Emulating Randomized Trials by Observational Database Studies: The RCT-DUPLICATE Initiative in COPD and Asthma.

Observational studies are increasingly used to provide real-world evidence in regulatory decision-making. The RCT-DUPLICATE initiative conducted observational studies emulating two trials in patients with asthma and three in COPD. For each trial, new-user cohorts were constructed from two US healthcare claims databases, comparing initiators of the study and comparator drugs, matched on propensity scores. Proportional hazards models were used to compare the treatments on study outcomes. The observational studies involved more subjects than the corresponding trials, with treatment arms well-matched on baseline characteristics. An asthma example involved emulation of the 26-week FDA-mandated D5896 trial. With 6,494 asthma patients per arm, the hazard ratio (HR) of a serious asthma-related event with budesonide-formoterol versus budesonide was 1.29 (95% CI: 0.63-2.65), compared with 1.07 (95% CI: 0.70-1.65) in the trial. A COPD example is the emulation of the one-year IMPACT trial. With 4,365 COPD patients per arm, the HR of a COPD exacerbation with triple therapy versus dual bronchodilators was 1.08 (95% CI: 1.00-1.17), compared with 0.84 (95% CI: 0.78-0.91) in the trial. We found mainly discordant results between observational analyses and randomized trials, likely from the forced discontinuation of treatments prior to randomization in the trials, not mimicable in the observational analyses.

Improving prevalence estimates of mental health and well-being indicators among Sexual Minority Men: A Propensity Weighting Approach.

The prevalence and relative disparities of mental health outcomes and well-being indicators are often inconsistent across studies of Sexual Minority Men (SMM) due to selection biases in community-based surveys (non-probability sample), as well as misclassification biases in population-based surveys where some SMM often conceal their sexual orientation identities. The current paper estimated the prevalence of mental health related outcomes (depressive symptoms, mental health service use [MHSU], anxiety) and well-being indicators (loneliness and self-rated mental health) among SMM, broken down by sexual orientation using the Adjusted Logistic Propensity score (ALP) weighting. We applied the ALP to correct for selection biases in the 2019 Sex Now data (a community-based survey of SMMs in Canada) by reweighting it to the 2015-2018 Canadian Community Health Survey (a population survey from Statistics Canada). For all SMMs, the ALP-weighted prevalence of depressive symptoms is 15.96% (95% CI: 11.36%, 23.83%), while for MHSU, it is 32.13% (95% CI: 26.09, 41.20). The ALP estimates lie in between the crude estimates from the two surveys. This method was successful in providing a more accurate estimate than relying on results from one survey alone. We recommend to the use of ALP on other minority populations under certain assumptions.

Semi-supervised mixture multi-source exchangeability model for leveraging real-world data in clinical trials.

The traditional trial paradigm is often criticized as being slow, inefficient, and costly. Statistical approaches that leverage external trial data have emerged to make trials more efficient by augmenting the sample size. However, these approaches assume that external data are from previously conducted trials, leaving a rich source of untapped real-world data (RWD) that cannot yet be effectively leveraged. We propose a semi-supervised mixture (SS-MIX) multisource exchangeability model (MEM); a flexible, two-step Bayesian approach for incorporating RWD into randomized controlled trial analyses. The first step is a SS-MIX model on a modified propensity score and the second step is a MEM. The first step targets a representative subgroup of individuals from the trial population and the second step avoids borrowing when there are substantial differences in outcomes among the trial sample and the representative observational sample. When comparing the proposed approach to competing borrowing approaches in a simulation study, we find that our approach borrows efficiently when the trial and RWD are consistent, while mitigating bias when the trial and external data differ on either measured or unmeasured covariates. We illustrate the proposed approach with an application to a randomized controlled trial investigating intravenous hyperimmune immunoglobulin in hospitalized patients with influenza, while leveraging data from an external observational study to supplement a subgroup analysis by influenza subtype.

Sampling weighting strategies in causal mediation analysis.

