Pharmacoepidemiology and Drug Safety最新文献

{"title":"Treatment Persistence and Variations in Prescribing Oral, Injectable, and Inhaled Corticosteroids: A Population-Based Drug Utilisation Study.","authors":"Eleni Domzaridou, Matthew J Carr, David M Williams, Anthony J Avery, Tjeerd van Staa, D Aled Rees, Darren M Ashcroft","doi":"10.1002/pds.70153","DOIUrl":"https://doi.org/10.1002/pds.70153","url":null,"abstract":"<p><strong>Purpose: </strong>To examine variation in oral, injectable, and inhaled corticosteroid (CS) prescribing in primary care, exploring treatment persistence and coverage.</p><p><strong>Methods: </strong>We examined patient-level electronic health records from English general practices in the Clinical Practice Research Datalink Aurum database. We delineated a cohort of new users of oral, injectable, or inhaled CS with prescriptions issued between January 1, 2000, and June 30, 2021. Lorenz curves assessed potential prescribing skewness, and Kaplan-Meier (KM) plots estimated treatment persistence. The Proportion of Patients Covered (PPC) method estimated the proportion of patients still covered by treatment 1 year after initiation.</p><p><strong>Results: </strong>We observed 1 942 571 CS users across 1471 general practices, with 20% of oral and inhaled CS users accounting for almost 80% of total CS use. Older patients with comorbidities including respiratory diseases (13.5%), skin conditions (5.8%), or inflammatory bowel diseases (1.6%) were more likely to be prescribed higher doses. The KM plots showed that 20% of oral and 50% of inhaled CS users were persistent after one and 2 months, respectively. The PPC method indicated that 30% of oral and 60% of inhaled CS users were covered by treatment 6 months post-initiation. Some variation was observed when different grace periods were applied. Combined use of oral and inhaled CS was observed for 6.9% of patients.</p><p><strong>Conclusion: </strong>A fifth of patients receiving CS accounted for over 80% of oral and inhaled CS prescribing in primary care. Identifying these patients is crucial for targeting future interventions to promote patient safety and cost-effective CS use.</p>","PeriodicalId":19782,"journal":{"name":"Pharmacoepidemiology and Drug Safety","volume":"34 5","pages":"e70153"},"PeriodicalIF":2.4,"publicationDate":"2025-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12042156/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144036998","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Proton Pump Inhibitor Use Exceeding the U.S. Food and Drug Administration Approved Treatment Duration for Patients With Peptic Ulcer Disease: A Retrospective Cohort Study.","authors":"Jordan A Villars, Timothy S Anderson, Jonathan G Yabes, Robert E Schoen, Ravy K Vajravelu","doi":"10.1002/pds.70152","DOIUrl":"https://doi.org/10.1002/pds.70152","url":null,"abstract":"<p><strong>Background: </strong>Proton-pump inhibitors (PPIs) are effective in treating peptic ulcer disease (PUD), but they are often prescribed beyond the approved duration. Because PPIs are associated with adverse effects, there is a need for effective stewardship.</p><p><strong>Objective: </strong>To identify the frequency of and healthcare factors associated with PPI prescriptions exceeding the approved eight-week treatment duration for PUD.</p><p><strong>Methods: </strong>We conducted a retrospective cohort study of patients diagnosed with acute PUD without other indications for PPI use using data from the Veterans Health Administration in the United States. Exposures were patient, provider, and facility factors that could influence PPI prescribing. The outcome was time to a filled PPI prescription exceeding the approved treatment duration for PUD. Associations were assessed using a multivariable time-to-recurrent-event model to calculate adjusted hazard ratios (aHR) and population-attributable fractions. Patients who developed indications for long-term PPI use were censored.</p><p><strong>Results: </strong>We identified 7708 patients with PUD who met eligibility criteria and received PUD treatment (median age 79 [IQR 71-85], 7% female). Thirty-five percent had PPI prescriptions exceeding the approved duration for a median of 346 days (IQR 165-643) of overuse. On the patient level, inpatient PUD diagnosis (aHR 1.32, 95% CI 1.25-1.39), use of nonsteroidal anti-inflammatory drugs (NSAIDs) (aHR 1.26, 95% CI 1.18-1.34), use of anticoagulants (aHR 1.25, 95% CI 1.13-1.38), and moderate frailty (1.15, 95% CI 1.06-1.26) had the strongest associations with filled PPI prescriptions exceeding the approved duration. On the health-system level, inpatient PUD diagnosis had the highest peak population attributable fraction at 0.26, followed by NSAIDs and anticoagulants at 0.18.