A review of covariate selection for non-experimental comparative effectiveness research

A methodological review of the high-dimensional propensity score in comparative-effectiveness and safety-of-interventions research finds incomplete reporting relative to algorithm development and robustness

OBJECTIVES

The use of secondary databases has become popular for evaluating the effectiveness and safety of interventions in real-life settings. However, the absence of important confounders in these databases is challenging. To address this issue, the high-dimensional propensity score (hdPS) algorithm was developed in 2009. This algorithm uses proxy variables for mitigating confounding by combining information available across several healthcare dimensions. This study assessed the methodology and reporting of the hdPS in comparative effectiveness and safety research.

STUDY DESIGN AND SETTING

In this methodological review, we searched PubMed and Google Scholar from July 2009 to May 2022 for studies that used the hdPS for evaluating the effectiveness or safety of healthcare interventions. Two reviewers independently extracted study characteristics and assessed how the hdPS was applied and reported. Risk of bias was evaluated with the Risk Of Bias In Non-randomised Studies - of Interventions (ROBINS-I) tool.

RESULTS

In total, 136 studies met the inclusion criteria; the median publication year was 2018 (Q1-Q3 2016-2020). The studies included 192 datasets, mostly North American databases (n = 132, 69%). The hdPS was used in primary analysis in 120 studies (88%). Dimensions were defined in 101 studies (74%), with a median of 5 (Q1-Q3 4-6) dimensions included. A median of 500 (Q1-Q3 200-500) empirically identified covariates were selected. Regarding hdPS reporting, only 11 studies (8%) reported all recommended items. Most studies (n = 81, 60%) had a moderate overall risk of bias.

CONCLUSION

There is room for improvement in the reporting of hdPS studies, especially regarding the transparency of methodological choices that underpin the construction of the hdPS.

Use of virtual care in ophthalmology in Ontario, Canada in 2020 during the COVID-19 pandemic

OBJECTIVE

To investigate the use and trends of virtual care in ophthalmology and examine associated factors in a universal health care system during the COVID-19 pandemic in 2020.

DESIGN

Cross-sectional study.

PARTICIPANTS

Ontarians eligible for the Ontario Health Insurance Plan.

METHODS

We used physician billing data from 2017-2020 to describe the use of virtual versus in-person care. We used logistic regression to examine factors associated with virtual care use.

RESULTS

The uptake of ophthalmic virtual visits increased immediately following the government's directive to ramp down clinic activities and institution of a new virtual fee code (17.6%), peaked 2 weeks later (55.8%), and decreased immediately after the directive was lifted (24.2%). In March-December 2020, virtual visits were higher in female (11.6%) versus male (10.3%) patients and in patients <20 years of age (16.4%) and 20-39 years of age (12.3%) versus those aged 40-64 years (10.8%) and 65+ years (10.6%). Patients residing in the poorest/poorer neighbourhoods (10.9%) had similar use as their counterparts (11.1%). Patients with an acute infectious disease (14.2%) or nonurgent diagnosis (16.2%) had the highest use. Those with retinal disease had the lowest use (4.2%). Female ophthalmologists (15.4%) provided virtual care more often than male ophthalmologists (9.9%). Ophthalmologists aged 60-69 years (13.1%) provided virtual care more often than any other age groups (<40 years: 11.3%; 40-49 years: 11.0%; 50-59 years: 10.0%; and 70+ years: 7.7%). Multiple logistic regression models revealed similar results.

CONCLUSION

Virtual care in ophthalmology increased significantly during the initial phase of the pandemic and decreased thereafter. There were significant variations in virtual care use by patient and ophthalmologist characteristics.

Real-world data analysis on effectiveness of integrative therapies: A practical guide to study design and data analysis using healthcare databases

Real world data (RWD) is increasingly used to investigate health outcomes and treatment efficacy in the field of integrative medicine. Due to the fact that the majority of RWDs are not intended for research, their secondary use in research necessitates complex study designs to account for bias and confounding. To conduct a robust analysis of RWD in integrative medicine, a comprehensive study design process that reflects the characteristics of integrative therapies is necessary. In this paper, we present a guide for designing comparative effectiveness RWE research in integrative medicine. We discuss key factors to consider when selecting RWDs for research on integrative medicine. We provide practical steps for developing a research question, formulating the PICOT objectives (population, intervention, comparator, outcome, and time horizon), and selecting and defining covariates with a summary table. Specific study designs are depicted with corresponding diagrams. Finally, data analysis procedures are introduced. We hope this article clarifies the importance of RWE research design and related processes in order to improve the rigor of RWD studies in the field of integrative medicine research.

Treatment Disparities Partially Mediate Socioeconomic- and Race/Ethnicity-Based Survival Disparities in Stage I-II Hepatocellular Carcinoma

BACKGROUND

Low socioeconomic status (SES) patients with early-stage hepatocellular carcinoma (HCC) receive procedural treatments less often and have shorter survival. Little is known about the extent to which these survival disparities result from treatment-related disparities versus other causal pathways. We aimed to estimate the proportion of SES-based survival disparities that are mediated by treatment- and facility-related factors among patients with stage I-II HCC.

METHODS

We analyzed patients aged 18-75 years diagnosed with stage I-II HCC in 2008-2016 using the National Cancer Database. Inverse odds weighting mediation analysis was used to calculate the proportion mediated by three mediators: procedure type, facility volume, and facility procedural interventions offered. Intersectional analyses were performed to determine whether treatment disparities played a larger role in survival disparities among Black and Hispanic patients.

RESULTS

Among 46,003 patients, 15.0% had low SES, 71.6% had middle SES, and 13.4% had high SES. Five-year overall survival was 46.9%, 39.9%, and 35.7% among high, middle, and low SES patients, respectively. Procedure type mediated 45.9% (95% confidence interval [CI] 31.1-60.7%) and 36.7% (95% CI 25.7-47.7%) of overall survival disparities for low and middle SES patients, respectively, which was more than was mediated by the two facility-level mediators. Procedure type mediated a larger proportion of survival disparities among low-middle SES Black (46.6-48.2%) and Hispanic patients (92.9-93.7%) than in White patients (29.5-29.7%).

CONCLUSIONS

SES-based disparities in use of procedural interventions mediate a large proportion of survival disparities, particularly among Black and Hispanic patients. Initiatives aimed at attenuating these treatment disparities should be pursued.

Schnider and Eleveld Models for Propofol Target-Controlled Infusion Anesthesia: A Clinical Comparison

BACKGROUND

Various pharmacokinetic/pharmacodynamic (PK/PD) models have been developed to accurately dose propofol administration during total intravenous anesthesia with target-controlled infusion (TIVA-TCI). We aim to clinically compare the performance of the Schnider model and the new and general-purpose Eleveld PK/PD model during TIVA-TCI.

METHODS

We conducted a prospective observational study at a single center, enrolling 78 female patients, including 37 adults (aged < 65 years) and 41 elderly patients (aged ≥ 65 years). These patients underwent breast surgery with propofol-remifentanil TIVA-TCI guided by the bispectral index (BIS) for depth of anesthesia monitoring (target value 40-60) and the surgical plethysmographic index (SPI) for antinociception monitoring (target value 20-50) without neuromuscular blockade. The concentration at the effect site of propofol (CeP) at loss of responsiveness (LoR) during anesthesia maintenance (MA) and at return of responsiveness (RoR), the duration of surgery and anesthesia (min), the time to RoR (min), the propofol total dose (mg), the deepening of anesthesia events (DAEs), burst suppression events (BSEs), light anesthesia events (LAEs) and unwanted spontaneous responsiveness events (USREs) were considered to compare the two PK/PD models.

RESULTS

Patients undergoing BIS-SPI-guided TIVA-TCI with the Eleveld PK/PD model showed a lower CeP at LoR (1.7 (1.36-2.25) vs. 3.60 (3.00-4.18) μg/mL, p < 0.001), higher CePMA (2.80 (2.55-3.40) vs. 2.30 (1.80-2.50) μg/mL, p < 0.001) and at RoR (1.48 (1.08-1.80) vs. 0.64 (0.55-0.81) μg/mL, p < 0.001) than with the Schnider PK/PD model. Anesthetic hysteresis was observed only in the Schnider PK/PD model group (p < 0.001). DAEs (69.2% vs. 30.8%, p = 0.001) and BSEs (28.2% vs. 5.1%, p = 0.013) were more frequent with the Eleveld PK/PD model than with the Schnider PK/PD model in the general patient population. DAEs (63.2% vs. 27.3%, p = 0.030) and BSEs (31.6% vs. 4.5%, p = 0.036) were more frequent with the Eleveld PK/PD model than with the Schnider PK/PD model in the elderly.

CONCLUSIONS

The Schnider and Eleveld PK/PD models impact CePs differently. A greater incidence of DAEs and BSEs in the elderly suggests more attention is necessary in this group of patients undergoing BIS-SPI-guided TIVA-TCI with the Eleveld PK/PD than with the Schnider model.

