Objective:To compare the risk of incident cardiovascular disease (CVD) events following sleeve gastrectomy (SG) and Roux-en-Y gastric bypass (RYGB).Background:Bariatric surgery is associated with reduced CVD risk, but the differential effect of contemporary bariatric procedures is unclear.Methods:We used insurance claims to conduct a retrospective cohort study of CVD outcomes for patients who underwent RYGB versus SG between 2010 and 2021. Patients were followed for up to 5 years for a primary composite major adverse cardiovascular event (MACE) outcome as well as individual outcomes, including myocardial infarction, stroke, heart failure, and arrhythmia. We compared cumulative risks of CVD events using multivariable Cox proportional hazards modeling in overall cohorts and subcohorts of older adults and those with type 2 diabetes (T2D) or pre-existing CVD and elevated morbidity.Results:Matched, weighted cohorts of 13,545 SG and RYGB patients were observed for an average of 2.5 years after surgery, with 26.2% not lost to follow-up by the end of 5 years. There was no difference in MACE risk between procedures [adjusted hazard ratio (aHR): 1.01 for RYGB vs. SG (95% CI: 0.90, 1.12)] in the overall cohort or among the subgroup of older adults [aHR: 0.97 for RYGB vs. SG (95% CI: 0.85, 1.10)]. Patients with T2D experienced a lower risk of MACE following RYGB compared with SG [aHR: 0.78 (95% CI: 0.66, 0.92)], as did those with pre-existing CVD or elevated morbidity before surgery [aHR: 0.81 (95% CI: 0.70, 0.93)].Conclusions:These findings further support the preferential use of RYGB over SG for patients with T2D or who have pre-existing CVD. However, among other groups of patients, including older adults, we did not observe a relative benefit of RYGB during the time horizon in this study.

Substantial observational evidence exists in support of bariatric surgery being associated with a reduction in risk for a wide range of outcomes, including cardiovascular disease (CVD) in patients with diabetes. Two recent studies, however, argued that much of that prior work suffers from various sources of underappreciated bias as well as design decisions that compromise whether one can conceive of a corresponding target trial. Furthermore, results based on analyses of claims data from Optum and electronic health record data from the Veterans Administration are presented as providing evidence of no CVD benefit for bariatric surgery in patients with diabetes. In this paper, we use data from a prior Kaiser Permanente study to emulate a trial that mimics the methods employed in the Veterans Administration study. This new analysis finds a reduction in risk of CVD in patients with diabetes, consistent with preexisting evidence. We discuss possible mechanisms by which the discrepant results can be reconciled, including issues of statistical validity that arise from small samples, whether recent work on transportability indicates that we should not always expect results to always be concordant, and the role of conservatism associated with "clinical trial thinking." We conclude with a discussion of what standards should be used when considering the work of others in the literature and the role that evidence triangulation may be play in the future.

OBJECTIVE:This study compared changes in health care costs and use across cohorts initiating four obesity medications (OMs). METHODS:Commercial insurance claims were used to identify new initiators of phentermine (n = 136,788), phentermine-topiramate-ER (n = 13,888), naltrexone-bupropion-SR (n = 28,712), or liraglutide (n = 49,266) for obesity. Multivariable difference-in-differences analyses compared the three newer drugs to phentermine for change in total annual prescription drug and health care costs, outpatient visits, and acute care use (combined emergency department visits and hospital stays). RESULTS:Up to 3 years after OM initiation, prescription and total health care costs were higher for all three newer OM cohorts relative to phentermine, with liraglutide having the greatest comparative increase in annual costs (e.g., total costs 73.6% [70.9%, 76.3%] higher in year 1). Primary care physician visit frequency was slightly lower for the newer OMs (e.g., -12.6% [-13.6%, -11.6%] in year 1 for liraglutide). CONCLUSIONS:The cost of OM treatment, even for medications that preceded highly effective weekly incretin analogues, appeared to increase overall prescription spending and total health care costs. Yet, other than a relative decrease in primary care physician visit frequency, more costly OMs were not associated with favorable changes in health care use patterns compared to generic phentermine. Reductions in OM cost and other interventions to support their longer-term use may be needed to fully realize the promise of these medications.

