Proton Pump Inhibitors
Conditions
Keywords
proton pump inhibitors, program evaluation, low-value care, de-implementation, patient safety
Brief summary
Proton pump inhibitors (PPIs) are medications used to treat acid-related stomach disorders, such as chronic heartburn. These medications are widely used by Veterans, with over 11 million 30-day prescriptions being filled each year. Though they are highly effective, long-term use of PPIs may be harmful. For this reason, experts recommend that PPIs be stopped in patients who do not have a clear need for these medications. Unfortunately, PPIs continue to be overused. To address this issue, the VA is implementing a national program to de-prescribe (i.e., reduce the dose of, or stop) PPIs. In this study, the investigators will be evaluating this national program by assessing: (a) how successfully the program was implemented; (b) understanding how effective the program was in improving appropriate use of PPIs; and, (c) ensuring no unintended consequences (such as peptic ulcer bleeding) occurred with PPI de-prescribing. This study addresses a potential safety concern for Veterans and aligns with VA's broader goal of de-implementing low-value care.
Detailed description
Background: Proton pump inhibitors (PPIs) are among the most commonly prescribed medications in the Veterans Health Administration (VHA), accounting for over 11 million 30-day prescriptions and nearly $50 million in medication costs annually. Though effective for treatment of acid-related disorders such as gastroesophageal reflux disease, PPIs have been associated with a number of potential harms in observational studies (e.g., dementia, chronic kidney disease, fractures), and increased mortality in Veterans. Nonetheless, PPIs continue to be used without an appropriate indication or for longer and at higher doses than necessary. Accordingly, VHA Pharmacy Benefits Management Services (PBM) will deploy RaPPID - a national Randomized PPI De-prescribing program - in Fiscal Year 2018 targeting patients for whom a short course of PPI is likely sufficient. This program will comprise activation of Clinical Pharmacy Specialists, provider education and academic detailing, and patient education. In partnership with PBM, the investigators propose to conduct an evaluation of this national program in a cluster-randomized design. Objectives: (1) assess the impact of the de-prescribing program on important clinical outcomes, and to understand how and why these outcomes were achieved or not achieved (outcomes and process evaluation); (2) assess the economic effects of the de-prescribing program (economic evaluation). Methods: The investigators will then assess the impact of RaPPID on PPI use (primary outcome) in a cluster randomized design (cluster = Veterans Integrated Service Network (VISN). The investigators will also assess a variety of unintended effects, including impact of reduced PPI use on upper GI symptoms and complications such as upper GI bleeding. Furthermore, the investigators will use process evaluation approaches to understand why and how the program was effective or ineffective in specific contexts. Finally, the investigators will use data from the outcomes evaluation of this proposal to estimate the budget impact of RaPPID, taking into account the impact of the program on VHA and non-VHA healthcare utilization. Impact: RaPPID will be among the largest concerted efforts at de-prescribing ever undertaken in VHA. Prospective evaluation of the program therefore presents a unique opportunity not only to enhance the program itself, but also to gain insights about how to reduce the use of low-value services more broadly, a key VHA priority for the coming decade. Importantly, the prospective, controlled study design the investigators propose will also allow us to make strong claims about whether PPIs cause the putative adverse effects to which they have been linked. Ultimately, this evaluation will provide not only valuable insight into the benefits and harms of a national effort to appropriately de-prescribe PPIs, but also broader lessons about how to effectively undertake other such interventions to de-implement entrenched clinical practices in the future.
Interventions
The PPI de-prescribing program includes alerts to clinical pharmacy specialists and primary care providers informing them of individual patients scheduled for upcoming primary care visits who meet criteria for PPI de-prescription; activation of clinical pharmacy specialists; education of primary care providers; and patient education.
Sponsors
Study design
Intervention model description
A cluster-randomized (by Veterans Integrated Service Networks or VISN) pragmatic trial of the provider-centered intervention versus pragmatic control.
Eligibility
Inclusion criteria
Chronic PPI users defined as 90-day prescription during the 120-day period prior to a scheduled VA primary care visit who receive: 1. Once-daily PPI with * No clear indication for PPIs, OR * Uncomplicated Gastroesophageal reflux disease (GERD) OR 2. Twice-daily PPI for any indication except Zollinger-Ellison
Exclusion criteria
Patients taking once-daily PPIs will be excluded if they have one or more of the following characteristics: * Eosinophilic esophagitis * Esophagitis * Esophageal ulcer * Esophageal stenosis/stricture * Dysphagia (other than oropharyngeal) * Barrett's esophagus * Peptic ulcer * Zollinger-Ellison * Idiopathic pulmonary fibrosis * NSAID + age \> 65 yrs, 2nd NSAID, aspirin, anti-thrombotic, OR corticosteroid * Aspirin + age 60 yrs, NSAID, anti-thrombotic, OR corticosteroid * Pancreatic enzyme replacement
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| PPI Prescribing (H1) | 12 months | The proportion of days proton pump inhibitors are prescribed in the 12 months following the index visit. |
Countries
United States
Participant flow
Recruitment details
This was a national Veterans Health Administration (VHA) program. Randomization was at the level of geographical regions called Veterans Integrated Service Networks (VISNs). At the date of randomization (8/27/2019), there were 18 VISNs. One VISN was excluded because it was the model for the national program. Out of 17 VISNs, 8 were randomized to the control arm and 9 to the intervention arm. Subjects were identified via an electronic algorithm during the recruitment year 9/16/2019 - 9/15/2020.
