Journal of Medical Economics最新文献

{"title":"TIME-trend and forecasting of the economic impact of inflammatory bowel disease in Catalonia: a population-based analysis.","authors":"Belén Garcia-Sagué, Eduard Brunet-Mas, Emili Vela, Stephanie Coward, Xavier Calvet, Caridad Pontes, Luigi Melcarne, Luis Enrique Frisancho, Laura Patricia Llovet, Patricia Pedregal-Pascual, Anna Puy, Maria José Ramirez-Lazaro, Albert Villoria, Sergio Lario, Gilaad G Kaplan","doi":"10.1080/13696998.2026.2692254","DOIUrl":"https://doi.org/10.1080/13696998.2026.2692254","url":null,"abstract":"<p><strong>Introduction: </strong>Few studies have assessed the distribution and temporal trends of healthcare expenditures related to inflammatory bowel disease (IBD). The aims of the study were to analyze the expenditure trends for IBD patients in Catalonia from 2011 to 2024, to identify key cost drivers, and to forecast future costs through 2036.</p><p><strong>Methods: </strong>All patients with a diagnosis of IBD included in the Catalan Health Surveillance System (CHSS) were eligible. CHSS compiles prospective data from public healthcare coverage of 8 million people in 2024. Healthcare utilization was analyzed, and expenditures were calculated using standard costs defined by the Catalan Department of Health. All expenditures were adjusted to 2024 euros using the Consumer Price Index to enable comparisons across years. Costs were estimated for IBD overall, Crohn's disease (CD) and ulcerative colitis (UC). An autoregressive integrated moving average model was used to forecast total costs up to 2036.</p><p><strong>Results: </strong>IBD-related healthcare expenditure tripled from €67.4 M in 2011 to €201.6 M in 2024, while per-patient costs rose from €3,981 to €4,753. Biologic therapies were the main cost driver, especially in CD. Mean per-patient biologic costs decreased by 16.6% in CD but increased by 44.5% in UC. CD patients consistently incurred higher per capita costs. Forecasts indicate continued growth in total expenditure, reaching €319.0 M by 2036.</p><p><strong>Conclusion: </strong>Overall, IBD-related healthcare expenditures in Catalonia markedly increased from 2011 to 2024, driven mainly by the increase in IBD prevalence. Per-patient cost moderately increased. Per-patient cost containment was observed in pharmaceutical costs, probably due to a strict policy favoring the use of biosimilars. These findings may be of help for designing future healthcare policies.</p>","PeriodicalId":16229,"journal":{"name":"Journal of Medical Economics","volume":"29 1","pages":"1849-1863"},"PeriodicalIF":3.1,"publicationDate":"2026-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148375990","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":"Cost-effectiveness of vagus nerve-mediated neuroimmune modulation for the treatment of moderate-to-severe rheumatoid arthritis.","authors":"Ellen L Air, Anne M Ryschon, Joshua M Rosenow, Abigail M Garner, Robert W Levin, Peter E Konrad, Jan B Pietzsch","doi":"10.1080/13696998.2026.2662122","DOIUrl":"10.1080/13696998.2026.2662122","url":null,"abstract":"<p><strong>Aims: </strong>Rheumatoid arthritis (RA) is associated with increased healthcare resource utilization and high-cost pharmacologic treatment. The objective of the current analysis was to evaluate the cost-effectiveness of vagus nerve-mediated neuroimmune modulation therapy (NIMT) using a novel implantable neurostimulation technology compared to status-quo pharmacologic treatment of moderate-to-severe rheumatoid arthritis patients in the US.</p><p><strong>Materials and methods: </strong>A decision-analytic Markov model was utilized to project strategy-specific costs and outcomes over 2-, 10-year, and lifetime horizons for NIMT and the status-quo. Clinical data from the RESET-RA study informed key model inputs including cohort characteristics, utilities, event rates, and medication utilization. Costs were derived from published literature and current reimbursement rates. The analysis was conducted from the US payer perspective, with both costs and outcomes discounted 3% in accordance with US cost-effectiveness guidelines. Cost-effectiveness was evaluated using established US willingness-to-pay thresholds of $50,000 (highly cost-effective) and $150,000 (cost-effective) per quality-adjusted life year (QALY) gained. Deterministic and probabilistic sensitivity analyses (PSA) were performed.