</p><p><strong>Conclusions: </strong>Markers of patient complexity and medication use not meeting gastroprotection guidelines are associated with inappropriate PPI persistence among patients with PUD. These data may inform future targeted PPI deprescribing programs.</p>","PeriodicalId":19782,"journal":{"name":"Pharmacoepidemiology and Drug Safety","volume":"34 5","pages":"e70152"},"PeriodicalIF":2.4,"publicationDate":"2025-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12038380/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144006873","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Validity of Diagnostic Codes and Laboratory Tests to Identify Cholangiocarcinoma and Its Subtypes.","authors":"Nicole D Ferrante, Rebecca A Hubbard, Kelley Weinfurtner, Anya I Mezina, Craig W Newcomb, Emma E Furth, Debika Bhattacharya, Basile Njei, Tamar H Taddei, Amit Singal, Maarouf A Hoteit, Lesley S Park, David Kaplan, Vincent Lo Re","doi":"10.1002/pds.70154","DOIUrl":"https://doi.org/10.1002/pds.70154","url":null,"abstract":"<p><strong>Background: </strong>The absence of validated methods to identify cholangiocarcinoma in real-world data has prevented the conduct of pharmacoepidemiologic studies to evaluate determinants of this malignancy and examine the effectiveness of cholangiocarcinoma treatments.</p><p><strong>Objective: </strong>To determine the accuracy of International Classification of Diseases for Oncology, Third Edition (ICD-O-3)-based algorithms to identify cholangiocarcinoma and its subtype (intrahepatic or extrahepatic) within US Veterans Health Administration (VA) data.</p><p><strong>Methods: </strong>We identified patients with cholangiocarcinoma ICD-O-3 diagnosis codes from January 2000-December 2019 in VA data. We developed eight algorithms utilizing ICD-O-3 histology codes for cholangiocarcinoma and further used ICD-O-3 topography codes for location (liver, intrahepatic bile duct, extrahepatic bile duct) plus maximum total bilirubin (≥ 3 mg/dL vs. < 3 mg/dL) within ± 45 days of diagnosis to identify cholangiocarcinoma subtype. Up to 80 patients were randomly selected for each algorithm, and their records were reviewed by two hepatologists. The positive predictive values (PPV) and 95% confidence interval (CI) for each algorithm were estimated.</p><p><strong>Results: </strong>Among 2934 unique patients who met inclusion criteria, 574 were randomly selected for validation. All eight algorithms had high PPV for definite or probable cholangiocarcinoma, ranging from 83.8% (95% CI, 73.8%-91.1%) to 100.0% (95% CI, 95.5%-100.0%). Among three algorithms to identify intrahepatic cholangiocarcinoma, two had PPV ≥ 80% (range: 88.8% [95% CI, 79.7%-94.7%]-91.3% [95% CI, 82.8%-96.4%]). Among five algorithms to identify extrahepatic cholangiocarcinoma, four had PPV ≥ 80% (range: 80.0% [95% CI, 69.6%-88.1%]-94.0% [83.5%-98.7%]).</p><p><strong>Conclusion: </strong>These algorithms can be used in future pharmacoepidemiologic studies to evaluate medications associated with intrahepatic or extrahepatic cholangiocarcinoma.</p>","PeriodicalId":19782,"journal":{"name":"Pharmacoepidemiology and Drug Safety","volume":"34 5","pages":"e70154"},"PeriodicalIF":2.4,"publicationDate":"2025-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12055315/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144025325","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Changes in Hormonal Contraceptive Dispensing Trends Among Commercially Insured Kentucky Females During the COVID-19 Pandemic.","authors":"Dustin K Miracle, Lindsey R Hammerslag, Svetla Slavova, Feitong Lei, Jeffery Talbert, Daniela C Moga, Patricia R Freeman","doi":"10.1002/pds.70159","DOIUrl":"https://doi.org/10.1002/pds.70159","url":null,"abstract":"<p><strong>Purpose: </strong>To evaluate the impact of the COVID-19 national emergency declaration on contraceptive dispensing trends among commercially insured Kentucky females.</p><p><strong>Methods: </strong>Data ranging from 1/7/2019 through 12/27/2020 for female enrollees aged 19-44 with a primary residence in Kentucky were extracted from the Merative Marketscan Commercial Claims and Encounters Database. A segmented regression analysis was used for statistical modeling of an interrupted time series design to describe changes in weekly contraceptive (oral, transdermal, and vaginal) dispensing rates and days' supply following the COVID-19 national emergency.