Value in Health最新文献

{"title":"Cost Savings and Improved Clinical Outcomes From a Mobile Health Cardiovascular Disease Self-Management Program.","authors":"Walter Roberts, Helena Lyson, Clint Speer, Elena Tovar, Edo Paz, Eyal Zimlichman","doi":"10.1016/j.jval.2025.01.025","DOIUrl":"10.1016/j.jval.2025.01.025","url":null,"abstract":"<p><strong>Objectives: </strong>This study evaluated the impact of a mobile health (mHealth) cardiovascular disease (CVD) self-management program on medical spending, healthcare utilization, and clinical outcomes.</p><p><strong>Methods: </strong>We conducted a retrospective cohort, pre-post observational analysis of medical claims data from February 2018 to September 2023. Participants enrolled in the mHealth CVD self-management program were compared with matched nonparticipants to assess changes in medical spending and utilization, using a difference-in-differences (DiD) approach. Moderation analyses tested whether the effect of participation on medical costs differed according to participants' demographic and employer characteristics. Linear mixed models were also used to evaluate changes in participant blood pressure.</p><p><strong>Results: </strong>There were 7112 participants and an equal number of matched nonparticipants from 14 employers included in the cost-savings analysis. Participation in the mHealth CVD self-management program was associated with an annualized per member savings of $1709 as compared with matched nonparticipants. Although savings were consistent across participant sex and employer type, older participants experienced greater savings. Program participation was associated with fewer inpatient hospital days and increased primary care utilization after enrollment. Program participants experienced significant reductions in blood pressure, with the largest reductions seen among those with stage 2 hypertension at baseline.</p><p><strong>Conclusions: </strong>Participation in an mHealth CVD self-management program was associated with significant cost savings and clinical improvements across diverse populations, highlighting its effectiveness as a cost-efficient tool for managing CVD and improving health outcomes. Further research is needed to explore the long-term benefits and cost-effectiveness of mHealth programs to support broader adoption.</p>","PeriodicalId":23508,"journal":{"name":"Value in Health","volume":" ","pages":""},"PeriodicalIF":4.9,"publicationDate":"2025-02-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143426300","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A Scoping Review of Interventions to Reduce the Environmental Footprint of Healthcare.","authors":"Anna-Veera Seppänen, Zeynep Or","doi":"10.1016/j.jval.2025.01.021","DOIUrl":"10.1016/j.jval.2025.01.021","url":null,"abstract":"<p><strong>Objectives: </strong>The healthcare sector contributes significantly to global warming; yet, strategies for reducing its impact are not well integrated into health policy. This scoping review aimed to identify the range of effective interventions that can reduce the environmental footprint of healthcare and to provide an overview of their impact.</p><p><strong>Methods: </strong>We searched for peer-reviewed articles published in English, French, or Swedish between 2010 and September 2024 in Medline and Web of Science, following the Joanna Briggs Institute guidelines and the Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Reviews. Publications were selected by 2 researchers and a documentalist. Data from included studies were extracted and synthesized in tables and described in a narrative synthesis.</p><p><strong>Results: </strong>We identified 7 systematic reviews and 44 original research articles. Most of the effective interventions targeted hospitals and varied from energy saving practices and reducing potent anesthetic gases to changing care protocols and improving waste management. The measured impact of interventions was context specific and depended on national energy sources. Only a few studies reported on the impact of structural and strategic changes in healthcare provision across care settings.</p><p><strong>Conclusions: </strong>There is an urgent need for better understanding the costs and benefits of diffusing effective green interventions across care providers and developing more systemic approaches for optimizing care provision and utilisation to achieve a meaningful impact.</p>","PeriodicalId":23508,"journal":{"name":"Value in Health","volume":" ","pages":""},"PeriodicalIF":4.9,"publicationDate":"2025-02-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143426207","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Statistical Methods for Analyzing EQ-5D in Randomized Clinical Trials: A Systematic Literature Review.","authors":"Jiajun Yan, Brittany Humphries, Ruinan Xie, Ziran Yin, Zhenyan Bo, Sha Diao, Jing Cai, Preston Tse, Meixuan Li, Eleanor Pullenayegum, Shun Fu Lee, Feng Xie","doi":"10.1016/j.jval.2025.02.001","DOIUrl":"10.1016/j.jval.2025.02.001","url":null,"abstract":"<p><strong>Objectives: </strong>We conducted a systematic literature review to summarize the application of statistical methods for analyzing treatment effect on EQ-5D in randomized clinical trials (RCTs).