</p><p><strong>Results: </strong>Under the base case assumptions, NIMT was cost-saving in less than 2 years. Over lifetime, NIMT was associated with incremental cost-savings of $350,052 and QALY gain of 0.87, and led to cost-savings of $25,397 and $197,062, and concurrent incremental QALY gains of 0.10 and 0.46 at 2 and 10 years, rendering NIMT the dominant strategy across all three horizons. In the PSA, NIMT was cost-effective at thresholds of $50,000 and $150,000 per QALY gained in 96.25% and 99.20% of simulations at 2 years, and 100% of simulations at 10 years and lifetime.</p><p><strong>Conclusion: </strong>In this model-based analysis of a recent randomized trial, the use of NIMT therapy corresponded to a reduction in total costs of care, improved quality of life, and was found to be a highly cost-effective or dominant treatment option for moderate-to-severe rheumatoid arthritis in a US setting.</p>","PeriodicalId":16229,"journal":{"name":"Journal of Medical Economics","volume":"29 1","pages":"1462-1473"},"PeriodicalIF":3.1,"publicationDate":"2026-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147930563","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":"Real-world burden of hypertrophic cardiomyopathy on healthcare use and health-related quality of life in Italy, Spain, and the United States.","authors":"Paulos Gebrehiwet, James Jackson, Lucy Hargreaves, Sophie Barlow, Laura LeBrocq, Johan Brekke, Michael Butzner, Elyce Freeman, Sanatan Shreay, Keitaro Akita, Yuichi Shimada","doi":"10.1080/13696998.2026.2671565","DOIUrl":"10.1080/13696998.2026.2671565","url":null,"abstract":"<p><strong>Background and aims: </strong>Hypertrophic cardiomyopathy (HCM) is a chronic, progressive, genetic heart disease characterized by hypercontractility and hypertrophy. As treatment options evolve, characterizing the existing healthcare needs and health-related quality of life (HRQoL) for patients with HCM is important.</p><p><strong>Methods: </strong>Data were derived from the Adelphi Real World HCM Disease Specific Programme, a cross-sectional survey in Italy, Spain, and the United States. Physicians reported demographic and clinical characteristics; patients completed the Kansas City Cardiomyopathy Questionnaire (KCCQ-23), EQ-5D-5L and EQ-VAS, and the Work Productivity and Activity Impairment (WPAI) questionnaire. Outcomes were statistically compared by New York Heart Association (NYHA) class or KCCQ clinical summary score (KCCQ-CSS) to assess impact of disease severity.</p><p><strong>Results: </strong>Patient mean age was 56.5 ± 14.8 years; 60.7% were male, and 78.3% had NYHA class II disease. Patients most commonly received beta blockers (88.4%) and diuretics (31.8%). HCM-related hospitalizations were reported for 25.8% of patients, increasing with higher NYHA or lower KCCQ-CSS. Mean EQ-5D-5L was 0.85, while EQ-VAS was 75.0, both decreasing with higher NYHA class or lower KCCQ-CSS. Overall work impairment and activity impairment (WPAI) were 16.1% and 26.6%, respectively, increasing with higher NYHA class or lower KCCQ-CSS.</p><p><strong>Conclusion: </strong>Our data show a significant remaining symptom burden and healthcare resource utilization (HCRU), as well as impact on HRQoL, work productivity, and daily activities, despite treatment intervention across a commonly applied measure of disease severity in HCM. These findings underscore the need for new treatments that can reduce HCRU and improve HRQoL of patients with HCM.</p>","PeriodicalId":16229,"journal":{"name":"Journal of Medical Economics","volume":"29 1","pages":"1537-1554"},"PeriodicalIF":3.1,"publicationDate":"2026-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148138340","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":"Rett syndrome and real-world treatment patterns of trofinetide in the United States.","authors":"Nazia Rashid, Vinod Kumar Yakkala, Safiuddin Shoeb Syed, Krithika Rajagopalan","doi":"10.1080/13696998.2026.2682668","DOIUrl":"10.1080/13696998.2026.2682668","url":null,"abstract":"<p><strong>Background: </strong>Trofinetide (TROF) remains the first and only FDA-approved pharmacologic treatment for Rett syndrome (RTT). There is limited real-world evidence on TROF use, restarts, dosing patterns and predictors of non-persistence. This study evaluated long-term treatment patterns and baseline predictors of non-persistence among RTT individuals initiating TROF.