Handing Off Electronic Prescription Data From Prescribers to Community Pharmacies: A Qualitative Analysis of Pharmacy Staff Perspectives.

OBJECTIVES: The aims of the study are to understand the process of how community pharmacies handle electronic prescriptions (e-prescriptions) and learn about different errors or potential errors encountered. METHODS: Fifteen remote, semistructured interviews were conducted with community pharmacy staff. Interview analysis was done with two adapted Systems Engineering Initiative for Patient Safety methods to understand the workflow and an affinity wall, which led to key words that were tallied to understand the frequency of different issues. RESULTS: Data entry in community pharmacies is a process that varies based on the different software platforms receiving e-prescriptions. Data entry of a medication product is typically a human-reliant process matching an e-prescription with an equivalent medication product. Current automated safety supports focus on matching the dispensed medication to the medication chosen at data entry. Substitutions may be required for a variety of reasons, however, pharmacists' comfort and permissions in doing so without provider involvement fluctuates. CONCLUSIONS: Prescription errors remain that could be prevented with additional support at the data entry step of e-prescriptions. Few studies demonstrate where these errors originate and what role current technology plays in contributing to or preventing these errors. Future work must consider how these matches between prescribed medications and pharmacy fulfilled medications occur. There is a need to identify potential tools to support data entry and prevent medication errors.

Designing Human-Centered AI to Prevent Medication Dispensing Errors: Focus Group Study With Pharmacists.

BACKGROUND: Medication errors, including dispensing errors, represent a substantial worldwide health risk with significant implications in terms of morbidity, mortality, and financial costs. Although pharmacists use methods like barcode scanning and double-checking for dispensing verification, these measures exhibit limitations. The application of artificial intelligence (AI) in pharmacy verification emerges as a potential solution, offering precision, rapid data analysis, and the ability to recognize medications through computer vision. For AI to be embraced, it must be designed with the end user in mind, fostering trust, clear communication, and seamless collaboration between AI and pharmacists. OBJECTIVE: This study aimed to gather pharmacists' feedback in a focus group setting to help inform the initial design of the user interface and iterative designs of the AI prototype. METHODS: A multidisciplinary research team engaged pharmacists in a 3-stage process to develop a human-centered AI system for medication dispensing verification. To design the AI model, we used a Bayesian neural network that predicts the dispensed pills' National Drug Code (NDC). Discussion scripts regarding how to design the system and feedback in focus groups were collected through audio recordings and professionally transcribed, followed by a content analysis guided by the Systems Engineering Initiative for Patient Safety and Human-Machine Teaming theoretical frameworks. RESULTS: A total of 8 pharmacists participated in 3 rounds of focus groups to identify current challenges in medication dispensing verification, brainstorm solutions, and provide feedback on our AI prototype. Participants considered several teaming scenarios, generally favoring a hybrid teaming model where the AI assists in the verification process and a pharmacist intervenes based on medication risk level and the AI's confidence level. Pharmacists highlighted the need for improving the interpretability of AI systems, such as adding stepwise checkmarks, probability scores, and details about drugs the AI model frequently confuses with the target drug. Pharmacists emphasized the need for simplicity and accessibility. They favored displaying only essential information to prevent overwhelming users with excessive data. Specific design features, such as juxtaposing pill images with their packaging for quick comparisons, were requested. Pharmacists preferred accept, reject, or unsure options. The final prototype interface included (1) checkmarks to compare pill characteristics between the AI-predicted NDC and the prescription's expected NDC, (2) a histogram showing predicted probabilities for the AI-identified NDC, (3) an image of an AI-provided "confused" pill, and (4) an NDC match status (ie, match, unmatched, or unsure). CONCLUSIONS: In partnership with pharmacists, we developed a human-centered AI prototype designed to enhance AI interpretability and foster trust. This initiative emphasized human-machine collaboration and positioned AI as an augmentative tool rather than a replacement. This study highlights the process of designing a human-centered AI for dispensing verification, emphasizing its interpretability, confidence visualization, and collaborative human-machine teaming styles.

Impact of an Area Agency on Aging pharmacist-led Community Care Transition Initiative.