</p><p><strong>Method: </strong>We searched 2 electronic databases (MEDLINE and EMBASE, from inception through 2021) and www.</p><p><strong>Clinicaltrial: </strong>gov. Eligible studies were RCTs that analyzed postbaseline EQ-5D data by treatment group. Information on trial characteristics, EQ-5D data characteristics, and statistical methods were extracted. Descriptive statistics were used to summarize results by dimension response, EQ visual analog scale (EQ VAS), and EQ-5D utility.</p><p><strong>Results: </strong>A total of 2125 trials met the eligibility criteria. EQ-5D was commonly considered a secondary (n = 1219, 57.4%) or exploratory (n = 775, 36.5%) endpoint in RCTs. EQ-5D utilities were the most analyzed. Both utilities and EQ VAS were primarily analyzed in numerical format. The most common statistical models for analyzing utilities were the linear fixed-effect model for single postbaseline (192/589, 32.6%) and the linear mixed-effect model for multiple post-baselines (338/984, 34.3%). Of the 2054 studies that analyzed numerical EQ-5D, 221 (10.8%) examined model assumptions and 438 (21.3%) adjusted for the baseline score. Missing data were explicitly assessed in 661 trials, among which 347 (52.5% of 661) applied imputations, with the 2 most used imputation methods being multiple imputations (n = 200, 57.6% of 347) and last observation carried forward (n = 106, 30.5% of 347).</p><p><strong>Conclusions: </strong>This review found that health utilities are the most frequently analyzed EQ-5D data collected in clinical trials, followed by EQ VAS. Significant variation was observed in the selection of models, with most trials lacking adjustments for baseline data and appropriate methods for handling missing data.</p>","PeriodicalId":23508,"journal":{"name":"Value in Health","volume":" ","pages":""},"PeriodicalIF":4.9,"publicationDate":"2025-02-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143426312","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Mapping the Landscape of Open Source Health Economic Models: A Systematic Database Review and Analysis: An ISPOR Special Interest Group Report.","authors":"Raymond H Henderson, Chris Sampson, Xavier G L V Pouwels, Stephanie Harvard, Ron Handels, Talitha Feenstra, Ramesh Bhandari, Aryana Sepassi, Renée Arnold","doi":"10.1016/j.jval.2025.01.019","DOIUrl":"10.1016/j.jval.2025.01.019","url":null,"abstract":"<p><strong>Objectives: </strong>Health economic models are crucial for health technology assessments to evaluate the value of medical interventions. Open-source models (OSMs), in which source code and calculations are publicly accessible, enhance transparency, efficiency, credibility, and reproducibility. This study systematically reviewed databases to map the landscape of available OSMs in health economics.</p><p><strong>Methods: </strong>A systematic database review was conducted, informed by guidance from ISPOR's OSM Special Interest Group. Eleven databases and specific OSM repositories were searched using predefined terms. Identified models were screened and duplicates were removed.</p><p><strong>Results: </strong>The search yielded 8664 hits, resulting in 182 unique OSMs. GitHub hosted the majority (74%), followed by Zenodo (11%). R was the predominant software platform (64%). Infectious disease was the most common application domain (29%). Markov models were the most frequent model type (49%). Licensing with Creative Commons was typical. Government and academic institutions were the primary sponsors, although many models lacked clear sponsorship.</p><p><strong>Conclusions: </strong>This review highlights the diversity and availability of open-source models (OSMs) in health economics, predominantly hosted on GitHub and developed using R. The models span various medical fields, with a strong focus on infectious diseases, oncology, and neurology. Ensuring clear licensing and standardized reporting is crucial to maximizing their impact. A combined approach of repository searches and traditional literature reviews provides a comprehensive method for identifying OSMs. Future efforts should enhance search strategies, improve reporting standards, and leverage OSMs to inform health policy decisions.