Glucagon-like peptide-1 receptor agonists (GLP-1RAs) have reshaped the clinical approach to managing obesity and type 2 diabetes. As approved indications have expanded, use of GLP-1RAs has increased rapidly in the United States. Although randomized trials demonstrate strong efficacy, many questions remain about their optimal use in clinical practice. Real-world data (RWD) from electronic health records, registries, insurance claims, and other sources offer a promising avenue to address these questions. However, concerns about data quality, selection bias, and incomplete ascertainment of medication use and outcomes pose significant challenges to the validity of the resulting evidence. In May 2025, the National Institute of Diabetes and Digestive and Kidney Diseases convened experts from regulatory agencies, payer organizations, and academia to explore these challenges. This second of 2 synopsis articles on the workshop summarizes the discussion around the strengths and limitations of various RWD sources and methodological approaches to strengthen causal inference and generalizability. Presenters highlighted pragmatic clinical trials and target trial emulation as strategies to generate stronger real-world evidence (RWE) that is relevant to both clinical practice and policy. The workshop underscored that careful attention to study design, data limitations, and analytic approach is essential to yield RWE that informs clinicians, patients, payers, and policymakers.

Background: Social determinants of health (SDOH) can affect metabolic health. Objective: To determine the effect of social vulnerability on the comparative effectiveness of metabolic bariatric surgery or medical and lifestyle intervention on glycemia and weight outcomes in people with type 2 diabetes (T2D). Design: Analysis of the effect modification of baseline Area Deprivation Index (ADI; a metric of social vulnerability) on longitudinal outcomes between randomized treatment groups using linear mixed-effects models. (ClinicalTrials.gov: NCT02328599) Setting: 4 U.S. academic centers. Participants: 258 participants with T2D enrolled in 4 randomized controlled trials of surgical versus medical management and a longitudinal observational follow-up study. Measurements: ADI linked to ZIP code data at randomization; weight loss and hemoglobin A(1c) (HbA(1c)) level at the end of the active intervention period (7 to 12 years). Results: Baseline characteristics were well balanced between the surgical and medical therapy groups after adjustment for study site and stratification by high versus low ADI. Surgery was more effective than medical therapy in reducing HbA(1c) level among persons with high ADI (net difference, -1.29% [95% CI, -1.95% to -0.63%]) and those with low ADI (net difference, -0.95% [CI, -1.29% to -0.62%]). Surgery was also more effective than medical therapy at producing weight loss across ADIs, with respective net differences of -10.6% (CI, -15.2% to -5.9%) for high ADI and -13.3% (CI, -15.7% to -10.9%) for low ADI. The interaction between ADI and intervention group was not significant for either HbA(1c) (P = 0.37) or weight loss (P = 0.31). Limitations: Small sample size; parent trials were not designed to address effect modification by ADI. Conclusion: Surgery was superior to medical therapy for people with T2D regardless of social deprivation. This study did not detect statistically significant differences in the comparative advantage of surgery over medical therapy by ADI. Primary Funding Source: National Institutes of Health.

Metabolic bariatric surgery (MBS) lowers long-term health expenditures among patients with type 2 diabetes (T2D). MBS resources are scarce however, and it is unknown which subgroups of patients with diabetes are most likely to see reduced expenditures. We conducted stratified analyses of total, medication, outpatient, and inpatient expenditures from 3 years pre- to 5.5 years post-MBS in a retrospective cohort study of 6,657 patients with T2D and obesity from 2012 to 2019 and 19,059 matched non-surgical patients. Patients were stratified into three groups based on diabetes severity, using the validated Advanced-DiaRem measure. Advanced-DiaRem scores range from 0 to 21, with higher scores indicating greater severity. Expenditures were estimated in 6-month intervals and analyzed using generalized linear models. Patients with high Advanced-DiaRem scores (13–21) had higher hemoglobin A1c, greater insulin use, and longer diabetes duration compared to those with moderate (8–12) and low scores (0–7). In the 6 months pre-surgery, patients with high scores had higher mean total expenditures (US6,418 non-surgical,6,896 surgical) than patients with moderate scores (4,332 non-surgical,5,373 surgical) and low scores (2,995 non-surgical,3,469 surgical). Post-MBS, estimated total expenditures for surgical patients declined across all Advanced-DiaRem subgroups, with the greatest reductions observed in patients with high scores. At 5.5 years post-surgery, total expenditures for surgical patients were 1,240/6-months lower than for non-surgical patients (95