BACKGROUND: Rural older adults are at risk of readmissions and medication-related problems after hospital discharge. OBJECTIVES: This study aimed to compare 30-day hospital readmissions between participants and nonparticipants and describe medication therapy problems (MTPs) and barriers to care, self-management, and social needs among participants. PRACTICE DESCRIPTION: The Michigan Region VII Area Agency on Aging (AAA) Community Care Transition Initiative (CCTI) for rural older adults after hospitalization. PRACTICE INNOVATION: Eligible AAA CCTI participants were identified by an AAA community health worker (CHW) trained as a pharmacy technician. Eligibility criteria were Medicare insurance; diagnoses at risk of readmission; length of stay, acuity of admission, comorbidities, and emergency department visits score more than 4; and discharge to home from January 2018 to December 2019. The AAA CCTI included a CHW home visit, telehealth pharmacist comprehensive medication review (CMR), and follow-up for up to 1 year. EVALUATION METHODS: A retrospective cohort study examined the primary outcomes of 30-day hospital readmissions and MTPs, categorized by the Pharmacy Quality Alliance MTP Framework. Primary care provider (PCP) visit completion, barriers to self-management, health, and social needs were collected. Descriptive statistics, Mann-Whitney U, and chi-square analyses were used. RESULTS: Of 825 eligible discharges, 477 (57.8%) enrolled in the AAA CCTI; differences in 30-day readmissions between participants and nonparticipants were not statistically significant (11.5% vs. 16.1%, P = 0.07). More than one-third of participants (34.6%) completed their PCP visit within 7 days. MTPs were identified in 76.1% of the pharmacist visits (mean MTP 2.1 [SD 1.4]). Adherence (38.2%) and safety-related (32.0%) MTPs were common. Physical health and financial issues were barriers to self-management. CONCLUSION: AAA CCTI participants did not have lower hospital readmission rates. The AAA CCTI identified and addressed barriers to self-management and MTPs in participants after the care transition home. Community-based, patient-centered strategies to improve medication use and meet rural adults' health and social needs after care transitions are warranted.

A Computable Algorithm for Medication Optimization in Heart Failure With Reduced Ejection Fraction.

Background: Guideline-directed medical therapy (GDMT) optimization can improve outcomes in heart failure with reduced ejection fraction. Objectives: The objective of this study was to determine if a novel computable algorithm appropriately recommended GDMT. Methods: Clinical trial data from the GUIDE-IT (Guiding Evidence-Based Therapy Using Biomarker Intensified Treatment in Heart Failure) and HF-ACTION (Heart Failure: A Controlled Trial Investigating Outcomes of Exercise Training) trials were evaluated with a computable medication optimization algorithm that outputs GDMT recommendations and a medication optimization score (MOS). Algorithm-based recommendations were compared to medication changes. A Cox proportional-hazards model was used to estimate the associations between MOS and the composite primary end point for both trials. Results: The algorithm recommended initiation of angiotensin-converting enzyme inhibitor/angiotensin receptor blocker, beta-blockers, and mineralocorticoid receptor antagonists in 52.8%, 34.9%, and 68.1% of GUIDE-IT visits, respectively, when not prescribed the drug. Initiation only occurred in 20.8%, 56.9%, and 15.8% of subsequent visits. The algorithm also identified dose titration in 48.8% of visits for angiotensin-converting enzyme inhibitor/angiotensin receptor blockers and 39.4% of visits for beta-blockers. Those increases only occurred in 24.3% and 36.8% of subsequent visits. A higher baseline MOS was associated with a lower risk of cardiovascular death or heart failure hospitalization (HR: 0.41; 95% CI: 0.21-0.80; P = 0.009) in GUIDE-IT and all-cause death and hospitalization (HR: 0.61; 95% CI: 0.44-0.84; P = 0.003) in HF-ACTION. Conclusions: The algorithm accurately identified patients for GDMT optimization. Even in a clinical trial with robust protocols, GDMT could have been further optimized in a meaningful number of visits. The algorithm-generated MOS was associated with a lower risk of clinical outcomes. Implementation into clinical care may identify and address suboptimal GDMT in patients with heart failure with reduced ejection fraction.

Comparing the variability of ingredient, strength, and dose form information from electronic prescriptions with RxNorm drug product descriptions.