BACKGROUND: Causal mediation analysis plays a crucial role in examining causal effects and causal mechanisms. Yet, limited work has taken into consideration the use of sampling weights in causal mediation analysis. In this study, we compared different strategies of incorporating sampling weights into causal mediation analysis. METHODS: We conducted a simulation study to assess 4 different sampling weighting strategies-1) not using sampling weights, 2) incorporating sampling weights into mediation "cross-world" weights, 3) using sampling weights when estimating the outcome model, and 4) using sampling weights in both stages. We generated 8 simulated population scenarios comprising an exposure (A), an outcome (Y), a mediator (M), and six covariates (C), all of which were binary. The data were generated so that the true model of A given C and the true model of A given M and C were both logit models. We crossed these 8 population scenarios with 4 different sampling methods to obtain 32 total simulation conditions. For each simulation condition, we assessed the performance of 4 sampling weighting strategies when calculating sample-based estimates of the total, direct, and indirect effects. We also applied the four sampling weighting strategies to a case study using data from the National Survey on Drug Use and Health (NSDUH). RESULTS: Using sampling weights in both stages (mediation weight estimation and outcome models) had the lowest bias under most simulation conditions examined. Using sampling weights in only one stage led to greater bias for multiple simulation conditions. DISCUSSION: Using sampling weights in both stages is an effective approach to reduce bias in causal mediation analyses under a variety of conditions regarding the structure of the population data and sampling methods.

Optimization-Based Stable Balancing Weights Versus Propensity Score Weighting for Samples With High Covariate Imbalance.

PURPOSE: To compare the performance (covariate balance, effective sample size [ESS]) of stable balancing weights (SBW) versus propensity score weighting (PSW). Two applied cases were used to compare performance: (Case 1) extreme imbalance in baseline covariates between groups and (Case 2) substantial discrepancy in sample size between groups. METHODS: Using the Premier Healthcare Database, we selected patients who (Case 1) underwent a surgical procedure with one of two different bipolar forceps between January 2000 and June 2020, or (Case 2) a neurological procedure using one of two different nonabsorbable surgical sutures between January 2000 and March 2020. Average treatment effects on the treated (ATT) weights were generated based on selected covariates. SBW was implemented using two techniques: (1) "grid search" to find weights of minimum variance at the lowest target absolute standardized mean difference (SMD); (2) finding weights of minimum variance at prespecified SMD tolerance. PSW and SBW methods were compared on postweighting SMDs, the number of imbalanced covariates, and ESS of the ATT-weighted control group. RESULTS: In both studies, improved covariate balance was achieved with both SBW techniques. All methods suffered from postweighting ESS that was lower than the unweighted control group's original sample size; however, SBW methods achieved higher ESS for the control groups. Sensitivity analyses using SBW to apply variable-specific SMD thresholds increased ESS, outperforming PSW. CONCLUSIONS: In this applied example, the optimization-based SBW method provided ample flexibility with respect to prespecification of covariate balance goals and resulted in better postweighting covariate balance and larger ESS as compared with PSW.

Evaluating the impact of a trial of labor after cesarean section on labor duration: a retrospective cohort study.

BACKGROUND: Cesarean section (C-section) rates are increasing globally, and repeated C-sections are associated with increased maternal morbidity. Trial of labor after C-section (TOLAC) is an approach to reduce the recurrence of C-sections. However, limited research exists on the impact of cesarean scars on labor duration in TOLAC, considering the termination of labor through C-section and selection bias. This study aimed to investigate the impact of cesarean scars on labor duration in TOLAC participants, accounting for potential confounding factors and biases. METHODS: This retrospective cohort study included 2,964 women who attempted vaginal birth at a single center in Japan from 2012 to 2021. The study categorized participants into TOLAC (n = 187) and non-TOLAC (n = 2,777) groups. Propensity scores were calculated based on 14 factors that could influence labor duration, and inverse probability of treatment weighting (IPTW) was applied. Cox proportional hazards regression analysis estimated hazard ratios (HRs) for labor duration, with and without IPTW adjustment. Sensitivity analyses used propensity score matching, bootstrapping, and interval censoring to address potential biases, including recall bias in the reported onset of labor. RESULTS: The unadjusted HR for labor duration in the TOLAC group compared to the non-TOLAC group was 0.83 (95% CI: 0.70-0.98, P = 0.027), indicating a longer labor duration in the TOLAC group. After adjusting for confounding factors using IPTW, the HR was 0.98 (95% CI: 0.74-1.30, P = 0.91), suggesting no significant difference in labor duration between the groups. Sensitivity analyses using propensity score matching, bootstrapping, and interval censoring yielded consistent results. These findings suggested that the apparent association between TOLAC and longer labor duration was because of confounding factors rather than TOLAC itself. CONCLUSIONS: After adjusting for confounding factors and addressing potential biases, cesarean scars had a limited impact on labor duration in TOLAC participants. Maternal and fetal characteristics may have a more substantial influence on labor duration.

