Dataset Information

Use of power-law analysis to predict abuse or diversion of prescribed medications: proof-of-concept mathematical exploration.

ABSTRACT: To conduct a proof-of-concept study comparing Lorenz-curve analysis (LCA) with power-law (exponential function) analysis (PLA), by applying segmented regression modeling to 1-year prescription claims data for three medications-alprazolam, opioids, and gabapentin-to predict abuse and/or diversion using power-law zone (PLZ) classification.In 1-year baseline observation, patients classified into the top PLZ groups (PLGs) were demographically and diagnostically similar to those in Lorenz-1 (top 1% of utilizers) and Lorenz-25 (top 25%). For prediction of follow-up (6-month post-baseline) Lorenz-1 use of alprazolam and opioids (i.e., potential abuse/diversion), PLA had somewhat lower sensitivity compared with LCA (83.5-95.4% vs. 99.5-99.9%, respectively) but better specificity (98.2-98.8% vs. 75.5%) and much better positive predictive value (PPV; 34.5-45.3% vs. 4.0-4.6%). Of top-PLG alprazolam- and opioid-treated patients, respectively, 20.7 and 9.9% developed incident (new) Lorenz-1 in followup, compared with?

SUBMITTER: Fairman KA

PROVIDER: S-EPMC6069871 | biostudies-other | 2018 Jul

REPOSITORIES: biostudies-other

ACCESS DATA

Publications

Use of power-law analysis to predict abuse or diversion of prescribed medications: proof-of-concept mathematical exploration.

Fairman Kathleen A KA Peckham Alyssa M AM Rucker Michael L ML Rucker Jonah H JH Sclar David A DA

BMC research notes 20180731 1

<h4>Objective</h4>To conduct a proof-of-concept study comparing Lorenz-curve analysis (LCA) with power-law (exponential function) analysis (PLA), by applying segmented regression modeling to 1-year prescription claims data for three medications-alprazolam, opioids, and gabapentin-to predict abuse and/or diversion using power-law zone (PLZ) classification.<h4>Results</h4>In 1-year baseline observation, patients classified into the top PLZ groups (PLGs) were demographically and diagnostically simi ...[more]

PMID: 30064519

Similar Datasets

Project description:BackgroundMany people want to build good habits to become healthier, live longer, or become happier but struggle to change their behavior. Gamification can make behavior change easier by awarding points for the desired behavior and deducting points for its omission.ObjectiveIn this study, we introduced a principled mathematical method for determining how many points should be awarded or deducted for the enactment or omission of the desired behavior, depending on when and how often the person has succeeded versus failed to enact it in the past. We called this approach optimized gamification of behavior change.MethodsAs a proof of concept, we designed a chatbot that applies our optimized gamification method to help people build healthy water-drinking habits. We evaluated the effectiveness of this gamified intervention in a 40-day field experiment with 1 experimental group (n=43) that used the chatbot with optimized gamification and 2 active control groups for which the chatbot's optimized gamification feature was disabled. For the first control group (n=48), all other features were available, including verbal feedback. The second control group (n=51) received no feedback or reminders. We measured the strength of all participants' water-drinking habits before, during, and after the intervention using the Self-Report Habit Index and by asking participants on how many days of the previous week they enacted the desired habit. In addition, all participants provided daily reports on whether they enacted their water-drinking intention that day.ResultsA Poisson regression analysis revealed that, during the intervention, users who received feedback based on optimized gamification enacted the desired behavior more often (mean 14.71, SD 6.57 times) than the active (mean 11.64, SD 6.38 times; P<.001; incidence rate ratio=0.80, 95% CI 0.71-0.91) or passive (mean 11.64, SD 5.43 times; P=.001; incidence rate ratio=0.78, 95% CI 0.69-0.89) control groups. The Self-Report Habit Index score significantly increased in all conditions (P<.001 in all cases) but did not differ between the experimental and control conditions (P>.11 in all cases). After the intervention, the experimental group performed the desired behavior as often as the 2 control groups (P≥.17 in all cases).ConclusionsOur findings suggest that optimized gamification can be used to make digital behavior change interventions more effective.Trial registrationOpen Science Framework (OSF) H7JN8; https://osf.io/h7jn8.

Project description:ObjectivesThe aim of this study was to demonstrate the existence of systematic associations in drug prescription that lead to the establishment of patterns of polypharmacy, and the clinical interpretation of the associations found in each pattern.MethodsA cross-sectional study was conducted based on information obtained from electronic medical records and the primary care pharmacy database in 2008. An exploratory factor analysis of drug dispensing information regarding 79,089 adult patients was performed to identify the patterns of polypharmacy. The analysis was stratified by age and sex.ResultsSeven patterns of polypharmacy were identified, which may be classified depending on the type of disease they are intended to treat: cardiovascular, depression-anxiety, acute respiratory infection (ARI), chronic obstructive pulmonary disease (COPD), rhinitis-asthma, pain, and menopause. Some of these patterns revealed a clear clinical consistency and included drugs that are prescribed together for the same clinical indication (i.e., ARI and COPD patterns). Other patterns were more complex but also clinically consistent: in the cardiovascular pattern, drugs for the treatment of known risk factors-such as hypertension or dyslipidemia-were combined with other medications for the treatment of diabetes or established cardiovascular pathology (e.g., antiplatelet agents). Almost all of the patterns included drugs for preventing or treating potential side effects of other drugs in the same pattern.ConclusionsThe present study demonstrated the existence of non-random associations in drug prescription, resulting in patterns of polypharmacy that are sound from the pharmacological and clinical viewpoints and that exist in a significant proportion of the population. This finding necessitates future longitudinal studies to confirm some of the proposed causal associations. The information discovered would further the development and/or adaptation of clinical patient guidelines to patients with multimorbidity who are taking multiple drugs.

Dataset Information

Use of power-law analysis to predict abuse or diversion of prescribed medications: proof-of-concept mathematical exploration.

Publications

Use of power-law analysis to predict abuse or diversion of prescribed medications: proof-of-concept mathematical exploration.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets