Medical Care最新文献

{"title":"Postoperative Complications and Readmission Rates in Robotic-Assisted and Manual Total Hip Arthroplasty: A Large, Multi-Hospital Study.","authors":"Cole Howell, Sietske Witvoet, Laura Scholl, Andrea Coppolecchia, Manoshi Bhowmik-Stoker, Antonia F Chen","doi":"10.1097/MLR.0000000000002082","DOIUrl":"10.1097/MLR.0000000000002082","url":null,"abstract":"<p><strong>Objective: </strong>This study aims to compare 90-day postoperative complications, readmissions, and emergency department (ED) visits between robotic-assisted (RA-THA) and manual (M-THA) total hip arthroplasty.</p><p><strong>Methods: </strong>A retrospective review of a multi-hospital database identified primary total hip arthroplasty patients between January 2016 and December 2021. The cohorts were 1-to-1 matched based on patient sex, age, and body mass index resulting in 8033 patients in each cohort (N = 16,066). Odds of 90-day revisits, readmission with >23 hours of observation, and ED visits were compared between cohorts. Complications reported during revisits and readmission were classified according to the Clinical Classification Software schema, using the International Classification of Diseases, 10th Revision codes, and compared using mixed-effect models.</p><p><strong>Results: </strong>This study found an overall 90-day revisit rate of 8.3%. RA-THA was associated with significantly reduced odds of revisit within 90 days [odds ratio (OR): 0.71, 95% CI: 0.58-0.89, P = 0.002] and readmissions with >23 hours of observation (OR: 0.61, 95% CI: 0.48-0.77, P < 0.001). RA-THA patients had fewer readmissions with >23 hours of observation due to dislocations (RA-THA: 0.09%; M-THA: 0.39%, P < 0.001), surgical site infections (RA-THA: 0.04%; M-THA: 0.20%, P = 0.004), and wound infections/cellulitis (RA-THA: 0.01%; M-THA: 0.11%, P = 0.021). No difference in ED visits was observed between cohorts (OR: 0.92, 95% CI: 0.77-1.09, P = 0.3). RA-THA patients had more ED visits for dyspnea without pulmonary embolism (RA-THA: 0.20%; M-THA: 0.06%, P = 0.03).</p><p><strong>Conclusion: </strong>RA-THA showed significantly lower odds of overall 90-day revisit rates and readmissions with >23 hours of observation, most notably for readmissions due to dislocation and surgical site infection/wound infections. There was no significant difference in the odds of ED visits between cohorts.</p>","PeriodicalId":18364,"journal":{"name":"Medical Care","volume":"63 7","pages":"465-471"},"PeriodicalIF":2.8,"publicationDate":"2025-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144266549","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":"Chronic Condition Burden by Veteran Status, Veterans Health Administration Enrollment, and Age Using Nationally Representative Survey Data.","authors":"Neetu Chawla, Diana J Govier, Claire Than, Danielle Rose, Denise M Hynes, Katherine J Hoggatt, Elizabeth M Yano","doi":"10.1097/MLR.0000000000002159","DOIUrl":"10.1097/MLR.0000000000002159","url":null,"abstract":"<p><strong>Background: </strong>Historically, US Veterans have reported higher chronic disease burden than non-Veterans. However, whether Veteran and Veterans Affairs (VA) coverage status continue to be associated with chronic disease burden or how these associations vary by age, especially among younger Veterans, is unknown.</p><p><strong>Objective: </strong>To examine the number of chronic conditions among male Veterans with and without VA coverage, and male non-Veterans, overall and by age group.</p><p><strong>Design and participants: </strong>Using 2018 National Health Interview Survey data, our sample included 2301 male Veterans and 9243 male non-Veterans.</p><p><strong>Main measures: </strong>The primary outcome was a number of chronic conditions, measured as a count (range 0-15) and categorically (0, 1, 2, 3+). We created a 3-category main independent variable (Veteran with VA coverage, Veteran without VA coverage, non-Veteran). Generalized linear regression models were used to estimate relationships between Veteran and VA coverage status and count of chronic conditions, overall and by age group (18-44, 45-64, 65+), adjusting for sociodemographic characteristics.</p><p><strong>Key results: </strong>Veterans with VA coverage, Veterans without VA coverage, and non-Veterans had an age-standardized mean of 1.44, 1.16, and 1.09 chronic conditions, respectively. In adjusted analyses, Veterans with VA coverage had 0.36 (95% CI: 0.25-0.46) more conditions and Veterans without VA coverage had 0.12 (95% CI: 0.04-0.21) more conditions compared with non-Veterans. In age-stratified analyses, Veterans with VA coverage aged 18-44 had 0.22 (95% CI: 0.06-0.38) more conditions; 45-64, 0.71 (95% CI: 0.41-0.99) more conditions; and 65+, 0.38 (95% CI: 0.18-0.57) more conditions compared with similar-aged non-Veterans.</p><p><strong>Conclusions: </strong>Veterans with VA coverage had the greatest number of chronic conditions, including when stratified by age group.