Prescribing Responsibly: Navigating the Tides of Deprescribing in Proton Pump Inhibitor Stewardship

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

A very interesting paper. A lot of important information has been gathered in one article. I have no comments on the content of the article.

Comments to improve the paper:
1. My main comment concerns the need to supplement the methodology of the scientific work.

In my opinion, this requires necessary supplementation. Dear authors, please indicate how you selected the scientific articles that you chose for your article.

For example, if the PRISMA method or another method was used, this should be described. What were the criteria for selecting the literature and the search strategies?

2. I suggest supplementing the section on research methodology in the article, for example: Data Sources and Literature Searches, Study Selection and Eligibility Criteria, Data Extraction, Quality Assessment and Certainty of Evidence. I also consider it important to supplement the information on the results of the search strategy, which articles were selected and why (Included Publications in article).

3. Clarifying the article with information on the strategy for selecting the cited works is my main recommendation in terms of expected additions and corrections to the article.

4. I also have a question: Do the authors recognise the possibility of involving pharmacists in the PPI deprescription process?

Author Response

Comment 1: “My main comment concerns the need to supplement the methodology of the scientific work.

In my opinion, this requires necessary supplementation. Dear authors, please indicate how you selected the scientific articles that you chose for your article.

For example, if the PRISMA method or another method was used, this should be described. What were the criteria for selecting the literature and the search strategies?”

Respond 1: We thank the reviewer for this important suggestion. In response, we have added a new Section 2 titled “Literature Selection Strategy” to clearly describe our approach to identifying and selecting the literature cited in this narrative review. This section outlines our targeted search strategies for clinical guidelines, randomized controlled trials, systematic reviews, meta-analyses, and observational studies relevant to appropriate PPI use, adverse effects, and deprescribing strategies. We also note that reference lists of key articles were reviewed to ensure completeness. We are grateful for the reviewer’s feedback, which helped improve the transparency and rigor of our methodology.

Comment 2: “I suggest supplementing the section on research methodology in the article, for example: Data Sources and Literature Searches, Study Selection and Eligibility Criteria, Data Extraction, Quality Assessment and Certainty of Evidence. I also consider it important to supplement the information on the results of the search strategy, which articles were selected and why (Included Publications in article).”

Respond 2: We thank the reviewer for their thoughtful recommendation to expand the methodology section. In response, we have enhanced Section 2 (“Literature Selection Strategy”) to more clearly describe our approach, including the data sources searched, keywords used, general inclusion criteria, and the rationale for selecting studies. While formal quality assessment tools and evidence grading were not applied given the narrative review format, we explicitly state that we prioritized high-quality evidence such as clinical guidelines, systematic reviews, meta-analyses, and randomized controlled trials. We also clarified that reference lists were reviewed to supplement database searches. We appreciate the reviewer’s suggestion, which has improved the transparency and methodological clarity of our review.

Comment 3: “Clarifying the article with information on the strategy for selecting the cited works is my main recommendation in terms of expected additions and corrections to the article.”

Respond 3: We appreciate the reviewer’s emphasis on the importance of clearly describing the strategy used to select cited works. In response, we have added and expanded Section 2 (“Literature Selection Strategy”) to provide a transparent overview of our approach to identifying and prioritizing relevant studies. To further assist readers, we have also added a sentence at the end of the Introduction to direct them to this section. We thank the reviewer for highlighting this key aspect, which has strengthened the clarity and rigor of our manuscript.

Comment 4: “I also have a question: Do the authors recognise the possibility of involving pharmacists in the PPI deprescription process?”

Respond 4: We thank the reviewer for this insightful question regarding pharmacist involvement in the deprescribing process. We fully agree that pharmacists play a critical role in identifying inappropriate PPI use, educating patients, collaborating with prescribers, and facilitating safe tapering strategies. In response, we have expanded Sections 5.1 (“Deprescribing Strategies”) and 5.4 (“Challenges and Barriers”) to explicitly describe the pharmacist’s contributions and the potential barriers to broader implementation of pharmacist-led deprescribing initiatives. We have also cited recent literature demonstrating both the impact of pharmacist-led interventions and the challenges related to staffing, reimbursement, and documentation access. We appreciate the reviewer’s thoughtful query, which helped us highlight an important and often underutilized resource in deprescribing efforts.

