Proton-pump inhibitors (PPI)
| Type | Recommendation |
|---|---|
| When to deprescribe | |
| CBR |
Given the potential adverse effects associated with prolonged use, we suggest deprescribing be offered to older people taking PPIs:
|
| Ongoing treatment | |
| CBR |
In people with complicated gastrointestinal pathologies (e.g. Barrett’s oesophagus, severe erosive disease) or those with a high risk of gastrointestinal complications using PPI for gastroprotection, we suggest continuing PPI therapy on the lowest effective dose according to the clinical indication, individual tolerance, and responses, provided this aligns with the individual's goals and preferences, following informed consent. |
| CBR |
If deprescribing is unsuccessful despite multiple attempts, taking into account the possibility of rebound acid hypersecretion (occurred typically within four to eight weeks after PPI discontinuation), we suggest maintaining the person on the lowest effective dose; however, reassessing the need for long-term therapy periodically. |
| How to deprescribe | |
| CBR |
We suggest individualising the tapering schedule and adjusting it according to the individual's response. In general, we suggest reducing the dose by 50% every two weeks (noting that enteric-coated formulations should not be broken*), ensuring individuals remain symptom-free before initiating each tapering. Once half the lowest standard dose formulation is reached, we suggest switching to alternate day dosing or discontinuing PPI therapy completely and switching to on-demand or intermittent use of PPIs, antacids, alginates, or H2 receptor antagonists (H2RAs) at the lowest effective dose. We suggest a slower tapering for people taking a high dose (e.g. 20mg omeprazole twice daily) prior to deprescribing to minimise rebound acid hypersecretion. If symptoms recur during tapering, we suggest restarting PPIs at the previously tolerated dose until symptoms resolve or using an antacid, alginate, or H2RAs as a “rescue therapy” for occasional symptoms, delaying further dose reductions by an agreed interval for stabilisation, and planning for a more gradual taper. * Marketed PPI dose forms that may be simpler to titrate include dispersible enteric tablets (omeprazole and esomeprazole), orally disintegrating tablets (lansoprazole), oral liquids (omeprazole) and granules (pantoprazole). |
| CBR |
For people on combination therapy of PPI and either an antacid or H2RA, we suggest deprescribing one at a time, prioritising antacids and then H2RAs as these are typically associated with a lower risk of harm when discontinued compared to PPIs and can be used as “rescue therapy” for occasional symptoms while tapering PPIs. Tapering of H2RAs can generally follow the same approach as PPI tapering (i.e. halving the dose every two weeks); however, we suggest individualising the tapering schedule and adjusting it according to the individual's response. Antacids and alginates typically do not require tapering unless used regularly and following patient preference. The dose for concomitant acid suppressants may also need to be adjusted temporarily to compensate for the lower dose of the other agent. |
| GPS |
Healthcare providers should offer education on diet and lifestyle modifications (or referral to other relevant healthcare providers) and clearly differentiate between symptom recurrence or exacerbation due to lifestyle factors and adverse drug withdrawal events (ungraded good practice statement). |
| Monitoring | |
| CBR |
We suggest closely monitoring for disease exacerbations (e.g. symptom recurrence) or adverse drug withdrawal events (e.g. rebound acid hypersecretion), every two weeks following each dose adjustment until at least eight weeks after the medicine has fully ceased, if practical. After this initial period, we suggest monthly monitoring for at least three months, followed by monitoring every six months thereafter. However, this should be tailored based on individual factors such as their preferences, responses and tolerance to deprescribing. If monitoring visits are impractical, we suggest advising people to report symptom recurrence (e.g. acid-related gastrointestinal symptoms) and/or any appearance of new symptoms as needed. |
| GPS |
In individuals with recurrent symptoms after deprescribing, healthcare providers should test for H. pylori infection and proceed with eradication if the infection is present (ungraded good practice statement). |
CBR, consensus-based recommendation; GPS, good practice statement
PPIs relieve symptoms in many conditions such as gastroesophageal reflux disease (GORD), dyspepsia, peptic ulcer disease, and hypersecretory conditions (e.g. Zollinger-Ellison syndrome), and as part of the eradication therapy for Helicobacter pylori (H. pylori) [143].