OBJECTIVE:To examine whether the economic benefits of bariatric surgery differ by patient subgroups, with the aim of identifying those that may yield the most favorable cost profile for improved return on investment in an exploratory analysis. STUDY SETTING AND DESIGN:To identify patient subgroups via the "mTree" (matching + decision tree) method, we conducted analyses of total expenditures 3 years after surgery in a retrospective cohort of 16,538 bariatric patients and 16,538 matched non-surgical patients. DATA SOURCES AND ANALYTIC SAMPLE:This study used electronic health records from Kaiser Permanente, an integrated health system, from 1/1/2012 to 12/31/2019. The cohort was randomly divided into training and test samples, and differences in median total expenditures within each subgroup were then estimated in a held-out test sample. We adapted a novel causal machine learning method mTree, previously developed for randomized trial data, to characterize heterogeneous treatment effects of bariatric surgery on total healthcare expenditures in a non-randomized observational study. This approach combines pair-matching and conditional inference trees to identify subgroups of patients with differential treatment effects while maintaining within-subgroup balance on important confounders. PRINCIPAL FINDINGS:Significant heterogeneity in the effect of bariatric surgery on total expenditures was observed across the eight identified patient subgroups, which were masked by a null average treatment effect. Five of eight subgroups identified in the training sample were replicated in the test sample, and patients using insulin with a Gagne score ≤ 4 had some of the greatest post-surgical expenditure reductions (-$4311, 95% CI: [-$6154, -$2504]). CONCLUSIONS:Subgroup identification is critical for providing context to average treatment effects by identifying patients who may generate a more (or less) promising return on investment. Patients from subgroups with more favorable post-surgical cost profiles may inform prioritization for bariatric surgery if these subgroups are validated in independent cohorts.

Abstract INTRODUCTION Age‐related macular degeneration (AMD) is associated with higher risk of Alzheimer's disease (AD), but how AD risk changes as AMD progresses remains unclear. METHODS Among participants in the Adult Changes in Thought study, we characterized the worst dry and wet AMD severities experienced to date on a scale of 0 to 10 and assessed whether they were associated with AD risk. RESULTS Among 2611 participants (550 AD cases, median 6.8 years follow‐up), AD hazard was estimated to increase by 4% per one unit increase in dry and wet scores (dry: 95% confidence interval [CI] for hazard ratio [HR] 1.01 to 1.07, p = 0.01; wet: 95% CI 1.00 to 1.08, p = 0.03). HR estimates were similar among people without diagnosed AMD and were attenuated among those with diagnosed AMD. DISCUSSION Worse AMD severity was associated with higher AD risk, with severity scores potentially offering greater clinical utility for AD risk characterization among individuals without diagnosed AMD.

Glucagon-like peptide-1 receptor agonists (GLP-1RAs) have transformed obesity and diabetes management, with rapidly expanding indications and use among U.S. adults. Despite their promise, key questions remain about optimal treatment pathways, long-term safety, effectiveness across diverse populations, adherence, and economic impact. Real-world evidence (RWE) derived from electronic health records, claims, and other data sources could address these gaps, but unique challenges complicate its use, such as inconsistent insurance coverage, high discontinuation rates, medication shortages, compounded formulations, and off-label prescribing. To explore these issues, the National Institute of Diabetes and Digestive and Kidney Diseases convened a workshop in May 2025 with experts from regulatory agencies, guideline committees, payers, and academia. Discussions focused on identifying knowledge gaps in GLP-1RA use, evaluating how RWE informs practice, assessing limitations of real-world data, and strategies to reduce bias in RWE. Presentations emphasized RWE's potential to complement randomized trials by capturing rare adverse events, long-term outcomes, and effectiveness in routine care. However, persistent challenges include data reliability, confounding, and incomplete capture of medication use and outcomes, particularly in pediatric and underserved populations. Coverage decisions remain heterogeneous across Medicare, Medicaid, and private payers and across time, underscoring the need for rigorous cost-benefit analyses. The workshop concluded that robust RWE is essential for developing value-based coverage policies and optimizing GLP-1RA use. Continued investment in high-quality data infrastructure and analytic methods will be critical to inform regulatory, clinical, and economic decisions as utilization expands.