OBJECTIVE: To determine the variability of ingredient, strength, and dose form information from drug product descriptions in real-world electronic prescription (e-prescription) data. MATERIALS AND METHODS: A sample of 10 399 324 e-prescriptions from 2019 to 2021 were obtained. Drug product descriptions were analyzed with a named entity extraction model and National Drug Codes (NDCs) were used to get RxNorm Concept Unique Identifiers (RxCUI) via RxNorm. The number of drug product description variants for each RxCUI was determined. Variants identified were compared to RxNorm to determine the extent of matching terminology used. RESULTS: A total of 353 002 unique pairs of drug product descriptions and NDCs were analyzed. The median (1st-3rd quartile) number of variants extracted for each standardized expression in RxNorm, was 3 (2-7) for ingredients, 4 (2-8) for strength, and 41 (11-122) for dosage forms. Of the pairs, 42.35% of ingredients (n = 328 032), 51.23% of strengths (n = 321 706), and 10.60% of dose forms (n = 326 653) used matching terminology, while 16.31%, 24.85%, and 13.05% contained nonmatching terminology, respectively. DISCUSSION: A wide variety of drug product descriptions makes it difficult to determine whether 2 drug product descriptions describe the same drug product (eg, using abbreviations to describe an active ingredient or using different units to represent a concentration). This results in patient safety risks that lead to incorrect drug products being ordered, dispensed, and used by patients. Implementation and use of standardized terminology may reduce these risks. CONCLUSION: Drug product descriptions on real-world e-prescriptions exhibit large variation resulting in unnecessary ambiguity and potential patient safety risks.

The Effects of the ManageHF4Life Mobile App on Patients With Chronic Heart Failure: Randomized Controlled Trial.

BACKGROUND: The successful management of heart failure (HF) involves guideline-based medical therapy as well as self-management behavior. As a result, the management of HF is moving toward a proactive real-time technological model of assisting patients with monitoring and self-management. OBJECTIVE: The aim of this paper was to evaluate the efficacy of enhanced self-management via a mobile app intervention on health-related quality of life, self-management, and HF readmissions. METHODS: A single-center randomized controlled trial was performed. Participants older than 45 years and admitted for acute decompensated HF or recently discharged in the past 4 weeks were included. The intervention group ("app group") used a mobile app, and the intervention prompted daily self-monitoring and promoted self-management. The control group ("no-app group") received usual care. The primary outcome was the change in Minnesota Living with Heart Failure Questionnaire (MLHFQ) score from baseline to 6 and 12 weeks. Secondary outcomes were the Self-Care Heart Failure Index (SCHFI) questionnaire score and recurrent HF admissions. RESULTS: A total of 83 participants were enrolled and completed all baseline assessments. Baseline characteristics were similar between the groups except for the prevalence of ischemic HF. The app group had a reduced MLHFQ at 6 weeks (mean 37.5, SD 3.5 vs mean 48.2, SD 3.7; P=.04) but not at 12 weeks (mean 44.2, SD 4 vs mean 45.9, SD 4; P=.78), compared to the no-app group. There was no effect of the app on the SCHFI at 6 or 12 weeks. The time to first HF readmission was not statistically different between the app group and the no-app group (app group 11/42, 26% vs no-app group 12/41, 29%; hazard ratio 0.89, 95% CI 0.39-2.02; P=.78) over 12 weeks. CONCLUSIONS: The adaptive mobile app intervention, which focused on promoting self-monitoring and self-management, improved the MLHFQ at 6 weeks but did not sustain its effects at 12 weeks. No effect was seen on HF self-management measured by self-report. Further research is needed to enhance engagement in the app for a longer period and to determine if the app can reduce HF readmissions in a larger study. TRIAL REGISTRATION: ClinicalTrials.gov NCT03149510; https://clinicaltrials.gov/ct2/show/NCT03149510.

Performance evaluation of a prescription medication image classification model: an observational cohort.

Technology assistance of pharmacist verification tasks through the use of machine intelligence has the potential to detect dangerous and costly pharmacy dispensing errors. National Drug Codes (NDC) are unique numeric identifiers of prescription drug products for the United States Food and Drug Administration. The physical form of the medication, often tablets and capsules, captures the unique features of the NDC product to help ensure patients receive the same medication product inside their prescription bottle as is found on the label from a pharmacy. We report and evaluate using an automated check to predict the shape, color, and NDC for images showing a pile of pills inside a prescription bottle. In a test set containing 65,274 images of 345 NDC classes, overall macro-average precision was 98.5%. Patterns of incorrect NDC predictions based on similar colors, shapes, and imprints of pills were identified and recommendations to improve the model are provided.

Human versus machine editing of electronic prescription directions.