Existing studies suggest that older people are commonly prescribed PPI therapy for prolonged periods without an appropriate clinical indication [144-146]. According to a 2017 Cochrane review, approximately 25% to 70% of people are prescribed a PPI inappropriately [147]. Both overprescribing and underprescribing of PPIs were reported in older people at hospital discharge [148]. Overprescribing was found to be associated with younger age and a lower burden of depression whereas underprescribing in people requiring gastroprotection due to increased risk of bleeding is more frequent in older age and those with greater comorbidities and polypharmacy [148]. Another study using administrative health claims data from the Australian Government Department of Veterans' Affairs revealed that 31% of veterans included in the study received a PPI, suggesting there is a scope to further improve the use of PPIs among older Australians [149]. A clear plan for periodic evaluation of the possibility of reducing the dose of PPI to the lowest effective dose is required. Chronic PPI use without ongoing reassessment contributes to polypharmacy and increases the risk of drug-drug interactions and adverse events.
Refer to the narrative evidence summary, the GRADE Summary of Findings table in the guidelines, and the Technical Report for a complete presentation of the deprescribing evidence based on the GRADE framework (including other factors considered in developing the recommendations).
Although PPIs are relatively safe when used in accordance with guideline duration, long-term use of PPIs has been associated with adverse events in several observational studies, including infections (pneumonia, enteric infection), nutritional deficiencies (hypomagnesaemia, vitamin B12 deficiency, iron deficiency, hypocalcaemia), fractures, dementia, gastric cancer, cardiovascular disease (ischemic heart disease, stroke), hepatic diseases (hepatic encephalopathy) and renal diseases (acute kidney injury, chronic kidney disease) [150-155]. While more high-quality studies are required to establish causation between the adverse effects and long-term PPI use, potential safety concerns warrant careful consideration and individualised risk-benefit assessment before continuing prolonged therapy [156]. However, the decision to discontinue PPI should not be based solely on PPI-associated adverse effects as a direct causation cannot be confirmed from existing observational studies [143]. Some individuals may prefer to continue taking PPIs due to a variety of reasons stemming from personal beliefs about the necessity of treatment [157]. In this instance, healthcare providers play a critical role in assessing and understanding their beliefs through person-centred discussions to manage any potential issues with suboptimal medicine use [158].
PPIs are typically indicated for short-term use of up to eight to 12 weeks except when a complicated gastrointestinal pathology is present, or gastroprotection is required in the presence of significant risk factors for gastrointestinal bleeding [144, 145, 159].
1. Complicated gastrointestinal pathologies
People with complicated gastrointestinal pathologies were typically excluded from deprescribing trials as long-term PPI therapy is indicated. These conditions include:
- Barrett's oesophagus
- Endoscopically confirmed severe erosive disease (e.g. severe erosive oesophagitis)
-
Gastroprotection in the following situations
- Secondary prevention of complicated peptic ulcer or uncomplicated peptic ulcer with concurrent treatment with NSAIDs, antiplatelets, oral anticoagulants or corticosteroids
- People receiving dual antiplatelet therapy (DAPT) or dual/triple antithrombotic therapy
- People receiving single antiplatelet therapy, either with a history of peptic ulcer disease, concomitant treatment with NSAIDs or steroids, or two of the following: 65 years or older, gastrointestinal reflux symptoms, or dyspepsia symptoms
- People receiving single anticoagulant therapy with at least one of the following risk factors: 75 years or older, history of peptic ulcer disease, or concomitant use of NSAIDs
- People with an increased risk of gastrointestinal bleeding who are receiving concomitant treatment with NSAIDs, corticosteroids or an SSRI in combination with an NSAID or anticoagulant
-
Recurring, uncontrolled, or persistent symptoms in the following situations:
- Endoscopy-negative GORD
- Functional dyspepsia
- Upper airway symptoms associated with laryngopharyngeal reflux (e.g. cough, dysphonia)
2. Gastroprotection
2a) Non-steroidal anti-inflammatory drugs (NSAIDs)
The American College of Gastroenterology (ACG) provides guidance on risk stratification for NSAID gastrointestinal toxicity and recommendations for prevention of NSAID-related ulcer complications (see Table 6) [160]. In addition to the risk factors for NSAID-related gastrointestinal complications outlined in Table 6 below, H. pylori infection is an independent and additive risk factor that further increases the likelihood of these complications. The ACG recommends testing for H. pylori infection before initiating long-term NSAID therapy and, if the infection is present, proceeding with eradication. The need for co-therapy with a gastroprotective agent (e.g. PPI) after eradication depends on the individual's underlying gastrointestinal risk.