AIMS:To characterize trends in utilization of glucagon-like peptide‑1 receptor agonists (GLP‑1RAs), sodium-glucose cotransporter‑2 inhibitors (SGLT2is), and metabolic/bariatric surgery among US adults with type 2 diabetes (T2D) and obesity. METHODS:We conducted a cohort study using electronic health record and claims data from Optum Labs Data Warehouse for US private and Medicare Advantage plan enrollees aged 18-79 with T2D and obesity initiating GLP‑1RAs or SGLT2is or undergoing MBS (2010-2024). We evaluated initiation, adherence (≥80% of days with medication), and persistence. RESULTS:Among 342,932 treatment initiations (335,605 individuals), 60.8% initiated GLP‑1RAs, 35.7% SGLT2is, and 3.6% MBS. Use of both medication classes increased over time (GLP-1RAs: 1,988 to 44,852; SGLT2is: 861 to 19,892). In 2024, semaglutide (62.2%) and tirzepatide (32.6%) were the most initiated GLP-1RAs; MBS was rare (0.7% of treatments). Median treatment duration was 240 days for GLP‑1RAs and 254 days for SGLT2is; ∼50% were adherent. Tirzepatide had the highest adherence and persistence. At one-year, 53.0% of GLP‑1RA and 56.0% of SGLT2i initiators were using their initiated treatment; transition to other therapies was uncommon. CONCLUSIONS:Initiation of GLP‑1RAs and SGLT2is, but not MBS utilization, has expanded in recent years, although adherence and persistence remain low. Further studies should evaluate how these shifting patterns affect health outcomes.

Objective:To enhance shared decision-making (SDM) about metabolic and bariatric surgery (MBS) through the systematic implementation and evaluation of SDM tools across two large healthcare systems. Methods:The project involved developing and integrating SDM tools into clinical workflows at two health systems and training clinicians to use two decision aids with eligible patients. Data were collected using patient surveys, electronic health record (EHR) data, and qualitative interviews with clinicians. The primary outcome was change in patient-reported measures of SDM using the collaboRATE items. Results:Decision aids were delivered to 1675 patients. Implementation was associated with an increase in referrals to MBS and the number of patients undergoing surgery at KPWA but not at UPMC. Baseline levels of SDM quality were high. Post-implementation collaboRATE scores decreased significantly, indicating a perceived decline in SDM quality, but not in two other SDM measures. Qualitative feedback highlighted the challenges and successes of integrating SDM tools into clinical practice. Conclusions:Implementation of SDM tools had mixed results across the sites. Implementation was associated with an increase in referrals and surgery at KPWA but not at UPMC. The perceived quality of SDM conversations also declined. Future efforts should focus on incorporating all weight management options, including medications and lifestyle interventions.