Understanding the effects of Haemophilus influenzae colonization on bronchiectasis: a retrospective cohort study.

BACKGROUND: Bacterial colonization is an essential aspect of bronchiectasis. Although Haemophilus influenzae is a frequent colonizer in some regions, its clinical impacts are poorly understood. This study aimed to elucidate the impact of H. influenzae colonization in patients with bronchiectasis. METHODS: This retrospective study screened adult patients diagnosed with bronchiectasis at a tertiary referral center between April 1, 2003, and May 16, 2021, in South Korea. Propensity score matching was used to match patients with and without H. influenzae colonization. We assessed the severity of bronchiectasis as per the bronchiectasis severity index, the incidence of exacerbation, differences in lung function, and all-cause mortality. RESULTS: Out of the 4,453 patients with bronchiectasis, 79 (1.8%) were colonized by H. influenzae. After 1:2 propensity score matching, 78 and 154 patients were selected from the H. influenzae colonizer and non-colonizer groups, respectively. Although there were no significant differences between the groups regarding baseline demographics, patients colonized with H. influenzae had a higher bronchiectasis severity index (median 6 [interquartile range 4-8] vs. 4 [2-7], p = 0.002), associated with extensive radiographic involvement (52.2% vs. 37.2%, p = 0.045) and mild exacerbation without hospitalization (adjusted incidence rate ratio 0.15; 95% confidence interval 0.12-0.24). Lung function and mortality rates did not reveal significant differences, regardless of H. influenzae colonization. CONCLUSION: H. influenzae colonization in bronchiectasis was associated with more severe disease and greater incidence of mild exacerbation, but not lung function and mortality. Attention should be paid to patients with bronchiectasis with H. influenzae colonization.

Real-World Effectiveness of Single-Inhaler Triple Therapy for COPD: Impact of Diabetes Comorbidity.

Type 2 diabetes is a frequent comorbidity in chronic obstructive pulmonary disease (COPD) patients, with the GOLD treatment recommendations asserting that the presence of diabetes be disregarded in the choice of treatment.In a cohort of COPD patients with frequent exacerbations, initiators of single-inhaler triple therapy or dual bronchodilators were compared on the incidence of COPD exacerbation and pneumonia over one year, adjusted by propensity score weighting and stratified by type 2 diabetes.The COPD cohort included 1,114 initiators of triple inhalers and 4,233 of dual bronchodilators (28% with type 2 diabetes). The adjusted hazard ratio (HR) of exacerbation with triple therapy was 1.04 (95% CI: 0.86-1.25) among COPD patients with type 2 diabetes and 0.74 (0.65-0.85) in those without. The incidence of severe pneumonia was elevated with triple therapy among patients with type 2 diabetes (HR 1.77; 1.14-2.75).Triple therapy in COPD is effective among those without, but not those with, type 2 diabetes. Future therapeutic trials in COPD should consider diabetes comorbidity.

Triple therapy for frequent COPD exacerbators is effective in patients without type 2 diabetes but not in those with type 2 diabetes. The impact of comorbidities should be considered in future COPD therapeutic trials.

Parametric g-formula for Testing Time-Varying Causal Effects: What It Is, Why It Matters, and How to Implement It in Lavaan.