</p>","PeriodicalId":18364,"journal":{"name":"Medical Care","volume":"63 7","pages":"507-513"},"PeriodicalIF":2.8,"publicationDate":"2025-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144789533","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":"Top-Rated Health Care and Ease of Access to Medications Linked to Lower Medicare and ADRD Costs.","authors":"Jie Chen, Seyeon Jang","doi":"10.1097/MLR.0000000000002140","DOIUrl":"https://doi.org/10.1097/MLR.0000000000002140","url":null,"abstract":"<p><strong>Importance: </strong>Little is known about the extent to which patient self-perception of care experience is associated with costs, especially for people with Alzheimer disease and related dementias (ADRD).</p><p><strong>Objective: </strong>This study explores the relationship between self-reported quality measures and Medicare costs and examines whether the ease of obtaining prescribed medications is associated with reduced overall Medicare costs, focusing on Medicare beneficiaries with ADRD.</p><p><strong>Design, setting, and participants: </strong>In this cross-sectional study, Medicare Beneficiary Summary File data from 2018, 2019, and 2021 were linked to the Medicare Consumer Assessment of Health Care Providers and Systems (CAHPS) Survey using beneficiary IDs. The study sample included community-dwelling Medicare fee-for-service beneficiaries.</p><p><strong>Exposures: </strong>Five quality measures were used as key exposure variables: (1) beneficiary's rating on health care; (2) ease of getting care/tests/treatment through the health plan; (3) whether the doctor always explained, listened, respected; and spent enough time with the patient; (4) ease of obtaining prescribed medications; and (5) whether doctor always talked about all the prescription medicines the beneficiary was taking.</p><p><strong>Main outcome and measure: </strong>Annual total Medicare payments per person.</p><p><strong>Results: </strong>The study included 230,617 Medicare fee-for-service beneficiaries aged 65 and older, including 16,452 beneficiaries with ADRD. Among the total beneficiaries, 53% were females (vs. 56% of ADRD beneficiaries), with a mean (SD) age of 75.8 (SD 7.27) years [vs. 82.5 (SD 7.97) years for ADRD beneficiaries]. Fully adjusted analyses showed significant negative associations between quality measures and total per-capita payments, with more pronounced cost reductions among patients with ADRD. Specifically, patients with ADRD who reported it was always easy to get care had reductions of $1,922.0 (95% CI, -$3304.8 to -$539.2), while those who reported it was always easy to get prescribed medications had reductions of $2964.5 (95% CI, -$4518.8 to -$1410.1). In addition, beneficiaries who reported that doctors always discussed the medicines experienced cost reductions of $2299.7 (95% CI, -$3800.5 to -$799.0) in medicare costs.</p><p><strong>Conclusion and relevance: </strong>Our findings suggest that high-quality care is not necessarily associated with high costs. Meanwhile, focusing on the ease of access to needed care, obtaining prescription drugs, and effective communication about medication is critical in improving care quality while reducing costs.</p>","PeriodicalId":18364,"journal":{"name":"Medical Care","volume":"63 6","pages":"405-412"},"PeriodicalIF":3.3,"publicationDate":"2025-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12061373/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144024931","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":"Risk Adjustment in Capitation Payments to Primary Care Providers: Does It Matter How We Account for Patients' Socioeconomic Status?","authors":"Lina Maria Ellegård, Maude Laberge","doi":"10.1097/MLR.0000000000002141","DOIUrl":"https://doi.org/10.1097/MLR.0000000000002141","url":null,"abstract":"<p><strong>Background: </strong>One of the critical challenges with capitation payment to primary care providers is ensuring that the fixed payments are equitable and adjusted for expected care needs. Patients of lower socioeconomic status (SES) generally have higher health care need. Sweden developed a Care Needs Index, which is used in the capitation payments to primary care providers to account for patient SES.</p><p><strong>Objectives: </strong>We aim to examine the potential value of collecting individual-level rather than geographic-level socioeconomic data to support an equitable payment to primary care providers.</p><p><strong>Research design: </strong>We used data from 3 regional administrative care registers, which cover all consultations in publicly funded health care, and Statistics Sweden's registers covering individual background characteristics. We estimated linear regression models and evaluated the model fit using the adjusted R2 with the Care Needs Index at the individual and at the district level. The population consisted of the 3,490,943 individuals residing in the 3 study regions for whom we had complete data.