Directed acyclic graphs for clinical research: a tutorial

Directed acyclic graphs (DAGs) are useful tools for visualizing the hypothesized causal structures in an intuitive way and selecting relevant confounders in causal inference. However, in spite of their increasing use in clinical and surgical research, the causal graphs might also be misused by a lack of understanding of the central principles. In this article, we aim to introduce the basic terminology and fundamental rules of DAGs, and DAGitty, a user-friendly program that easily displays DAGs. Specifically, we describe how to determine variables that should or should not be adjusted based on the backdoor criterion with examples. In addition, the occurrence of the various types of biases is discussed with caveats, including the problem caused by the traditional approach using p-values for confounder selection. Moreover, a detailed guide to DAGitty is provided with practical examples regarding minimally invasive surgery. Essentially, the primary benefit of DAGs is to aid researchers in clarifying the research questions and the corresponding designs based on the domain knowledge. With these strengths, we propose that the use of DAGs may contribute to rigorous research designs, and lead to transparency and reproducibility in research on minimally invasive surgery.

Estimating the Individual Treatment Effect on Survival Time Based on Prior Knowledge and Counterfactual Prediction

The estimation of the Individual Treatment Effect (ITE) on survival time is an important research topic in clinics-based causal inference. Various representation learning methods have been proposed to deal with its three key problems, i.e., reducing selection bias, handling censored survival data, and avoiding balancing non-confounders. However, none of them consider all three problems in a single method. In this study, by combining the Counterfactual Survival Analysis (CSA) model and Dragonnet from the literature, we first propose a CSA–Dragonnet to deal with the three problems simultaneously. Moreover, we found that conclusions from traditional Randomized Controlled Trials (RCTs) or Retrospective Cohort Studies (RCSs) can offer valuable bound information to the counterfactual learning of ITE, which has never been used by existing ITE estimation methods. Hence, we further propose a CSA–Dragonnet with Embedded Prior Knowledge (CDNEPK) by formulating a unified expression of the prior knowledge given by RCTs or RCSs, inserting counterfactual prediction nets into CSA–Dragonnet and defining loss items based on the bounds for the ITE extracted from prior knowledge. Semi-synthetic data experiments showed that CDNEPK has superior performance. Real-world experiments indicated that CDNEPK can offer meaningful treatment advice.

Synthetic Negative Controls: Using Simulation to Screen Large-scale Propensity Score Analyses

The propensity score has become a standard tool to control for large numbers of variables in healthcare database studies. However, little has been written on the challenge of comparing large-scale propensity score analyses that use different methods for confounder selection and adjustment. In these settings, balance diagnostics are useful but do not inform researchers on which variables balance should be assessed or quantify the impact of residual covariate imbalance on bias. Here, we propose a framework to supplement balance diagnostics when comparing large-scale propensity score analyses. Instead of focusing on results from any single analysis, we suggest conducting and reporting results for many analytic choices and using both balance diagnostics and synthetically generated control studies to screen analyses that show signals of bias caused by measured confounding. To generate synthetic datasets, the framework does not require simulating the outcome-generating process. In healthcare database studies, outcome events are often rare, making it difficult to identify and model all predictors of the outcome to simulate a confounding structure closely resembling the given study. Therefore, the framework uses a model for treatment assignment to divide the comparator population into pseudo-treatment groups where covariate differences resemble those in the study cohort. The partially simulated datasets have a confounding structure approximating the study population under the null (synthetic negative control studies). The framework is used to screen analyses that likely violate partial exchangeability due to lack of control for measured confounding. We illustrate the framework using simulations and an empirical example.

Research and scholarly methods: Propensity scores

Propensity score methods are increasingly used as a tool to control for confounding bias in pharmacoepidemiologic studies. The propensity score is a dimension reducing balancing score, creating treatment and reference groups that have comparable distributions of measured covariates. The purpose of this methods review is to provide an overview of the use of propensity score methods, including a summary of important data assumptions, various applications of the propensity score, and how to evaluate covariate balance. This article is intended for pharmacists and researchers who wish to receive an introduction to propensity score methods and be able to engage in high-level discussions on application and reporting.

Assessments of risk of bias in systematic reviews of observational nutritional epidemiologic studies are often not appropriate or comprehensive: a methodological study

BACKGROUND

An essential component of systematic reviews is the assessment of risk of bias. To date, there has been no investigation of how reviews of non-randomised studies of nutritional exposures (called 'nutritional epidemiologic studies') assess risk of bias.

OBJECTIVE

To describe methods for the assessment of risk of bias in reviews of nutritional epidemiologic studies.

METHODS

We searched MEDLINE, EMBASE and the Cochrane Database of Systematic Reviews (Jan 2018-Aug 2019) and sampled 150 systematic reviews of nutritional epidemiologic studies.

RESULTS

Most reviews (n=131/150; 87.3%) attempted to assess risk of bias. Commonly used tools neglected to address all important sources of bias, such as selective reporting (n=25/28; 89.3%), and frequently included constructs unrelated to risk of bias, such as reporting (n=14/28; 50.0%). Most reviews (n=66/101; 65.3%) did not incorporate risk of bias in the synthesis. While more than half of reviews considered biases due to confounding and misclassification of the exposure in their interpretation of findings, other biases, such as selective reporting, were rarely considered (n=1/150; 0.7%).

CONCLUSION

Reviews of nutritional epidemiologic studies have important limitations in their assessment of risk of bias.

Lower Cardiorespiratory Fitness Is Associated With Right Ventricular Geometry and Function - The Sedentary's Heart: SHIP

Background

Lower cardiorespiratory fitness (CRF) is associated with an increased risk for cardiovascular disease. However, very little information is available about the association between lower CRF and right ventricular (RV) remodeling. We investigated the relationship between CRF and RV structure and function in a large, aging, and largely sedentary adult population–based cohort.

Methods and Results

We used cross‐sectional data of 2844 subjects (1486 women; median age, 51 years; interquartile range, 40–62 years) from the population‐based cohort SHIP (Study of Health in Pomerania) with echocardiography, of which 941 also had cardiac magnetic resonance imaging. We analyzed the associations of peak oxygen uptake with RV parameters determined by both imaging techniques using multivariable‐adjusted linear regression models. In echocardiography, a 1 L/min lower peak oxygen uptake was associated with a 1.18 mm (95% CI, 0.66–1.71; P<0.001) smaller RV end‐diastolic diameter and a 1.41 mm (95% CI, 0.90–1.92; P<0.001) narrower RV end‐diastolic outflow tract diameter. Similarly, using cardiac magnetic resonance imaging measurements, a 1 L/min lower peak oxygen uptake was associated with a 23.5 mL (95% CI, 18.7–28.4; P<0.001) smaller RV end‐diastolic volume, a 13.0 mL (95% CI, 9.81–16.2; P<0.001) lower RV end‐systolic volume, and a 10.7 mL/beat (95% CI, 8.10–13.3; P<0.001) lower RV stroke volume.

Conclusions

Our results indicate a significant association between CRF and RV remodeling. Lower CRF was associated with smaller RV chamber and lower RV systolic function, stroke volume, and cardiac output.

Early pregnancy exposure to metal mixture and birth outcomes - A prospective study in Project Viva

BACKGROUND

Prenatal exposure to metals has been individually associated with birth outcomes. However, little is known about the effect of metal mixture, particularly at low exposure levels.

OBJECTIVES

To estimate individual and joint effects of metal mixture components on birth outcomes.

METHODS

We used data from 1,391 mother-infant pairs in Project Viva (1999-2002). We measured 11 metals in maternal 1st trimester erythrocyte; abstracted birth weight from medical records; calculated gestational age from last menstrual period or ultrasound; and obtained birth length (n = 729) and head circumference (n = 791) from research measurements. We estimated individual and joint effects of metals using multivariable linear and Bayesian kernel machine regressions.

RESULTS

In both single metal and metal mixture analyses, exposure to higher concentrations of arsenic was associated with lower birth weight in males, zinc with higher head circumference in females, and manganese with higher birth length in sex-combined analysis. We also observed sex-specific metal interactions with birth outcomes. Arsenic and manganese showed a synergistic association with birth weight in males, in whom an interquartile range (IQR) increase in arsenic was associated with 25.3 g (95% CI: -79.9, 29.3), 47.9 g (95% CI: -98.0, 2.1), and 72.2 g (95% CI: -129.8, -14.7) lower birth weight when manganese concentrations were at 25th, 50th, and 75th percentiles, respectively. Lead and zinc showed an antagonistic association with head circumference in males, where an IQR increase in lead was associated with 0.18 cm (95% CI: -0.35, -0.02), 0.10 cm (95% CI: -0.25, 0.04), 0.03 cm (95% CI: -0.2, 0.14) smaller head circumference when zinc concentrations were at 25th, 50th, and 75th percentiles, respectively. Exposure to higher concentrations of arsenic was also associated with lower gestational age in males when concentrations of manganese and lead were higher.

DISCUSSION

Maternal erythrocyte concentrations of arsenic, manganese, lead, and zinc were individually and interactively associated with birth outcomes. The associations varied by infant sex and exposure level of other mixture components.

Association of Cardiovascular Outcomes and Mortality With Sustained Long-Acting Insulin Only vs Long-Acting Plus Short-Acting Insulin Treatment

IMPORTANCE

Cardiovascular events and mortality are the principal causes of excess mortality and health care costs for people with type 2 diabetes. No large studies have specifically compared long-acting insulin alone with long-acting plus short-acting insulin with regard to cardiovascular outcomes.

OBJECTIVE

To compare cardiovascular events and mortality in adults with type 2 diabetes receiving long-acting insulin who do or do not add short-acting insulin.