Psychologists leverage longitudinal designs to examine the causal effects of a focal predictor (i.e., treatment or exposure) over time. But causal inference of naturally observed time-varying treatments is complicated by treatment-dependent confounding in which earlier treatments affect confounders of later treatments. In this tutorial article, we introduce psychologists to an established solution to this problem from the causal inference literature: the parametric g-computation formula. We explain why the g-formula is effective at handling treatment-dependent confounding. We demonstrate that the parametric g-formula is conceptually intuitive, easy to implement, and well-suited for psychological research. We first clarify that the parametric g-formula essentially utilizes a series of statistical models to estimate the joint distribution of all post-treatment variables. These statistical models can be readily specified as standard multiple linear regression functions. We leverage this insight to implement the parametric g-formula using lavaan, a widely adopted R package for structural equation modeling. Moreover, we describe how the parametric g-formula may be used to estimate a marginal structural model whose causal parameters parsimoniously encode time-varying treatment effects. We hope this accessible introduction to the parametric g-formula will equip psychologists with an analytic tool to address their causal inquiries using longitudinal data.

Propensity scores as a novel method to guide sample allocation and minimize batch effects during the design of high throughput experiments.

BACKGROUND: We developed a novel approach to minimize batch effects when assigning samples to batches. Our algorithm selects a batch allocation, among all possible ways of assigning samples to batches, that minimizes differences in average propensity score between batches. This strategy was compared to randomization and stratified randomization in a case-control study (30 per group) with a covariate (case vs control, represented as ß1, set to be null) and two biologically relevant confounding variables (age, represented as ß2, and hemoglobin A1c (HbA1c), represented as ß3). Gene expression values were obtained from a publicly available dataset of expression data obtained from pancreas islet cells. Batch effects were simulated as twice the median biological variation across the gene expression dataset and were added to the publicly available dataset to simulate a batch effect condition. Bias was calculated as the absolute difference between observed betas under the batch allocation strategies and the true beta (no batch effects). Bias was also evaluated after adjustment for batch effects using ComBat as well as a linear regression model. In order to understand performance of our optimal allocation strategy under the alternative hypothesis, we also evaluated bias at a single gene associated with both age and HbA1c levels in the 'true' dataset (CAPN13 gene). RESULTS: Pre-batch correction, under the null hypothesis (ß1), maximum absolute bias and root mean square (RMS) of maximum absolute bias, were minimized using the optimal allocation strategy. Under the alternative hypothesis (ß2 and ß3 for the CAPN13 gene), maximum absolute bias and RMS of maximum absolute bias were also consistently lower using the optimal allocation strategy. ComBat and the regression batch adjustment methods performed well as the bias estimates moved towards the true values in all conditions under both the null and alternative hypotheses. Although the differences between methods were less pronounced following batch correction, estimates of bias (average and RMS) were consistently lower using the optimal allocation strategy under both the null and alternative hypotheses. CONCLUSIONS: Our algorithm provides an extremely flexible and effective method for assigning samples to batches by exploiting knowledge of covariates prior to sample allocation.

Federated causal inference in heterogeneous observational data.

We are interested in estimating the effect of a treatment applied to individuals at multiple sites, where data is stored locally for each site. Due to privacy constraints, individual-level data cannot be shared across sites; the sites may also have heterogeneous populations and treatment assignment mechanisms. Motivated by these considerations, we develop federated methods to draw inferences on the average treatment effects of combined data across sites. Our methods first compute summary statistics locally using propensity scores and then aggregate these statistics across sites to obtain point and variance estimators of average treatment effects. We show that these estimators are consistent and asymptotically normal. To achieve these asymptotic properties, we find that the aggregation schemes need to account for the heterogeneity in treatment assignments and in outcomes across sites. We demonstrate the validity of our federated methods through a comparative study of two large medical claims databases.

Frameworks for estimating causal effects in observational settings: comparing confounder adjustment and instrumental variables.