Introduction and Objective: The real-world frequency and associated indicators of type 2 diabetes (T2D) remission (normoglycemia after cessation of glucose-lowering drugs for ≥3 months) are underexplored. This study aimed to evaluate indicators of remission in a diverse cohort of adults with T2D. Methods: Retrospective cohort study of adults (≥18 years) with T2D using 2014-2020 electronic health record data from 6 large US healthcare delivery systems. At baseline, all adults had HbA1c ≥6.5% and a history of glucose-lowering medication use. Pregnant women or adults who underwent bariatric surgery before or during the study were excluded. T2D remission was defined as a return of HbA1c to <6.5% that persisted for ≥3 months after cessation of glucose-lowering medications. Multivariate logistic regression was used to evaluate the association between T2D remission and clinical and demographic characteristics. Results: Our cohort included 656,819 adults (mean [SD] age 60 (13) years, 46.6% female, 17.7% Asian, 10.8% Black, 31.9% Hispanic, 1.5% Native Hawaiian Pacific Islander, 49.5% White). Over 3 years of follow-up, 4% (26,935 adults) achieved T2D remission. The strongest indicators of remission were not receiving glucose-lowering medications at baseline vs. receiving ≥3 medications (OR 19.5, 95% CI: 15.3-25.0), baseline T2D duration <1 year vs. ≥4 years (OR 3.6, 3.5-3.8), baseline HbA1c <7% vs ≥11% (OR 3.5, 3.3-3.7), and weight loss >10% vs. weight stable (OR 2.8, 2.7-2.9) during follow-up. At baseline, 0.5% of adults had received a GLP-1 RA or SGLT2i, and their use was not associated with remission (OR 0.9, 0.7-1.1). Conclusion: T2D remission was observed in 4% of adults with T2D in this cohort, with low use of GLP-1 RA and no bariatric surgery. Clinicians can use these characteristics associated with remission to identify patients who may benefit most from tailored, intensive interventions. Future research should evaluate if achieving T2D remission can reduce morbidity. L.A. Rodriguez: None. B.B. Thapa: None. W.T. Dyer: None. D. Arterburn: None. R. Neugebauer: None. P.J. O'Connor: None. J. An: Research Support; Bayer Pharmaceuticals, Inc, Merck & Co., Inc, AstraZeneca. L.K. Gilliam: None. S.A. Hooker: None. M. Nolan: Other Relationship; GlaxoSmithKline plc, Pfizer Inc. C.E. Oshiro: None. T. Thomas: None. G. Simonson: Other Relationship; Abbott, Sanofi-Aventis U.S. J. Schmittdiel: Research Support; Gilead Sciences, Inc. Patient-Centered Outcomes Research Institute (PCORI) Award (DB-2020C2-20318). National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) Diabetes Research for Equity through Advanced Multilevel Science Center for Diabetes Translational Research (DREAMS-CDTR) (P30 DK092924). NIDDK (1K01DK138122-01).

BACKGROUND:Both sedentary behavior and mental health issues are highly prevalent among older adults. Observational data indicate positive associations between high sedentary behavior and conditions like depression and anxiety. However, it is not known if reducing sedentary behavior leads to changes in mental health. PURPOSE:We conducted an exploratory analysis of a sedentary behavior reduction intervention's effect on self-reported mental health outcomes and quality of life in older adults with obesity. METHODS:The Health Aging Resources to Thrive trial (clinicaltrials.gov # NCT03739762) enrolled older adults aged 60-89 years with a body mass index of 30-50 kg/m2, who self-reported sitting >6 hr/day. Participants were randomized to a sedentary behavior reduction intervention or a healthy living attention-control group. The intervention included 10 behavioral coaching sessions, sitting reduction goals, a standing desk, and a fitness tracker. The attention-control group received 10 coaching contacts to set general healthy living goals excluding physical activity or sedentary behavior. Mental health and quality of life outcomes (depressive symptoms, anxiety symptoms, loneliness, quality of life, sleep disturbance, and pain interference) were measured at baseline, 3 months, and 6 months. We used linear regression models with generalized estimating equations to estimate adjusted mean changes from baseline for each outcome and mean change difference between groups. RESULTS:Participants (N = 283, n = 140 intervention, n = 143 control; mean age = 68.8 years, 65.7% women, 68.9% White non-Hispanic, 55.5% retired) in the intervention group had significant reductions in perceived loneliness score compared with the control group at 6 months (-0.27; 95% confidence interval [-0.52, -0.02]). Depressive symptoms, pain, sleep disturbance, and quality of life improved in both the intervention and control groups, but differences between groups were not statistically significant at 3 or 6 months. CONCLUSIONS:Sedentary behavior reduction may positively impact loneliness in older adults.

