Deprescribing Proton Pump Inhibitors: What the Evidence Supports
A synthesis of 66 deprescribing studies, multiple systematic reviews, and current clinical guidelines evaluating when and how to safely discontinue long-term PPI therapy — and when to continue.
This is an original evidence synthesis. See References for source studies.
One-Minute Takeaway
One-Minute Takeaway
- Evidence from 25 RCTs and 32 pragmatic trials supports deprescribing proton pump inhibitors in patients without ongoing high-risk indications — successful discontinuation or dose reduction is achievable in the majority, with on-demand therapy controlling symptoms in most who experience recurrence (Rossi et al. 2024 systematic review).
- Long-term PPI harms remain predominantly observational: a large RCT of 17,598 patients (pantoprazole vs placebo over 3 years) found a statistically significant increased risk of enteric infections (OR 1.33) but no significant increase in CKD, dementia, fractures, or cardiovascular events.
- Rebound acid hypersecretion is a real physiological phenomenon that causes transient symptom recurrence after PPI withdrawal — gradual tapering over 2–4 weeks with step-down to on-demand therapy may mitigate this, though no single tapering protocol has demonstrated clear superiority.
- Definite indications for continued long-term PPI use include Barrett’s oesophagus, severe (Los Angeles grade C/D) oesophagitis, Zollinger-Ellison syndrome, and gastroprotection in patients on chronic NSAIDs with GI risk factors — these patients should not be deprescribed. Through 2025.
Why Deprescribing PPIs Matters
Proton pump inhibitors are among the most widely prescribed medications globally. While highly effective for acid-related disorders, a substantial proportion of long-term users lack a continuing indication. A 2024 systematic review and meta-analysis estimated that inappropriate PPI use affects up to 50–70% of users in various clinical settings. This represents a significant burden of unnecessary medication — contributing to polypharmacy, drug interactions, potential harm, and avoidable healthcare cost.
The clinical challenge is nuanced: while observational studies have linked chronic PPI use to a range of adverse outcomes, RCT data have largely failed to confirm causality. Simultaneously, abrupt PPI cessation can trigger rebound acid hypersecretion, creating a cycle of symptom recurrence that reinforces continued use. Effective deprescribing requires understanding who should stop, who should continue, and how to manage the transition.
What Evidence Was Reviewed
This synthesis draws on the 2024 Rossi et al. systematic review (66 studies), the 2022 AGA Clinical Practice Update on PPI deprescribing, the 2017 Canadian GRADE-based deprescribing guideline, the Windsor Workshop consensus, and the primary RCT and observational data informing these documents. The evidence spans deprescribing strategies, long-term PPI harms, rebound acid hypersecretion, and patient-level factors that influence deprescribing success.
| Source | Design | n | Population | Key Finding | Quality |
|---|---|---|---|---|---|
| Rossi et al. 2024 SR | Systematic review (66 studies) | NR | Adult PPI users across settings | 25 RCTs + 32 pragmatic trials; collaborative multi-professional approaches most effective; on-demand therapy controls symptoms in most | High |
| AGA Clinical Practice Update 2022 | Expert review / BPA | NR | Ambulatory adult PPI users | 10 Best Practice Advice statements; indications for continuation vs deprescribing clearly stratified | High |
| Farrell et al. 2017 (Canadian guideline) | GRADE-based guideline + SR | NR | Adults on long-term PPIs | Recommends deprescribing in absence of high-risk indication; tapering or on-demand over abrupt stop | High |
| Windsor Workshop 2024 | Consensus + evidence review | NR | Primary care + gastroenterology | Algorithm for rational PPI use-reduction; dose step-down before cessation; lifestyle first | Moderate |
| Moayyedi et al. (Cochrane) 2017 | Meta-analysis of RCTs | 3,408 | GERD patients on maintenance PPI | On-demand PPI effective for symptom control; dose reduction led to small non-significant symptom increase | High |
| Reimer et al. 2009 (RCT) | Double-blind RCT | 120 | Healthy volunteers given PPI then withdrawn | PPI withdrawal after 8 weeks caused rebound acid-related symptoms in previously asymptomatic subjects | High |
| Moayyedi et al. 2019 (COMPASS pantoprazole RCT) | RCT | 17,598 | Adults ≥65, stable CVD, on antithrombotics | Enteric infections OR 1.33 (significant); no significant CKD, dementia, fractures, CV events at 3.1 years | High |
| Xie et al. 2024 (Nature Comms) | Prospective cohort (5 cohorts) | >2,000,000 | US, UK, and Chinese populations | PPI use associated with 15 diseases; dose-response observed; 82% of excess risk concentrated in top 40% baseline risk | Moderate |
Key Findings: Evidence for Deprescribing Proton Pump Inhibitors
Who Should Continue Long-Term PPI Therapy
The AGA 2022 Clinical Practice Update clearly delineates definite indications for continued PPI use. These include Barrett’s oesophagus (regardless of symptoms), severe erosive oesophagitis (Los Angeles grade C or D), eosinophilic oesophagitis responsive to PPI, Zollinger-Ellison syndrome, and gastroprotection for patients on chronic NSAIDs who carry GI bleeding risk factors (age ≥65, history of peptic ulcer, concomitant anticoagulant or corticosteroid use). Patients taking dual antiplatelet therapy following coronary stenting who have additional GI risk factors also warrant continuation. These patients should not be deprescribed, as the evidence clearly favours ongoing therapy.
