© The Author(s) 2023
Article reuse guidelines:
sagepub.com/journals-permissionsDOI: 10.1177/17151635231203214
The authors are keen on providing patients with the best evidencebased medicine, with a focus on preventing prescribing cascades in complex patients, and feel passionate about improving patient outcomes through deprescription of unnecessary medications.
Les auteurs sont désireux de fournir aux patients la meilleure médecine factuelle en mettant l’accent sur la prévention des cascades de prescriptions chez les patients complexes, et se passionnent pour l’amélioration des résultats chez les patients grâce à la déprescription de médicaments non nécessaires.
Background: Proton pump inhibitors are considered potentially inappropriate medications in the geriatric population. The use of proton pump inhibitors by older adults has increased over the past several decades; however, existing literature suggests that only one-third of patients prescribed these medications have a valid indication for use. The aim of this study was to assess the appropriateness of proton pump inhibitor therapy in a population of ambulatory geriatric patients and to determine the impact of an interdisciplinary clinic on their use.
Methods: This was a retrospective, cross-sectional study of patients referred to an ambulatory geriatrics clinic between October 2017 and March 2021. Participants were eligible for inclusion if they were taking a proton pump inhibitor at the time of clinic admission. Appropriateness of proton pump inhibitor therapy was assessed based on indication and dosing regimen.
Results: In total, 280 participants were included. The mean age was 79.6 years (SD 7.7) and 60% were female. Overall, 94 of 280 (33.6%) participants had a documented valid indication for their proton pump inhibitor. The clinic team intervened to reduce the dose or discontinue 76 of 186 (40.8%) inappropriate proton pump inhibitors.
Interpretation: Only one-third of proton pump inhibitors used by ambulatory geriatric patients have a documented appropriate indication for use. Given the medical complexity of this population and their vulnerability to adverse effects, the ambulatory geriatric population should be a priority target for proton pump inhibitor deprescribing initiatives. Can Pharm J (Ott) 2023;156:331-336.
The use of proton pump inhibitors (PPIs) has increased worldwide over the past several decades, with a greater increase in older adults compared to younger populations.1-7 In Canada, the use of PPIs by older adults increased from 26.7% in 2011 to 29.1% in 2016, the majority of which included long-term treatment.8,9 Long-term PPI therapy, defined as treatment greater than 8 weeks in duration, is indicated for clinically significant erosive esophagitis and related conditions (i.e., Barrett’s esophagus, Zollinger-Ellison syndrome), those with complications of gastroesophageal reflux disease (GERD) (i.e., esophageal strictures) and gastroprotection for long-term users of nonsteroidal anti-inflammatory drugs (NSAIDs) who are at high risk of gastrointestinal (GI) bleeding.10,11 Short-term therapy, less than 8 weeks in duration, is recommended for uninvestigated GERD, treatment of NSAID-induced peptic ulcer disease (PUD) and Helicobacter pylori eradication.10,11 Despite the available guidance on appropriate PPI use, literature published over the past decade has consistently demonstrated that approximately twothirds of prescribed PPIs do not have a valid indication for use.4,12-17 However, much of the data on appropriateness have been derived from the broader adult population, with a limited amount of information on the appropriateness of PPI therapy in older adults or those who are frail.