</p><p><strong>Methods: </strong>A retrospective claims analysis of linked IQVIA Anonymized Patient Level Data and TROF pharmacy data (1 January 2021 to 30 September 2024) was performed. RTT individuals initiating 1^st TROF prescription (RX) between 1 April 2023 and 31 March 2024 with ≥6 months of pre- and post-index continuous enrollment were classified as persistent (gap ≤90 days) and non-persistent. Median (IQR) time on treatment was analyzed, and Kaplan-Meier assessed time to non-persistence. Restarts among non-persistent were examined. Dosing patterns such as mean dose in milligrams (mg) and percentage target daily dose (%TDD) at each RX were evaluated. Predictors of non-persistence were assessed using logistic regression.</p><p><strong>Results: </strong>Among 1,175 TROF initiators, 54.9% were persistent and 45.1% were non-persistent; 7.6% of non-persistent individuals restarted. Mean ± SD age was 14.7 ± 10.8 vs. 16.5 ± 11.6 years and median ± IQR time on TROF was 14.3 ± 4.6 vs. 3.0 ± 3.8 months, respectively. Over 75% remained on treatment beyond 3 months, and more than half continued through study follow-up. Mean twice daily (BID) TROF dose was lower among persistent (7422.0 mg) vs. non-persistent (7850.8 mg) at RX1 and increased to 7953.7 mg vs 8423.4 mg, respectively at RX2. Thereafter, the persistent individuals maintained stable dosing, while the non-persistent showed greater fluctuation over time. Similar patterns were observed for %TDD. Older age and history of infectious disorders were significantly associated with non-persistence.</p><p><strong>Conclusions: </strong>In this real-world analysis, more than half of RTT individuals remained persistent on TROF for ≥14 months. Stable dosing patterns were observed among persistent individuals, while greater dose variability was observed among non-persistent individuals. Thus, these findings may provide guidance to clinicians about long-term TROF use in real-world practice.</p>","PeriodicalId":16229,"journal":{"name":"Journal of Medical Economics","volume":"29 1","pages":"1715-1727"},"PeriodicalIF":3.0,"publicationDate":"2026-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148239402","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":"Casemix-based budget impact analysis of ferric derisomaltose versus ferric carboxymaltose in patients with inflammatory bowel disease and iron deficiency anaemia in the Netherlands.","authors":"U H Oron, G Koek, R Fijn","doi":"10.1080/13696998.2026.2692303","DOIUrl":"https://doi.org/10.1080/13696998.2026.2692303","url":null,"abstract":"<p><strong>Aims: </strong>The prevalence of inflammatory bowel disease (IBD) in the Netherlands is approximately 100,000 individuals currently (578 per 100,000 inhabitant), and is rising. Iron deficiency anaemia (IDA) is common in patients with IBD, secondary to chronic blood loss and impaired iron absorption. Intravenous iron is recommended as first-line treatment in patients with clinically active IBD, with previous intolerance to oral iron and in patients who need erythropoietin stimulating agents. The objective was to evaluate the difference in treatment costs and consequential costs of ferric derisomaltose (FDI) versus ferric carboxymaltose (FCM) in patients with IBD and IDA in the Netherlands, based on a casemix model using hospital data from the DBC (<i>Diagnose Behandel Combinatie</i>) system. This study aims to elucidate the differences in budget impact, from a health insurers perspective.</p><p><strong>Materials and methods: </strong>In order to model the hospital business, a coding analysis was done, to identify the relevant codes, tariffs, and volumes. In a casemix scenario analysis, the clinical differences between the use of FCM and FDI have been modelled, with a time horizon of 5 years.</p><p><strong>Results: </strong>The analysis showed that with FDI the overall costs were lower than with FCM. With full use of FDI in the eligible population, the potential annual savings excluding complications in the national health insurer perspective is €2.9 million (excl. VAT) on iron infusion cost only. Including related costs and complications, the potential annual savings are €28.2 million. The sensitivity analyses proved the outcome to be robust.