Table 6. Recommendations for preventing NSAID-related ulcer complications based on risk stratification for NSAID gastrointestinal toxicity and cardiovascular risk (adapted from Lanza et al., 2009 [160])
| Cardiovascular (CV) risk |
Risk factors for NSAID gastrointestinal (GI) toxicity
|
||
|---|---|---|---|
|
High GI risk History of ulcer complications or more than two of the above risk factors |
Moderate GI risk One or two of the above risk factors |
Low GI risk No risk factors |
|
| High CV risk* | Avoid NSAIDs or COX-2 inhibitors and use alternative therapy | Naproxen + PPI/misoprostol | Naproxen + PPI/misoprostol |
| Low CV risk | Alternative therapy if possible, or COX-2-specific NSAID + PPI/misoprostol | NSAID+ PPI/misoprostol | NSAID alone (the least ulcerogenic NSAID at the lowest effective dose) |
* arbitrarily defined as low-dose aspirin required
2b) Antithrombotic agents
Current guidelines provide clear recommendations for co-therapy with a PPI for gastroprotection in the following situations [161]
- Dual antiplatelet therapy (DAPT)
- Dual and triple antithrombotic therapy (i.e. combination antiplatelet and anticoagulant therapy)
-
Single antiplatelet therapy in the presence of at least one risk factor:
- History of peptic disease;
- Concomitant treatment with NSAIDs or steroids; or
- Two of the following: 65 years or older, gastrointestinal reflux symptoms, or dyspepsia symptoms.
-
Single anticoagulant therapy in the presence of at least one risk factor:
- 75 years or older;
- History of peptic ulcer disease;
- Concomitant use of NSAIDs.
2c) Corticosteroids
In people taking corticosteroids, routine use of PPI for gastroprotection is generally not required, unless in the presence of other risk factors for gastrointestinal complications such as active peptic ulcer disease or when the steroid therapy is combined with NSAIDs or anticoagulants [162].
2d) Selective serotonin reuptake inhibitors (SSRIs)
SSRIs have been associated with an increased risk of gastrointestinal bleeding, especially when combined with NSAID therapy [163]. The risks and benefits of concomitant therapy with SSRIs and NSAIDs should be reviewed periodically and communicated to people as appropriate. Gastroprotection can be considered if there is a clear indication to continue NSAIDs or SSRIs in people with a higher risk of gastrointestinal bleeding.
We identified 12 before-and-after studies related to PPI deprescribing from the systematic review and meta-analysis [164-175].
Overall, these studies suggest that deprescribing PPIs can be achieved but most studies did not report clinically meaningful outcomes (see Table 7). The current evidence on health-related outcomes is derived from only two single-arm studies of small sample sizes and evidence is very low in certainty. The evidence at this stage is insufficient to inform evidence-based recommendations.
Key study characteristics and results
A narrative summary of each study is provided below, highlighting key characteristics and main findings.
Lee 2017 recruited nursing home participants taking either pantoprazole or esomeprazole (for longer than six months) without an ongoing indication for long-term use and successfully discontinued PPIs in 19/27 (70%) of participants eight weeks after the intervention [167].
McDonald 2015 implemented an educational initiative paired with a web-based quality improvement tool to reduce inappropriate PPI discharge prescriptions in a hospital. The appropriateness of PPIs and suitability for trial withdrawal was determined based on a list of consensus indications. These indications were:
- Gastric or duodenal ulcer within the past three months;
- Pathological hypersecretory conditions;
- GORD within the last three months not responsive to H2 receptor antagonists (H2RAs) and non-pharmacological strategies;
- Erosive esophagitis;
- Recurring symptoms recently associated with severe indigestion within the last three months not responsive to H2RAs or non-pharmacological strategies;
- H. pylori eradication;
- Dual antiplatelet therapy;
- Antiplatelet therapy with anticoagulants;
- Antiplatelet or anticoagulant therapy with a history of previously complicated ulcer; and
- Antiplatelet or NSAID with two of the following: concomitant systematic corticosteroids, age > 60, previously complicated ulcer, concomitant NSAID, or antiplatelet/anticoagulant.
The study reported that 17/18 (94%) patients who had their inappropriate PPI therapy deprescribed during a hospital admission remained off treatment at three months follow-up, with one patient restarting the PPI due to reflux symptoms [170].