ABSTRACT Background Few prior interventions for obesity have focused on reward‐related eating. Researchers developed a mobile health mindfulness‐based intervention, Eat Right Now (ERN), for improving reward related eating; ERN has not yet been feasibility tested as a weight loss intervention. Objective To obtain 6‐month pilot and feasibility data in patients using the ERN intervention alone or with asynchronous coaching. Methods A 6‐month, two‐arm pilot and feasibility randomized trial was conducted involving 20 adults with a BMI ≥ 25 kg/m2 recruited from Kaiser Permanente Washington. Participants were randomized to ERN alone or ERN plus individualized coaching. Weight loss was assessed via a Wi‐Fi‐enabled digital scale. Results Among the 20 randomized participants, 17 (85%) started the intervention and remained enrolled in six months. Among these, 82% (14/17) had weight data captured by a Wi‐Fi scale and 70.5% (12/17) completed follow‐up questionnaires at the 6‐month time point. The coached ERN participants had 4.3% weight loss (95% confidence interval (CI): 2.3%, 6.3%) and uncoached participants had 3.6% weight loss (95% CI: 1.2%, 5.9%). Participants in both approaches reported reductions in reward‐related eating. Conclusions This pilot randomized trial of ERN demonstrates the feasibility and acceptability of the intervention for the majority of both the coached and uncoached groups. Further efforts are needed to ensure high retention and data capture in future studies.

Target trial emulation (TTE) is a popular framework for observational studies based on electronic health records (EHR). A key component of this framework is determining the patient population eligible for inclusion in both a target trial of interest and its observational emulation. Missingness in variables that define eligibility criteria, however, presents a major challenge towards determining the eligible population when emulating a target trial with an observational study. In practice, patients with incomplete data are almost always excluded from analysis despite the possibility of selection bias, which can arise when subjects with observed eligibility data are fundamentally different than excluded subjects. Despite this, to the best of our knowledge, very little work has been done to mitigate this concern. In this paper, we propose a novel conceptual framework to address selection bias in TTE studies, tailored towards time-to-event endpoints, and describe estimation and inferential procedures via inverse probability weighting (IPW). Under an EHR-based simulation infrastructure, developed to reflect the complexity of EHR data, we characterize common settings under which missing eligibility data poses the threat of selection bias and investigate the ability of the proposed methods to address it. Finally, using EHR databases from Kaiser Permanente, we demonstrate the use of our method to evaluate the effect of bariatric surgery on microvascular outcomes among a cohort of severely obese patients with Type II diabetes mellitus (T2DM).

Objective: Type 2 diabetes and obesity are associated with reduced health-related quality of life (HRQoL) and health utility (HU), but long-term effects of metabolic/bariatric surgery (MBS) compared to medical/lifestyle intervention (MLI) on these outcomes are unclear. Research Design and Methods: We studied 228 individuals with type 2 diabetes and obesity randomized to MBS (Roux-en-Y gastric bypass, sleeve gastrectomy, or gastric band; n=152) or MLI (n=76) in the ARMMS-T2D study. HRQoL (SF-36, including Physical Component Score [PCS] and Mental Component Score [MCS]) and HU (SF-6D) were measured annually up to 12 years. Results: At baseline, participant’s age was 49.2±8.0 years, 68.4% were female, BMI was 36.3±3.4 kg/m², and HbA1c was 8.7±1.6%. PCS improved significantly more in MBS vs. MLI over 12 years (+2.37±0.53 vs. -0.95±0.73; difference = 3.32±0.85, p<0.001). MBS was associated with better general health (p<0.001), physical function (p=0.001) and vitality (p=0.003). Reduction in BMI was greater after MBS vs. MLI (p<0.001) and correlated with improved PCS (r=-0.43, p<0.001). Change in PCS was not associated with change in HbA1c. MCS changed minimally from baseline and was similar between MBS and MLI during follow-up (-0.21±0.61 vs. -0.89±0.84; difference = 0.68±0.97, p=0.48). Improvements in HU were greater in MBS vs. MLI over 12 years (+0.02±0.01 vs. -0.01±0.01; difference = 0.03±0.01, p=0.003). Conclusions: Metabolic surgery produces sustained weight loss and improves PCS, general health, physical function, vitality, and health utility in individuals with type 2 diabetes and obesity compared to medical therapy up to 12 years after intervention.

