This meta-analysis synthesizes prevalence data from studies examining inappropriate acid suppressor agent use across clinical settings and geographies. The pooled evidence consistently shows that the majority of prescriptions do not meet established guideline criteria, and that this pattern holds across patient populations, care settings, and continents — pointing to a systemic rather than isolated problem.Documentation Index
Fetch the complete documentation index at: https://mintlify.com/namakala/inappropriate-acid-suppressor-agent-use/llms.txt
Use this file to discover all available pages before exploring further.
Executive summary
The overall findings from this meta-analysis can be summarized in four key observations:- Inappropriate use is highly prevalent. Approximately seven in ten prescriptions globally were classified as inappropriate, representing a substantial and persistent burden across healthcare systems.
- Heterogeneity is extreme. I² values approaching 99% were observed across all analyses, indicating that nearly all observed variability reflects true differences across studies rather than sampling error.
- Pooled estimates are robust to publication bias. The Copas selection model produced estimates essentially identical to the unadjusted pooled prevalence, and tests for residual selection bias were generally non-significant.
- Guideline definitions drive heterogeneity. Variation in how appropriateness is defined across studies emerged as a primary source of between-study variability.
Call to actions
The findings carry direct implications across four domains:- Clinical Practice: There is a substantial and persistent overuse or misuse of acid suppressing agents across settings and regions. Clinicians should review prescribing practices against current guideline criteria, particularly at transitions of care.
- Policy: Standardized definitions of appropriateness are needed to improve comparability across studies and to guide deprescribing initiatives. Without consensus definitions, cross-study synthesis will remain methodologically limited.
- Research: Future studies should clearly define guideline criteria and report context-specific drivers of inappropriate prescribing. Transparent reporting of the guideline version applied is essential for reproducibility.
- Quality Improvement: Interventions targeting prescribing behavior are warranted, particularly in high-prevalence settings. Audit-and-feedback cycles and clinical decision support tools are candidate strategies.
Subgroup findings
Subgroup analyses were conducted for age, continent, study quality, and clinical setting. High prevalence and extreme heterogeneity persisted across all strata.Age
Pooled prevalence of inappropriate use remained high across all age strata, with estimates exceeding 80% in several subgroups. Heterogeneity was substantial within each age group, with I² values exceeding 99%. Copas modeling suggested possible small study effects in some age groups, but adjusted and unadjusted estimates were not significantly different.Continent
Pooled prevalence estimates varied across continents, reflecting geographic differences in prescribing practices and guideline implementation. However, substantial heterogeneity persisted within continental subgroups, indicating that geography alone does not fully explain the observed variability.Study quality
When stratified by methodological quality using JBI classification, prevalence estimates differed modestly between strata, but inappropriate use remained consistently high across all quality levels. Adjustment for potential selection bias did not substantially alter subgroup-specific estimates.Clinical setting
Differences in pooled prevalence were observed between inpatient and outpatient settings, with some settings demonstrating higher rates of inappropriate use. Even in the subgroup with the lowest prevalence, inappropriate prescribing remained common.Meta-regression findings
Univariable and multivariable meta-regression models examined study-level covariates including publication year, sample size, guideline use, JBI classification, setting, and continent. Although certain moderators were associated with differences in pooled prevalence, substantial residual heterogeneity remained after accounting for all included covariates. No single factor adequately accounted for the extreme variability observed across studies.These findings are from the interim report. Final prevalence estimates depend on the complete dataset and may differ from interim values reported here.
Visualizations
Forest plots, funnel plots, Doi plots, and Copas model outputs for the overall and subgroup analyses.
Quarto report
How the full PDF report is generated, its structure, and key helper functions used throughout.