BACKGROUND: Pharmacy staff are responsible for editing poor-quality and difficult-to-read electronic prescription (e-prescription) directions. Machine translation (MT) models are capable of translating free text from 1 sequence into another. However, the quality of MTs of e-prescriptions into pharmacy label directions is unknown. OBJECTIVE: To determine the types and frequencies of e-prescription direction component errors made by an MT model, pharmacy staff, and prescribers. METHODS: A prospective evaluation was conducted on a random sample of 300 patient directions in a test set of e-prescriptions from a mail-order pharmacy. Each row included directions produced by (1) prescribers on e-prescriptions, (2) pharmacy staff on prescription labels, and (3) an open neural MT model. Annotators labeled direction sets for missing direction components, use of abbreviations and medical jargon, and incorrect information (e.g., changing the number of tablets to be taken). The longest common subsequence (LCS) compared the amount of pharmacy staff editing with and without MT. RESULTS: Out of 279 direction sets labeled, the MT model directions contained no quality issues in 196 (70.3%) samples compared with 187 (67.0%) and 83 (29.8%) samples for pharmacy staff directions and prescriber directions, respectively. The MT model directions contained more incorrect components (n = 23). Median LCS was greater without MT (30.0 vs. 18.5, P < 0.01, Wilcoxon signed-rank test), indicating more editing was needed. CONCLUSION: MT could be used to improve the quality of e-prescription directions; however, MT makes high-risk mistakes such as incorrectly predicting the tapering regimen for prednisone. The use of semiautomated MT, where pharmacy staff can review model predictions to detect and resolve quality issues, should be considered to improve safety and decrease total work time compared with current practice. MT has strengths and weaknesses for improving the editing process of the patient directions compared with pharmacy staff alone.

Implementation of Evidence-Based Practice for Benign Paroxysmal Positional Vertigo in the Emergency Department: A Stepped-Wedge Randomized Trial.

STUDY OBJECTIVE: We evaluated a strategy to increase use of the test (Dix-Hallpike's test [DHT]) and treatment (canalith repositioning maneuver [CRM]) for benign paroxysmal positional vertigo in emergency department (ED) dizziness visits. METHODS: We conducted a stepped-wedge randomized trial in 6 EDs. The population was visits with dizziness as a principal reason for the visit. The intervention included educational sessions and decision aid materials. Outcomes were DHT or CRM documentation (primary), head computed tomography (CT) use, length of stay, admission, and 90-day stroke events. The analysis was multilevel logistic regression with intervention, month, and hospital as fixed effects and provider as a random effect. We assessed fidelity with monitoring intervention use and semistructured interviews. RESULTS: We identified 7,635 dizziness visits during 18 months. The DHT or CRM was documented in 1.5% of control visits (45/3,077; 95% confidence interval 1% to 1.9%) and 3.5% of intervention visits (159/4,558; 95% confidence interval 3% to 4%; difference 2%, 95% confidence interval 1.3% to 2.7%). Head CT use was lower in intervention visits compared with control visits (44.0% [1,352/3,077] versus 36.9% [1,682/4,558]). No differences were observed in admission or 90-day subsequent stroke risk. In fidelity evaluations, providers who used the materials typically reported positive clinical experiences but provider engagement was low at facilities without an emergency medicine residency program. CONCLUSION: These findings provide evidence that an implementation strategy of a benign paroxysmal positional vertigo-focused approach to ED dizziness visits can be successful and safe in promoting evidence-based care. Absolute rates of DHT and CRM use, however, were still low, which relates in part to our broad inclusion criteria for dizziness visits.

An Educational Intervention for Acute Dizziness Care: A Randomized, Vignette-based Study.

IMPORTANCE: Benign paroxysmal positional vertigo (BPPV) is a common cause of acute dizziness. Strong evidence exists for diagnosing BPPV using the Dix-Hallpike Test (DHT) and treating it with the canalith repositioning maneuver (CRM). Despite this, both are infrequently used in the emergency department (ED). OBJECTIVE: As an early method to evaluate a BPPV-focused educational intervention, we evaluated whether an educational intervention improved ED provider performance on hypothetical stroke and BPPV cases delivered by vignette. DESIGN: A randomized, controlled, educational intervention study in ED physicians. The intervention aimed to promote the appropriate use of the DHT and CRM. A BPPV vignette, a stroke-dizziness (safety) vignette, and vignette scoring schemes (higher scores indicating more optimal care) used previously established vignette methodology. SETTING: We recruited participants at the exhibitor hall of an emergency medicine annual meeting. PARTICIPANTS: We recruited 48 emergency physicians. All were board certified or residency trained and board eligible. All were engaged in the active practice of emergency medicine. None were trainees. INTERVENTIONS: Intervention group: a narrated, educational presentation by computer followed by the clinical vignettes. CONTROL GROUP: Received no educational intervention and completed the clinical vignettes-intended to mirror current clinician practice. MAIN OUTCOME MEASURE: Primary endpoint: total score (out of 200 points) on a vignette-based scoring instrument assessing the performance of history, physical, and diagnostic testing on hypothetical stroke and BPPV cases. RESULTS: The efficacy threshold was crossed at the interim analysis. The intervention group had higher performance scores compared with controls (113.2 versus 68.6, p < 0.00001). BPPV and safety subscores were both significantly higher in the intervention group. Sixty-two percent of the intervention group planned to use the DHT versus 29% of controls. After the vignette described characteristic BPPV nystagmus, 100% of the intervention group planned to use the CRM versus 17% of controls. CONCLUSIONS AND RELEVANCE: The educational intervention increased provider performance in dizziness vignettes, including more frequent appropriate use of the DHT/CRM. These findings indicate the intervention positively influenced planned behavior. Future work is needed to implement and evaluate this intervention in clinical practice.

