongly disagree, 7 = strongly agree). Five statements focused on specific knowledge: My knowledge of correct initiation of PPI therapy is good. My knowledge of correct initiation of PPI therapy is good. My knowledge of rebound acid hypersecretion associated with PPI therapy is good. My knowledge of rebound acid hypersecretion associated with PPI therapy is good. My knowledge of long-term PPI therapy is good. My knowledge of long-term PPI therapy is good. My knowledge of risks associated with long-term PPI therapy is good. My knowledge of risks associated with long-term PPI therapy is good. My knowledge of correct withdrawal of PPI therapy is good. My knowledge of correct withdrawal of PPI therapy is good. Two statements assessed general knowledge and prescribing confidence: My overall knowledge of PPI therapy is very good. My overall knowledge of PPI therapy is very good. I feel confident in prescribing PPIs. I feel confident in prescribing PPIs. Responses were matched by identical email addresses to pair pre- and post-AD. Three additional statements were included exclusively in post-AD to assess the perceived impact of the AD visit: The AD visit answered all my questions regarding PPI therapy. The AD visit answered all my questions regarding PPI therapy. The AD visit has changed my practice regarding PPI use. The AD visit has changed my practice regarding PPI use. The AD visit has confirmed my current practice regarding PPI use. The AD visit has confirmed my current practice regarding PPI use. Statistical analysis Only GPs who completed both the pre- and post-AD were included in the analysis. Mean composite scores were calculated for the five specific knowledge items and the two general knowledge and skills items, for both pre- and post-AD. The difference in these composite scores represented the perceived change in learning outcomes. Cronbach’s alpha was used to assess the reliability and internal consistency of the Likert items. The Shapiro–Wilk test was applied to evaluate the normality of the score differences. Depending on the results of the normality test, either parametric or non-parametric statistical tests were employed for further analysis. Ratings of the AD visit’s impact were reported as mean, median and range. Statistical significance was defined as p < 0.01. Results A total of 1,018 general practitioners (GPs) received AD visits during the study period. Of these, 138 GPs (14%) completed both the pre- and post-AD. The proportion of GPs who completed both varied across the four Norwegian health regions: North (6%), Central (8%), South-East (21%), and West (26%). Overall, 33% of GPs completed the pre-AD, and 30% completed the post-AD. Specific knowledge The internal consistency between the five Likert items measuring perceived specific knowledge in PPI therapy was high, with Cronbach’s alpha values of 0.809 (pre-AD) and 0.890 (post-AD). The Shapiro–Wilk test indicated normal distribution of the change in perception (statistic = 0.987, p = 0.210). No significant differences were observed between the four regions using either parametric (ANOVA, p = 0.649) or non-parametric (Kruskal–Wallis, p = 0.697) tests. Figure 1(a) illustrates the change in perceived specific knowledge. The median positive change on the Likert scale was two points for specific knowledge. Seventy-four percent of GPs reported a 2–4 point improvement in specific knowledge. Change in perception (mean and range) of specific knowledge (a) and general knowledge and skills (b) in PPI therapy among GPs (n = 138) from four Norwegian health regions. Bold label on the x-axis in (a) and (b) is the median change for all 4 regions. Change in perception represents the difference between the mean composite scores of five statements (specific knowledge) and two statements (general knowledge and skills), rated before and after AD using a Likert scale ranging from 1: strongly disagree to 7: strongly agree. AD: academic detailing; PPI: proton pump inhibitors. General knowledge and skills The internal consistency of the two Likert items assessing general knowledge and skills was acceptable, with Cronbach’s alpha values of 0.756 (pre-AD) and 0.767 (post-AD). The Shapiro–Wilk test indicated a non-normal distribution of the change in perception (statistic = 0.964, p = 0.001). No significant regional differences were found using the Kruskal–Wallis test (p = 0.697). Figure 1(b) presents the change in perceived general knowledge and skills. The median positive change on the Likert scale was two points for general knowledge and skills. Sixty-four percent of GPs reported a 2–4 point improvement in general knowledge and skills. Perceived impact of AD Three additional statements assessing the impact of the AD visit were included in post-AD. None of these items were normally distributed (Shapiro–Wilk statistics: 0.744, 0.870, and 0.932; all p < 0.01). Table 1 summarizes the mean, median, and range of responses to the statements. No significant regional differences were found using non-parametric tests (Kruskal–Wallis: p = 0.223, p = 0.539, p = 0.028). Table 1.Rating of statements concerning the effect of AD among 138 Norwegian GPs.StatementsThe AD visit answered all my questions regarding PPI therapyThe AD visit has changed my practice with PPIThe AD visit has confirmed my practice with PPIMean6,285,544,92Median765Range3–71–72–7AD: academic detailing; PPI: proton pump inhibitors; GPs: general practitioners.A seven-point Likert scale ranging from 1: strongly disagree to 7: strongly agree was used to measure the effect of AD. Rating of statements concerning the effect of AD among 