</p><p><strong>Measures: </strong>The main outcome variable was the number of face-to-face consultations with a GP or a nurse at a primary care practice. We use the R2 to compare the predictive power of the models.</p><p><strong>Results: </strong>The share of the variation explained did not depend on whether the Care Needs Index was measured at the individual level or the small area level.</p><p><strong>Conclusions: </strong>SES explains very little variation in primary care visits, and there is no gain from having individual-level information about the individual's SES compared with having district-level information only.</p>","PeriodicalId":18364,"journal":{"name":"Medical Care","volume":"63 6","pages":"430-435"},"PeriodicalIF":3.3,"publicationDate":"2025-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12061383/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144033232","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":"Perinatal Resources and Wildfire Smoke.","authors":"Michel Boudreaux, Maranna Yoder, Evan Ellicott, Molly Passarella, Scott A Lorch","doi":"10.1097/MLR.0000000000002130","DOIUrl":"10.1097/MLR.0000000000002130","url":null,"abstract":"<p><strong>Background: </strong>Pregnant people and infants are vulnerable to wildfire smoke. However, the availability of perinatal resources in communities impacted by smoke is unknown.</p><p><strong>Objective: </strong>Describe perinatal resources in counties prone to wildfire smoke.</p><p><strong>Research study design: </strong>Smoke data came from the Hazard Mapping System and perinatal resources were gathered from various sources. Choropleth maps described the geographic distribution of smoke. Unadjusted associations and multivariable regressions compared perinatal resource levels by smoke risk. Subgroup analysis of the most rural counties was conducted.</p><p><strong>Subjects: </strong>Counties in the contiguous United States (n=3108) during the 2016-2020 period.</p><p><strong>Measures: </strong>Relative smoke risk was defined as the bottom, middle, and top third of the average annual smoke-days distribution. Perinatal resources included driving distance to the nearest maternity care hospital and NICU, the volume and geographic isolation of the nearest maternity care hospital, and county-based measures of OB-GYN and family medicine physicians.</p><p><strong>Results: </strong>Average annual smoke-days ranged from 3.8 (SD=2.0) in low-risk to 15.3 (SD=5.5) in high-risk counties. Compared with low-risk counties, high-risk counties had fewer OB-GYNs per 10,000 births (-32.2, 95% CI: -45.7 to -20.6; P<0.001) and were farther to the nearest maternity hospital (10.1 miles, 95% CI: 8.7-11.5; P<0.001). High-risk counties were also farther to the nearest NICU. Associations were not explained by sociodemographics and were observed in the subset of the most rural counties.</p><p><strong>Conclusions: </strong>Communities prone to wildfire smoke often lack geographic access to the health care resources needed to treat pregnant people and infants in a timely manner.</p>","PeriodicalId":18364,"journal":{"name":"Medical Care","volume":"63 6","pages":"396-404"},"PeriodicalIF":2.8,"publicationDate":"2025-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144003831","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":"The Impact of a Bidirectional Clinic to Community Social Care Referral Program.","authors":"Shari D Bolen, Jonathan Lever, Chris Mundorf, Alvonta Jenkins, Rachel Waitzman, Samantha Smith, Matthew Finley, Joseph Daprano, Eva Johnson, Marie Masotya, Shivani Joshi, Anandhi Gunder, Melissa E Lohr, David Bar-Shain, David C Kaelber, Tatyana Khaled, Dieter Sumerauer, Heidi Gullet, Kurt C Stange","doi":"10.1097/MLR.0000000000002144","DOIUrl":"https://doi.org/10.1097/MLR.0000000000002144","url":null,"abstract":"<p><strong>Background: </strong>Practical knowledge of how to address patients' social needs could have a large health impact.</p><p><strong>Objective: </strong>Describe a scalable electronic health record (EHR)-facilitated, clinic-to-community linkage (CCL) program that addresses social needs at 6 clinics in 4 health systems.</p><p><strong>Research design: </strong>Primary care teams referred eligible patients to United Way 211 (UW 211) via a point-of-care EHR referral between 2018 and 2023. Patients were eligible if they were adults with uncontrolled blood pressure or blood sugar or 2-17 years old with overweight/obesity or asthma. UW 211 referred patients to assess and connect them with community resources and provided electronic feedback to the EHR. We conducted descriptive analyses of process measures (eg, patients referred, needs identified, need resolution). We then conducted pre-post analyses of selected health outcomes (ie, blood pressure, weight, and asthma exacerbations) versus comparison clinics.