</p><p><strong>Results: </strong>A total of 90 541 enrollees met study inclusion criteria. The estimated weekly contraceptive dispensing rate per 100 reproductive-aged female enrollees was 3.22 (95% confidence interval [CI] 3.16-3.28) at the beginning of the pre-pandemic period. Following the national emergency, an immediate estimated rate increase of 0.11 (95% CI 0.01-0.21; p = 0.030) was seen with no change in trend. At the beginning of the pre-pandemic period, the estimated weekly percentage of days' supply > 28 days was 29.2% (95% CI 28.8-29.6) with an increasing trend of 1.1% every 10 weeks (slope 0.11 [95% CI 0.09-0.12; p < 0.001]). Following the national emergency, an immediate decrease of 1.5% (95% CI -2.2 to -0.8; p < 0.001) was observed, followed by sustainment of the pre-pandemic trend. No differential impacts were seen with regard to age group (19-26 vs. 27-44) or rural-urban classification.</p><p><strong>Conclusions: </strong>Following the COVID-19 national emergency declaration, trends in both contraceptive dispensing and days' supply among commercially insured Kentucky females were relatively stable, suggesting multiple behavioral and policy-related factors potentially overshadowing changes in access.</p>","PeriodicalId":19782,"journal":{"name":"Pharmacoepidemiology and Drug Safety","volume":"34 5","pages":"e70159"},"PeriodicalIF":2.4,"publicationDate":"2025-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144079303","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Implementing Negative Control Outcomes to Assess Comparability of Treatments for Psoriasis and Psoriatic Arthritis.","authors":"Mary E Horner, Alexis Ogdie, Kate K Orroth, Shia T Kent, Kathy V Tran, Cynthia Deignan, Myriam Cordey, M Alan Brookhart","doi":"10.1002/pds.70156","DOIUrl":"10.1002/pds.70156","url":null,"abstract":"<p><strong>Purpose: </strong>Treatment selection is typically associated with prognosis, leading to potential confounding in comparative studies. We used negative control outcomes (NCOs) to identify potential residual confounding when comparing apremilast initiators to other psoriasis (PsO) or psoriatic arthritis (PsA) treatment initiators.</p><p><strong>Methods: </strong>Adults with PsO/PsA who initiated treatment from September 23/March 21, 2016, respectively, with apremilast, topicals, methotrexate, interleukin (IL)-17 inhibitor (i), IL-12/23i, or tumor necrosis factor inhibitor (TNFi) were identified in the OPTUM Clinformatics DataMart database. Follow-up ended at treatment switch/discontinuation, NCO, end of enrollment, or September 30, 2022. NCOs addressed confounding for healthy users (wellness visit, herpes zoster vaccine, colon cancer screening, pelvic screening), functional status (accidents), and channeling. The 1-year relative risk (RR) for each NCO was estimated for all treatment comparisons using an inverse probability of treatment and censoring weighted estimator.</p><p><strong>Results: </strong>In PsO, potential healthy user bias was detected in apremilast vs. IL-17i initiators, with a higher likelihood of herpes zoster vaccine and colon cancer screening (RR [95% CI]: 2.01 [1.41, 2.88] and 1.42 [1.13, 1.77], respectively). Wellness visits and pelvic exams were less likely among apremilast vs. topical initiators (0.84 [0.72, 0.98] and 0.83 [0.70, 0.98], respectively). The wellness visit RR was attenuated in individuals with ≥ 1 pre-index topical prescription (0.90 [0.78, 1.04]). In PsA, minimal residual confounding was observed between apremilast and other treatments.</p><p><strong>Conclusions: </strong>Eligibility criteria (prior topicals) and weighting reduced residual confounding when comparing apremilast vs. other treatments for PsO and PsA. Integration of NCOs into comparative effectiveness/safety studies of PsO/PsA treatments may help identify unmeasured confounding.