Discontinuation Categories Underlying Gaps in Dispensing for Six Medication Groups

Pharmacoepidemiology and Drug Safety, Journal Year: 2025, Volume and Issue: 34(4)

Published: April 1, 2025

ABSTRACT Purpose Accurately identifying medication discontinuations at scale is important for developing evidence about deprescribing. Gaps in dispensing often serve as proxies discontinuation but are imprecise. We categorize reasons gaps to inform data‐based methods accurately identify discontinuations. Methods Using pharmacy data, we purposively sampled from a population of adults age 65+ with 2+ chronic conditions who experienced 90‐day gap dispensing—with and without subsequent fills—of oral diabetes drugs, statins, proton pump inhibitors, drugs anticholinergic properties, anticoagulants antiplatelet or antihypertensives. reviewed clinical documentation (e.g., visit notes, communications, orders) last through the plus 120 days classify true (clinically intended) non‐discontinuations (no intent discontinue), then into subcategories. Medications no documented explanation continued listing on patient's list were classified non‐discontinuations. Results Of N = 1906 records reviewed, there 1068 (56%) 838 (44%) Subcategories within included provider discontinue, substitutions, intentional stops followed by restarts, agreeing colleague's decision discontinue. Non‐discontinuations low adherence, changes dose, formulary, drug formulation. Proportions categories subcategories varied group. Conclusion proxy measures may introduce bias misclassification, complicate causal interpretations.

Language: Английский

---

Establishing a Validation Framework of Treatment Discontinuation in Claims Data Using Natural Language Processing and Electronic Health Records

Clinical Pharmacology & Therapeutics, Journal Year: 2025, Volume and Issue: unknown

Published: April 8, 2025

Measuring medication discontinuation in claims data primarily relies on the gaps between prescription fills, but such definitions are rarely validated. This study aimed to establish a natural language processing (NLP)‐based validation framework evaluate performance of claims‐based algorithms for commonly used medications against NLP‐based reference standards from electronic health records (EHRs). A total 36,656 patients receiving antipsychotic (APMs), benzodiazepines (BZDs), warfarin, or direct oral anticoagulants (DOACs) were identified Mass General Brigham EHRs 2007–2020. These EHR linked with 97,900 Medicare Part D claims. An NLP‐aided chart review was applied determine (reference standard). In data, defined by having gap larger than 15–90 days (claims‐based algorithms). Sensitivity, specificity, and predictive values standard measured. The sensitivity specificity 90‐day‐gap‐based 0.46 0.79 haloperidol, 0.41 0.85 atypical APMs, 0.47 0.75 BZDs, 0.33 0.80 0.38 0.87 DOACs, respectively. corresponding estimates 15‐day‐gap‐based 0.68 0.55 0.59 0.62 0.71 0.45 0.61 0.49 0.58 0.64 Positive affected rates less lengths. overall accuracy differs medications. demonstrates scalability utility multiple

Language: Английский

---

Assessing causality in deprescribing studies: A focus on adverse drug events and adverse drug withdrawal events

Journal of the American Geriatrics Society, Journal Year: 2024, Volume and Issue: unknown

Published: Oct. 24, 2024

Abstract Generating real‐world evidence about the effect of medication discontinuation or dose reduction on outcomes, such as adverse drug effects (ADE; intended benefit) and occurrence withdrawal events (ADWE; unintended harm), is crucial to informing deprescribing decisions. Determining causal difficult for many reasons, including lack randomization in study designs other design measurement issues that pose threats internal validity. The inherent challenge how identify effects, both benefits harms, a new stoppage when implemented patients with potential clinical social risks may influence likelihood well outcomes. We discuss methodological estimating risk ADEs ADWEs considering: (1) sampling populations sufficient size demonstrate clinically meaningful quantifiable (2) accurate appropriately timed covariates, discontinuation, (3) statistical approaches managing confounding biases long‐term use by individuals multiple morbidities. Designing rigorous studies address validity will support generation improving ability assess harms exposure interest absence medication. Iterative learnings data quality, variable definition, measurement, exposure‐outcome associations inform strategies improve inferences possible studies.

Language: Английский

---

Defining key deprescribing measures from electronic health data: A multisite data harmonization project

Journal of the American Geriatrics Society, Journal Year: 2024, Volume and Issue: unknown

Published: Nov. 28, 2024

Abstract Background Stopping or reducing risky unneeded medications (“deprescribing”) could improve older adults' health. Electronic health data can support observational and intervention studies of deprescribing, but there are no standardized measures for key variables, healthcare systems have differing types availability. We developed definitions chronic medication use discontinuation based on electronic applied them in a case study benzodiazepines Z‐drugs five diverse US systems. Methods conducted retrospective cohort adults age 65+ from 2017 to 2019 with benzodiazepine Z‐drug use. determined whether sites had access orders and/or dispensings. using both types. Discontinuation were (1) gaps availability during follow‐up (2) not having available at fixed time point. examined the impact varying gap length requiring 30‐day period without orders/dispensings (“halo”) around compared results derived versus dispensings one site. Results Approximately 1.6%–2.6% benzodiazepine/Z‐drug (total N = 6775, ranging 431 2122 across sites). Depending definition site, proportion discontinuing 12 months ranged 6% 49%. Requiring longer “halo” resulted lower estimates. At only 56% those defined also qualified dispensings, rate 180 days was 20% 32% Conclusions ≥90 point may more accurately capture than shorter halo. Orders underestimate Work is needed adapt these other drug classes settings.

Language: Английский

---