Implementation of evidence-based practice for benign paroxysmal positional vertigo: DIZZTINCT- A study protocol for an exploratory stepped-wedge randomized trial.

BACKGROUND: Benign paroxysmal positional vertigo (BPPV) is the most common peripheral vestibular disorder, and accounts for 8% of individuals with moderate or severe dizziness. BPPV patients experience substantial inconveniences and disabilities during symptomatic periods. BPPV therapeutic processes - the Dix-Hallpike Test (DHT) and the Canalith Repositioning Maneuver (CRM) - have an evidence base that is at the clinical practice guideline level. The most commonly used CRM is the modified Epley maneuver. The DHT is the gold standard test for BPPV and the CRM is supported by numerous randomized controlled trials and systematic reviews. Despite this, BPPV care processes are underutilized. METHODS/DESIGN: This is a stepped-wedge, randomized clinical trial of a multi-faceted educational and care-process-based intervention designed to improve the guideline-concordant care of patients with BPPV presenting to the emergency department (ED) with dizziness. The unit of randomization and target of intervention is the hospital. After an initial observation period, the six hospitals will undergo the intervention in five waves (two closely integrated hospitals will be paired). The order will be randomized. The primary endpoint is measured at the individual patient level, and is the presence of documentation of either the Dix-Hallpike Test or CRM. The secondary endpoints are referral to a health care provider qualified to treat dizziness for CRM and 90-day stroke rates following an ED dizziness visit. Formative evaluations are also performed to monitor and identify potential and actual influences on the progress and effectiveness of the implementation efforts. DISCUSSION: If this study safely increases documentation of the DHT/CRM, this will be an important step in implementing the use of these evidenced-based processes of care. Positive results will support conducting larger-scale follow-up studies that assess patient outcomes. The data collection also enables evaluation of potential and actual influences on the progress and effectiveness of the implementation efforts. TRIAL REGISTRATION: ClinicalTrials.gov, ID: NCT02809599 . The record was first available to the public on 22 June 2016 prior to the enrollment of the first patients in October 2016.

Black Cigarette Smokers Report More Attention to Smoking Cues Than White Smokers: Implications for Smoking Cessation.

INTRODUCTION: Black cigarette smokers have lower rates of smoking cessation compared with Whites. However, the mechanisms underlying these differences are not clear. Many Blacks live in communities saturated by tobacco advertisements. These cue-rich environments may undermine cessation attempts by provoking smoking. Moreover, attentional bias to smoking cues (attention capture by smoking cues) has been linked to lower cessation outcomes. Cessation attempts among Blacks may be compromised by attentional bias to smoking cues and a cue-rich environment. METHOD: Attention to smoking cues in Black and White smokers was examined in 2 studies. In both studies, assessments were completed during 2 laboratory visits: a nonabstinent session and an abstinent session. In study 1, nontreatment-seeking smokers (99 Whites, 104 Blacks) completed the Subjective Attentional Bias Questionnaire (SABQ; a self-report measure of attention to cues) and the Smoking Stroop task (a reaction time measure of attentional bias to smoking cues). In study 2, 110 White and 74 Black treatment-seeking smokers completed these assessments and attempted to quit. RESULTS: In study 1, Blacks reported higher ratings than Whites on the SABQ (p = .005). In study 2, Blacks also reported higher ratings than Whites on the SABQ (p = .003). In study 2, Blacks had lower biochemical-verified point prevalence abstinence than Whites, and the between-race difference in outcome was partially mediated by SABQ ratings. CONCLUSION: Blacks reported greater attention to smoking cues than Whites, possibly due to between-race differences in environments. Greater attention to smoking cues may undermine cessation attempts.