</p>","PeriodicalId":23508,"journal":{"name":"Value in Health","volume":" ","pages":""},"PeriodicalIF":4.9,"publicationDate":"2025-02-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143426305","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Newer Glucagon-Like Peptide-1 Receptor Agonists Are Associated With Improved Glycemic Control in US Adults With Type 2 Diabetes: A Population-Level Time Series Analysis.","authors":"Lei Lv, Yan Wang, Lin Xie, Josh Noone, Sara Alvarez, Yuehan Zhang, Yan Song, Daniel M Rotroff","doi":"10.1016/j.jval.2025.01.018","DOIUrl":"10.1016/j.jval.2025.01.018","url":null,"abstract":"<p><strong>Objectives: </strong>The population-level impact of newer glucagon-like peptide-1 receptor agonist (GLP-1 RA) treatments on patients with type 2 diabetes remains unclear. Therefore, this study investigated the association of newer GLP-1 RAs with population-level diabetes-related outcomes in US adults with type 2 diabetes.</p><p><strong>Methods: </strong>This was a noninterventional, retrospective analysis of administrative insurance claims data obtained from Optum's deidentified Clinformatics® Data Mart Database between January 2004 and December 2022. Individual patient data were aggregated into monthly time series at the population level. Associations between the proportion of patients on newer GLP-1 RAs of interest (dulaglutide, once-weekly or oral semaglutide, and tirzepatide) and average hemoglobin A1c (HbA1c), HbA1c <7.0%, and HbA1c <8.0% were assessed using time series regressions. Time-based trends in population-level outcomes were assessed using interrupted time series analyses. All analyses were adjusted for potential confounders.</p><p><strong>Results: </strong>Time series regression analyses demonstrated that a higher proportion of patients taking newer GLP-1 RAs was associated with a lower average HbA1c and a higher proportion of patients with HbA1c <7.0% (after 5 months of treatment) and <8.0% (after 2 months of treatment). Results from the interrupted time series analyses demonstrated that population-level glycemic control has been improving since the first newer GLP-1 RA (dulaglutide) was approved. This trend has persisted, even after the COVID-19 emergency declaration.</p><p><strong>Conclusions: </strong>A higher percentage of patients receiving newer GLP-1 RAs was associated with significantly improved population-level glycemic control. The newer GLP-1 RA period displayed improved population-level glycemic control compared with the older GLP-1 RA period.</p>","PeriodicalId":23508,"journal":{"name":"Value in Health","volume":" ","pages":""},"PeriodicalIF":4.9,"publicationDate":"2025-02-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143415371","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Role of Health Equity in Health Technology Assessment Processes: A Landscape Analysis of 13 Health Systems in Asia","authors":"Chanthawat Patikorn PharmD, PhD ,&nbsp;Chia Jie Tan PhD ,&nbsp;Jeong-Yeon Cho PharmD, PhD ,&nbsp;Sarayuth Khuntha BSc ,&nbsp;Nguyen Thi Ha PhD ,&nbsp;Rini Noviyani PhD ,&nbsp;Mac Ardy J. Gloria PhD ,&nbsp;Anton L.V. Avanceña PhD ,&nbsp;Sitaporn Youngkong PhD ,&nbsp;Kyoko Shimamoto PhD ,&nbsp;Nathorn Chaiyakunapruk PharmD, PhD","doi":"10.1016/j.jval.2025.01.012","DOIUrl":"10.1016/j.jval.2025.01.012","url":null,"abstract":"<div><h3>Objectives</h3><div>This landscape analysis aimed to summarize the role of health equity in the health technology assessment (HTA) process (topic nomination, topic prioritization, assessment, appraisal, and decision making) in Asia.</div></div><div><h3>Methods</h3><div>A comprehensive literature review was conducted, followed by in-depth interviews with key informants. Content analysis was performed to summarize the role of health equity in HTA in 13 health systems in Asia, including Brunei Darussalam, Cambodia, China, Indonesia, Japan, Malaysia, Myanmar, Philippines, Singapore, South Korea, Taiwan, Thailand, and Vietnam.</div></div><div><h3>Results</h3><div>Health equity was reported to be considered in most health systems’ HTA processes, except for Cambodia and Myanmar, which do not have an established HTA process. Interviews revealed that health equity has been more frequently considered to address the unmet medical needs of specific diseases (eg, high disease burden or severity, rare diseases, cancer, and diseases affecting children and the elderly) in Brunei Darussalam, China, Japan, Malaysia, Singapore, South Korea, Taiwan, Thailand, and Vietnam or inequities in socially disadvantaged groups (eg, socioeconomic status and geographical location) in Indonesia and the Philippines. Equity-informative economic evaluation was still in the early stages, with only 3 health systems reporting their use.</div></div><div><h3>Conclusions</h3><div>Health equity is considered in the HTA process in most Asian health systems. However, quantitative evaluation of health equity impact is still in its infancy because few health systems have just begun to perform equity-informative economic evaluations.