Search

Background: Little is known about whether machine-learning algorithms developed to predict opioid overdose using earlier years and from a single state will perform as well when applied to other populations. We aimed to develop a machine-learning algorithm to predict 3-month risk of opioid overdose using Pennsylvania Medicaid data and externally validated it in two data sources (ie, later years of Pennsylvania Medicaid data and data from a different state)., Methods: This prognostic modelling study developed and validated a machine-learning algorithm to predict overdose in Medicaid beneficiaries with one or more opioid prescription in Pennsylvania and Arizona, USA. To predict risk of hospital or emergency department visits for overdose in the subsequent 3 months, we measured 284 potential predictors from pharmaceutical and health-care encounter claims data in 3-month periods, starting 3 months before the first opioid prescription and continuing until loss to follow-up or study end. We developed and internally validated a gradient-boosting machine algorithm to predict overdose using 2013-16 Pennsylvania Medicaid data (n=639 693). We externally validated the model using (1) 2017-18 Pennsylvania Medicaid data (n=318 585) and (2) 2015-17 Arizona Medicaid data (n=391 959). We reported several prediction performance metrics (eg, C-statistic, positive predictive value). Beneficiaries were stratified into risk-score subgroups to support clinical use., Findings: A total of 8641 (1·35%) 2013-16 Pennsylvania Medicaid beneficiaries, 2705 (0·85%) 2017-18 Pennsylvania Medicaid beneficiaries, and 2410 (0·61%) 2015-17 Arizona beneficiaries had one or more overdose during the study period. C-statistics for the algorithm predicting 3-month overdoses developed from the 2013-16 Pennsylvania training dataset and validated on the 2013-16 Pennsylvania internal validation dataset, 2017-18 Pennsylvania external validation dataset, and 2015-17 Arizona external validation dataset were 0·841 (95% CI 0·835-0·847), 0·828 (0·822-0·834), and 0·817 (0·807-0·826), respectively. In external validation datasets, 71 361 (22·4%) of 318 585 2017-18 Pennsylvania beneficiaries were in high-risk subgroups (positive predictive value of 0·38-4·08%; capturing 73% of overdoses in the subsequent 3 months) and 40 041 (10%) of 391 959 2015-17 Arizona beneficiaries were in high-risk subgroups (positive predictive value of 0·19-1·97%; capturing 55% of overdoses). Lower risk subgroups in both validation datasets had few individuals (≤0·2%) with an overdose., Interpretation: A machine-learning algorithm predicting opioid overdose derived from Pennsylvania Medicaid data performed well in external validation with more recent Pennsylvania data and with Arizona Medicaid data. The algorithm might be valuable for overdose risk prediction and stratification in Medicaid beneficiaries., Funding: National Institute of Health, National Institute on Drug Abuse, National Institute on Aging., Competing Interests: Declaration of interests W-HL-C and WFG are named as inventors in one preliminary patent (U1195.70174US00) filing from the University of Florida and University of Pittsburgh for use of the machine learning algorithm for opioid risk prediction in Medicare described in this Article. W-HL-C , WFG, DLW, and C-YC are recipients of a grant from the National Institute on Aging (NIA; R21 AG060308). W-HL-C , WFG , JMD, AJG, GC, JLH, CCK, DCM, QY, JW, and HHZ are recipients of a grant from the US National Institute on Drug Abuse (NIDA; R01DA044985). W-HL-C declares grants from the Richard King Mellon Foundation–University of Pittsburgh, University of Florida Clinical and Translational Science Institute, the US National Institute of Mental Health (1R03MH114503-01 and R01MH121907), Pharmaceutical Research and Manufacturers of America Foundation, NIDA (1R01DA050676-01A1 and R01DA044985), Veterans Affairs (VA) Merit 1 (I01HX002191-01A2), and Merck, Sharp & Dohme and Bristol Myers Squibb. CKK declares grants from Lilly, Pfizer, GSK, Cumberland Pharmaceuticals, AbbVie, and EMD Serono; consultant fees from EMD Serono, Express Scripts, and Regeneron; an advisory board role with EMD Serono, Thusane, Regeneron, Taiwan Lipisome Company, Amzell, LG Chem, and Novartis; payment or honoraria from Focus Communications and PRIME Education; and being on the Data and Safety Monitoring Committee for Kolon Tissue Gene and on the board of directors for the International Chinese Osteoarthritis Research Society. WFG declares grant or contract to his institution from Richard King Mellon Foundation. JMD declares salary support from the Pennsylvania Department of Human Services, Richard King Mellon Foundation, and the US National Institute of Health (NIH)–NIDA (R01DA048019). AJG declares grants or contracts from NIH and the VA; royalties from UpToDate; and other financial and non-financial interests from American Society of Addiction Medicine, Association for Multidisciplinary Education and Research in Substance Use and Addiction, and the International Society of Addiction Journal Editors. JCW declares grants or contracts from Allegheny Health Network, Carnegie Mellon University, and University of Pittsburgh Medical Center; payment or honoraria from St Jude's; and receipt of equipment, materials, or other services from Amazon Web Services and Azure. DLW declares grant funding from Merck, Sharp & Dohme, and NIH–NIDA (1R01DA050676-01A1). All other authors declare no competing interests., (Copyright © 2022 The Author(s). Published by Elsevier Ltd. This is an Open Access article under the CC BY-NC-ND 4.0 license. Published by Elsevier Ltd.. All rights reserved.)