</p>","PeriodicalId":19782,"journal":{"name":"Pharmacoepidemiology and Drug Safety","volume":"34 5","pages":"e70156"},"PeriodicalIF":2.4,"publicationDate":"2025-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12087269/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144094545","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Safety Monitoring of Bivalent COVID-19 mRNA Vaccines Among Recipients 6 Months and Older in the United States.","authors":"Patricia C Lloyd, Elizabeth R Smith, Joann F Gruber, Michelle Ondari, Hui Lee Wong, Mao Hu, Tainya C Clarke, Rowan McEvoy, Kandace L Amend, Daniel C Beachler, Cheryl N McMahill-Walraven, John D Seeger, Alex Secora, Djeneba Audrey Djibo, Jennifer Song, Nandini Selvam, Jonathan P DeShazo, Robin Clifford, Eugenio Abente, Yoganand Chillarige, Richard A Forshee, Steven A Anderson, Azadeh Shoaibi","doi":"10.1002/pds.70151","DOIUrl":"https://doi.org/10.1002/pds.70151","url":null,"abstract":"<p><strong>Purpose: </strong>Active monitoring of health outcomes after COVID-19 vaccination provides early detection of rare outcomes post-licensure. We evaluated health outcomes following bivalent COVID-19 Pfizer-BioNTech (BNT162b2) and Moderna (mRNA-1273.222) vaccination in the United States.</p><p><strong>Methods: </strong>Multiple health outcomes were monitored monthly from August 2022 to July 2023 in four administrative claims databases (CVS Health, Carelon Research, Optum, and Medicare). The study included individuals 6 months and older who received a bivalent COVID-19 BNT162b2 or mRNA-1273.222 vaccination during the study period and met a minimum continuous enrollment requirement in a medical insurance plan prior to COVID-19 vaccination. Descriptive analyses monitored counts of vaccinations, outcomes, and concomitant COVID-19 and influenza vaccination. Maximized Sequential Probability Ratio Testing (MaxSPRT) tested for elevations in the observed incidence rate of outcomes post-vaccination compared to annual historical rates estimated from 2019 or 2020, adjusted for claims delay in the observed rate. Where case counts permitted, historical rates were standardized by age and/or sex for all persons, and race and/or nursing home residency status for Medicare persons only.</p><p><strong>Results: </strong>Overall, 13.9 million individuals 6 months and older received a bivalent COVID-19 vaccine. A statistical signal occurred for two outcomes in one database (significance level of 1%): anaphylaxis following both bivalent COVID-19 vaccines in persons 18-64 years and myocarditis/pericarditis following bivalent BNT162b2 vaccines in individuals 18-35 years. Among 642 142 vaccinated children 6 months-17 years, no signals were identified.</p><p><strong>Conclusions: </strong>Results were consistent with published COVID-19 vaccine safety studies and support the safety profile of bivalent COVID-19 mRNA vaccines.</p>","PeriodicalId":19782,"journal":{"name":"Pharmacoepidemiology and Drug Safety","volume":"34 5","pages":"e70151"},"PeriodicalIF":2.4,"publicationDate":"2025-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144014639","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Chronic Disease Management to Enhance Medication Adherence Trajectories in Long-Term Survivors of Stroke: A Population-Based Cohort Study.","authors":"Lachlan L Dalli, Monique F Kilkenny, Muideen T Olaiya, David Ung, Joosup Kim, Leonid Churilov, Dominique A Cadilhac, Vijaya Sundararajan, Amanda G Thrift, Mark R Nelson, Natasha A Lannin, Rebecca Barnden, Velandai Srikanth, Nadine E Andrew","doi":"10.1002/pds.70148","DOIUrl":"10.1002/pds.70148","url":null,"abstract":"<p><strong>Purpose: </strong>Although chronic disease management (CDM) has been reported to improve medication adherence after stroke or transient ischaemic attack (TIA), the impact on specific patterns of medication adherence is unclear. We aimed to evaluate the population effect of receiving a CDM claim on trajectories of medication adherence in long-term survivors of stroke/TIA.</p><p><strong>Methods: </strong>A cohort study was undertaken using observational data from PRECISE (42 Australian Stroke Clinical Registry hospitals [Victoria and Queensland; 2012-2015] linked with medication dispensing and primary care claims). Community-dwelling adults with ≥ 1 primary care visit were included. The exposure was a CDM claim (versus no claim) in primary care within 7-18 months post-stroke/TIA. Medication adherence (antihypertensive, antithrombotic, lipid-lowering) was assessed between 19 and 30 months post-stroke/TIA, using group-based trajectory models. Average treatment effects were estimated using multi-level logistic regression with inverse probability treatment weights.</p><p><strong>Results: </strong>Among 11 580 survivors of stroke/TIA (median age 70 years, 42% female; 45% with CDM claim), four distinct adherence patterns were identified: near-perfect adherence, high adherence, declining adherence, and non-use. After adjustment, having a CDM claim (vs no claim) promoted near-perfect adherence (odds ratio [OR]: 1.16 [95% CI 1.08-1.25]) for antithrombotic medications. Whereas, having a CDM claim (vs no CDM claim) promoted high adherence for antihypertensive (OR: 1.33 [95% CI 1.24-1.44]) or lipid-lowering (OR: 1.26 [95% CI 1.16-1.37]) medications. The odds of non-use were also reduced by 17%-23% in those with (vs without) a CDM claim.