Reeve 2015 recruited six people from a hospital outpatient clinic who were taking PPIs with complex polypharmacy and consented to participate in a PPI deprescribing trial. PPI appropriateness was determined using study-specific PPI assessment guidelines developed via a literature review and independent assessment by a gastroenterologist. Appropriate indications included Barrett's oesophagus, secondary prevention of peptic ulcer disease in high-risk patients, primary prevention of peptic ulcer disease in NSAID users, and GORD with currently uncontrolled symptoms. Likely appropriate indications included PPI initiated by a gastroenterologist and primary prevention of peptic ulcer disease in high aspirin or corticosteroid users. All six participants either ceased (n=3) or reduced (n=3) their PPIs during the trial, although at six months, only four patients remained without their PPIs [171].
Wahking 2018 reported a pharmacist-led inpatient PPI stewardship program to reduce PPI use, both during hospitalisation and upon discharge. PPI appropriateness was determined using study-specific criteria for continuation developed by the hospital PPI stewardship team. Criteria for inpatient continuation of PPI include Barrett's oesophagus, upper gastrointestinal bleeding, erosive esophagitis, ulcer diagnosed in the past eight weeks, or longer but with documented persistent GORD symptoms, H pylori eradication, oesophageal strictures secondary to acid reflux, hypersecretory disorder, gastric malignancy or previous oesophageal or gastric surgery (excluding total gastrectomy), chronic kidney disease IV or Va, current diagnosis of acid-related disorder, active cancer, PPI initiated by a gastroenterologist, and previously failed attempts at deprescribing. In the study, inpatient PPI therapy was successfully discontinued in 211/220 (96%) patients. Upon discharge, among the patients who had their inpatient PPI discontinued, 24/42 (57%) maintained PPI discontinuation at three months, while 18/22 (82%) patients maintained dose reduction at three months [174].
Bhardwaj 2022 conducted a PPI deprescribing telehealth program led by student pharmacists to evaluate PPI appropriateness in veterans via remote chart reviews based on the study-specific PPI deprescribing protocol listing appropriate long-term indications. Out of the 24 veterans who consented to attempt deprescribing and lacked an appropriate indication for their long-term PPI therapy, 13/24 (54%) had their PPI discontinued and 4/24 (17%) had their dose reduced at study completion (duration was not stated) [164].
Calvo 2021 applied a deprescribing algorithm in hospitalised patients with inappropriate long-term PPI use (daily PPI use of 8 weeks or more) before admission, did not meet the criteria of the current clinical practice guidelines, and asymptomatic and reported 61/75 (81%) remaining off their PPIs at week four with 54/75 (72%) remaining off at week 24 [165].
Leszcynski 2023 reported that inappropriate PPI use significantly reduced from 84% to 44% at 12 months (p < 0.0001) after the implementation of a pharmacist-led deprescribing algorithm in a primary care geriatric ambulatory office [168]. The study determined PPI appropriateness based on a study-specific algorithm developed using components of the PPI deprescribing clinical practice guideline and information from previously published PPI deprescribing trials.
Czikk 2022 conducted a deprescribing trial informed by absolute indications from a PPI deprescribing clinical practice guideline and the Choosing Wisely guideline. Absolute indications for PPIs include:
- Erosive esophagitis
- Barrett's oesophagus
- Gastrointestinal bleeding secondary to an ulcer
- NSAID user plus one other risk factor (age > 65 years, prior ulcer, concomitant anticoagulant, antiplatelet, or prednisone)
- Antiplatelet user plus one other risk factor (history of ulcer, concomitant anticoagulant or NSAID) or two other risk factors (age > 65 years, concurrent use of prednisone, GORD)
- Dual antiplatelet therapy plus one other risk factor (age > 65, concomitant anticoagulants, prednisone or NSAIDs).
There were 29 patients with end-stage kidney disease in a haemodialysis unit of a hospital who did not have an absolute indication for a PPI. At eight weeks, 14 restarted their PPI due to reoccurrence of GORD (n = 10), gastrointestinal bleeding (n = 2, of which one case was fatal), gastrointestinal complaints (n = 1), and initiation of dual antiplatelet therapy (n = 1) [166].
Tandun 2019 conducted a pharmacist-led PPI deprescribing intervention in a long-term care facility and reported that 24/30 (80%) residents who received an active order of either pantoprazole or esomeprazole at any dose had their PPI successfully deprescribed (complete discontinuation or maintained on reduced dose) by the end of the four months study period [172]. The method of determining the appropriateness of PPI and suitability for trial withdrawal was not mentioned.