To estimate causal effects, analysts performing observational studies in health settings utilize several strategies to mitigate bias due to confounding by indication. There are two broad classes of approaches for these purposes: use of confounders and instrumental variables (IVs). Because such approaches are largely characterized by untestable assumptions, analysts must operate under an indefinite paradigm that these methods will work imperfectly. In this tutorial, we formalize a set of general principles and heuristics for estimating causal effects in the two approaches when the assumptions are potentially violated. This crucially requires reframing the process of observational studies as hypothesizing potential scenarios where the estimates from one approach are less inconsistent than the other. While most of our discussion of methodology centers around the linear setting, we touch upon complexities in non-linear settings and flexible procedures such as target minimum loss-based estimation and double machine learning. To demonstrate the application of our principles, we investigate the use of donepezil off-label for mild cognitive impairment. We compare and contrast results from confounder and IV methods, traditional and flexible, within our analysis and to a similar observational study and clinical trial.

Confounding-adjustment methods for the causal difference in medians.

BACKGROUND: With continuous outcomes, the average causal effect is typically defined using a contrast of expected potential outcomes. However, in the presence of skewed outcome data, the expectation (population mean) may no longer be meaningful. In practice the typical approach is to continue defining the estimand this way or transform the outcome to obtain a more symmetric distribution, although neither approach may be entirely satisfactory. Alternatively the causal effect can be redefined as a contrast of median potential outcomes, yet discussion of confounding-adjustment methods to estimate the causal difference in medians is limited. In this study we described and compared confounding-adjustment methods to address this gap. METHODS: The methods considered were multivariable quantile regression, an inverse probability weighted (IPW) estimator, weighted quantile regression (another form of IPW) and two little-known implementations of g-computation for this problem. Methods were evaluated within a simulation study under varying degrees of skewness in the outcome and applied to an empirical study using data from the Longitudinal Study of Australian Children. RESULTS: Simulation results indicated the IPW estimator, weighted quantile regression and g-computation implementations minimised bias across all settings when the relevant models were correctly specified, with g-computation additionally minimising the variance. Multivariable quantile regression, which relies on a constant-effect assumption, consistently yielded biased results. Application to the empirical study illustrated the practical value of these methods. CONCLUSION: The presented methods provide appealing avenues for estimating the causal difference in medians.

Bayesian approaches to include real-world data in clinical studies.

Randomized clinical trials have been the mainstay of clinical research, but are prohibitively expensive and subject to increasingly difficult patient recruitment. Recently, there is a movement to use real-world data (RWD) from electronic health records, patient registries, claims data and other sources in lieu of or supplementing controlled clinical trials. This process of combining information from diverse sources calls for inference under a Bayesian paradigm. We review some of the currently used methods and a novel non-parametric Bayesian (BNP) method. Carrying out the desired adjustment for differences in patient populations is naturally done with BNP priors that facilitate understanding of and adjustment for population heterogeneities across different data sources. We discuss the particular problem of using RWD to create a synthetic control arm to supplement single-arm treatment only studies. At the core of the proposed approach is the model-based adjustment to achieve equivalent patient populations in the current study and the (adjusted) RWD. This is implemented using common atoms mixture models. The structure of such models greatly simplifies inference. The adjustment for differences in the populations can be reduced to ratios of weights in such mixtures. This article is part of the theme issue 'Bayesian inference: challenges, perspectives, and prospects'.

Utility of propensity score-based Bayesian borrowing of external adult data in pediatric trials: A pragmatic evaluation through a case study in acute lymphoblastic leukemia (ALL).

A fully powered randomized controlled cancer trial can be challenging to conduct in children because of difficulties in enrollment of pediatric patients due to low disease incidence. One way to improve the feasibility of trials in pediatric patients, when clinically appropriate, is through borrowing information from comparable external adult trials in the same disease. Bayesian analysis of a pediatric trial provides a way of seamlessly augmenting pediatric trial efficacy data with data from external adult trials. However, not all external adult trial subjects may be equally clinically relevant with respect to the baseline disease severity, prognostic factors, co-morbidities, and prior therapy observed in the pediatric trial of interest. The propensity score matching method provides a way of matching the external adult subjects to the pediatric trial subjects on a set of clinically determined baseline covariates, such as baseline disease severity, prognostic factors and prior therapy. The matching then allows Bayesian information borrowing from only the most clinically relevant external adult subjects. Through a case study in pediatric acute lymphoblastic leukemia (ALL), we examine the utility of propensity score matched mixture and power priors in bringing appropriate external adult efficacy information into pediatric trial efficacy assessment, and present considerations for scaling fixed borrowing from external adult data.