138 Norwegian GPs. AD: academic detailing; PPI: proton pump inhibitors; GPs: general practitioners. A seven-point Likert scale ranging from 1: strongly disagree to 7: strongly agree was used to measure the effect of AD. Discussion This study found positive self-perceived learning outcomes among Norwegian GPs following AD focused on rational therapy with PPIs. Improvements of 2–4 points on a 7 point Likert scale were observed among most GPs, both in specific knowledge and in general knowledge and skills related to PPI therapy. Additionally, the AD visits received positive ratings by the GPs in terms of their perceived impact. The variation in ratings across individual questionnaire items was to be expected. The five statements addressing specific knowledge directly reflected the core topics discussed during the AD sessions, which likely contributed to the positive learning outcomes. Similarly, the high rating for the statement ‘The AD visit answered all my questions regarding PPI therapy’ aligns with the targeted nature of the campaign. Ratings related to general knowledge and practice change may reflect that some GPs were already familiar with issues surrounding inappropriate PPI prescribing and the associated adverse effects of these drugs. The low overall participation rate (14%) raises concerns about the generalisability of our findings, and bias related to perceived relevance cannot be excluded. One possible explanation is dissatisfaction with the AD visit; however, we argue that the low response rate is primarily due to organizational factors. In the South-East and West regions, a dedicated secretary managed distribution of pre- and post-AD. The participation rates were 57% and 53% for pre-AD; 46% and 43% for post-AD. The West region, with the highest proportion of both completed questionnaires (26%), had prior experience from a similar study of rational migraine therapy where 25% of eligible GPs completed both questionnaires [15]. In contrast, the North and Central regions, lacked administrative support, showed participation rates around 20% for pre- and post-AD. In these regions, the detailers themselves were responsible for distributing the survey links, which may have led to inconsistencies. Additionally, some GPs provided personal email addresses, while others used work addresses, which may have resulted in survey links being missed or filtered as spam. Furthermore, we found 6 cases of duplicate pre-AD and 2 cases of duplicate post-AD questionnaires. Previous research has shown that web-based surveys distributed via email have significantly lower response rates compared to postal surveys [16]. Response rates among GPs to surveys are often below 50%, and reducing questionnaire length or offering a financial incentive does not guarantee improved response [17]. A challenge in the present study was that GPs had to answer two consecutive questionnaires. Importantly, among the 310 GPs who completed post-AD (30% of all visited), 100% expressed interest in receiving a subsequent AD visit on a different topic. Furthermore, the 172 GPs that only completed post-AD showed no difference in how they had assessed the perceived impact of the AD visit compared with the 138 GPs that had completed both questionnaires. Evaluation of the KUPP also shows that over 90% of GPs find AD to be a valuable and practice-relevant educational method [18]. In conclusion, the consistently positive learning outcomes among GPs who completed both questionnaires support the continuation of this AD campaign. The low response rate appears to be a consequence of study logistics rather than dissatisfaction with the intervention. Future efforts will focus on improving study organization and expanding participation. A follow-up study is planned to include both previously visited and new participants through the end of 2025. This study will utilize data from the Norwegian Prescribed Drug Registry to examine whether the perceived learning outcomes reported in the present study are associated with actual changes in PPI prescribing behaviour among Norwegian GPs after educational AD visits. Acknowledgements The authors would like to thank Inger Elise Engelund, cand. polit for her contribution to the development of the questionnaires and the organisation of the research data. We also acknowledge all the detailers, as well as the assistance with managing questionnaire distribution before and after visits, provided by Solmei Danielsen, health secretary. Funding Statement This study was supported only by the Regional Medicines Information and Pharmacovigilance Centres (RELIS) and clinical pharmacology units at the four regional university hospitals in Norway. Ethical approval The present study was applied to the data protection officer at Haukeland University Hospital and coordinated at the Centre for Patient-Reported Outcomes Data (PROM) at Haukeland University Hospital, Bergen, Norway. Generative artificial intelligence (AI) During the preparation of this work the author(s) used M365 Copilot in order to check language and readability of drafts. After using this tool/service, the author(s) reviewed and edited the content as needed and take(s) full responsibility for the content of the published article. Disclosure statement No potential conflict of interest was reported by the author(s). Data availability statement Data on GPs are not available in agreement with the data protection officer at Haukeland Hospital. References Associated Data Data Availability Statement Data on GPs are not available in agreement with the data protection officer at Haukeland Hospital.