</div></div>","PeriodicalId":23508,"journal":{"name":"Value in Health","volume":"28 4","pages":"Pages 582-590"},"PeriodicalIF":4.9,"publicationDate":"2025-02-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143373993","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Exploring a Patient-Centered Methodology to Identify Suitable Proxy Conditions as a Source for Utility Data, Using Alopecia Areata as a Case Example","authors":"Alice M. Biggane PhD ,&nbsp;Sofia Schiavo MSc ,&nbsp;Rowena Randall PhD ,&nbsp;Masara Elgares MSc ,&nbsp;Patrice Carter PhD","doi":"10.1016/j.jval.2025.01.017","DOIUrl":"10.1016/j.jval.2025.01.017","url":null,"abstract":"<div><h3>Objectives</h3><div>In the absence of EQ-5D utility data, the National Institute for Health and Care Excellence recommends sourcing utility values from a proxy health condition, provided that data indicate that the health-related quality of life (HRQoL) impact of the conditions is similar. Given that there is no standard practice for identifying proxy conditions, this study presents a patient-centered approach for selecting a proxy condition, using alopecia areata (AA) as an example.</div></div><div><h3>Methods</h3><div>A structured electronic search was conducted to identify conceptual models for conditions similar to AA with overlapping HRQoL domains. Conceptual models were compared to determine the most suitable proxy condition, defined as the condition with the most overlapping domains and reported similarities between patient HRQoL. A search and comparison of utility data between AA and the selected proxy condition was conducted to validate their comparability in HRQoL impact.</div></div><div><h3>Results</h3><div>Seven conditions with overlapping HRQoL domains with AA were identified using conceptual models. Atopic dermatitis (AD) had the most overlapping domains (n = 6) and the greatest evidence of similarities between patient HRQoL; thus, it was selected as the proxy condition. Conceptual models indicated that both conditions affected patients’ physical, emotional, and functional well-being. Furthermore, AD and AA utilities were comparable, supporting the choice of AD as the proxy condition.</div></div><div><h3>Conclusions</h3><div>Given the correlation observed between AA and the selected proxy condition using this methodology, this study presents a potential approach to identify proxy conditions for diseases that may be underrepresented in terms of HRQoL data.</div></div>","PeriodicalId":23508,"journal":{"name":"Value in Health","volume":"28 4","pages":"Pages 599-610"},"PeriodicalIF":4.9,"publicationDate":"2025-02-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143374862","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Cost-Effectiveness of a Digitally Supported Care Management Program for Caregivers of People With Dementia","authors":"Michelle Pfaff MSc ,&nbsp;Wolfgang Hoffmann MD, MPH ,&nbsp;Melanie Boekholt MA ,&nbsp;Olga Biernetzky PhD ,&nbsp;Iris Blotenberg PhD ,&nbsp;Dilshad Afrin MSc ,&nbsp;Moritz Platen PhD ,&nbsp;Stefan Teipel MD ,&nbsp;Jochen René Thyrian PhD ,&nbsp;Ingo Kilimann MD ,&nbsp;Bernhard Michalowsky PhD","doi":"10.1016/j.jval.2025.01.011","DOIUrl":"10.1016/j.jval.2025.01.011","url":null,"abstract":"<div><h3>Objectives</h3><div>To examine the cost-effectiveness of a digitally supported care management system (CMS) for caregivers of people with dementia (PwD) compared with usual care.</div></div><div><h3>Methods</h3><div>The analysis was based on 192 caregivers (<em>n</em> = 96 CMS, <em>n</em> = 96 usual care) of PwD in a cluster-randomized controlled trial testing a digitally supported CMS, aiming to identify and address caregivers’ unmet needs and develop and implement an individualized support and care plan over 6 months. Incremental costs from the public-payer and societal perspectives, quality-adjusted life years (QALY), and the incremental cost-effectiveness ratio 6 months after baseline were calculated using multivariate regression models. We assessed the probability of cost-effectiveness using a range of willingness-to-pay thresholds.</div></div><div><h3>Results</h3><div>Caregivers in the intervention group gained QALYs (+0.004 [95% CI −0.003 to 0.012], <em>P</em> value = .225) and had lower costs from the public payer (−378€ [1926-1168], <em>P</em> value = .630), but higher costs from the societal perspective (+1324 [−3634 to 6284], <em>P</em> value = .599). The intervention dominated usual care from the payer perspective, whereas the incremental cost-effectiveness ratio was €331 000/QALY from a societal perspective. The probability of cost-effectiveness was 72% and 79% from the public payer and 33% and 35% from a societal perspective at the willingness-to-pay thresholds threshold of €40 000 and €80 000/QALY gained.</div></div><div><h3>Conclusions</h3><div>CMS was likely cost-effective from the payer but not from a societal perspective, underlining the importance of informal care. The gain in QALY was marginal and could be due to the short observation period. Focusing on both the caregiver and the PwD, rather than assessing the PwD needs through the caregiver, could improve the cost-effectiveness results.