</p><p><strong>Results: </strong>Referral ranges varied by clinic from 3% to 43%, with 1224 total patients referred and 38% (n=461) reached by UW 211. All 461 had at least one need, and 87% (n=400) had one need resolved or a resolution in progress. Reached patients had an average of 2.9 (SD 1.3) needs and an average of 10.1 resource referrals provided (SD 6.1). Top needs included food, physical activity, housing and utilities. No differences were found pre to post within the intervention clinics except for improvements in blood pressure control. However, comparison clinics had greater improvements in blood pressure control during the same time frame.</p><p><strong>Conclusions: </strong>An EHR-facilitated, closed-loop CCL program to address patients' social needs is feasible. Further research on the comparative effectiveness and sustainability of models to address social needs will be critical in advancing health equity.</p>","PeriodicalId":18364,"journal":{"name":"Medical Care","volume":"63 6","pages":"449-457"},"PeriodicalIF":3.3,"publicationDate":"2025-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144031340","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":"Insights Into Perinatal Health Care Resources in United States Counties Affected by Wildfire Smoke.","authors":"Steven Babin","doi":"10.1097/MLR.0000000000002147","DOIUrl":"https://doi.org/10.1097/MLR.0000000000002147","url":null,"abstract":"","PeriodicalId":18364,"journal":{"name":"Medical Care","volume":"63 6","pages":"393-395"},"PeriodicalIF":3.3,"publicationDate":"2025-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144025008","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":"Associations of Co-Occurring Substance Use Disorder With Diabetes Care Quality, Complications, and Hospitalizations.","authors":"Margae J Knox, Dominic Hodgkin, Natalie E Slama, Stacy A Sterling, Lisa K Gilliam, Asma Asyyed, Esti Iturralde","doi":"10.1097/MLR.0000000000002138","DOIUrl":"10.1097/MLR.0000000000002138","url":null,"abstract":"<p><strong>Background: </strong>Substance use disorder (SUD) is a risk factor for diabetes complications and hospitalizations, though a full continuum of diabetes care quality and health outcomes has not been examined among patients with diabetes accessing substance use treatment.</p><p><strong>Objective: </strong>To improve care delivery, this study compared patients with diabetes and co-occurring SUD to those with diabetes and no SUD.</p><p><strong>Population: </strong>In all, 4325 patients with diabetes and a SUD specialty treatment visit versus 255,652 patients with diabetes and no SUD diagnosis in a large, integrated delivery system from 2016 to 2021 were included.</p><p><strong>Research design: </strong>Retrospective cohort study using electronic health record data. Modified Poisson regression models estimated relationships for co-occurring SUD and each outcome, adjusting for sociodemographic and clinical factors.</p><p><strong>Measures: </strong>Care quality measures included HbA1c, blood pressure, retinal and cholesterol screening, HbA1c < 8%, blood pressure < 140/90 mm Hg, and LDL-cholesterol < 100 mg/dL. Diabetes complications included cardiovascular, cerebrovascular, retinopathy, and lower limb conditions. Hospitalization types included diabetes-related and other conditions, for example, chronic liver disease, and psychiatric.</p><p><strong>Results: </strong>Patients with co-occurring SUD, compared with those without SUD, were more often male, younger, non-Hispanic White, and had a mood disorder. Co-occurring SUD was associated with more HbA1c screening and higher prevalence of HbA1c <8, yet also with elevated risks for nearly all complication types, and all but one hospitalization type, especially chronic liver disease and chronic pain-related hospitalization.</p><p><strong>Conclusions: </strong>Despite comparable or better diabetes care quality, elevated risk of complications and hospitalization persisted among patients with co-occurring SUD. Both biopsychosocial and system-based mechanisms likely contribute to these elevated risks. Silo-bridging care coordination may help address multifaceted health needs.</p>","PeriodicalId":18364,"journal":{"name":"Medical Care","volume":"63 6","pages":"443-448"},"PeriodicalIF":2.8,"publicationDate":"2025-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12442420/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145069899","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":"Probabilistically Linkage of California's Birth Certificate and Hospital Discharge Data.","authors":"Bharti Garg, Aaron B Caughey, Blair G Darney","doi":"10.1097/MLR.0000000000002139","DOIUrl":"10.1097/MLR.0000000000002139","url":null,"abstract":"<p><strong>Objective: </strong>To link California's birth certificate data with maternal and infant hospital discharge data to get a valuable database for epidemiological research.