Patient dissatisfaction with medication cessation, a lack of symptom control with abrupt discontinuation and the belief that PPIs are a benign class of medications are all reasons found to contribute to the increasing use of PPIs.18-20 However, the long-term use of PPIs has been linked to increased falls and fractures, Clostridioides difficile infections, pneumonia, acute interstitial nephritis, acute kidney injury and electrolyte and micronutrient deficiencies.11,21 The geriatric population is at greater risk for adverse effects of PPIs due to their inherently higher risk of falls, infections associated with age-related immunosenescence, declining kidney function and polypharmacy.22,23 As such, PPIs have been included in multiple tools for identification and management of potentially inappropriate medications in older adults, and both geriatric and gastroenterology guidelines alike recommend attempting to deprescribe PPIs in those without a valid indication for longterm use.10,24-27
The purpose of this study was to assess the appropriateness of PPIs prescribed to a population of ambulatory geriatric patients and determine the impact of an interdisciplinary health care team on their use. The primary objective was to determine the proportion of ambulatory older adults taking a PPI that had a valid, documented indication for use and appropriate dosing regimen. We hypothesized that PPI appropriateness in this population would be lower than the approximate one-third seen in the general adult population.4,13-17 Secondary objectives were to characterize the population of older adults prescribed PPIs; determine if age, sex, polypharmacy, osteoporosis or frailty affected the inappropriate use of PPIs; and assess the proportion of patients taking inappropriate PPIs that underwent deprescribing at an interdisciplinary geriatric clinic.
This was a retrospective, cross-sectional analysis using electronic medical records (EMRs) of individuals referred to an ambulatory clinic for older adults living with frailty or geriatric syndromes in British Columbia. The clinic consists of an interdisciplinary team of geriatricians, family physicians, clinical pharmacists, nurses and other health care professionals. It provides short-term (3–6 months) assessment and intervention to older adults living independently or in assisted living facilities but does not provide care to residents of institutional or long-term care facilities.
Potential participants were identified from a preexisting database of all patients referred to the study clinic. Individuals were eligible for study inclusion if they received care at the study clinic between October 4, 2017, and March 11, 2021, and had documented use of a PPI on a pharmacist-obtained best possible medication history (BPMH) upon clinic intake. For patients referred to the study clinic more than once during the study period, only their first admission was eligible for inclusion.
The principal investigator (RD) extracted data from Profile (Intrahealth Canada Ltd, North Vancouver, BC) and Meditech (Medical Information Technology, Westwood, MA), the EMRs used at the study clinic. Extracted data were entered into the electronic data capture tool REDCap (Research Electronic Data Capture; REDCap Consortium, Nashville, TN). Information extracted from the EMRs included participant demographics, medical history, pharmacist-documented BPMH upon clinic admission, reason(s) for referral to the study clinic, Rockwood Clinical Frailty Score, laboratory values and PPI- related interventions by the clinic team members.
Each participant was assessed for PPI appropriateness using the following definition: “A PPI prescribed at the correct dose for 1 of the following indications: gastroesophageal reflux disease (GERD) with refractory symptoms upon discontinuation, GERD with inadequate control with a histamine-2-receptor antagonist (H2RA), erosive esophagitis, current peptic ulcer disease (PUD) treatment, recurrent PUD, or past PUD with risk factors for recurrence (nonsteroidal anti-inflammatory, anticoagulant, dual antiplatelet therapy or corticosteroid use).”27-29 Any recommendation by the interdisciplinary team to stop, taper or reduce the dose of a PPI was recorded as a deprescribing intervention.27 No deprescribing interventions (i.e., education, audit and feedback) were provided to the interdisciplinary team prior to the data collection period.
We predicted that the proportion of participants receiving a PPI appropriately would be slightly less than one-third, the approximate appropriateness rate in previous studies of the general adult population.4,13-17 Using the assumption that approximately 25% of participants would be using PPIs appropriately, we calculated a required sample size of 289 in order to achieve a precision of 5%.30
Continuous data were described with means and standard deviations (SDs), ordinal data were described with medians and interquartile ranges (IQRs) and categorical data were described as percentages. Stepwise multivariate logistic regression was used to determine the impact of age, sex, polypharmacy, osteoporosis and frailty on inappropriate PPI use. Any factor with a p-value <0.10, as well as age and sex, was included in the logistic regression model. A p-value of <0.05 was considered statistically significant. Statistical analysis was completed using IBM SPSS Statistics, Version 28 (SPSS, Inc., Chicago, IL).