We sincerely thank the reviewer for their thoughtful, constructive, and collegial feedback. Your comments helped us improve the clarity, methodological transparency, and practical relevance of our manuscript. We appreciate the opportunity to revise our work in response to your insights and hope the updated version addresses your concerns and suggestions effectively.

Reviewer 2 Report

Comments and Suggestions for Authors

This narrative review examines the appropriate use of proton pump inhibitors (PPIs) and the emerging need for deprescribing practices. The authors summarize indications for PPI therapy (e.g., GERD, peptic ulcers) and then review evidence of long-term risks – including vitamin/mineral deficiencies, infections, renal and cardiovascular effects, and certain malignancies – largely drawn from observational studies.
They emphasize that many patients remain on PPIs without a clear indication and outline deprescribing strategies (tapering dose, on-demand use, switching to H₂-blockers, lifestyle changes) to mitigate harm. The review concludes that clinicians should regularly reassess PPI necessity and apply evidence-based deprescribing protocols to improve patient safety. Overall, the manuscript aims to bridge current prescribing practices with patient-centered stewardship of PPI therapy.

Major Comments

• Scope and Relevance: The topic of PPI overuse and deprescribing is timely and relevant to a pharmacoepidemiology audience. The manuscript comprehensively addresses known issues in PPI use. However, it reads largely as a synthesis of existing guidelines and studies without presenting new analyses. The authors should clarify the unique contribution or perspective of this review (beyond summarizing literature) to justify its publication. For example, if this is meant as a policy piece, emphasizing novel insights (e.g., on implementation strategies) would strengthen its appeal. The content level is appropriate for the journal, but the authors might deepen the analysis to meet scholarly expectations.

• Validation of Claims and References: The review generally cites relevant sources, which is a strength. Nonetheless, some statements need clearer support or completion. For instance, Table 3 reports a relative risk of 2.56 for iron deficiency anemia with long-term PPI use, but omits the sample size (“n=?”). The authors should fill in such missing details to maintain credibility. Similarly, the manuscript could incorporate recent evidence; for example, an emulated trial in cirrhosis found that deprescribing PPIs was associated with reduced ascites and encephalopathy. 

  • Including such data would validate the recommendation for deprescribing in certain populations. Overall, every key assertion should clearly map to an appropriate reference, and any omissions (like the missing “n” in Table 3) should be corrected.

  • Statistical Rigor: While the authors do not perform new analyses, they summarize many meta-analyses and trials. Table 3 is helpful, but it could better contextualize heterogeneity and study design. For example, the gastric cancer risk meta-analysis shows a high heterogeneity (I²=95%), yet the manuscript treats the association at face value. The authors should highlight such heterogeneity and its implications (i.e., inconsistency across studies) so readers understand the strength of evidence. Also, the table mixes odds ratios (OR) and relative risks (RR) from different study types without always clarifying context. The authors should specify when pooled estimates come from observational cohorts vs. RCTs, and interpret the clinical significance accordingly. For example, many reported ORs are modest (1.2–1.7) and often from observational data; the text should caution against over-interpreting these as causal effects without RCT confirmation.

  • Figures and Tables: Figure 1 (Evidence Pyramid) is standard and reinforces the text’s point about evidence levels. Ensure it is clearly labeled and legible. Table 3 provides a broad summary of systematic reviews, which is valuable. However, readability can improve: columns should consistently align (e.g. put all effect estimates in the same format), and footnotes should explain any symbols (e.g. “↑” for increased risk). Crucially, the missing sample size in the iron-deficiency row (Table 3, “n=?”) must be corrected. Without it, the entry is incomplete. The authors should double-check all tables and figures for completeness and clarity (for instance, explicitly labeling which populations are studied, such as “clopidogrel users” or “general GERD patients” in Table 3 for context).

  • Organization and Clarity: The manuscript is well-structured into logical sections (Indications, Risks, Deprescribing, etc.). Writing is generally clear, and headings are used appropriately. However, some sections (particularly the various “Risks” subsections) are dense. It may help to begin each risk subsection with a brief summary statement. For example, a one-sentence summary of evidence (e.g., “Multiple observational studies have linked long-term PPI use to increased risk of chronic kidney disease”) would orient readers before delving into details. Similarly, consider using bulleted lists or subheadings within lengthy sections to break up text (the depression of the polypharmacy barriers section could, for instance, list barriers by stakeholder). Shorter paragraphs (3–5 sentences) and occasional bullet lists for key points would enhance readability.