</p><p><strong>Limitations: </strong>Whilst using the DBC data would seem to reflect the real-life clinical behavior, it is known that the DBC system may show some underreporting. The analysis timeframe is limited to 5 years.</p><p><strong>Conclusions: </strong>Results showed that FDI significantly reduces healthcare expenditures versus FCM in patients with IBD and IDA in the Netherlands.</p>","PeriodicalId":16229,"journal":{"name":"Journal of Medical Economics","volume":"29 1","pages":"1984-1997"},"PeriodicalIF":3.1,"publicationDate":"2026-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148471148","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":"Cost-effectiveness of treatment and care of patients with gastrointestinal inflammatory diseases: a systematic review.","authors":"Kristina Randlová, Petra Marešová, Lukáš Režný, Jan Hruška, Sonja Srdanović, Guðmundur H Guðmundsson, Egill Másson, Kamil Kuča","doi":"10.1080/13696998.2026.2637360","DOIUrl":"10.1080/13696998.2026.2637360","url":null,"abstract":"<p><strong>Aims: </strong>The aim of this study was to systematically review and descriptively synthesize published cost-effectiveness evidence for biologic versus non-biologic treatments as well as comparisons among different biologic therapies and treatment sequencing strategies, in inflammatory bowel disease, highlighting patterns and variability across disease types, treatment strategies, and study designs.</p><p><strong>Materials and methods: </strong>We conducted a systematic review following the PRISMA guidelines, searching the Web of Science and PubMed databases for studies published between 2013 and 2024. Studies comparing the cost-effectiveness of treatments for inflammatory bowel disease were included.</p><p><strong>Results: </strong>Eighteen studies met the inclusion criteria, covering Crohn's disease and/or ulcerative colitis. Biologic therapies were generally associated with superior health outcomes compared with conventional treatments. Reported quality-adjusted life years (QALYs) ranged from 2.23 to 18.12 for biologic therapies and from 1.69 to 17.99 for non-biologic treatments. Although biologic therapies have higher costs, they are generally considered cost-effective. For Crohn's disease, infliximab was reported as a cost-effective biologic option, while findings for ulcerative colitis varied. Surgical intervention (colectomy) was identified as a cost-effective in selected clinical scenarios.</p><p><strong>Conclusion: </strong>Biologic therapies for inflammatory bowel disease are cost-effective, providing significant health benefits that offset higher costs. Substantial methodological heterogeneity and reliance on model-based economic evaluations limit direct comparability across studies. Future economic evaluations should focus on methodological consistency and transparency in assumptions to strengthen the interpretability of cost-effectiveness evidence.</p>","PeriodicalId":16229,"journal":{"name":"Journal of Medical Economics","volume":"29 1","pages":"919-939"},"PeriodicalIF":3.1,"publicationDate":"2026-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147491152","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":"Cost-effectiveness of exagamglogene autotemcel gene-edited therapy in patients with sickle cell disease with recurrent vaso-occlusive crises in the United States.","authors":"Andrea Lopez, Michael Gargano, Hongbo Yang, Foluso Joy Ogunsile, Nanxin Li, Yanwen Xie, Sushanth Jeyakumar, Chuka Udeze","doi":"10.1080/13696998.2026.2624971","DOIUrl":"10.1080/13696998.2026.2624971","url":null,"abstract":"<p><strong>Objective: </strong>Exagamglogene autotemcel (exa-cel) is a one-time nonviral gene-edited therapy approved in the United States (US) for treatment of patients aged ≥12 years with sickle cell disease (SCD) with recurrent vaso-occlusive crises (VOCs). Standard of care (SOC) for SCD includes symptomatic care, hydroxyurea and/or red blood cell transfusions. This study estimated the long-term clinical outcomes and cost-effectiveness of exa-cel relative to SOC among patients with SCD with recurrent VOCs.