Who Should Be Considered for Deprescribing
The largest candidate group for deprescribing comprises patients initiated on PPIs for short-term indications — uncomplicated GERD with symptom resolution, empiric dyspepsia, ICU stress ulcer prophylaxis continued post-discharge, or uncomplicated peptic ulcer disease after H. pylori eradication — where treatment has been extended without re-evaluation. The AGA update notes that PPIs empirically prescribed for laryngopharyngeal symptoms (cough, throat clearing, globus, hoarseness) should be deprescribed, as RCT evidence has definitively demonstrated that PPIs offer no benefit for these symptoms.
Deprescribing Strategies: What Works
The 2024 Rossi et al. systematic review identified collaborative, multi-professional interventions combining clinical decision algorithms with patient education as the most effective deprescribing approaches. Successful strategies typically involved a pharmacist or nurse-led review using standardised criteria, followed by physician-approved dose reduction or discontinuation with structured follow-up. Across the 66 included studies, several approaches to medication reduction were tested — including abrupt cessation, gradual tapering, dose step-down (e.g., twice-daily to once-daily, then to half-dose), and switching to on-demand therapy.
Rebound acid hypersecretion (RAHS) is a well-documented physiological consequence of PPI withdrawal. During PPI therapy, profound acid suppression triggers compensatory hypergastrinaemia, which drives parietal cell hyperplasia and expansion of the acid-producing capacity of the stomach. When the PPI is stopped, this expanded acid-producing machinery is no longer suppressed, resulting in transient acid overproduction that exceeds pre-treatment levels. A landmark double-blind RCT by Reimer et al. (2009) demonstrated that even previously asymptomatic healthy volunteers developed dyspepsia, heartburn, and acid regurgitation after withdrawal from an 8-week PPI course — providing direct evidence that RAHS creates de novo symptoms.
The clinical significance is that patients may attribute rebound symptoms to recurrence of their original condition and reinitiate PPI therapy, perpetuating unnecessary use. Patient education about the transient nature of rebound symptoms (typically resolving within 2–4 weeks) is essential. Bridging with antacids, alginates, or short H2 receptor antagonist courses may help manage the transition. Gradual tapering may theoretically allow progressive normalisation of gastrin levels, though head-to-head data comparing tapering with abrupt cessation remain limited.
The evidence regarding long-term PPI harms requires careful stratification by study design. Observational studies and meta-analyses have reported associations between chronic PPI use and CKD (RR ~1.72), fractures (RR ~1.20), C. difficile infection, community-acquired pneumonia, hypomagnesaemia, vitamin B12 deficiency, and dementia. However, these associations are drawn almost exclusively from retrospective data vulnerable to confounding, protopathic bias, and residual indication bias.