  • Discussion of Limitations: The review alludes to limitations (e.g., noting conflicting results in cardiovascular outcomes), but it lacks a dedicated discussion of evidence limitations. The authors should explicitly state that much of the risk data comes from observational studies, which are subject to confounding and bias. For instance, many associations (e.g., PPI use with infection or fracture) may reflect underlying patient comorbidities rather than a direct drug effect. Highlighting this uncertainty (perhaps in a “Limitations” paragraph) will give balance. It would also be helpful to discuss generalizability – for example, whether most data come from Western populations and how that might affect applicability elsewhere. Admitting the evidence gaps (e.g., lack of large RCTs confirming harms) would make the review more nuanced.

  • Strength and Relevance of Evidence: The evidence cited is extensive, but the authors should better appraise its quality. As noted, many cited systematic reviews report significant heterogeneity (Table 3) and wide confidence intervals. The text should more consistently comment on this (e.g., “the pooled estimate was significant but heterogeneity was high, indicating variability among studies”). Conversely, where high-quality RCT data exist, it should be emphasized. For example, a meta-analysis of RCTs in patients on clopidogrel found no increase in cardiovascular events with PPIs, aligning with the COGENT trial – this nuance is mentioned briefly but could be clearer. The authors might adopt language grading evidence strength (e.g., “low-quality evidence suggests…”). They should also anchor their discussion to guideline statements when possible. For instance, the American Gastroenterological Association explicitly recommends deprescribing PPIs when no clear indication exists (PMID: 37303818);  integrating such guideline positions would reinforce the review’s conclusions with authoritative backing.

    • Justification of Conclusions and Recommendations: The conclusions generally follow from the evidence presented, but some claims could be tempered. The authors appropriately advocate for regular PPI review and targeted deprescribing strategies, which are in line with guidelines. However, phrases like “H₂RAs are generally safer” or implying PPIs definitively cause certain outcomes might overstate the case given inconsistent data. The conclusions should emphasize “considering deprescribing when appropriate” rather than suggesting an across-the-board change. It may be helpful to explicitly state that deprescribing aims for patients without ongoing indications (as per evidence-based guidelines PMID: 37303818) 

      • . In summary, the recommendations are sensible but should stress individualized decision-making, acknowledging that for high-risk patients (e.g., severe esophagitis), continued PPI use is appropriate.

      • Additional Literature and Context: There are a few relevant recent publications that could enrich the discussion. For example, a systematic review by Olmastroni et al. (2024) on deprescribing interventions highlights that multi-disciplinary approaches (algorithms, pharmacist involvement, patient engagement) are most effective (https://doi.org/10.3390/jcm13206283). Citing such findings could guide readers on how to implement deprescribing in practice. Also, mentioning ongoing research (e.g., the DROPIT randomized deprescribing trial) would illustrate that this is an active field. If space allows, summarizing the Cochrane review on PPI deprescribing (which found that on-demand therapy increases symptom recurrence) would be informative and balanced. Finally, the barriers section could note patient perspectives or health-system factors more explicitly, perhaps with a citation or two if available.

        Recommendation

        Major Revisions. The manuscript addresses an important clinical issue and is thoroughly researched, but it requires substantial strengthening before publication. Key claims should be tied more explicitly to evidence (with missing data filled in), and the discussion should better appraise the quality and heterogeneity of the cited studies. Organizational refinements (shorter paragraphs, summary statements) would improve clarity. Including recent studies and explicitly acknowledging limitations will increase the rigor and credibility of the recommendations. With these revisions and integration of up-to-date evidence (e.g., recent trials and guidelines), the paper could make a valuable contribution to pharmacoepidemiology.