Patient-reported health outcomes of SARS-CoV-2-tested patients presenting to emergency departments: a propensity score-matched prospective cohort study.

OBJECTIVE: This study aimed to compare the long-term physical and mental health outcomes of matched severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2)-positive and SARS-CoV-2-negative patients controlling for seasonal effects. STUDY DESIGN: This was a retrospective cohort study. METHODS: This study enrolled patients presenting to emergency departments participating in the Canadian COVID-19 Emergency Department Rapid Response Network. We enrolled consecutive eligible consenting patients who presented between March 1, 2020, and July 14, 2021, and were tested for SARS-CoV-2. Research assistants randomly selected four site and date-matched SARS-CoV-2-negative controls for every SARS-CoV-2-positive patient and interviewed them at least 30 days after discharge. We used propensity scores to match patients by baseline characteristics and used linear regression to compare Veterans RAND 12-item physical health component score (PCS) and mental health component scores (MCS), with higher scores indicating better self-reported health. RESULTS: We included 1170 SARS-CoV-2-positive patients and 3716 test-negative controls. The adjusted mean difference for PCS was 0.50 (95% confidence interval [CI]: -0.36, 1.36) and -1.01 (95% CI: -1.91, -0.11) for MCS. Severe disease was strongly associated with worse PCS (ß = -7.4; 95% CI: -9.8, -5.1), whereas prior mental health illness was strongly associated with worse MCS (ß = -5.4; 95% CI: -6.3, -4.5). CONCLUSION: Physical health, assessed by PCS, was similar between matched SARS-CoV-2-positive and SARS-CoV-2-negative patients, whereas mental health, assessed by MCS, was worse during a time when the public experienced barriers to care. These results may inform the development and prioritization of support programs for patients.

The Role of Pre-Treatment Traumatic Stress Symptoms in Adolescent Substance Use Treatment Outcomes.

Background: Prominent theories suggest that individuals with co-occurring traumatic stress symptoms (TSS) and substance use (SU) may be less responsive to SU treatment compared to those with SU only. However, empirical findings in adult samples are mixed, and there has been limited work among adolescents. This study assesses the association between TSS and SU treatment outcomes among trauma-exposed adolescents, using statistical methods to reduce potential confounding from important factors such as baseline SU severity. Method: 2,963 adolescents with lifetime history of victimization received evidence-based SU treatment in outpatient community settings. At baseline, 3- and 6-months, youth were assessed using the Global Appraisal of Individual Needs Traumatic Stress Scale and the Substance Frequency Scale. Propensity score weighting was used to mitigate potential confounding due to baseline differences in sociodemographic characteristics and SU across youth with varying levels of TSS. Results: Propensity score weighting successfully balanced baseline differences in sociodemographic factors and baseline SU across youth. Among all youth, mean SU was lower at both 3- and 6- month follow-up relative to baseline, indicating declining use. After adjusting for potential confounders, we observed no statistically significant relationship between TSS and SU at either 3- or 6-month follow-up. Conclusions: Based on this investigation, conducted among a large sample of trauma-exposed youth receiving evidence-based outpatient SU treatment, baseline TSS do not appear to be negatively associated with SU treatment outcomes. However, future research should examine whether youth with TSS achieve better outcomes through integrative treatment for both SU and TSS.

The results of this study provide keenly needed evidence that, among youth with prior victimization, presence and level of traumatic stress symptoms at substance use treatment initiation does not lead to significantly worse treatment outcomes for youth in outpatient treatment. This suggests that evidence-based outpatient substance use treatment modalities may be effective at improving substance use outcomes even when co-existing traumatic stress symptoms are present.

Identifying and Estimating Causal Moderation for Treated and Targeted Subgroups.