</p><p><strong>Background: </strong>Secondary data sources are widely used for epidemiological research. Although California's birth certificate and patient discharge data (PDD) are readily available separately, the linked data are only available till 2012. We obtained birth certificate data from the California Department of Public Health and hospital discharge data from the Department of Health Care Access and Information. In this study, we propose a methodology to link these 2 datasets, which can be used for perinatal epidemiological research. We utilized data from 2008 to 2019.</p><p><strong>Methods: </strong>We used probabilistic linkage methods to link birth certificates and hospital discharge data. Hospital discharge data was included as 2 datasets: maternal and infant discharge records. The linkage was a 2-step process: (1) Linkage of birth certificate with infant's hospital discharge data to form combined data. (2) Linkage of combined birth certificate-infant's discharge data with maternal discharge data.</p><p><strong>Results: </strong>We included 5,661,695 births from birth certificates and 5,617,921 infant discharge files. After linkage, we were able to link 92.2% of the birth certificate records with the infant's discharge files using variables: maternity hospital, infant's birth date, infant's sex, mother's residence zip code, and birth Hospital County. When the combined vital statistics-infant's PDD data were linked with maternal PDD data, 90.0% of vital statistics data linked with both infant and maternal PDD, 2.5% linked to only infant PDD, and 1.5% linked to only maternal PDD.</p><p><strong>Conclusion: </strong>Our linkage algorithm produces effective linked data that can be used for epidemiological research. This process is complex and needs to be evaluated every year as some of the variables change, or some added information becomes available in some files.</p>","PeriodicalId":18364,"journal":{"name":"Medical Care","volume":" ","pages":"458-463"},"PeriodicalIF":3.3,"publicationDate":"2025-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143753346","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":"Combining Machine Learning and Comparative Effectiveness Methodology to Study Primary Care Pharmacotherapy Pathways for Veterans With Depression.","authors":"Ozgur Ozmen, Everett Rush, Byung H Park, Makoto Jones, Jodie Trafton, Lisa Brenner, Randall W Rupper, Merry Ward, Jonathan R Nebeker, Steven D Pizer","doi":"10.1097/MLR.0000000000002145","DOIUrl":"https://doi.org/10.1097/MLR.0000000000002145","url":null,"abstract":"<p><strong>Objectives: </strong>To demonstrate an innovative method combining machine learning with comparative effectiveness research techniques and to investigate a hitherto unstudied question about the effectiveness of common prescribing patterns.</p><p><strong>Data sources: </strong>United States Veterans Health Administration Corporate Data Warehouse.</p><p><strong>Study design: </strong>For Operation Enduring Freedom/Operation Iraqi Freedom veterans with major depressive disorder, we generate pharmacotherapy pathways (of antidepressants) using process mining and machine learning. We select the medication episodes that were started at subtherapeutic doses by the first assigned primary care physician and observe the paths that those medication episodes follow. Using 2-stage least squares, we test the effectiveness of starting at a low dose and staying low for longer versus ramping up fast while balancing observable and unobservable characteristics of patients and providers through instrumental variables. We leverage predetermined provider practice patterns as instruments.</p><p><strong>Data collection: </strong>We collected outpatient pharmacy data for selective serotonin reuptake inhibitors and selective norepinephrine reuptake inhibitors, patient and provider characteristics (as control variables), and the instruments for our cohort. All data were extracted for the period between 2006 and 2020.</p><p><strong>Principal findings: </strong>There is a statistically significant positive effect (0.68, 95% CI 0.11-1.25) of \"ramping up fast\" on engagement in care. When we examine the effect of \"ramping up slow\", we see an insignificant negative impact on engagement in care (-0.82, 95% CI -1.89 to 0.25). As expected, the probability of drop-out also seems to have a negative effect on engagement in care (-0.39, 95% CI -0.94 to 0.17). We further validate these results by testing with medication possession ratios calculated periodically as an alternative engagement in care metric.</p><p><strong>Conclusions: </strong>Our findings contradict the \"Start low, go slow\" adage, indicating that ramping up the dose of an antidepressant faster has a significantly positive effect on engagement in care for our population.</p>","PeriodicalId":18364,"journal":{"name":"Medical Care","volume":"63 6","pages":"422-429"},"PeriodicalIF":3.3,"publicationDate":"2025-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144024487","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}