Participants by arm
| Arm | Count |
|---|---|
| De-prescribing Program The 9 Veterans Integrated Service Networks (VISNs) randomly assigned to the PPI de-prescribing program. VISNs are the 18 geographical regions that make up the VHA.
PPI De-prescribing Program: The PPI de-prescribing program included alerts to clinical pharmacy specialists and primary care providers informing them of individual patients scheduled for upcoming primary care visits who meet criteria for PPI de-prescription; activation of clinical pharmacy specialists; education of primary care providers; and patient education. | 119 |
| No De-prescribing Program The 8 Veterans Integrated Service Networks (VISNs) randomly assigned to usual care and will not receive the national de-prescribing program. VISNs are the 18 geographical regions that make up the VHA. | 100 |
| Total | 219 |
Baseline characteristics
| Characteristic | De-prescribing Program | Total | No De-prescribing Program |
|---|---|---|---|
| Age, Continuous | 64.74 years STANDARD_DEVIATION 13 | 64.69 years STANDARD_DEVIATION 13.09 | 64.63 years STANDARD_DEVIATION 13.19 |
| Charlson Comorbidity Index (CCI) Charlson Comorbidity Index = 0 | 44,970 Participants | 82680 Participants | 37,710 Participants |
| Charlson Comorbidity Index (CCI) Charlson Comorbidity Index = 1 | 30,391 Participants | 55172 Participants | 24,781 Participants |
| Charlson Comorbidity Index (CCI) Charlson Comorbidity Index = 3 | 11,982 Participants | 22294 Participants | 10,312 Participants |
| Charlson Comorbidity Index (CCI) Charlson Comorbidity Index = 4 | 6,329 Participants | 11741 Participants | 5,412 Participants |
| Charlson Comorbidity Index (CCI) Charlson Comorbidity Index >= 5 | 10,933 Participants | 20932 Participants | 9,999 Participants |
| Charlson Comorbidity Index (CCI) Charlson Comorbidity Score = 2 | 14,889 Participants | 27487 Participants | 12,598 Participants |
| Ethnicity (NIH/OMB) Hispanic or Latino | 2,624 Participants | 10381 Participants | 7,757 Participants |
| Ethnicity (NIH/OMB) Not Hispanic or Latino | 116,866 Participants | 209915 Participants | 93,049 Participants |
| Ethnicity (NIH/OMB) Unknown or Not Reported | 4 Participants | 10 Participants | 6 Participants |
| Race (NIH/OMB) American Indian or Alaska Native | 825 Participants | 1610 Participants | 785 Participants |
| Race (NIH/OMB) Asian | 716 Participants | 1263 Participants | 547 Participants |
| Race (NIH/OMB) Black or African American | 16,964 Participants | 30912 Participants | 13,948 Participants |
| Race (NIH/OMB) More than one race | 0 Participants | 0 Participants | 0 Participants |
| Race (NIH/OMB) Native Hawaiian or Other Pacific Islander | 816 Participants | 1485 Participants | 669 Participants |
| Race (NIH/OMB) Unknown or Not Reported | 6,197 Participants | 11070 Participants | 4,873 Participants |
| Race (NIH/OMB) White | 93,976 Participants | 173966 Participants | 79,990 Participants |
| Sex: Female, Male Sex Female | 8,101 Participants | 15137 Participants | 7,036 Participants |
| Sex: Female, Male Sex Male | 111,389 Participants | 205159 Participants | 93,770 Participants |
Adverse events
| Event type | EG000 affected / at risk | EG001 affected / at risk |
|---|---|---|
| deaths Total, all-cause mortality | 0 / 0 | 0 / 0 |
| other Total, other adverse events | 0 / 0 | 0 / 0 |
| serious Total, serious adverse events | 0 / 0 | 0 / 0 |
Outcome results
PPI Prescribing (H1)
The proportion of days proton pump inhibitors are prescribed in the 12 months following the index visit.
Time frame: 12 months
Population: All participants from the 17 randomized VISNs meeting study criteria within the 1-year recruitment period.
| Arm | Measure | Value (MEAN) | Dispersion |
|---|---|---|---|
| De-prescribing Program | PPI Prescribing (H1) | 73.69 percentage of days on PPI | Standard Deviation 0.27 |
| No De-prescribing Program | PPI Prescribing (H1) | 72.66 percentage of days on PPI | Standard Deviation 0.27 |