Background: Low-income subsidy/dual eligibility (LIS/DE) status and disability status may be associated with high-risk medication (HRM) use but are not usually accounted for in medication-use quality measures., Objective: To examine the association of: 1) LIS/DE status and HRM use; and 2) disability status and HRM use, while controlling for both health plan level effects and patient characteristics for Medicare beneficiaries enrolled in Medicare Advantage Prescription Drug Plans (MA-PD) and stand-alone Prescription Drug Plans (PDP)., Methods: This retrospective cross-sectional study used 2013 Medicare data to determine if LIS/DE status and disability status were independently associated with HRM use (using the Pharmacy Quality Alliance HRM measure) in MA-PDs and PDPs. Multivariable generalized linear mixed models assessed the association of LIS/DE and HRM use, and disability and HRM use, after adjusting for health plan effect and patient-level confounders for MA-PD and PDP beneficiaries., Results: Of 520,019 MA-PD beneficiaries, 88,693 (17.1%) were LIS/DE and 48,997 (9.4%) were disabled. Of 881,264 PDP beneficiaries, 213,096 (24.2%) were LIS/DE, and 83,593 (9.5%) were disabled. LIS/DE beneficiaries had a higher percent of HRM users compared to non-LIS/DE MA-PD (13.3% vs. 9.7%, p < 0.001) and PDP (17.1% vs. 13.2%, p < 0.001) beneficiaries. Disabled beneficiaries had a higher percent of HRM users compared to non-disabled MA-PD (17.0% vs. 9.6%, p < 0.001) and PDP (22.9% vs. 13.2%, p < 0.001) beneficiaries. Multivariable analyses showed LIS/DE (adjusted odds ratio [AOR] = 1.07; 95% CI = 1.04, 1.10) and disability (AOR = 1.38; 95% CI = 1.34, 1.42) were associated with HRM use among MA-PD and PDP beneficiaries (LIS/DE AOR = 1.14; 95% CI = 1.12, 1.16; disability AOR = 1.37; 95% CI = 1.34, 1.40)., Conclusions: The association of LIS/DE and disability with higher HRM use in both MA-PD and PDP beneficiaries, when controlling for health plan effects and patient characteristics, suggests these factors should be considered when comparing health plan performance on HRM measures., (Copyright © 2021 Elsevier Inc. All rights reserved.)

Introduction: Concurrent use of prescription opioids with gabapentinoids may pose risks of serious drug interactions. Yet, little is known about the trends in and patient characteristics associated with concurrent opioid-gabapentinoid use among older Medicare opioid users with chronic noncancer pain., Methods: A cross-sectional study was conducted among Medicare older beneficiaries (aged ≥65 years) with chronic noncancer pain who filled ≥1 opioid prescription within 3 months after a randomly selected chronic noncancer pain diagnosis (index date) in a calendar year between 2011 and 2018. Patient characteristics were measured in the 6-month baseline before the index date, and concurrent opioid-gabapentinoid use for ≥1 day was measured in the 3-month follow-up after the index date. Multivariable modified Poisson regression hwas used to assess the trends and characteristics of concurrent opioid-gabapentinoid use. Analyses were conducted from January to June 2021., Results: Among 464,721 eligible older beneficiaries with chronic noncancer pain and prescription opioids, the prevalence of concurrent opioid-gabapentinoid use increased from 17.0% in 2011 to 23.5% in 2018 (adjusted prevalence ratio=1.48, 95% CI=1.45, 1.53). Concurrent users versus opioid-only users tended to be non-Black, low-income subsidy recipients, and Southern residents. The clinical factors associated with concurrent opioid-gabapentinoid use included having a diagnosis of neuropathic pain, polypharmacy, and risk factors for opioid-related adverse events., Conclusions: Concurrent opioid-gabapentinoid use among older Medicare beneficiaries with chronic noncancer pain and prescription opioids has increased significantly between 2011 and 2018. Future studies are warranted to investigate the impact of concurrent use on outcomes in older patients. Interventions that reduce inappropriate concurrent use may target older patients with identified characteristics., (Copyright © 2021 American Journal of Preventive Medicine. Published by Elsevier Inc. All rights reserved.)