</div></div>","PeriodicalId":23508,"journal":{"name":"Value in Health","volume":"28 4","pages":"Pages 527-535"},"PeriodicalIF":4.9,"publicationDate":"2025-02-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143374788","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Are Updated COVID-19 Vaccines Still Relevant for All Adult Age Groups? An Economic Evaluation of the Monovalent XBB.1.5 Vaccine in Australia.","authors":"Charles Ebuka Okafor, Syed Afroz Keramat, Namal N Balasooriya, Echezona H Dioji","doi":"10.1016/j.jval.2025.01.014","DOIUrl":"10.1016/j.jval.2025.01.014","url":null,"abstract":"<p><strong>Objectives: </strong>There is a progressive decline in the relative effectiveness of updated COVID-19 vaccines. Surveillance reports in Australia have also shown a wide variation in the disease severity and mortality across age groups. This study aimed to perform a cost-utility analysis of the monovalent XBB.1.5 vaccine relative to no updated booster vaccine between September 2023 and August 2024.</p><p><strong>Methods: </strong>A cost-utility analysis was performed using a Markov model from the healthcare system perspective for 3 different age groups (18-64, 65-74, and ≥75 years). Costs and outcomes with the monovalent XBB.1.5 vaccine were compared with no updated booster dose for 1 year. Health outcomes were expressed as quality-adjusted life years (QALYs), whereas costs were presented in 2023 Australian dollars. Aggregate distributional cost-effectiveness analysis and sensitivity analyses were performed. The willingness-to-pay threshold was set at A$50 000/QALY.</p><p><strong>Results: </strong>The updated vaccine was dominant for the 18 to 64 years group and cost-effective for the 65 to 74 years group (incremental cost-effectiveness ratio [ICER] = A$10 786/QALY) and ≥75 years group (ICER = A$36,531/QALY) relative to no updated booster vaccine. The major determinants of the ICER was the vaccine uptake rate. There was inequality in health benefits between the older First Nations versus nonindigenous Australians. The results were robust to simultaneous changes in the parameters' values.</p><p><strong>Conclusions: </strong>The monovalent XBB.1.5 vaccine is cost-effective and still relevant for all adult age groups in Australia. Based on current evidence, the study findings support the promotion of booster vaccination for Australian adults.</p>","PeriodicalId":23508,"journal":{"name":"Value in Health","volume":" ","pages":""},"PeriodicalIF":4.9,"publicationDate":"2025-02-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143374782","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Do Machine Learning Approaches Perform Better Than Regression Models in Mapping Studies? A Systematic Review.","authors":"Tianqi Hong, Shitong Xie, Xinran Liu, Jing Wu, Gang Chen","doi":"10.1016/j.jval.2024.12.010","DOIUrl":"10.1016/j.jval.2024.12.010","url":null,"abstract":"<p><strong>Objectives: </strong>To identify how machine learning (ML) approaches were implemented in mapping studies and to determine the extent to which ML improved performance compared with regression models (RMs).</p><p><strong>Methods: </strong>A systematic literature search was conducted in 12 databases from inception to December 2023 to identify studies that applied ML to develop mapping algorithms. A data template was applied to extract data set information, source and target measures, ML approaches and RMs, mapping types (direct vs indirect), goodness-of-fit indicators (mean absolute error, mean squared error, root mean squared error, R-squared, and intraclass correlation coefficient), and validation methods. Differences in goodness-of-fit indicators between ML and RMs were summarized. Potential advantages and challenges for ML were further discussed.</p><p><strong>Results: </strong>Thirteen mapping studies were identified, in which both ML and RM were adopted. Bayesian networks were the most frequently used ML approach (n = 6), followed by the least absolute shrinkage and selection operator (n = 4). The ordinary least square model was the most used RM (n = 8), followed by the censored least absolute deviation and multinomial logit models (n = 5 each). The average improvement in the goodness-of-fit of ML compared with that of RMs by indicators were 0.007 (mean absolute error), 0.004 (mean squared error), 0.058 (R-squared), 0.016 (intraclass correlation coefficient), and -0.0004 (root mean squared error).</p><p><strong>Conclusions: </strong>There is an increasing number of studies using ML in developing mapping algorithms. Generally, a minor improvement of goodness-of-fit was observed compared with RMs when using mean-based comparisons. Issues such as how to interpret, apply, and externally validate the ML-based outputs would affect their implementation. Future studies are warranted to verify advantages of ML approaches.</p>","PeriodicalId":23508,"journal":{"name":"Value in Health","volume":" ","pages":""},"PeriodicalIF":4.9,"publicationDate":"2025-02-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143374860","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}