This study was approved by the Interior Health Authority Research Ethics Board (ref. number 2021-22-071-H) and University of British Columbia Clinical Research Ethics Board (ref. number H21-03280) on December 8, 2021.
A total of 1971 potential participants were screened for study inclusion, of whom 978 had a pharmacist-documented BPMH on clinic admission during the study period. Of these, 280 were taking a PPI and included in the study. The mean age was 79.6 years (SD 7.7 years) and 168 (60%) were female (Table 1). Participants were taking a mean of 7.8 (SD 3.3) medications and the median Rockwood Clinical Frailty Score was 5 (IQR 4, 5). Osteoporosis or osteopenia was documented in 37.5% of participants, and 58.9% of the population had experienced a self-reported fall in the past year. The most commonly prescribed PPI was pantoprazole (59.6%), followed by rabeprazole (23.3%). Most patients had been using a PPI for over 8 weeks (61.4%), but documentation of duration was unavailable for 33.9% of participants.
PPI therapy met the criteria for appropriateness in 94 of 280 (33.6%) participants (Table 2). The most common indications for appropriate PPI therapy were refractory GERD upon discontinuation (45/280, 16.1%) and history of PUD with risk factors for recurrence (28/280, 10%). PPI therapy did not meet the criteria for appropriateness in 186 of 280 (65.4%) participants. The most common reasons for inappropriateness were GERD without an adequate documented trial of discontinuation (135/280, 48.2%), hiatus hernia without trial of discontinuation (15/280, 5.4%) and no documented indication for PPI therapy (16/280, 5.7%). Regression analysis found that only age was associated with the inappropriate use of a PPI (odds ratio, 1.042; 95% confidence interval, 1.007–1.079).
The interdisciplinary team within the study clinic intervened on 75 of 186 (40.3%) inappropriate PPIs by reducing the PPI dose for 43 of 186 (23.1%) participants, discontinuing PPI therapy in 32 of 186 (17.2%) participants and changing PPI therapy to use as needed in 1 of 186 (0.5%) participants.
This retrospective, cross-sectional review assessed the appropriateness of PPI prescribing in an ambulatory geriatric population and described the impact of an interdisciplinary health care team on inappropriate PPI use. The study population was approximately 80 years of age and living with mild frailty, as indicated by the median Rockwood Clinical Frailty Score of 5.31 Only about one-third of participants had a valid indication for PPI use, and the interdisciplinary team intervened to de-escalate approximately 40% of the inappropriate PPI prescriptions.
Our findings are consistent with previously published literature suggesting that approximately two-thirds of PPIs used in ambulatory adults are inappropriate. These findings are consistent across various age categories, community settings and study designs, which increases the confidence in our findings.12-17 With a mean age of 80 years, the present study adds to the existing literature in that it is the oldest ambulatory population published to date assessing the appropriateness of PPI therapy. Older age was associated with inappropriate PPI use in this study, which has been previously documented.4,17,32 Other previously documented risk factors for inappropriate PPI use include male sex, multimorbidity/medical complexity, number of medications and medications associated with bleeding.4,32 Our study did not find a statistically significant impact of sex or polypharmacy on inappropriate PPI use, although it may have been unpowered for these analyses.
Within the present study, the interdisciplinary team intervened to reduce or discontinue approximately 40% of inappropriate PPIs. While not extensively described in the literature, the success rate of PPI deprescribing in ambulatory patients appears to be high in individuals who are willing to attempt deescalation. Several small studies found success rates between 79% and 100% in individuals motivated to stop or reduce their PPIs.33-35 Broader systems-based interventions, such as audit and feedback, EMR notifications and provision of education and educational resources to prescribers, have been shown to reduce inappropriate PPI use by approximately 50%.36-38 The lower 40% deprescribing rate in this study may be explained by the inclusion of anyone taking a PPI, not just those who were motivated to deprescribe, and the absence of a structured intervention to the health care team. Nevertheless, this study supports previous evidence that PPI deprescribing is possible in older adults and highlights the role that an interdisciplinary team providing comprehensive geriatric care can have on reducing inappropriate PPI prescribing.