Background: Survivors of opioid overdose have substantially increased mortality risk, although this risk is not evenly distributed across individuals. No study has focused on predicting an individual's risk of death after a nonfatal opioid overdose., Objective: To predict risk of death after a nonfatal opioid overdose., Design and Participants: This retrospective cohort study included 9686 Pennsylvania Medicaid beneficiaries with an emergency department or inpatient claim for nonfatal opioid overdose in 2014-2016. The index date was the first overdose claim during this period., Exposures, Main Outcome, and Measures: Predictor candidates were measured in the 180 days before the index overdose. Primary outcome was 180-day all-cause mortality. Using a gradient boosting machine model, we classified beneficiaries into six subgroups according to their risk of mortality (< 25th percentile of the risk score, 25th to < 50th, 50th to < 75th, 75th to < 90th, 90th to < 98th, ≥ 98th). We then measured receipt of medication for opioid use disorder (OUD), risk mitigation interventions (e.g., prescriptions for naloxone), and prescription opioids filled in the 180 days after the index overdose, by risk subgroup., Key Results: Of eligible beneficiaries, 347 (3.6%) died within 180 days after the index overdose. The C-statistic of the mortality prediction model was 0.71. In the highest risk subgroup, the observed 180-day mortality rate was 20.3%, while in the lowest risk subgroup, it was 1.5%. Medication for OUD and risk mitigation interventions after overdose were more commonly seen in lower risk groups, while opioid prescriptions were more likely to be used in higher risk groups (both p trends < .001)., Conclusions: A risk prediction model performed well for classifying mortality risk after a nonfatal opioid overdose. This prediction score can identify high-risk subgroups to target interventions to improve outcomes among overdose survivors.

Because they contain detailed individual-level data on various patient characteristics including their medical conditions and treatment histories, electronic health record (EHR) systems have been widely adopted as an efficient source for health research. Compared to data from a single health system, real-world data (RWD) from multiple clinical sites provide a larger and more generalizable population for accurate estimation, leading to better decision making for health care. However, due to concerns over protecting patient privacy, it is challenging to share individual patient-level data across sites in practice. To tackle this issue, many distributed algorithms have been developed to transfer summary-level statistics to derive accurate estimates. Nevertheless, many of these algorithms require multiple rounds of communication to exchange intermediate results across different sites. Among them, the One-shot Distributed Algorithm for Logistic regression (termed ODAL) was developed to reduce communication overhead while protecting patient privacy. In this paper, we applied the ODAL algorithm to RWD from a large clinical data research network-the OneFlorida Clinical Research Consortium and estimated the associations between risk factors and the diagnosis of opioid use disorder (OUD) among individuals who received at least one opioid prescription. The ODAL algorithm provided consistent findings of the associated risk factors and yielded better estimates than meta-analysis., (©2020 AMIA - All rights reserved.)

Objective: To quantify the extent and identify predictors of potentially inappropriate antidepressant use among older adults with dementia and newly diagnosed major depressive disorders (MDD)., Methods: This retrospective cohort study included older adults (aged ≥65 years) with dementia and newly diagnosed MDD using Medicare 5% sample claims data (2012-2013). Based on Healthcare Effectiveness Data and Information Set guidelines, intake period for new antidepressant medication use was from May 1, 2012, through April 30, 2013. Index prescription start date was the first date of antidepressant prescription claim during the intake period. Dependent variable of this study was potentially inappropriate antidepressant use as defined by the Beers Criteria and the Screening Tool of Older Persons' potentially inappropriate Prescriptions criteria. The authors conducted multiple logistic regression analysis to identify individual-level predictors of potentially inappropriate antidepressant use., Results: The authors' final study sample consisted of 7,625 older adults with dementia and newly diagnosed MDD, among which 7.59% (N = 579) initiated treatment with a potentially inappropriate antidepressant. Paroxetine (N = 394) was the most commonly initiated potentially inappropriate antidepressant followed by amitriptyline (N = 104), nortriptyline (N = 35), and doxepin (N = 32). Initiation of a potentially inappropriate antidepressant was associated with age and baseline use of anxiolytic medications., Conclusion: More than 7% of older adults in the study sample initiated a potentially inappropriate antidepressant, and the authors identified a few individual-level factors significantly associated with it. Appropriately tailored interventions to address modifiable and nonmodifiable factors significantly associated with potentially inappropriate antidepressant prescribing are required to minimize risks in this vulnerable population., (Copyright © 2019 American Association for Geriatric Psychiatry. Published by Elsevier Inc. All rights reserved.)