DESIGN, SETTING, AND PARTICIPANTS

This retrospective cohort study emulated a randomized experiment in which adults with type 2 diabetes who experienced a qualifying glycated hemoglobin A1c (HbA1c) level of 6.8% to 8.5% with long-acting insulin were randomized to continuing treatment with long-acting insulin (LA group) or adding short-acting insulin within 1 year of the qualifying HbA1c level (LA plus SA group). Retrospective data in 4 integrated health care delivery systems from the Health Care Systems Research Network from January 1, 2005, to December 31, 2013, were used. Analysis used inverse probability weighting estimation with Super Learner for propensity score estimation. Analyses took place from April 1, 2018, to June 30, 2019.

EXPOSURES

Long-acting insulin alone or with added short-acting insulin within 1 year from the qualifying HbA1c level.

MAIN OUTCOMES AND MEASURES

Mortality, cardiovascular mortality, acute myocardial infarction, stroke, and hospitalization for heart failure.

RESULTS

Among 57 278 individuals (39 279 with data on cardiovascular mortality) with a mean (SD) age of 60.6 (11.5) years, 53.6% men, 43.5% non-Hispanic White individuals, and 4 years of follow-up (median follow-up of 11 [interquartile range, 5-20] calendar quarters), the LA plus SA group was associated with increased all-cause mortality compared with the LA group (hazard ratio, 1.27; 95% CI, 1.05-1.49) and a decreased risk of acute myocardial infarction (hazard ratio, 0.89; 95% CI, 0.81-0.97). Treatment with long-acting plus short-acting insulin was not associated with increased risks of congestive heart failure, stroke, or cardiovascular mortality.

CONCLUSIONS AND RELEVANCE

Findings of this retrospective cohort study suggested an increased risk of all-cause mortality and a decreased risk of acute myocardial infarction for the LA plus SA group compared with the LA group. Given the lack of an increase in major cardiovascular events or cardiovascular mortality, the increased all-cause mortality with long-acting plus short-acting insulin may be explained by noncardiovascular events or unmeasured confounding.

Simulation-based design of pragmatic trials in psoriatic arthritis using propensity scores

BACKGROUND/AIMS

Design of clinical trials requires careful decision-making across several dimensions, including endpoints, eligibility criteria, and subgroup enrichment. Clinical trial simulation can be an informative tool in trial design, providing empirical evidence by which to evaluate and compare the results of hypothetical trials with varying designs. We introduce a novel simulation-based approach using observational data to inform the design of a future pragmatic trial.

METHODS

We utilize propensity score-adjusted models to simulate hypothetical trials under alternative endpoints and enrollment criteria. We apply our approach to the design of pragmatic trials in psoriatic arthritis, using observational data embedded within the Tight Control of Inflammation in Early Psoriatic Arthritis study to simulate hypothetical open-label trials comparing treatment with tumor necrosis factor-α inhibitors to methotrexate. We first validate our simulations of a trial with traditional enrollment criteria and endpoints against a recently published trial. Next, we compare simulated treatment effects in patient populations defined by traditional and broadened enrollment criteria, where the latter is consistent with a future pragmatic trial. In each trial, we also consider five candidate primary endpoints.

RESULTS

Our results highlight how changes in the enrolled population and primary endpoints may qualitatively alter study findings and the ability to detect heterogeneous treatment effects between clinical subgroups. For treatments of interest in the study of psoriatic arthritis, broadened enrollment criteria led to diluted estimated treatment effects. Endpoints with greater responsiveness to treatment compared with a traditionally used endpoint were identified. These considerations, among others, are important for designing a future pragmatic trial aimed at having high external validity with relevance for real-world clinical practice.

CONCLUSION

Observational data may be leveraged to inform design decisions in pragmatic trials. Our approach may be generalized to the study of other conditions where existing trial data are limited or do not generalize well to real-world clinical practice, but where observational data are available.

Lessons Learned Using Real-World Data to Emulate Randomized Trials: A Case Study of Treatment Effectiveness for Newly Diagnosed Immune Thrombocytopenia

Regulatory agencies are increasingly considering real-world evidence (RWE) to support label expansions of approved medicines. We conducted a comparative effectiveness study to emulate a proposed randomized trial of romiplostim vs. standard-of-care (SOC) therapy among patients with recently diagnosed (≤12 months) immune thrombocytopenia (ITP), that could support expansion of the romiplostim label. We discuss challenges that we encountered and solutions that were developed to address those challenges. Study size was a primary concern, particularly for romiplostim initiators, given the rarity of ITP and the stringent trial eligibility criteria. For this reason, we leveraged multiple data sources (Nordic Country Patient Registry for Romiplostim; chart review study of romiplostim initiators in Europe; Flatiron Health EMR linked with MarketScan claims). Additionally, unlike the strictly controlled clinical trial setting, platelet counts were not measured at regular intervals in the observational data sources, and therefore the end point of durable platelet response often used in trials could not be reliably measured. Instead, the median platelet count was chosen as the primary end point. Ultimately, while we observed a slightly higher median platelet count in the romiplostim group vs. SOC, precision was limited because of small study size (median difference was 11 × 10⁹ /L (95% CI: -59, 81)). We underscore the importance of conducting comprehensive feasibility assessments to identify fit-for-purpose data sources with sufficient sample size, data elements, and follow-up. Beyond technical challenges, we also discuss approaches to increase the credibility of RWE, including systematic incorporation of clinical expertise into study design decisions, and separation between decision makers and the data.

Quality and Location of the Surgical Episode Mediate a Large Proportion of Socioeconomic-Based Survival Disparities in Patients with Resected Stage I-III Colon Cancer

INTRODUCTION

Lower socioeconomic status (SES) is associated with shorter overall survival (OS) in patients with locoregional colon cancer. We aimed to estimate: (1) the proportion of SES-based OS disparities mediated by disparities in the quality and location of surgical treatment in patients with resected stage I-III colon cancer and (2) the relative importance of components of surgical quality.

PATIENTS AND METHODS

We examined patients ages 18-80 years with resected stage I-III colon adenocarcinoma using the 2010-2016 National Cancer Database. SES was defined at the zip code level. Inverse odds weighting mediation analysis was used to estimate the proportion mediated (PM) for nine treatment quality-related and facility-related factors and composite PMs in models including all nine mediators. Models compared high SES patients with each lower SES stratum.

RESULTS

Among 171,009 patients, 5-year OS increased from 70.4% in low SES patients to 78.1% in high SES. When high SES patients were compared with low, lower-middle, and upper-middle SES patients, PM ranges among lower SES strata were: minimally invasive surgery 16.0-16.6%, lymph nodes examined 7.7-9.6%, positive margins 3.8-6.5%, length of stay 16.7-28.1%, readmissions insignificant to 3.7%, treatment at > 1 CoC facility 2.7-3.1%, facility type insignificant to 7.3%, facility volume 2.9-8.2%, and adjusted facility 90-day mortality rates 33.2-42.8%. Composite PMs were 76.9% (95% CI 61.3%, 92.4%) for low SES, 68.7% (95% CI 56.4%, 81.1%) for lower-middle SES, and 60.9% (95% CI 43.1%, 78.6%) for upper-middle SES.

CONCLUSIONS

These data suggest that improving the quality of the surgical episode for disadvantaged patients undergoing resection for locoregional colon cancer could decrease SES-based survival disparities by over half.

Core concepts in pharmacoepidemiology: Violations of the positivity assumption in the causal analysis of observational data: Consequences and statistical approaches

In the causal analysis of observational data, the positivity assumption requires that all treatments of interest be observed in every patient subgroup. Violations of this assumption are indicated by nonoverlap in the data in the sense that patients with certain covariate combinations are not observed to receive a treatment of interest, which may arise from contraindications to treatment or small sample size. In this paper, we emphasize the importance and implications of this often-overlooked assumption. Further, we elaborate on the challenges nonoverlap poses to estimation and inference and discuss previously proposed methods. We distinguish between structural and practical violations and provide insight into which methods are appropriate for each. To demonstrate alternative approaches and relevant considerations (including how overlap is defined and the target population to which results may be generalized) when addressing positivity violations, we employ an electronic health record-derived data set to assess the effects of metformin on colon cancer recurrence among diabetic patients.

Current management of malignant bowel obstructions: a survey of acute care surgeons and surgical oncologists

Background

Malignant small bowel obstructions (MSBOs) are one of the most challenging problems surgeons encounter, and evidence-based treatment recommendations are lacking. We hypothesized that current opinions on MSBO management differ between acute care surgeons (ACSs) and surgical oncologists (SOs).

Methods

We developed three case scenarios describing patients with previously treated cancer who developed an MSBO. Each case had five to six alternate scenarios, intended to capture the heterogeneity of MSBO presentations. Members of the Society of Surgical Oncology, the American Society of Peritoneal Surface Malignancies, and the Eastern Association for the Surgery of Trauma were asked how likely they would be to offer surgical treatment in each scenario. Responses were analyzed for factors associated with the likelihood surgeons would offer surgical management.