The COMPASS RCT (2019) — the largest randomised safety assessment of PPIs — enrolled 17,598 patients aged ≥65 and found no significant increase in CKD (OR 1.17, 95% CI 0.94–1.45), fractures, pneumonia, or dementia over 3.1 years of pantoprazole vs placebo. Enteric infections were significantly increased (OR 1.33). A 2023 prospective cohort study and a 2024 Mendelian randomisation analysis found no causal link between PPI use and dementia. These findings suggest that while some harms are real (enteric infections, hypomagnesaemia), many widely cited associations may be confounded. The absence of RCT-confirmed causality should temper alarm but does not eliminate the value of deprescribing where PPIs are not indicated.
Older adults are disproportionately affected by inappropriate long-term PPI use and simultaneously bear the greatest theoretical risk from potential harms (fractures, CKD, infections, drug interactions). A 2023 observational study in multimorbid older adults found high rates of potentially inappropriate PPI continuation. The Canadian deprescribing guideline specifically targets this population, noting that pill burden reduction and decreased drug-interaction risk are independently valuable goals. However, deprescribing in older adults requires sensitivity to communication barriers, cognitive impairment limiting self-directed symptom management, and the need for simplified on-demand rescue strategies. Structured pharmacist-led interventions with carer involvement have shown particular promise in this population.
Despite the evidence supporting deprescribing, implementation remains inconsistent. The 2024 systematic review identified several recurring barriers: clinical inertia (physicians not re-evaluating established prescriptions), patient reluctance driven by fear of symptom recurrence, unclear documentation of original PPI indication, fragmented care between specialists and primary care, and insufficient systems for structured follow-up after deprescribing. Hospital-to-community transitions represent a particularly vulnerable point — PPIs initiated for ICU stress ulcer prophylaxis are frequently continued at discharge without explicit indication documentation. Interventions that address multiple barriers simultaneously — combining decision support algorithms, patient education, and follow-up reminders — appear most effective at closing the gap between evidence and practice.
Quality & Consistency of Evidence
What the Evidence Does NOT Show
Several clinically important questions about PPI deprescribing remain inadequately addressed by current evidence.
Optimal tapering protocol: No RCT has directly compared specific tapering schedules (e.g., 4-week vs 8-week taper, half-dose step-down vs alternate-day dosing) in a head-to-head design. Current recommendations are based on physiological reasoning and indirect evidence rather than definitive trial data.
Long-term outcomes of deprescribing: Most deprescribing trials report outcomes at 6–12 months. Whether successful deprescribers remain off PPIs at 2–5 years, and whether deprescribing translates into reduced long-term harms, has not been established.
PPI harms beyond 3 years: The COMPASS RCT had a median follow-up of only 3.1 years. Harms such as CKD progression, bone loss, and cancer risk may require longer exposure periods to manifest. Longer-term RCT safety data are unlikely to become available due to ethical and practical constraints.
Deprescribing in complex patients: Nearly all deprescribing research has been conducted in relatively healthy patients with mild-to-moderate GERD. Evidence for deprescribing strategies in patients with multiple comorbidities, cognitive impairment, or complex polypharmacy remains limited.
Practical Implications
Based on the current evidence, the following considerations emerge for clinicians reviewing long-term PPI prescriptions.
Both the AGA and NICE recommend annual reassessment of PPI need. The available evidence indicates that the majority of long-term users can safely reduce or stop PPI therapy when no high-risk indication persists. Documenting the original indication — or its absence — is the essential first step.
The available evidence supports a stepwise approach: reduce to the lowest dose over 2–4 weeks, then switch to on-demand use. Patient education about rebound symptoms — that they are transient, typically lasting 2–4 weeks, and do not indicate disease recurrence — is associated with higher deprescribing success rates.
Deprescribing should not be applied indiscriminately. Barrett’s oesophagus, severe erosive oesophagitis (LA grade C/D), Zollinger-Ellison syndrome, and gastroprotection for high-risk NSAID users require ongoing therapy. Inappropriate deprescribing in these patients carries real clinical risk.
The 2024 systematic review found that interventions involving pharmacists, clinical algorithms, and patient engagement tools achieved higher deprescribing rates than physician-only interventions. Integrating PPI review into medication reconciliation workflows, particularly at hospital discharge, addresses one of the highest-yield transition points.
Evidence Grade & Bottom Line
Evidence Grade: Moderate
The feasibility and safety of PPI deprescribing in appropriate candidates is well-supported by multiple RCTs and systematic reviews. However, the quality of evidence for specific tapering protocols is low, and the long-term harm-reduction benefit of deprescribing — while biologically plausible — has not been demonstrated in RCTs. This grade reflects evidence that may change with further study, particularly regarding optimal tapering strategies and long-term outcomes.