        
        
        

Author Response

Comment 1: “Scope and Relevance: The topic of PPI overuse and deprescribing is timely and relevant to a pharmacoepidemiology audience. The manuscript comprehensively addresses known issues in PPI use. However, it reads largely as a synthesis of existing guidelines and studies without presenting new analyses. The authors should clarify the unique contribution or perspective of this review (beyond summarizing literature) to justify its publication. For example, if this is meant as a policy piece, emphasizing novel insights (e.g., on implementation strategies) would strengthen its appeal. The content level is appropriate for the journal, but the authors might deepen the analysis to meet scholarly expectations.”

Response 1: We sincerely thank the reviewer for their thoughtful and constructive feedback. We agree that it is important to clearly articulate the unique contribution of our manuscript. In response, we have revised the abstract and introduction to better highlight our perspective on deprescribing as a stewardship opportunity. Specifically, we emphasize how our review extends beyond summarizing the literature by proposing a practical, implementation-oriented framework grounded in stewardship principles. Additionally, we have refined the conclusion to underscore the importance of integrating deprescribing strategies into clinical workflows through behavior change and system-level interventions. We are grateful for the reviewer’s insight, which has strengthened the clarity and scholarly contribution of our work.

Comment 2: “Validation of Claims and References: The review generally cites relevant sources, which is a strength. Nonetheless, some statements need clearer support or completion. For instance, Table 3 reports a relative risk of 2.56 for iron deficiency anemia with long-term PPI use, but omits the sample size (“n=?”). The authors should fill in such missing details to maintain credibility. Similarly, the manuscript could incorporate recent evidence; for example, an emulated trial in cirrhosis found that deprescribing PPIs was associated with reduced ascites and encephalopathy. Including such data would validate the recommendation for deprescribing in certain populations. Overall, every key assertion should clearly map to an appropriate reference, and any omissions (like the missing “n” in Table 3) should be corrected.”

Response 2: We sincerely appreciate the reviewer’s careful attention to the accuracy and completeness of our references and data presentation. In response, we have updated Table 3 to include the missing sample size information, ensuring greater clarity and transparency. Additionally, we have incorporated recent evidence from an emulated clinical trial by Tapper et al. (2024), which found that deprescribing PPIs in patients with compensated cirrhosis was associated with reduced risks of ascites and hepatic encephalopathy. This important study has been integrated into Section 4 to strengthen the evidence base supporting deprescribing in specific high-risk populations. We are grateful for the reviewer’s insightful suggestions, which have meaningfully improved the rigor and relevance of our manuscript.

Comment 3: “Statistical Rigor: While the authors do not perform new analyses, they summarize many meta-analyses and trials. Table 3 is helpful, but it could better contextualize heterogeneity and study design. For example, the gastric cancer risk meta-analysis shows a high heterogeneity (I²=95%), yet the manuscript treats the association at face value. The authors should highlight such heterogeneity and its implications (i.e., inconsistency across studies) so readers understand the strength of evidence. Also, the table mixes odds ratios (OR) and relative risks (RR) from different study types without always clarifying context. The authors should specify when pooled estimates come from observational cohorts vs. RCTs, and interpret the clinical significance accordingly. For example, many reported ORs are modest (1.2–1.7) and often from observational data; the text should caution against over-interpreting these as causal effects without RCT confirmation.”

Response 3: We thank the reviewer for this thoughtful and important comment regarding statistical interpretation. In response, we have revised the text immediately preceding Table 3 to provide greater context about the nature of the evidence presented. Specifically, we now clarify that the table summarizes findings from systematic reviews and meta-analyses (SR-MAs), most of which are based on observational studies. We also highlight the presence of substantial heterogeneity in several analyses (e.g., I² = 95% for the gastric cancer meta-analysis) and caution readers against overinterpreting modest effect sizes, particularly in the absence of randomized controlled trial data. These additions aim to improve the transparency, nuance, and clinical applicability of our synthesis. We are grateful for the reviewer’s guidance in strengthening the rigor and interpretive clarity of our manuscript.

Comment 4: “Figures and Tables: Figure 1 (Evidence Pyramid) is standard and reinforces the text’s point about evidence levels. Ensure it is clearly labeled and legible. Table 3 provides a broad summary of systematic reviews, which is valuable. However, readability can improve: columns should consistently align (e.g. put all effect estimates in the same format), and footnotes should explain any symbols (e.g. “↑” for increased risk). Crucially, the missing sample size in the iron-deficiency row (Table 3, “n=?”) must be corrected. Without it, the entry is incomplete. The authors should double-check all tables and figures for completeness and clarity (for instance, explicitly labeling which populations are studied, such as “clopidogrel users” or “general GERD patients” in Table 3 for context).”