Results

316 surgeons completed the survey: 119 (37.7%) SOs and 197 (62.3%) ACSs. Overall, SOs were nearly twice as likely as ACSs to recommend surgical management. The largest differences between provider groups were seen in patients with an increased metastatic burden. In a patient with MSBO with metastatic colon cancer, both SOs (95.8%) and ACSs (94.4%) were likely or very likely to offer an operation (p=0.587); however, this fell to 91.6% and 77.7%, respectively, when this patient had multiple hepatic metastases (p=0.001). All surgeons were less likely to offer surgery to patients with multiple sites of obstruction, recurrent MSBO, and shorter disease-free intervals.

Discussion

Opinions on MSBO management differ based on surgeon training and experience. Multidisciplinary management of patients with MSBO should be offered when available and increased emphasis placed on determining optimal management guidelines across specialties.

Level of evidence

Level IV Epidemiologic.

A Pharmacoepidemiologic Approach to Evaluate Real-world Effectiveness of Hormonal Contraceptives in the Presence of Drug-drug Interactions

Accurate estimation of conception is critical in the assessment of the effects of drugs used during pregnancy or to prevent pregnancy. In a novel application, we studied the effectiveness of oral contraceptives (OCs), where misclassification of conception relative to OC exposure may obscure effect estimates.

Treatment satisfaction mediates the association between perceived physician empathy and psychological distress in a community sample of individuals with chronic pain

OBJECTIVE

While the benefits of patient-centered care have been consistently demonstrated in the health literature, there exists a dearth of pathway research within health outcome research, especially within the chronic pain context. This study examined the relationship between perceived physician empathy and patient psychological distress and its underlying mechanism.

METHODS

A community sample of 259 adults with chronic pain completed online questionnaires measuring patient-perceived physician empathy, treatment satisfaction, depressive and anxiety symptoms. Analyses were conducted using correlational and mediation analyses.

RESULTS

Results revealed perceived empathy to be positively and strongly correlated with treatment satisfaction (r = .72, p < .001). A significant negative correlation was also demonstrated between perceived empathy and depressive symptoms (r = -.13, p < .05), but not between perceived empathy and anxious symptoms (r = .03, p = .65). Results revealed significant mediation models between perceived empathy and patient depressive symptoms (indirect effect: B = -.19, SE =.06, 95 % CI [-.31, -.09]) and anxious symptoms (indirect effect: B = -.24, SE = .06, 95 % CI [-.35, -.14]), via treatment satisfaction as mediator and including covariates.

CONCLUSION

Chronic pain patients who perceive greater levels of physician empathy experience fewer depressive and anxious symptoms, as mediated by treatment satisfaction.

PRACTICE IMPLICATIONS

Clinical training and practice should promote empathetic components of health communication within chronic pain treatment.

Decompositions of the Contribution of Treatment Disparities to Survival Disparities in Stage I-II Pancreatic Adenocarcinoma

BACKGROUND

Higher socioeconomic status (SES) and non-Hispanic White (NHW) race/ethnicity are associated with higher treatment rates and longer overall survival (OS) among US patients with stage I-II pancreatic ductal adenocarcinoma. The proportion of OS disparities mediated through treatment disparities (PM) and the proportion predicted to be eliminated (PE) if treatment disparities were eliminated are unknown.

METHODS

We analyzed 2007-2015 data from the Surveillance, Epidemiology, and End Results (SEER) census tract-level database and the National Cancer Database (NCDB) using causal mediation analysis methods to understand the extent to which treatment disparities mediate OS disparities. In the first set of decompositions, race/ethnicity was controlled for as a covariate proximal to SES, and lower SES strata were compared with the highest SES stratum. In the second set, an intersectional perspective was taken and each SES-race/ethnicity combination was compared with highest SES-NHW patients, who had the highest treatment rates and longest OS.

RESULTS

The SEER and NCDB cohorts contained 16,921 patients and 44,638 patients, respectively. When race/ethnicity was controlled for, PMs ranged from 43 to 48% and PEs ranged from 46 to 50% for various lower SES strata. When separately comparing each SES-race/ethnicity combination with the highest SES-NHW patients, results were similar for lower SES-NHW patients but differed markedly for non-Hispanic Black and Hispanic patients, for whom PMs ranged from 60 to 80% and PEs ranged from 55 to 75% for most lower SES strata.

CONCLUSIONS

These results suggest that efforts to reduce treatment disparities are worthwhile, particularly for NHB and Hispanic patients, and simultaneously point to the importance of non-treatment-related causal pathways.

Recurrence patterns in patients with Stage II melanoma: The evolving role of routine imaging for surveillance

BACKGROUND AND OBJECTIVES

The relatively recent availability of effective systemic therapies for metastatic melanoma necessitates reconsideration of current surveillance patterns. Evidence supporting surveillance guidelines for resected Stage II melanoma is lacking. Prior reports note routine imaging detects only 21% of recurrent disease. This study aims to define recurrence patterns for Stage II melanoma to inform future surveillance guidelines.

METHODS

This is a retrospective study of patients with Stage II melanoma. We analyzed risk factors for recurrence and methods of recurrence detection. We also assessed survival. Yearly hazards of recurrence were visualized.

RESULTS

With a median follow-up of 4.9 years, 158 per 580 patients (27.2%) recurred. Overall, most recurrences were patient-detected (60.7%) or imaging-detected (27.3%). Routine imaging was important in detecting recurrence in patients with distant recurrences (adjusted rate 43.1% vs. 9.4% for local/in-transit; p = .04) and with Stage IIC melanoma (42.5% vs. 18.5% for IIA; p = .01). Male patients also self-detected recurrent disease less than females (52.1% vs. 76.8%; p < .01).

CONCLUSIONS

Routine imaging surveillance played a larger role in detecting recurrent disease for select groups in this cohort than noted in prior studies. In an era of effective systemic therapy, routine imaging should be considered for detection of asymptomatic relapse for select, high-risk patient groups.

Evidence synthesis for constructing directed acyclic graphs (ESC-DAGs): a novel and systematic method for building directed acyclic graphs

Background

Directed acyclic graphs (DAGs) are popular tools for identifying appropriate adjustment strategies for epidemiological analysis. However, a lack of direction on how to build them is problematic. As a solution, we propose using a combination of evidence synthesis strategies and causal inference principles to integrate the DAG-building exercise within the review stages of research projects. We demonstrate this idea by introducing a novel protocol: ‘Evidence Synthesis for Constructing Directed Acyclic Graphs’ (ESC-DAGs)’.

Methods

ESC-DAGs operates on empirical studies identified by a literature search, ideally a novel systematic review or review of systematic reviews. It involves three key stages: (i) the conclusions of each study are ‘mapped’ into a DAG; (ii) the causal structures in these DAGs are systematically assessed using several causal inference principles and are corrected accordingly; (iii) the resulting DAGs are then synthesised into one or more ‘integrated DAGs’. This demonstration article didactically applies ESC-DAGs to the literature on parental influences on offspring alcohol use during adolescence.

Conclusions

ESC-DAGs is a practical, systematic and transparent approach for developing DAGs from background knowledge. These DAGs can then direct primary data analysis and DAG-based sensitivity analysis. ESC-DAGs has a modular design to allow researchers who are experienced DAG users to both use and improve upon the approach. It is also accessible to researchers with limited experience of DAGs or evidence synthesis.

Comparison of Mortality and Major Cardiovascular Events Among Adults With Type 2 Diabetes Using Human vs Analogue Insulins

Importance

The comparative cardiovascular safety of analogue and human insulins in adults with type 2 diabetes who initiate insulin therapy in usual care settings has not been carefully evaluated using machine learning and other rigorous analytic methods.

Objective

To examine the association of analogue vs human insulin use with mortality and major cardiovascular events.

Design, Setting, and Participants

This retrospective cohort study included 127 600 adults aged 21 to 89 years with type 2 diabetes at 4 health care delivery systems who initiated insulin therapy from January 1, 2000, through December 31, 2013. Machine learning and rigorous inference methods with time-varying exposures were used to evaluate associations of continuous exposure to analogue vs human insulins with mortality and major cardiovascular events. Data were analyzed from September 1, 2017, through June 30, 2018.

Exposures

On the index date (first insulin dispensing), participants were classified as using analogue insulin with or without human insulin or human insulin only.

Main Outcomes and Measures

Overall mortality, mortality due to cardiovascular disease (CVD), myocardial infarction (MI), stroke or cerebrovascular accident (CVA), and hospitalization for congestive heart failure (CHF) were evaluated. Marginal structural modeling (MSM) with inverse probability weighting was used to compare event-free survival in separate per-protocol analyses. Adjusted and unadjusted hazard ratios and cumulative risk differences were based on logistic MSM parameterizations for counterfactual hazards. Propensity scores were estimated using a data-adaptive approach (machine learning) based on 3 nested covariate adjustment sets. Sensitivity analyses were conducted to address potential residual confounding from unmeasured differences in risk factors across delivery systems.

Results

The 127 600 participants (mean [SD] age, 59.4 [12.6] years; 68 588 men [53.8%]; mean [SD] body mass index, 32.3 [7.1]) had a median follow-up of 4 quarters (interquartile range, 3-9 quarters) and experienced 5464 deaths overall (4.3%), 1729 MIs (1.4%), 1301 CVAs (1.0%), and 3082 CHF hospitalizations (2.4%). There were no differences in adjusted hazard ratios for continuous analogue vs human insulin exposure during 10 quarters for overall mortality (1.15; 95% CI, 0.97-1.34), CVD mortality (1.26; 95% CI, 0.86-1.66), MI (1.11; 95% CI, 0.77-1.45), CVA (1.30; 95% CI, 0.81-1.78), or CHF hospitalization (0.93; 95% CI, 0.75-1.11).