</p><p><strong>Methods: </strong>A Markov model was used to compare the expected lifetime costs and clinical outcomes of patients with SCD with recurrent VOCs treated with exa-cel versus SOC from the US payer and societal perspectives. The model structure is based on disease severity, characterized by VOC frequency, which impacts the risk of developing SCD-related complications and mortality. The model incorporated data from the phase 3 pivotal CLIMB SCD-121 trial alongside published literature. Model outcomes included number of VOCs and other acute complications, proportion of patients developing chronic complications, life years (LYs), quality-adjusted LYs (QALYs), costs, and incremental cost-effectiveness ratios (ICERs).</p><p><strong>Results: </strong>Over a lifetime horizon, exa-cel was projected to improve survival by 30.8 years (mean age of death, exa-cel: 74.5 vs. SOC: 43.6), reduce the number of VOC events by 77 (7 vs. 84), and reduce undiscounted disease-related costs by $3.34 M ($0.55 M vs. $3.89 M) compared to treatment with SOC. Patients treated with exa-cel also were less likely to experience acute complications or develop chronic complications compared to SOC. The ICER per discounted QALY for exa-cel versus SOC was $16,800 from the payer perspective; exa-cel was dominant (less costly, more effective than SOC) from the societal perspective.</p><p><strong>Conclusions: </strong>Compared to SOC, exa-cel was projected to considerably reduce the number of VOCs, improve survival, and reduce disease-related costs in patients with SCD. Exa-cel was projected to be a cost-effective treatment option.</p>","PeriodicalId":16229,"journal":{"name":"Journal of Medical Economics","volume":"29 1","pages":"547-562"},"PeriodicalIF":3.1,"publicationDate":"2026-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147276172","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":"Breaking barriers in women's pelvic health: claims-based economic analysis and healthcare utilization of an AI care program compared to usual care.","authors":"Ana P Pereira, Andrea M Seet, Beatriz Domingues, Dora Janela, Akshat Pradhan, Anabela C Areias, Xin Tong, Virgílio Bento, Vijay Yanamadala, Steven P Cohen, Jennesa Atherton, Fernando Dias Correia, Luke Belz, Fabíola Costa","doi":"10.1080/13696998.2026.2627835","DOIUrl":"10.1080/13696998.2026.2627835","url":null,"abstract":"<p><strong>Aim: </strong>To compare healthcare utilization and spending among women enrolled in an employer-sponsored, artificial intelligence (AI) structured pelvic care program with those receiving usual in-person care for pelvic floor dysfunction (PFD) in routine clinical settings.</p><p><strong>Methods: </strong>This retrospective payor-perspective economic evaluation used exact and propensity score-matched cohorts derived from a third-party U.S. nationwide claims database from July 2022 to May 2025. Eligible participants were adult females with a pelvic-related condition, at least 24 months of continuous health-insurance coverage, and a minimum of one pelvic claim in the prior year. Intervention group (IG) comprised women who participated in the AI pelvic care program (consisting of biofeedback-mediated pelvic floor muscle training asynchronously monitored by a physical therapist specialized in pelvic health). Comparator group (CG) included women who sought a medical or physical therapy evaluation visit for PFD. Self-reported clinical outcomes available for the IG were assessed using latent-basis growth analysis.</p><p><strong>Results: </strong>The matched cohort included 602 women (301 per group). Relative to CG, IG patients had substantially lower healthcare spending over 12 months, with mean gross per-person pelvic-related savings of $3,082.4 (95% CI $1,270.2 to $4,894.7, <i>p</i><.001). Savings were primarily associated with fewer surgical procedures (per-person difference of $2,534.2; 95% CI $831.2 to $4,237.2, <i>p</i>=.004), with differences also noted in medical office visits and imaging utilization. IG participants demonstrated significant improvements in pelvic floor symptom burden, work productivity, and mental health.</p><p><strong>Limitations: </strong>Claims-based analyses cannot exclude unmeasured confounding, misclassification, or selection bias. The one-year follow-up limits assessment of long-term economic impact.