A “Moderate” evidence grade means that deprescribing is supported for patients without ongoing high-risk indications, but clinicians should use clinical judgement when selecting a tapering approach and should provide patients with a symptom management strategy. The evidence does not support deprescribing driven solely by concern over unproven harms when a clear indication for continued PPI use exists.
- The majority of long-term PPI users without a high-risk indication can successfully reduce or stop therapy with structured deprescribing
- Rebound acid hypersecretion is a real phenomenon that causes transient symptoms — patient education and gradual tapering mitigate its impact
- On-demand PPI therapy provides adequate symptom control for most patients after deprescribing
- RCT data do not confirm most observational PPI harm associations (except enteric infections), but the absence of indication remains the strongest reason to deprescribe
- Barrett’s oesophagus, severe oesophagitis, Zollinger-Ellison syndrome, and NSAID gastroprotection in high-risk patients are definite indications for continued therapy — deprescribing should not override these
Article Information
For Educational Purposes Only. This is an original evidence synthesis informed by the studies listed below. It does not replace clinical judgement. Drug dosages should be verified against current prescribing information.
References
- Rossi A, Perrella L, Scotti S, et al. Approaches to deprescribing proton pump inhibitors in clinical practice: a systematic review. J Clin Med. 2024;13(20):6283. DOI: 10.3390/jcm13206283
- Targownik LE, Fisher DA, Saini SD. AGA Clinical Practice Update on de-prescribing of proton pump inhibitors: expert review. Gastroenterology. 2022;162(4):1334-1342. DOI: 10.1053/j.gastro.2021.12.258
- Farrell B, Pottie K, Thompson W, et al. Deprescribing proton pump inhibitors: evidence-based clinical practice guideline. Can Fam Physician. 2017;63(5):354-364. PMID: 28500192
- Hungin APS, Scarpignato C, Keenan J, et al. Proton pump inhibitors: rational use and use-reduction — the Windsor Workshop. Digestion. 2024;105(4):285-299. DOI: 10.1159/000538962
- Reimer C, Søndergaard B, Hilsted L, Bytzer P. Proton-pump inhibitor therapy induces acid-related symptoms in healthy volunteers after withdrawal of therapy. Gastroenterology. 2009;137(1):80-87. DOI: 10.1053/j.gastro.2009.03.058
- Moayyedi P, Eikelboom JW, Bosch J, et al. Safety of proton pump inhibitors based on a large, multi-year, randomized trial of patients receiving rivaroxaban or aspirin. Gastroenterology. 2019;157(3):682-691. DOI: 10.1053/j.gastro.2019.05.056
- Xie Y, Choi T, Al-Aly Z. Individualized prevention of proton pump inhibitor related adverse events by risk stratification. Nat Commun. 2024;15:3584. DOI: 10.1038/s41467-024-48007-8
- Dutta AK, Sharma V, Jain A, et al. Inappropriate use of proton pump inhibitors in clinical practice globally: a systematic review and meta-analysis. Gut. 2024;74:e5. DOI: 10.1136/gutjnl-2024-332154
- Jankovic K, Gralnek IM, Awadie H. Emerging long-term risks of the use of proton pump inhibitors and potassium-competitive acid blockers. Annu Rev Med. 2025;76:143-153. DOI: 10.1146/annurev-med-050223-112834
- Chaudhry M, Ali U, Bukhari SHA, et al. Long-term proton pump inhibitor use and the risk of kidney disease, dementia, and fractures: a systematic review. Cureus. 2025;17(8):e90627. DOI: 10.7759/cureus.90627
- Nochaiwong S, Ruengorn C, Awiphan R, et al. The association between proton pump inhibitor use and the risk of adverse kidney outcomes: a systematic review and meta-analysis. Nephrol Dial Transplant. 2018;33(2):331-342. DOI: 10.1093/ndt/gfw470
- Dutta AK, Jain A, Jearth V, et al. Guidelines on optimizing the use of proton pump inhibitors: PPI stewardship. Indian J Gastroenterol. 2023;42:601-628. DOI: 10.1007/s12664-023-01428-7