Conclusions and Relevance

Insulin-naive adults with type 2 diabetes who initiate and continue treatment with human vs analogue insulins had similar observed rates of major cardiovascular events, CVD mortality, and overall mortality.

Associations Between Difficulties in Emotion Regulation and Post-Traumatic Stress Disorder in Deployed Service Members of the German Armed Forces

BACKGROUND

Experiencing a traumatic event can lead to post-traumatic stress disorder (PTSD), but not every traumatized person develops PTSD. Several protective and risk factors have been identified in civilians and veterans to explain why some individuals develop PTSD and others do not. However, no research has confirmed the relationship between emotion regulation and PTSD in deployed German Armed Forces service members after a foreign assignment. Previous studies have identified some protective factors, such as social support, social acknowledgment, specific personal values, and posttraumatic growth, as well as risk factors, like moral injury and emotion regulation. Thus, the aim of the present study is to confirm the relationship between emotion regulation and PTSD and to test for factors that are associated with higher severity of PTSD symptoms in such a sample.

METHODS

A post-hoc secondary analysis was conducted on data collected in a randomized controlled trial. Participants (N = 72) were male active and former military service members that have returned from deployment and were recruited from the German Armed Forces. These participants were separated into two groups according to PTSD diagnosis based on the results of a structured diagnostic interview. Data from evaluation questionnaires administered upon entry into the study were subjected to a cross-sectional analysis. The measures included the severity of PTSD symptoms, clusters of PTSD symptoms, clinical measures, and several measures assessing PTSD-related constructs. Analyses included the Spearman rank correlation coefficient, X² tests for nominal data, Mann-Whitney U-tests for non-parametric data, and a mediation analysis.

RESULTS

The results of the mediation analysis revealed that difficulties in emotion regulation were significantly associated with the severity of PTSD symptoms, which was mediated by social acknowledgment and experimental avoidance but not by moral injury. The analyses showed that the severity of PTSD symptoms and all clusters of PTSD symptoms were significantly associated with most of the measured constructs in expectable directions. Participants in the PTSD group showed significantly higher mean scores on questionnaires measuring constructs that have been associated with PTSD, like emotion regulation and moral injury. They also showed lower mean scores in questionnaires for social support and social acknowledgment as a victim or survivor than participants in the non-PTSD group.

CONCLUSION

The present results show that difficulties in emotion regulation are directly associated with the severity of PTSD symptoms in service members of the German Armed Forces. This association is mediated by social acknowledgment and experimental avoidance, but not by moral injury. Thus, future studies should investigate these potentially crucial factors for better understanding of the development and maintenance of PTSD in service members of the German Armed Forces after deployment to create possible treatment adaptions.

CLINICAL TRIAL REGISTRATION

Australian Clinical Trials Registry, identifier ACTRN 12616000956404 http://www.anzctr.org.au/Trial/Registration/TrialReview.aspx?id=370924.

Methods for the Selection of Covariates in Nutritional Epidemiology Studies: A Meta-Epidemiological Review

BACKGROUND

Observational studies provide important information about the effects of exposures that cannot be easily studied in clinical trials, such as nutritional exposures, but are subject to confounding. Investigators adjust for confounders by entering them as covariates in analytic models.

OBJECTIVE

The aim of this study was to evaluate the reporting and credibility of methods for selection of covariates in nutritional epidemiology studies.

METHODS

We sampled 150 nutritional epidemiology studies published in 2007/2008 and 2017/2018 from the top 5 high-impact nutrition and medical journals and extracted information on methods for selection of covariates.

RESULTS

Most studies did not report selecting covariates a priori (94.0%) or criteria for selection of covariates (63.3%). There was general inconsistency in choice of covariates, even among studies investigating similar questions. One-third of studies did not acknowledge potential for residual confounding in their discussion.

CONCLUSION

Studies often do not report methods for selection of covariates, follow available guidance for selection of covariates, nor discuss potential for residual confounding.

Effectiveness of two-drug therapy versus monotherapy as initial regimen in hypertension: A propensity score-matched cohort study in the UK Clinical Practice Research Datalink

Purpose

To compare the effectiveness on blood pressure (BP) of initial two‐drug therapy versus monotherapy in hypertensive patients.

Methods

Using the Clinical Practice Research Datalink, linked with Hospital Episode Statistics and Office for National Statistics, we identified a cohort of adults with uncontrolled hypertension, initiating one or two antihypertensive drug classes between 2006 and 2014. New users of two drugs and monotherapy were matched 1:2 by propensity score. Main exposure was “as‐treated,” ie, until first regimen change. Primary and secondary endpoints were systolic and diastolic BP control and major adverse cardiovascular event (MACE), respectively. Hazard ratios (HR) and 95% confidence intervals (CI) were estimated using Cox proportional hazard models.

Results

Of 54 523 eligible patients, 3256 (6.0%) were initiated to a two‐drug combination. Of these, 2807 were matched to 5614 monotherapy users. Mean exposure duration was 12.7 months, with 76.5% patients changing their initial regimen. Two‐drug therapy was associated with a clinically significant BP control increase in all hypertensive patients (HR = 1.17 [95%CI: 1.09‐1.26]), more so in patients with grade 2‐3 hypertension (HR = 1.28 [1.17‐1.41]). An increase of 27% in BP control (HR = 1.27 [1.08‐1.49]) was observed in patients initiating an ACEi+CCB combination compared with initiators of either single class. No significant association was found between two‐drug therapy and MACE. Several sensitivity analyses confirmed the main findings.

Conclusions

Few patients initiated therapy with two drugs, reflecting UK guidelines' recommendation to start with monotherapy. This study supports the greater effectiveness of two‐drug therapy as the initial regimen for BP control.

Drug price, dosage and safety: Real-world evidence of oral hypoglycemic agents

OBJECTIVES

Drug price reduction is one of the major policies to restrain pharmaceutical expenses worldwide. This study explores whether there is a relationship between drug price and clinical quality using real-world data.

METHODS

Patients with newly-diagnosed type 2 diabetes receiving metformin or sulfonylureas during 2001 and 2010 were identified using the claim database of the Taiwan universal health insurance system. Propensity score matching was performed to obtain comparable subjects for analysis. Pharmaceutical products were categorized as brand-name agents (BD), highpriced generics (HP) or low-priced generics (LP). Indicators of clinical quality were defined as the dosage of cumulative oral hypoglycemic agents (OHA), exposure to other pharmacological classes of OHA, hospitalization or urgent visit for hypoglycemia or hyperglycemia, insulin utilization and diagnosis of diabetic complications within 1 year after diagnosis.

RESULTS

A total of 40,152 study subjects were identified. A generalized linear mix model showed that HP and BD users received similar OHA dosages with comparable clinical outcomes. By contrast, LP users had similar outcomes to BD users but received a 39% greater OHA dosage. A marginally higher risk of poor glycemic control in LP users was also observed.

CONCLUSIONS

Drug price is related to indicators of clinical quality. Clinicians and health authorities should monitor the utilization, effectiveness and clinical safety indicators of generic drugs, especially those with remarkably low prices.

Lymph Node Ratio in Pancreatic Adenocarcinoma After Preoperative Chemotherapy vs. Preoperative Chemoradiation and Its Utility in Decisions About Postoperative Chemotherapy

BACKGROUND

Single-center studies in pancreatic adenocarcinoma have suggested that preoperative chemotherapy (PCT) is associated with higher lymph node ratio (LNR) than preoperative chemoradiation (PCRT). The association of postoperative chemotherapy with overall survival (OS) in patients treated with PCT and PCRT remains unclear. Our objectives were to investigate whether (1) PCT is associated with higher LNR than PCRT and (2) postoperative chemotherapy is associated with longer OS after PCT and PCRT in LNR-stratified cohorts.

METHODS

A retrospective cohort study was performed of patients with pancreatic adenocarcinoma treated with PCT or PCRT followed by resection between 2006 and 2014 in the National Cancer Database. Temporal trends were evaluated with Cuzick's test. OS was evaluated with multivariable Cox regression and inverse probability weighted (IPW) Cox regression.

RESULTS

Of 4187 patients, 1993 (47.6%) received PCT. PCT rates were stable at approximately 30% in 2006-2010 (p = 0.33) but increased to 64.9% by 2014 (p < 0.001). Node positivity rates were higher after PCT than PCRT (62.7 vs. 41.8%, P < 0.001) and mean LNR was higher (0.10 [95% CI 0.096, 0.11] vs. 0.058 [95% CI 0.052, 0.063], P < 0.001). Postoperative chemotherapy was associated with longer OS in patients with LNR 0.01-0.149 after PCT by univariate analysis (median OS 34.5 vs. 26.5 months, P = 0.002), multivariable Cox regression (HR 0.64, 95% CI 0.48, 0.84), and IPW Cox regression (HR 0.72, 95% CI 0.55, 0.94). Postoperative chemotherapy was not associated with longer OS for patients who were node-negative or who had LNR ≥ 0.15 after PCT or for any patient subgroups after PCRT.