Response 4: We sincerely appreciate the reviewer’s detailed and helpful feedback on the figures and tables. In response, we have updated Figure 1 to enhance clarity and expanded the figure caption to better describe the hierarchy of evidence. For Table 3, we corrected the previously missing sample size, standardized the formatting of all effect estimates, and added a new column to specify the patient populations studied in each meta-analysis. We also verified that all footnotes are complete and explain any symbols used. These improvements have strengthened the readability, transparency, and utility of our data presentation, and we are grateful for the reviewer’s thoughtful suggestions.

Comment 5: “Organization and Clarity: The manuscript is well-structured into logical sections (Indications, Risks, Deprescribing, etc.). Writing is generally clear, and headings are used appropriately. However, some sections (particularly the various “Risks” subsections) are dense. It may help to begin each risk subsection with a brief summary statement. For example, a one-sentence summary of evidence (e.g., “Multiple observational studies have linked long-term PPI use to increased risk of chronic kidney disease”) would orient readers before delving into details. Similarly, consider using bulleted lists or subheadings within lengthy sections to break up text (the depression of the polypharmacy barriers section could, for instance, list barriers by stakeholder). Shorter paragraphs (3–5 sentences) and occasional bullet lists for key points would enhance readability.”

Response 5: We thank the reviewer for the kind comments on the manuscript’s overall structure and organization, as well as the helpful suggestions to enhance clarity and readability. In response, we have revised Section 3 to ensure that each risk-related subsection now begins with a concise summary statement to orient the reader. We also shortened and streamlined several of the longer subsections to improve flow and reduce cognitive load. These changes have improved the clarity and accessibility of the content, and we are grateful for the reviewer’s insightful recommendations.

Comment 6: “Discussion of Limitations: The review alludes to limitations (e.g., noting conflicting results in cardiovascular outcomes), but it lacks a dedicated discussion of evidence limitations. The authors should explicitly state that much of the risk data comes from observational studies, which are subject to confounding and bias. For instance, many associations (e.g., PPI use with infection or fracture) may reflect underlying patient comorbidities rather than a direct drug effect. Highlighting this uncertainty (perhaps in a “Limitations” paragraph) will give balance. It would also be helpful to discuss generalizability – for example, whether most data come from Western populations and how that might affect applicability elsewhere. Admitting the evidence gaps (e.g., lack of large RCTs confirming harms) would make the review more nuanced.”

Response 6: We thank the reviewer for this insightful recommendation. In response, we have added a dedicated “Limitations of the Evidence Base” subsection at the end of Section 3 to explicitly address the methodological limitations of the studies cited in our review. This new paragraph highlights that most of the evidence comes from observational studies, discusses the potential for confounding and bias, and emphasizes the modest effect sizes and substantial heterogeneity in several meta-analyses. We also address the limited generalizability of the findings, given that most data are drawn from Western populations. We appreciate the reviewer’s suggestion, which has strengthened the critical appraisal and nuance of our manuscript.

Comment 7: “Strength and Relevance of Evidence: The evidence cited is extensive, but the authors should better appraise its quality. As noted, many cited systematic reviews report significant heterogeneity (Table 3) and wide confidence intervals. The text should more consistently comment on this (e.g., “the pooled estimate was significant but heterogeneity was high, indicating variability among studies”). Conversely, where high-quality RCT data exist, it should be emphasized. For example, a meta-analysis of RCTs in patients on clopidogrel found no increase in cardiovascular events with PPIs, aligning with the COGENT trial – this nuance is mentioned briefly but could be clearer. The authors might adopt language grading evidence strength (e.g., “low-quality evidence suggests…”). They should also anchor their discussion to guideline statements when possible. For instance, the American Gastroenterological Association explicitly recommends deprescribing PPIs when no clear indication exists (PMID: 37303818);  integrating such guideline positions would reinforce the review’s conclusions with authoritative backing.”