Competing health priorities and the unwillingness of patients to try deprescribing their PPI may have also contributed to the low deprescribing rate in the current study. In a scoping review, Thompson et al.19 determined that 40% of patients taking a PPI would be willing to reduce their PPI use if their physician believed it was “a good idea.” Patients reported symptom recurrence as their greatest concern, with approximately 68% of patients stating they would not even tolerate minor symptoms returning.19 Patients may also be unaware of the expected duration of PPI use. From the same scoping review, approximately 60% of PPI users did not recall receiving any information on the expected duration of PPI use, while 80% to 90% of those using PPIs for GERD or upper GI symptoms believe that their PPI was required long term.19 This suggests an opportunity for health care providers to enhance patient education on desired outcomes, duration of use and risks associated with PPIs.
A large proportion of participants in this study had a medical history including at least 1 risk factor that might increase the risk of PPI-associated adverse events, including falls, osteoporosis or past infections. Considering the high proportion of PPIs that were deemed inappropriate in this study, patients with these associated comorbidities may be prime targets for reassessment of therapy and discontinuation initiatives. Future research should focus on increasing awareness of PPI adverse effects, appropriate durations of therapy and the effectiveness of deprescribing strategies across various populations.
This study did have several methodological challenges. As with any retrospective chart review, undocumented or ambiguously documented information may have contributed to uncertainty of results. Some data, such as participant comorbidities or previous trials of PPI deprescribing, may have been undocumented in the medical record, and as a result, some PPIs may have been incorrectly recorded as inappropriate when an undocumented valid indication existed. This may have been the case for duration of PPI therapy, for which data were missing for one-third of participants. However, given that only 12% of geriatric PPI users in our province use therapy for less than 90 days, we can assume that the majority of undocumented PPI durations were likely over 8 weeks in duration.9 Our definition of PPI appropriateness was broad, including several indications that are not universally considered appropriate, such as refractory GERD upon PPI discontinuation or use of a corticosteroid or dual antiplatelet therapy in individuals with a history of PUD.10 As such, we may have overestimated the appropriateness of PPI therapy in this population. We did underestimate the appropriateness of PPIs in our sample size calculation and did not achieve our required minimal sample size of 289 participants. Despite this, a post hoc power assessment found that we still achieved a precision of 6% with our sample size of 280 participants around our point estimate of 33.4%.30 Finally, this study captured data from urban ambulatory geriatric patients with mild frailty and may not be applicable to other populations, such as younger adults, those who with moderate to severe frailty or those who are hospitalized or residing in long-term care facilities.
Approximately two-thirds of PPIs lacked a valid documented indication for use in this sample of ambulatory geriatric patients. The interdisciplinary health care team within the study clinic intervened to reduce or discontinue PPI therapy in approximately 40% of inappropriate PPI users. Data from this study highlight an opportunity for structured interventions to enhance deprescribing initiatives and the need for further research to determine the most effective strategies.
From the Interior Health Authority, Department of Pharmacy Services (Dixon, Bolt), Kelowna, British Columbia; and the University of British Columbia Faculty of Pharmaceutical Sciences (Bolt), Kelowna, British Columbia. Contact Jennifer.bolt@ubc.ca.
Author Contributions: Both authors contributed to the study conception and design. Data collection was performed by R. Dixon and data analysis was performed by J. Bolt. The first draft of the manuscript was written by R. Dixon. Both authors commented on previous versions of the manuscript. Both authors read and approved the final manuscript.
Funding: The authors declare that no funds, grants or other support were received during the preparation of this manuscript.
Industry sponsorship: The authors declare that no industry sponsorship was received during the preparation of this manuscript.
Conflicts of interest: The authors have no financial or nonfinancial conflicts of interests to disclose.
ORCID iD: Jennifer Bolt https://orcid.org/0000-0001-7597-8036