CONCLUSIONS

PCT is associated with a higher LNR and higher rates of node positivity than PCRT. Postoperative chemotherapy is associated with longer OS than observation in patients with a LNR of 0.01-0.149 after PCT.

Facility Variation in Local Staging of Rectal Adenocarcinoma and its Contribution to Underutilization of Neoadjuvant Therapy

BACKGROUND

Guidelines recommend neoadjuvant therapy (NT) for clinical stage II-III (locally advanced) rectal adenocarcinoma, but utilization remains suboptimal. The causes of NT omission remain poorly understood.

METHODS

The main outcomes in this study of patients with resected clinically non-metastatic rectal adenocarcinoma in the 2010-2015 National Cancer Database were local staging utilization in patients with non-metastatic tumors (i.e., undocumented clinical stage/pathologic stage I-III) and NT utilization for locally advanced tumors. Multivariable regression was used to examine predictors of these outcomes. Facility-specific risk- and reliability-adjusted local staging and NT rates were calculated. Positive margins and overall survival (OS) were examined as secondary outcomes.

RESULTS

Local staging was omitted in 7737/43,819 (17.7%) patients with clinically non-metastatic tumors and NT was omitted in 5199/31,632 (16.4%) patients with locally advanced tumors. NT was utilized in 24,826 (91.1%) locally advanced patients who had local staging vs. 1607 (36.6%) patients who did not; 2785 (53.6%) locally advanced patients with NT omitted also had local staging omitted. Treatment at facilities with lowest quintile local staging rates was associated with NT omission (relative risk 2.41, 95% confidence interval 2.11, 2.75). Adjusted facility local staging rates varied sixfold (16.1-98.0%), facility NT rates varied twofold (43.9-95.9%), and they were correlated (r = 0.58; P < 0.001). Local staging omission and NT omission were independently associated with positive margins and decreased OS.

CONCLUSIONS

Local staging omission is a common care process in over half of cases of omitted NT. These data emphasize the need for quality improvement efforts directed at providing facilities feedback about their local staging rates.

Urologic care and progression to end-stage kidney disease: a Chronic Kidney Disease in Children (CKiD) nested case-control study

INTRODUCTION

Children with chronic kidney disease (CKD) risk progressing to end-stage kidney disease (ESKD). The majority of CKD causes in children are related to congenital anomalies of the kidney and urinary tract, which may be treated by urologic care.

OBJECTIVE

To examine the association of ESKD with urologic care in children with CKD.

STUDY DESIGN

This was a nested case-control study within the Chronic Kidney Disease in Children (CKiD) prospective cohort study that included children aged 1-16 years with non-glomerular causes of CKD. The primary exposure was prior urologic referral with or without surgical intervention. Incidence density sampling matched each case of ESKD to up to three controls on duration of time from CKD onset, sex, race, age at baseline visit, and history of low birth weight. Conditional logistic regression analysis was performed to estimate rate ratios (RRs) for the incidence of ESKD.

RESULTS

Sixty-six cases of ESKD were matched to 153 controls. Median age at baseline study visit was 12 years; 67% were male, and 7% were black. Median follow-up time from CKD onset was 14.9 years. Seventy percent received urologic care, including 100% of obstructive uropathy and 96% of reflux nephropathy diagnoses. Cases had worse renal function at their baseline visit and were less likely to have received prior urologic care. After adjusting for income, education, and insurance status, urology referral with surgery was associated with 50% lower risk of ESKD (RR 0.50 [95% confidence interval [CI] 0.26-0.997), compared to no prior urologic care (Figure). After excluding obstructive uropathy and reflux nephropathy diagnoses, which were highly correlated with urologic surgery, the association was attenuated (RR 0.72, 95% CI 0.24-2.18).

DISCUSSION

In this study, urologic care was commonly but not uniformly provided to children with non-glomerular causes of CKD. Underlying specific diagnoses play an important role in both the risk of ESKD and potential benefits of urologic surgery.

CONCLUSION

Within the CKiD cohort, children with non-glomerular causes of CKD often received urologic care. Urology referral with surgery was associated with lower risk of ESKD compared to no prior urologic care but depended on specific underlying diagnoses.

County-level Variation in Use of Surgery and Cancer-specific Survival for Stage I-II Pancreatic Adenocarcinoma

OBJECTIVE

The aim of the study was to describe county-level variation in use of surgery for stage I-II pancreatic ductal adenocarcinoma (PDAC) and the association between county surgery rates and cancer-specific survival (CSS).

BACKGROUND

The degree of small geographic area variation in use of surgery for stage I-II PDAC and the association between area surgery rates and CSS remain incompletely defined.

METHODS

This is a retrospective cohort study of patients aged 18 to 80 years in the 2007 to 2015 Surveillance, Epidemiology, and End Results database with stage I-II PDAC without contraindications to surgery or refusal. Multilevel models were used to characterize county-level variation in use of surgery and CSS. County-specific risk- and reliability-adjusted surgery rates and CSS rates were calculated.

RESULTS

Of 18,100 patients living in 581 counties, 10,944 (60.5%) underwent surgery. Adjusted county-specific surgery rates varied 1.5-fold from 49.9% to 74.6%. Median CSS increased in a graded fashion from 13 months [interquartile range (IQR) 13-14] in counties with surgery rates of 49.9% to 56.9% to 18 months (IQR 17-19) in counties with surgery rates of 68.0% to 74.6%. Results were similar in multivariable analyses. Adjusted county 18-month CSS rates varied 1.6-fold from 32.7% to 53.7%. Adjusted county surgery and 18-month CSS rates were correlated (r = 0.54; P < 0.001) and county surgery rates explained approximately half of county-level variation in CSS. Only 18 (3.1%) counties had adjusted surgery rates of 68.0% to 74.6%, which was associated with the longest CSS.

CONCLUSIONS

County-specific rates of surgery varied substantially, and patients living in areas with higher surgery rates lived longer. These data suggest that increasing use of surgery in stage I-II PDAC could lead to improvements in survival.

Persistent organic pollutants and gestational diabetes: A multi-center prospective cohort study of healthy US women

BACKGROUND

Persistent organic pollutants (POPs) are linked with insulin resistance and type-2 diabetes (T2D) in the general population. However, their associations with gestational diabetes (GDM) are inconsistent.

OBJECTIVE

We prospectively evaluated the associations of POPs measured in early pregnancy with GDM risk. We also assessed whether pre-pregnancy BMI (ppBMI) and family history of T2D modify this risk.

METHODS

In NICHD Fetal Growth Study, Singletons, we measured plasma concentration of 76 POPs, including 11 organochlorine pesticides (OCPs), 9 polybrominated diphenylethers (PBDEs), 44 polychlorinated biphenyls (PCBs), and 11 per-and polyfluoroalkyl substances (PFAS) among 2334 healthy non-obese women at 8-13 weeks of gestation. GDM was diagnosed by Carpenter and Coustan criteria. We constructed chemical networks using a weighted-correlation algorithm and examined the associations of individual chemical and chemical networks with GDM using multivariate Poisson regression with robust variance.

RESULTS

Higher concentrations of PCBs with six or more chlorine atoms were associated with increased risk of GDM in the overall cohort (risk ratios [RRs] range: 1.08-1.13 per 1-standard deviation [SD] increment) and among women with a family history of T2D (RRs range: 1.08-1.48 per 1-SD increment) or normal ppBMI (RRs range: 1.08-1.22 per 1-SD increment). Similar associations were observed for the chemical network comprised of PCBs with ≥6 chlorine atoms and the summary measure of total PCBs and non-dioxin like PCBs (138, 153, 170, 180). Furthermore, four PFAS congeners - perfluorononanoic acid (PFNA), perfluorooctanoic acid (PFOA), perfluoroheptanoic acid (PFHpA), and perfluorododecanoic acid (PFDoDA) - showed significant positive associations with GDM among women with a family history of T2D (RRs range:1.22-3.18 per 1-SD increment), whereas BDE47 and BDE153 showed significant positive associations among women without a family history of T2D.

CONCLUSIONS

Environmentally relevant levels of heavily chlorinated PCBs and some PFAS and PBDEs were positively associated with GDM with suggestive effect modifications by family history of T2D and body adiposity status.

Disparities in utilization of treatment for clinical stage I-II pancreatic adenocarcinoma by area socioeconomic status and race/ethnicity

BACKGROUND

Utilization of multimodality therapy for clinical stage I-II pancreatic ductal adenocarcinoma is associated with meaningful prolongation of survival. Although the qualitative existence of disparities in treatment utilization by socioeconomic status and race/ethnicity is well documented, the absolute magnitudes of these disparities have not been previously quantified.

METHODS

The exposures in this retrospective cohort study of the 2010-2015 National Cancer Database were a 7-value area-level socioeconomic status index and race/ethnicity. Main outcomes were surgery, chemotherapy, and multimodality therapy (surgery and chemotherapy). Adjusted rate differences were calculated after logistic regression. Models excluded intermediate variables. Overall survival was evaluated in unadjusted and adjusted analyses.