Response 7: We thank the reviewer for their thoughtful suggestions to strengthen the appraisal of evidence quality and relevance. In response, we have revised the limitations section to explicitly note that wide confidence intervals and high between-study heterogeneity were common across many meta-analyses, reflecting variability in study populations, methodologies, and outcome definitions. This addition reinforces the need for cautious interpretation of pooled estimates. We have also clarified our discussion of cardiovascular outcomes by emphasizing that although early observational studies raised concerns, high-quality RCT evidence—including a meta-analysis of randomized trials and the landmark COGENT trial—found no significant increase in cardiovascular events with concomitant PPI and clopidogrel use. Finally, we have incorporated a citation to the American Gastroenterological Association’s Best Practice Advice [PMID: 37303818] to support deprescribing recommendations in patients without a clear ongoing indication. We are grateful for the reviewer’s detailed feedback, which has meaningfully improved the nuance, clarity, and strength of our manuscript.

Comment 8: “Justification of Conclusions and Recommendations: The conclusions generally follow from the evidence presented, but some claims could be tempered. The authors appropriately advocate for regular PPI review and targeted deprescribing strategies, which are in line with guidelines. However, phrases like “H₂RAs are generally safer” or implying PPIs definitively cause certain outcomes might overstate the case given inconsistent data. The conclusions should emphasize “considering deprescribing when appropriate” rather than suggesting an across-the-board change. It may be helpful to explicitly state that deprescribing aims for patients without ongoing indications (as per evidence-based guidelines PMID: 37303818). In summary, the recommendations are sensible but should stress individualized decision-making, acknowledging that for high-risk patients (e.g., severe esophagitis), continued PPI use is appropriate.”

Response 8: We thank the reviewer for this valuable feedback regarding the tone and justification of our conclusions and recommendations. In response, we have reviewed the manuscript carefully to ensure that our language reflects appropriate caution, particularly in describing associations rather than implying causality. For example, we revised phrases that could overstate the evidence, such as those implying that H₂RAs are definitively safer or that PPIs cause specific outcomes. We also updated the conclusion to emphasize that deprescribing should be considered when clinically appropriate and in alignment with evidence-based guidelines, including the AGA’s Best Practice Advice [PMID: 37303818]. Our revised framing underscores individualized decision-making and acknowledges that continued PPI therapy remains appropriate for high-risk patients. We appreciate the reviewer’s thoughtful comments, which have improved the clarity, balance, and clinical relevance of our manuscript.

Comment 9: “Additional Literature and Context: There are a few relevant recent publications that could enrich the discussion. For example, a systematic review by Olmastroni et al. (2024) on deprescribing interventions highlights that multi-disciplinary approaches (algorithms, pharmacist involvement, patient engagement) are most effective (https://doi.org/10.3390/jcm13206283). Citing such findings could guide readers on how to implement deprescribing in practice. Also, mentioning ongoing research (e.g., the DROPIT randomized deprescribing trial) would illustrate that this is an active field. If space allows, summarizing the Cochrane review on PPI deprescribing (which found that on-demand therapy increases symptom recurrence) would be informative and balanced. Finally, the barriers section could note patient perspectives or health-system factors more explicitly, perhaps with a citation or two if available.”

Response 9: We are grateful to the reviewer for suggesting additional literature to enrich the context and strengthen the implementation aspects of our manuscript. In response, we have incorporated findings from the 2024 systematic review by Rossi et al., which emphasizes the effectiveness of multifaceted deprescribing interventions involving pharmacists, deprescribing algorithms, and patient engagement. This citation has been integrated into the “Challenges and Barriers” section to illustrate evidence-based solutions to common obstacles. We also added a reference to the ongoing DROPIT randomized trial in the “Future Directions” section to highlight that PPI deprescribing remains an active area of research. We appreciate the reviewer’s thoughtful recommendations, which have enhanced the practical relevance and timeliness of our manuscript.

We are deeply grateful to the reviewer for the time, care, and expertise reflected in their comprehensive feedback. Your insightful comments helped us strengthen the manuscript’s clarity, critical appraisal of evidence, methodological transparency, and practical application. We have carefully addressed each point and believe the revised version is significantly improved as a result of your thoughtful suggestions.

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

The article has become suitable for publication.