</p><p><strong>Conclusions: </strong>CDM claims were associated with favourable trajectories of medication adherence in long-term survivors of stroke/TIA.</p>","PeriodicalId":19782,"journal":{"name":"Pharmacoepidemiology and Drug Safety","volume":"34 5","pages":"e70148"},"PeriodicalIF":2.4,"publicationDate":"2025-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12050132/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143986776","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Trends in Opioid and Gabapentinoid Utilization: A Time-Series Analysis Across 72 Countries From 2012 to 2023.","authors":"Yilei Liu, Scott D Rothenberger, Mina Tadrous, Bryant Shuey, Shanzeh Chaudhry, Katie J Suda","doi":"10.1002/pds.70149","DOIUrl":"https://doi.org/10.1002/pds.70149","url":null,"abstract":"<p><strong>Purpose: </strong>We compare trends in gabapentinoid and opioid utilization overall and by economic development category. We also sought to predict future trends and assess correlations in gabapentinoid and opioid utilization.</p><p><strong>Methods: </strong>We conducted a repeated cross-sectional analysis of retail prescriptions for 72 countries from Q1 2012 to Q3 2023. We measured standardized units/1000 population for gabapentinoid and opioid sales, stratified by development category, and used time-series models to predict trends for the following 3 years. Granger causality tests examined predictive relationships between gabapentinoid and opioid sales.</p><p><strong>Results: </strong>Global gabapentinoid annual sales rose by 114.5% from 2012 to 2022, with a higher increase in developing (180.9%) than developed economies (110.0%). In contrast, annual opioid sales declined globally by 25.4%, with a 27.9% decrease in developed and a 16.8% increase in developing economies. Assuming current trends persist over the following 3 years, gabapentinoid quarterly sales are forecasted to rise by 7.7% in developed and 18.6% in developing economies, while opioid quarterly sales are expected to decrease by 9.5% and increase by 15.1%, respectively. Granger causality tests indicated that gabapentinoids may predict opioid sales globally for the following year, but opioids did not predict gabapentinoid sales.</p><p><strong>Conclusion: </strong>We evaluated the global trends in gabapentinoid and opioid sales, suggesting important differences in pain management practices across developed and developing economies. Our findings highlight the need to ensure the safe use of gabapentinoids and opioids while balancing proper pain management.</p>","PeriodicalId":19782,"journal":{"name":"Pharmacoepidemiology and Drug Safety","volume":"34 5","pages":"e70149"},"PeriodicalIF":2.4,"publicationDate":"2025-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143974363","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"How Effective Are Machine Learning and Doubly Robust Estimators in Incorporating High-Dimensional Proxies to Reduce Residual Confounding?","authors":"Mohammad Ehsanul Karim, Yang Lei","doi":"10.1002/pds.70155","DOIUrl":"10.1002/pds.70155","url":null,"abstract":"<p><strong>Background: </strong>Residual confounding presents a persistent challenge in observational studies, particularly in high-dimensional settings. High-dimensional proxy adjustment methods, such as the high-dimensional propensity score (hdPS), are widely used to address confounding bias by incorporating proxies for unmeasured confounders. Extensions of hdPS have integrated machine learning, such as LASSO and super learner (SL), and doubly robust estimators, such as targeted maximum likelihood estimation (TMLE). However, the comparative performance of these methods, especially under different learner configurations and high-dimensional proxies, remains unclear.</p><p><strong>Method: </strong>We conducted plasmode simulations to evaluate the performance of standard methods, SL, TMLE, and double cross-fit TMLE (DC-TMLE) under varying exposure and outcome prevalence scenarios. Learner libraries included: 1 learner (logistic regression), 3 learners (logistic regression, MARS, and LASSO), and 4 learners (adding XGBoost, a non-Donsker learner). Metrics included bias, coverage, and variability.