RESULTS

Of 43,760 patients, 63.4% underwent surgery. Of 39,808 patients without chemotherapy contraindications, refusal, or missing data, 75.1% received chemotherapy and 51.4% received multimodality therapy. Adjusted rate differences for utilization of surgery, chemotherapy, and multimodality therapy in the lowest socioeconomic status patients were -10.0 (95% confidence interval [CI] -12.4 to -7.5), -12.7 (95% CI -16.3 to -9.1), and -15.4 (95% CI -18.8 to -12.0), respectively, versus the highest socioeconomic status patients. Adjusted rate differences for multimodality therapy utilization in non-Hispanic Black and Hispanic patients were -10.1 (95% CI -13.6 to -6.7) and -11.8 (95% CI -14.3 to -9.2), respectively, versus non-Hispanic White patients. Median overall survival increased in a graded fashion from 14.1 (95% CI 13.4-14.8) months in the lowest socioeconomic status patients to 20.2 months (95% CI 19.6-20.8) in the highest socioeconomic status patients. Survival differences were attenuated but not eliminated in multivariable Cox models.

CONCLUSION

Socioeconomic status and race/ethnicity are more powerful determinants of whether patients receive treatment for clinical stage I-II pancreatic ductal adenocarcinoma than previously appreciated. Nationwide quality improvement efforts aimed at addressing these inequities are warranted.

Prospective health associations of drive for muscularity in young adult males

OBJECTIVE

The objective of the study was to estimate prospective associations of drive for muscularity measured in 2013 and related health outcomes (depressive symptoms, overeating, binge eating, purging, binge drinking, and use of muscle-building products [e.g., creatine and steroids]) measured in 2014.

METHOD

The data come from a U.S. national large prospective cohort study, the Growing Up Today Study (GUTS) and included 2,460 males aged 18-32 years. Muscularity concerns were assessed with The Drive for Muscularity Scale (item responses ranging from 1 = Never to 6 = Always), which measures the degree of the respondents' preoccupation with increasing their muscularity.

RESULTS

Gay and bisexual males presented with higher drive for muscularity compared to heterosexual males (β = 0.30, 95% CI = 0.17, 0.43, p < .0001). Prospective analysis using generalized estimating equations indicated that each unit increase in drive for muscularity was associated with increased odds of exhibiting significant depressive symptoms (OR = 1.23, 95% CI = 1.05-1.44, p = .01), binge drinking (OR = 1.21, 95% CI = 1.02-1.45, p = .03), dieting (OR = 1.17, 95% CI = 1.01, 1.35, p = .04), and use of muscle-building products (OR = 4.49, 95% CI = 3.74, 5.40, p < .0001).

DISCUSSION

The drive for muscularity appears to be relatively prevalent across adolescence and early adulthood in males and is associated with broad and substantial health consequences. Muscularity-oriented concerns should be carefully considered when assessing and treating males with body image dissatisfaction.

Outcomes of laparoscopic and open surgery in children with and without congenital heart disease

BACKGROUND

Children with congenital heart disease (CHD) often require noncardiac surgery. We compared outcomes following open and laparoscopic intraabdominal surgery among children with and without CHD.

METHODS

We performed a retrospective cohort study using the 2013-2015 National Surgical Quality Improvement Project-Pediatrics. We matched 45,012 children <18years old who underwent laparoscopic surgery to 45,012 children who underwent open surgery. We determined the associations between laparoscopic (versus open) surgery and 30-day mortality, in-hospital mortality, 30-day morbidity, and postoperative length-of-stay.

RESULTS

Among children with minor CHD, laparoscopic surgery was associated with lower 30-day mortality (Odds Ratio [OR] 0.34 [95% Confidence Interval 0.15-0.79]), inhospital mortality (OR 0.42 [0.22-0.81]) and 30-day morbidity (OR 0.61 [0.50-0.73]). As CHD severity increased, this benefit of laparoscopic surgery decreased for 30-day morbidity (ptrend=0.01) and in-hospital mortality (ptrend=0.05), but not for 30-day mortality (ptrend=0.27). Length-of-stay was shorter for laparoscopic approaches for children at cost of higher readmissions. On subgroup analysis, laparoscopy was associated with lower odds of postoperative blood transfusion in all children.

CONCLUSIONS

Intraabdominal laparoscopic surgery compared to open surgery is associated with decreased morbidity in patients with no CHD and lower morbidity and mortality in patients with minor CHD, but not in those with more severe CHD.

LEVEL-OF-EVIDENCE

Level III: Treatment Study.

Outcomes When Using Adjunct Dexmedetomidine with Propofol Sedation in Mechanically Ventilated Surgical Intensive Care Patients

Objective: Compare the duration of mechanical ventilation between patients receiving sedation with continuous infusions of propofol alone or combination with the use of dexmedetomidine and propofol. Design: Retrospective, propensity matched (1:1) cohort study, employing eight variables chosen a priori for matching. Timing of exposure to dexmedetomidine initiation was incorporated into a matching algorithm. Setting: Level 1, university-based, 32-bed, adult, mixed trauma and surgical intensive care unit (SICU). Continuous sedation was delivered according to a protocol methodology with daily sedation vacation and spontaneous breathing trials. Choice of sedation agent was physician directed. Patients: Between 2010 and 2014, 149 SICU patients receiving mechanical ventilation for >24 h received dexmedetomidine with propofol. Propensity matching resulted in 143 pair cohorts. Interventions: Dexmedetomidine with propofol or propofol alone. Measurements and Main Results: There was no statistical difference in SICU length of stay (LOS), with a median absolute difference of 5.3 h for propofol alone group (p = 0.43). The SICU mortality was not statistically different (RR = 1.002, p = 0.88). Examining a 14-day period post-treatment with dexmedetomidine, on any given day (excluding days 1 and 14), dexmedetomidine with propofol-treated patients had a 0.5% to 22.5% greater likelihood of being delirious (CAM-ICU positive). In addition, dexmedetomidine with propofol-treated patients had a 4.5% to 18.8% higher likelihood of being above the target sedation score (more agitated) compared to propofol-alone patients. Conclusions: In this propensity matched cohort study, adjunct use of dexmedetomidine to propofol did not show a statistically significant reduction with respect to mechanical ventilation (MV) duration, SICU LOS, or SICU mortality, despite a trend toward receiving fewer hours of propofol. There was no evidence that dexmedetomidine with propofol improved sedation scores or reduced delirium.

Outcomes When Using Adjunct Dexmedetomidine with Propofol Sedation in Mechanically Ventilated Surgical Intensive Care Patients

Objective: Compare the duration of mechanical ventilation between patients receiving sedation with continuous infusions of propofol alone or combination with the use of dexmedetomidine and propofol. Design: Retrospective, propensity matched (1:1) cohort study, employing eight variables chosen a priori for matching. Timing of exposure to dexmedetomidine initiation was incorporated into a matching algorithm. Setting: Level 1, university-based, 32-bed, adult, mixed trauma and surgical intensive care unit (SICU). Continuous sedation was delivered according to a protocol methodology with daily sedation vacation and spontaneous breathing trials. Choice of sedation agent was physician directed. Patients: Between 2010 and 2014, 149 SICU patients receiving mechanical ventilation for >24 h received dexmedetomidine with propofol. Propensity matching resulted in 143 pair cohorts. Interventions: Dexmedetomidine with propofol or propofol alone. Measurements and Main Results: There was no statistical difference in SICU length of stay (LOS), with a median absolute difference of 5.3 h for propofol alone group (p = 0.43). The SICU mortality was not statistically different (RR = 1.002, p = 0.88). Examining a 14-day period post-treatment with dexmedetomidine, on any given day (excluding days 1 and 14), dexmedetomidine with propofol-treated patients had a 0.5% to 22.5% greater likelihood of being delirious (CAM-ICU positive). In addition, dexmedetomidine with propofol-treated patients had a 4.5% to 18.8% higher likelihood of being above the target sedation score (more agitated) compared to propofol-alone patients. Conclusions: In this propensity matched cohort study, adjunct use of dexmedetomidine to propofol did not show a statistically significant reduction with respect to mechanical ventilation (MV) duration, SICU LOS, or SICU mortality, despite a trend toward receiving fewer hours of propofol. There was no evidence that dexmedetomidine with propofol improved sedation scores or reduced delirium.

Studying additive interaction in a healthcare database: Case study of NSAIDs, cardiovascular profiles, and acute myocardial infarction

Purpose

There are clinical trial data on risk of acute myocardial infarction (MI) with nonsteroidal anti-inflammatory drugs (NSAIDs) in patients at increased cardiovascular (CV) risk requiring chronic daily treatment. This study investigated whether risks of acute MI with real-world prescription NSAIDs, such as low-dose or intermittent use, vary according to an individual’s CV profile.

Methods

Nested case-control analyses were carried out on an administrative health cohort from Quebec, Canada by randomly selecting 10 controls per case matched on age ± 1 year, sex, and month and year of cohort entry. We measured the additive joint effects on acute MI of current NSAID use and presence of hypertension, coronary heart disease (CHD), history of previous MI, or concomitant use of cardioprotective aspirin. The endpoint was the relative excess risk due to interaction (RERI). To verify the robustness of interaction findings, we performed sensitivity analyses with varying specifications of NSAID exposure-related variables.