Visser 2021 targeted both statins and PPIs by applying study-specific evidence-based implicit deprescribing algorithms in nursing home residents [173]. In this study, 34/66 (52%) of the residents had their PPI and/or statin dosage either successfully reduced or discontinued after three months which were maintained at six months. Of the 31 residents who were using a PPI, 22 (71%) had their PPIs discontinued, five (16%) had their dose reduced by 50%, two (6%) continued their PPIs, two (6%) had their PPIs initially discontinued but restarted due to withdrawal effects which resolved after restarting PPI at a lower dose than they initially had.
Mati 2024 reassessed PPI use in 97 patients within the long-term care department of a geriatric hospital. All patients had been on PPIs for over eight weeks, with the mean treatment duration being four years. During the reassessment, the initial indication for PPI, use of antithrombotic agents or NSAIDs, risk of gastrointestinal bleeding, risk of infection, risk of bone fractures, history of hyponatremia and hepatic risk were recorded. The reassessment resulted in one of three outcomes: PPI continuation, dose adjustment, or gradual discontinuation. Among the 97 patients, 53 (55%) underwent gradual PPI discontinuation, three (3%) had their dose adjusted, and 41 (42%) continued PPI therapy. At the three-month follow-up, 38 of the 53 patients (72%) who discontinued PPIs remained off treatment, nine (17%) resumed PPI therapy, and six (11%) had died with causes of death not stated. Of those who restarted PPIs, six (67%) did so within the first month of discontinuation, while the remaining three (33%) resumed within three months. The primary reasons for PPI resumption were recurrent gastro-oesophageal reflux disease (GORD) with epigastric pain (n=5, 56%) and suspected peptic ulcer with acute anaemia (n=4, 44%) [175].
Linsky 2022 targeted diabetes medicines and PPIs by mailing patient-centred educational brochures to veterans two weeks prior to their scheduled primary care appointments. Targeted veterans were either taking a PPI for at least 90 consecutive days or were at an increased hypoglycaemia risk (diabetes diagnosis with a prescription for insulin or sulfonylurea, most recent HbA1c < 7%, and either aged 65 or over, had renal insufficiency, or cognitive impairment). PPI appropriateness was not determined from administrative data as the goal of the study was to promote discussion of deprescribing. Compared to a historical control group, targeted veterans (i.e. intervention group participants) were more likely to have their medicines discontinued or reduced (14% versus 4%, p = 0.009) and more likely to discuss with their healthcare providers about the target medicine (12% versus 1%, p = 0.001) [169].
Various methods were used for deprescribing in the included studies and there was no direct evidence that any particular method was associated with the greatest benefits and harms. However, compared to abrupt cessation, dose tapering is likely more acceptable and helpful in determining the lowest effective dose for some patients requiring dose reduction rather than complete cessation. There was also a lack of clear differentiation between rebound acid hypersecretion and relapse of the initial condition. Rebound acid hypersecretion may occur with abrupt discontinuation of prolonged PPI therapy due to reversal of long-term inhibition of gastric acid secretion, which can be mistaken as a need to restart PPI therapy even when it is not indicated. Rebound acid hypersecretion, if it occurs, is typically within four to eight weeks after PPI discontinuation, although in some cases it may last up to 26 weeks [176]. There are several strategies to manage rebound acid hypersecretion, such as using histamine type 2 receptor antagonists/blockers (H2 blockers) or over-the-counter antacids on demand. Another reasonable strategy is the use of PPI on-demand until symptoms are controlled. If symptoms are not controlled two months after deprescribing PPI, continuation of PPI therapy may be indicated.
Three non-controlled trials reported important or critical outcomes of very low certainty. In the study by Reeve 2015 (n=6) [171], the PPI dose was halved every two weeks, and if participants remained asymptomatic on the reduced dose, the daily dose was changed to as-required administration. In the study by Czikk 2022 (n=29) [166], PPI was withdrawn over two weeks. Finally, in the study by Mati 2024, PPI was gradually discontinued every two days for 3 weeks until the lowest possible marketed dose was reached [175].
Other studies did not report important or critical outcomes associated with deprescribing; however, the withdrawal schedules are summarised below:
- Slowly tapered according to a study-specific protocol (study=1, n=170) [164]
- Dose halved every two to four weeks until the lowest dose (study=1, n=228) [168]
- Abrupt discontinuation or dose reduction (study=1, n=220) [174]
- Abrupt cessation, gradual taper, or switching to “on-demand” dosing (study=1, n=75) [165]
- Abrupt cessation (study=1, n=28) [167]
- Individualised (studies=2, n=124) [172, 173]
- Not described (studies=2, n=500) [169, 170]