</p><p><strong>Results: </strong>Methods without proxies exhibited the highest bias and poorest coverage, highlighting the critical role of proxies in confounding adjustment. Standard methods incorporating high-dimensional proxies showed robust performance, achieving low bias and near-nominal coverage. TMLE and DC-TMLE reduced bias but exhibited worse coverage compared to standard methods, particularly with larger learner libraries. Notably, DC-TMLE, expected to address under-coverage issues, failed to perform adequately in high-dimensional settings with non-Donsker learners, further emphasizing the instability introduced by complex libraries.</p><p><strong>Conclusion: </strong>Our findings underscore the utility of high-dimensional proxies in standard methods and the importance of tailoring learner configurations in SL and TMLE to ensure reliable confounding adjustment in high-dimensional contexts.</p>","PeriodicalId":19782,"journal":{"name":"Pharmacoepidemiology and Drug Safety","volume":"34 5","pages":"e70155"},"PeriodicalIF":2.4,"publicationDate":"2025-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12076102/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144027811","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Association Between β-Adrenoreceptor Agonists and Antagonists and Parkinson's Disease: Systematic Review and Meta-Analysis.","authors":"Agnieszka Szmigiel, Miguel Monteiro da Rocha, Kate Browne, Daniel Morales, David Benee Olsen, Charlotte Warren-Gash, Ian Douglas, Krishnan Bhaskaran, Helena Carreira","doi":"10.1002/pds.70140","DOIUrl":"10.1002/pds.70140","url":null,"abstract":"<p><strong>Background: </strong>β-agonists and β-antagonists are among the most prescribed drugs worldwide. In 2018, studies suggesting a harmful association between propranolol and Parkinson's disease (PD) prompted a signal procedure by the European Medicines Agency's safety committee, which concluded with no update of product information. Several studies have been published since then. We aimed to systematically review, critically appraise, and meta-analyse all studies on the association between the use of β-antagonists (including propranolol) and β-agonists, and the risk of PD.</p><p><strong>Methods: </strong>We searched Embase and Medline up to December 2024 for observational and intervention studies that reported relative risk estimates of the association between use of these medicines and PD. Two reviewers screened the records, extracted the data, and assessed the risk of bias. The restricted maximum likelihood method was used to compute pooled effect estimates and 95% confidence intervals (CIs).</p><p><strong>Results: </strong>Twenty-two studies were eligible. Overall, 20 had a high risk of bias in at least one domain. Twelve studies had medium to high risk of outcome misclassification. Of the 14 studies concerning β-antagonists, eleven had an unclear or high risk of protopathic bias, as propranolol is indicated for the treatment of essential tremor. Control for confounding by socio-economic status, area of residence (urban/rural), and smoking (a protective factor against PD) was deficient or lacking in 9/22, 15/22, and 12/22 studies, respectively. Lag times were applied in 9/22 studies. In meta-analysis, the summary relative risk (RR) of PD was 1.41 (95% CI: 1.18-1.68) for the class of β-antagonists (12 studies) and 0.93 (0.84-1.03) for β2-agonists (11 studies). Among specific β-antagonists, the summary RR of PD was 2.36 (1.66-3.36) for propranolol (7 studies), 0.84 (0.80-0.88) for carvedilol (3 studies) and 1.02 (0.87-1.18) for metoprolol (4 studies). For specific β2-agonists, summary RR was 0.88 (0.77-1.01) for salbutamol (7 studies), 0.91 (0.88-0.95) for short-acting β2-agonists (6 studies), and 0.85 (0.76-0.96) for long-acting β2 agonists (5 studies). Restricting to subgroups based on quality criteria resulted in weaker or non-statistically significant associations.</p><p><strong>Conclusion: </strong>The quality and quantity of the available evidence do not support a causal association between use of β-adrenoreceptor modulators and PD. Significant associations are most likely explained by protopathic bias and confounding.</p>","PeriodicalId":19782,"journal":{"name":"Pharmacoepidemiology and Drug Safety","volume":"34 4","pages":"e70140"},"PeriodicalIF":2.4,"publicationDate":"2025-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11979683/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143811988","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}