Results

The cohort consisted of 233 816 elderly individuals, including 21 256 acute MI cases. For hypertension, CHD, and previous MI, we identified additive interactions on MI risk with some but not all NSAIDs, which also depended on the definition of NSAID exposure. Hypertension was sub-additive with naproxen but not with the other NSAIDs. Celecoxib and CHD were sub-additive in the primary analysis only (modelling NSAID dose on index date or up to 7 days before–best-fitting base model) whereas celecoxib and rofecoxib were super-additive with a history of previous MI in the secondary analysis only (modelling NSAID use on index date). For cardioprotective aspirin we found no evidence for an additive interaction with any of the NSAIDs.

Conclusions

Alternative specifications of NSAID exposure concurred in finding that concomitant use of cardioprotective aspirin does not attenuate the risks of acute MI with NSAIDs. However we were unable to demonstrate consistent interactions between an individual’s cardiovascular comorbidities and NSAID-associated acute MI. Our study highlights challenges of studying additive interactions in a healthcare database and underscores the need for sensitivity analyses.

Diagnostics for Confounding of Time-varying and Other Joint Exposures

The effects of joint exposures (or exposure regimes) include those of adhering to assigned treatment versus placebo in a randomized controlled trial, duration of exposure in a cohort study, interactions between exposures, and direct effects of exposure, among others. Unlike the setting of a single point exposure (e.g., propensity score matching), there are few tools to describe confounding for joint exposures or how well a method resolves it. Investigators need tools that describe confounding in ways that are conceptually grounded and intuitive for those who read, review, and use applied research to guide policy. We revisit the implications of exchangeability conditions that hold in sequentially randomized trials, and the bias structure that motivates the use of g-methods, such as marginal structural models. From these, we develop covariate balance diagnostics for joint exposures that can (1) describe time-varying confounding, (2) assess whether covariates are predicted by prior exposures given their past, the indication for g-methods, and (3) describe residual confounding after inverse probability weighting. For each diagnostic, we present time-specific metrics that encompass a wide class of joint exposures, including regimes of multivariate time-varying exposures in censored data, with multivariate point exposures as a special case. We outline how to estimate these directly or with regression and how to average them over person-time. Using a simulated example, we show how these metrics can be presented graphically. This conceptually grounded framework can potentially aid the transparent design, analysis, and reporting of studies that examine joint exposures. We provide easy-to-use tools to implement it.

Association of time-to-surgery with outcomes in clinical stage I-II pancreatic adenocarcinoma treated with upfront surgery

BACKGROUND

Time-to-surgery from cancer diagnosis has increased in the United States. We aimed to determine the association between time-to-surgery and oncologic outcomes in patients with resectable pancreatic ductal adenocarcinoma undergoing upfront surgery.

METHODS

The 2004-2012 National Cancer Database was reviewed for patients undergoing curative-intent surgery without neoadjuvant therapy for clinical stage I-II pancreatic ductal adenocarcinoma. A multivariable Cox model with restricted cubic splines was used to define time-to-surgery as short (1-14 days), medium (15-42), and long (43-120). Overall survival was examined using Cox shared frailty models. Secondary outcomes were examined using mixed-effects logistic regression models.

RESULTS

Of 16,763 patients, time-to-surgery was short in 34.4%, medium in 51.6%, and long in 14.0%. More short time-to-surgery patients were young, privately insured, healthy, and treated at low-volume hospitals. Adjusted hazards of mortality were lower for medium (hazard ratio 0.94, 95% confidence interval, .90, 0.97) and long time-to-surgery (hazard ratio 0.91, 95% confidence interval, 0.86, 0.96) than short. There were no differences in adjusted odds of node positivity, clinical to pathologic upstaging, being unresectable or stage IV at exploration, and positive margins. Medium time-to-surgery patients had higher adjusted odds (odds ratio 1.11, 95% confidence interval, 1.03, 1.20) of receiving an adequate lymphadenectomy than short. Ninety-day mortality was lower in medium (odds ratio 0.75, 95% confidence interval, 0.65, 0.85) and long time-to-surgery (odds ratio 0.72, 95% confidence interval, 0.60, 0.88) than short.

CONCLUSION

In this observational analysis, short time-to-surgery was associated with slightly shorter OS and higher perioperative mortality. These results may suggest that delays for medical optimization and referral to high volume surgeons are safe.

Propensity Scores in Pharmacoepidemiology: Beyond the Horizon

Purpose of review

Propensity score methods have become commonplace in pharmacoepidemiology over the past decade. Their adoption has confronted formidable obstacles that arise from pharmacoepidemiology's reliance on large healthcare databases of considerable heterogeneity and complexity. These include identifying clinically meaningful samples, defining treatment comparisons, and measuring covariates in ways that respect sound epidemiologic study design. Additional complexities involve correctly modeling treatment decisions in the face of variation in healthcare practice, and dealing with missing information and unmeasured confounding. In this review, we examine the application of propensity score methods in pharmacoepidemiology with particular attention to these and other issues, with an eye towards standards of practice, recent methodological advances, and opportunities for future progress.

Recent findings

Propensity score methods have matured in ways that can advance comparative effectiveness and safety research in pharmacoepidemiology. These include natural extensions for categorical treatments, matching algorithms that can optimize sample size given design constraints, weighting estimators that asymptotically target matched and overlap samples, and the incorporation of machine learning to aid in covariate selection and model building.

Summary

These recent and encouraging advances should be further evaluated through simulation and empirical studies, but nonetheless represent a bright path ahead for the observational study of treatment benefits and harms.

A threshold trajectory was revealed by isolating the effects of hemoglobin rate of rise in anemia of chronic kidney disease

BACKGROUND

To assess cardiovascular risk among various hemoglobin (Hb) rates of rise (RoR) in chronic kidney disease (CKD) patients with anemia who have initiated therapy with erythropoiesis stimulating agents (ESAs).

METHODS

Observational cohort of CKD patients initiating ESA therapy from the Centricity® database, 1990-2011. Proportional hazards models tested the hypothesis that a slower Hb RoR (0 < g/dl/month ⩽ 0.125) is associated with a lower cardiovascular (CV) incidence [composite of fatal/nonfatal myocardial infarction (MI) and stroke] compared with faster RoR (0.125 < g/dl/month ⩽ 2.0, and >2.0 g/dl/month).

RESULTS

A total of 9220 patients receiving ESAs were followed for an average of 3.1 years. Slow (group B) RoR versus medium (group C') and fast (group D') RoR in Hb, throughout all Hb milestones, was associated with lower risk of the composite endpoint [B (slow) versus D' (fast) [hazard ratio (HR) = 0.20 (0.11, 0.39), p < 0.0001]; B versus C' (medium) [HR = 0.34 (0.19, 0.62), p = 0.0004], and C' versus D' [HR = 0.60 (0.42, 0.85), p = 0.005]]. Within achieved Hb milestones, HRs were: B versus D' at milestone ⩾ 14.1 g/dl [HR = 0.17 (0.05, 0.56); p = 0.004] and at milestone 12.6-14.0 [HR = 0.18 (0.07, 0.46), p = 0.0004].

CONCLUSION

Rapid Hb rise is associated with adverse CV outcomes, with markedly lower risk for rates below a threshold trajectory of 0.125 g/dl/month, even with complete correction.

Risk of acute myocardial infarction with NSAIDs in real world use: bayesian meta-analysis of individual patient data

Objective To characterise the determinants, time course, and risks of acute myocardial infarction associated with use of oral non-steroidal anti-inflammatory drugs (NSAIDs).Design Systematic review followed by a one stage bayesian individual patient data meta-analysis.Data sources Studies from Canadian and European healthcare databases.Review methods Eligible studies were sourced from computerised drug prescription or medical databases, conducted in the general or an elderly population, documented acute myocardial infarction as specific outcome, studied selective cyclo-oxygenase-2 inhibitors (including rofecoxib) and traditional NSAIDs, compared risk of acute myocardial infarction in NSAID users with non-users, allowed for time dependent analyses, and minimised effects of confounding and misclassification bias. Exposure and outcomes Drug exposure was modelled as an indicator variable incorporating the specific NSAID, its recency, duration of use, and dose. The outcome measures were the summary adjusted odds ratios of first acute myocardial infarction after study entry for each category of NSAID use at index date (date of acute myocardial infarction for cases, matched date for controls) versus non-use in the preceding year and the posterior probability of acute myocardial infarction.Results A cohort of 446 763 individuals including 61 460 with acute myocardial infarction was acquired. Taking any dose of NSAIDs for one week, one month, or more than a month was associated with an increased risk of myocardial infarction. With use for one to seven days the probability of increased myocardial infarction risk (posterior probability of odds ratio >1.0) was 92% for celecoxib, 97% for ibuprofen, and 99% for diclofenac, naproxen, and rofecoxib. The corresponding odds ratios (95% credible intervals) were 1.24 (0.91 to 1.82) for celecoxib, 1.48 (1.00 to 2.26) for ibuprofen, 1.50 (1.06 to 2.04) for diclofenac, 1.53 (1.07 to 2.33) for naproxen, and 1.58 (1.07 to 2.17) for rofecoxib. Greater risk of myocardial infarction was documented for higher dose of NSAIDs. With use for longer than one month, risks did not appear to exceed those associated with shorter durations.Conclusions All NSAIDs, including naproxen, were found to be associated with an increased risk of acute myocardial infarction. Risk of myocardial infarction with celecoxib was comparable to that of traditional NSAIDS and was lower than for rofecoxib. Risk was greatest during the first month of NSAID use and with higher doses.