</p><p><strong>Conclusions: </strong>Participation in this AI pelvic care program was associated with markedly lower healthcare utilization and spending compared with usual care, largely linked to fewer surgical interventions. These findings highlight the potential of accessible, guideline-concordant AI pelvic care to lessen healthcare spending associated with PFD and inform payor-oriented care delivery models.</p>","PeriodicalId":16229,"journal":{"name":"Journal of Medical Economics","volume":"29 1","pages":"516-531"},"PeriodicalIF":3.1,"publicationDate":"2026-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147283933","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":"Comparing deep learning and classical regression approaches for predicting healthcare expenditure and spending: a systematic review.","authors":"John Tayu Lee, Melody Hsiao-San Yeh, Vincent Cheng-Sheng Li, Hsiao-Hui Chen, Yi-Hsuan Liu, Yu-Chun Chen, David Bin-Chia Wu","doi":"10.1080/13696998.2026.2630598","DOIUrl":"10.1080/13696998.2026.2630598","url":null,"abstract":"<p><strong>Aims: </strong>This study compares deep learning architectures with traditional regression and tree-based models for individual-level healthcare cost prediction, with particular attention to performance differences across data contexts.</p><p><strong>Methods: </strong>We conducted a preregistered systematic review (PROSPERO CRD420251129440). Web of Science, PubMed, Embase, and Scopus were searched through August 2025. Eligible studies used real-world individual-level data (claims, electronic health records, or registries) to predict cost-related outcomes with at least one deep learning architecture and one classical regression comparator, and reported quantitative performance. Data were extracted on population, predictors, outcome horizon, model type, validation strategy, performance metrics, calibration, and interpretability.</p><p><strong>Results: </strong>Eight studies met inclusion criteria, spanning the United States, Europe, and Asia. In longitudinal designs-such as multi-year claims prediction and medication or hospitalization time-series forecasting-sequential deep learning models, particularly LSTM and CNN-LSTM hybrids, consistently outperformed regression and tree-based algorithms. Reported gains included approximately 10-20% reductions in RMSE/MAE, R² improvements of 0.01-0.15, and AUROC values up to 0.78 for high-risk classification. Across studies, prior costs and utilization were consistently the strongest predictors, while social determinants and free-text features were rarely incorporated. In contrast, for low-dimensional, structured, cross-sectional medical data, generalized linear models and tree-based approaches remain robust baseline models due to their interpretability and calibration stability.</p><p><strong>Limitations: </strong>Evidence is based on a small and heterogeneous set of eight studies, with limited external or temporal validation, short prediction horizons, and sparse assessment of calibration, economic interpretability, and fairness, warranting cautious interpretation.</p><p><strong>Conclusions: </strong>Deep learning offers clear gains for longitudinal, sequence-rich cost forecasting, whereas tree-based methods remain highly competitive for cross-sectional tabular prediction. Overall, these findings are consistent with the proposed Complexity-Performance Hypothesis, which posits that the predictive advantages of deep learning emerge primarily when model capacity is well matched to data complexity.</p>","PeriodicalId":16229,"journal":{"name":"Journal of Medical Economics","volume":"29 1","pages":"654-671"},"PeriodicalIF":3.1,"publicationDate":"2026-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147355320","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":"Cost per event avoided: adopt or avoid?","authors":"P Petrou, Olga Pitsillidou, Constantinos Petrou, Maarten J Postma","doi":"10.1080/13696998.2026.2638716","DOIUrl":"https://doi.org/10.1080/13696998.2026.2638716","url":null,"abstract":"","PeriodicalId":16229,"journal":{"name":"Journal of Medical Economics","volume":"29 1","pages":"885-888"},"PeriodicalIF":3.0,"publicationDate":"2026-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147468227","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}