Missingness in variables that define study eligibility criteria is a seldom addressed challenge in electronic health record (EHR)-based settings. It is typically the case that patients with incomplete eligibility information are excluded from analysis without consideration of (implicit) assumptions that are being made, leaving study conclusions subject to potential selection bias. In an effort to ascertain eligibility for more patients, researchers may look back further in time prior to study baseline, and in using outdated values of eligibility-defining covariates may inappropriately be including individuals who, unbeknownst to the researcher, fail to meet eligibility at baseline. To the best of our knowledge, however, very little work has been done to mitigate these concerns. We propose a robust and efficient estimator of the causal average treatment effect on the treated, defined in the study eligible population, in cohort studies where eligibility-defining covariates are missing at random. The approach facilitates the use of flexible machine-learning strategies for component nuisance functions while maintaining appropriate convergence rates for valid asymptotic inference. This method is directly motivated by, and applied throughout to EHR data from Kaiser Permanente to analyze differences between two common bariatric surgical interventions for long-term weight and glycemic outcomes among a cohort of severely obese patients with type II diabetes mellitus.

BACKGROUND:Roux-en-Y gastric bypass (RYGB) and sleeve gastrectomy (SG) differ in their effects on body weight and risk for reoperation. However, it is unclear whether long-term health expenditures differ by procedure type in patients with diabetes. OBJECTIVE:To compare health expenditures 3 years before and 5.5 years after bariatric surgery between patients with diabetes undergoing RYGB versus SG. DESIGN:Retrospective cohort study using target trial emulation principles. SETTING:Integrated health system. PATIENTS:Patients with diabetes undergoing RYGB (n = 3147) or SG (n = 3510) from 2012 to 2019. MEASUREMENTS:Total, inpatient, outpatient, and medication expenditures. RESULTS:Characteristics of patients undergoing RYGB and SG were well balanced after weighting; 73% were female, average body mass index was 43.8 kg/m2, and average age was 50 years. Expenditures per 6-month period decreased by about 30% for both groups, from $4039.06 (95% CI, $3770.88 to $4326.31) 3 years before to $2441.13 (CI, $2151.07 to $2770.30) 5.5 years after RYGB and from $3918.37 (CI, $3658.75 to $4196.40) 3 years before to $2658.15 (CI, $2279.17 to $3100.16) 5.5 years after SG. Total expenditures after surgery did not differ between groups through 5.5 years (difference at 5.5 years, -$217.02 [CI, -$671.29 to $201.96]) except for the first 6 months, when expenditures were transiently higher in the RYGB group (difference, $564.32 [CI, $232.60 to $895.20]), driven by a higher inpatient admission rate. Otherwise, postsurgical outpatient and medication expenditures did not appear to differ between RYGB and SG. LIMITATION:Unobserved confounding. CONCLUSION:Overall expenditures decreased substantially in the postsurgical period, primarily due to reductions in pharmacy expenditures, with no differences between RYGB and SG except in the first 6 months after surgery. PRIMARY FUNDING SOURCE:National Institute of Diabetes and Digestive and Kidney Diseases.

Causal weighted quantile treatment effects (WQTE) are a useful complement to standard causal contrasts that focus on the mean when interest lies at the tails of the counterfactual distribution. To-date, however, methods for estimation and inference regarding causal WQTEs have assumed complete data on all relevant factors. In most practical settings, however, data will be missing or incomplete data, particularly when the data are not collected for research purposes, as is the case for electronic health records and disease registries. Furthermore, such data sources may be particularly susceptible to the outcome data being missing-not-at-random (MNAR). In this paper, we consider the use of double-sampling, through which the otherwise missing data are ascertained on a sub-sample of study units, as a strategy to mitigate bias due to MNAR data in the estimation of causal WQTEs. With the additional data in-hand, we present identifying conditions that do not require assumptions regarding missingness in the original data. We then propose a novel inverse-probability weighted estimator and derive its asymptotic properties, both pointwise at specific quantiles and uniformly across a range of quantiles over some compact subset of (0,1), allowing the propensity score and double-sampling probabilities to be estimated. For practical inference, we develop a bootstrap method that can be used for both pointwise and uniform inference. A simulation study is conducted to examine the finite sample performance of the proposed estimators. The proposed method is illustrated with data from an EHR-based study examining the relative effects of two bariatric surgery procedures on BMI loss at 3 years post-surgery.