Extant literature on moderation effects narrowly focuses on the average moderated treatment effect across the entire sample (AMTE). Missing is the average moderated treatment effect on the treated (AMTT) and other targeted subgroups (AMTS). Much like the average treatment effect on the treated (ATT) for main effects, the AMTS changes the target of inferences from the entire sample to targeted subgroups. Relative to the AMTE, the AMTS is identified under weaker assumptions and often captures more policy-relevant effects. We present a theoretical framework that introduces the AMTS under the potential outcomes framework and delineates the assumptions for causal identification. We then propose a generalized propensity score method as a tool to estimate the AMTS using weights derived with Bayes Theorem. We illustrate the results and differences among the estimands using data from the Early Childhood Longitudinal Study. We conclude with suggestions for future research.

The Mortality Risk of Proton Pump Inhibitors in 1.9 Million US Seniors: An Extended Cox Survival Analysis.

BACKGROUND & AIMS: Observational studies have linked proton pump inhibitors (PPIs) with increased risk of mortality and other safety outcomes, in contradiction with a recent PPI randomized controlled trial (RCT). Observational studies may be prone to reverse causality, where deaths are attributed to the treatment rather than the conditions that are treated (protopathic bias). METHODS: We analyzed an incident drug user cohort of 1,930,728 elderly Medicare fee-for-service beneficiaries to evaluate the PPI-associated risk of death with a Cox regression analysis with time-varying covariates and propensity score adjustments. To correct for protopathic bias which occurs when a given drug is associated with prodromal signs of death, we implemented a lag-time approach by which any study drug taken during a 90-day look-back window before each death was disregarded. RESULTS: Among 1,930,728 study individuals, 80,972 (4.2%) died during a median 3.8 years of follow-up, yielding an overall unadjusted death rate/1000 person-years of 9.85; 14.31 for PPI users and 7.93 for non- users. With no lag-time, PPI use (vs no use) was associated with 10% increased mortality risk (adjusted HR=1.10; 95% CI 1.08-1.12). However, with a lag-time of 90 days, mortality risk associated with PPI use was near zero (adjusted HR=1.01; 95% CI 0.99-1.02). CONCLUSION: Given the usage patterns of PPIs in patients with conditions that may presage death, protopathic bias may explain the association of PPIs with increased risk of death reported in observational studies.

Resource Use Among Diabetes Patients Who Mainly Visit Primary Care Physicians Versus Medical Specialists: a Retrospective Cohort Study.

BACKGROUND: It is not uncommon for medical specialists to predominantly care for patients with certain chronic conditions rather than primary care physicians (PCPs), yet the resource implications from such patterns of care are not well understood. OBJECTIVE: To assess resource use of diabetes patients who predominantly visit a PCP versus a medical specialist. DESIGN: Retrospective cohort study of diabetes patients aging into the traditional Medicare program. Patients were attributed to a PCP or medical specialist annually based on a preponderance of ambulatory care visits and categorized according to whether attribution changed year to year. Propensity score weighting was used to balance baseline demographic characteristics, diabetes complications, and underlying health conditions between patients attributed to PCPs and to medical specialists. Spending and utilization were measured up to 3 patient-years. SUBJECTS: A total of 141,558 patient-years. MAIN MEASURES: Total visits, unique physicians, hospital admissions, emergency department visits, procedures, imaging, and tests. KEY RESULTS: Each year, roughly 70% of patients maintained attribution to a PCP and 15% to a medical specialist relative to the previous year. After propensity weighting, patients continuously attributed to a PCP versus medical specialist from 1 year to the next had lower average total payer payments ($10,326 [SD $57,386] versus $14,971 [SD $74,112], P<0.0001) and lower total patient out-of-pocket payments ($1,707 [SD $6,020] versus $2,443 [SD $7,984], P<0.0001). Rates of hospitalization, emergency department visits, procedures, imaging, and tests were lower among patients attributed to PCPs as well. CONCLUSIONS: Older adults with diabetes who receive more of their ambulatory care from a PCP instead of a medical specialist show evidence of lower resource use.