Objectives: The US opioid medication epidemic has resulted in serious health consequences for patients. Formulary management tools adopted by payers, specifically prior authorization (PA) policies, may lower the rates of opioid medication abuse and overdose. We compared rates of opioid abuse and overdose among enrollees in plans that varied in their use of PA from "High PA" (ie, required PA for 17 to 74 opioids), with "Low PA" (ie, required PA for 1 opioid), and "No PA" policies for opioid medications., Study Design: Retrospective cohort study of patients initiating opioid treatment in Pennsylvania Medicaid from 2010 to 2012., Methods: Generalized linear models with generalized estimating equations were employed to assess the relationships between the presence of PA policies and opioid medication abuse and overdose, as measured in Medicaid claims data, adjusting for demographics, comorbid health conditions, benzodiazepine/muscle relaxant use, and emergency department use., Results: The study cohort included 297,634 enrollees with a total of 382,828 opioid treatment episodes. Compared with plans with No PA, enrollees in High PA (adjusted rate ratio [ARR], 0.89; 95% confidence interval [CI], 0.85-0.93; P <.001) and Low PA plans (ARR, 0.93; 95% CI, 0.87-1.00; P = .04) had lower rates of abuse. Enrollees in the Low PA plan had a lower rate of overdose than those within plans with No PA (ARR, 0.75; 95% CI, 0.59-0.95; P = .02). High PA plan enrollees were also less likely than No PA enrollees to experience an overdose, but this association was not statistically significant (ARR, 0.88; 95% CI, 0.76-1.02; P = .08)., Conclusions: Enrollees within Medicaid plans that utilize PA policies appear to have lower rates of abuse and overdose following initiation of opioid medication treatment.

Background: Evidence-based interventions exist for prevention of chronic disease in older adults. Partnering with community organizations may provide a mechanism for disseminating these interventions., Objective: To describe the partnership and program implementation by the Arthritis Foundation (AF) and the University of Pittsburgh., Methods: The AF Exercise Program (AFEP; an existing evidence-based program) was enhanced with the "10 Keys"™ to Healthy Aging (a prevention-focused program bundling the most common risk factors for chronic disease and disability in older adults and applies behavior change strategies to enhance prevention). The program was delivered in 20 sessions over 10 weeks by community health workers in a cluster-randomized trial., Lessons Learned: Partnering with an organization having an existing infrastructure supports program delivery at the community level. This partnership provided programming in 54 sites across Pittsburgh and surrounding communities., Conclusions: This collaborative partnership created a productive synergy maximizing strengths in both research and program delivery.

Genital powder use has been associated with risk of epithelial ovarian cancer in some, but not all, epidemiologic investigations, possibly reflecting the carcinogenic effects of talc particles found in most of these products. Whether risk increases with number of genital powder applications and for all histologic types of ovarian cancer also remains uncertain. Therefore, we estimated the association between self-reported genital powder use and epithelial ovarian cancer risk in eight population-based case-control studies. Individual data from each study were collected and harmonized. Lifetime number of genital powder applications was estimated from duration and frequency of use. Pooled ORs were calculated using conditional logistic regression matched on study and age and adjusted for potential confounders. Subtype-specific risks were estimated according to tumor behavior and histology. 8,525 cases and 9,859 controls were included in the analyses. Genital powder use was associated with a modest increased risk of epithelial ovarian cancer [OR, 1.24; 95% confidence interval (CI), 1.15-1.33] relative to women who never used powder. Risk was elevated for invasive serous (OR, 1.20; 95% CI, 1.09-1.32), endometrioid (OR, 1.22; 95% CI, 1.04-1.43), and clear cell (OR, 1.24; 95% CI, 1.01-1.52) tumors, and for borderline serous tumors (OR, 1.46; 95% CI, 1.24-1.72). Among genital powder users, we observed no significant trend (P = 0.17) in risk with increasing number of lifetime applications (assessed in quartiles). We noted no increase in risk among women who only reported nongenital powder use. In summary, genital powder use is a modifiable exposure associated with small-to-moderate increases in risk of most histologic subtypes of epithelial ovarian cancer.