Confounding and effect modification: distribution and measure

Variable selection when estimating effects in external target populations

External validity is an important part of epidemiologic research. To validly estimate effects in specific external target populations using a chosen effect measure (ie, "transport"), some methods require that one account for all effect measure modifiers (EMMs). However, little is known about how including other variables that are not EMMs (ie, non-EMMs) in adjustment sets affects estimates. Using simulations, we evaluated how inclusion of non-EMMs affected estimation of the transported risk difference (RD) by assessing the impacts of covariates that (1) differ (or not) between the trial and the target, (2) are associated with the outcome (or not), and (3) modify the RD (or not). We assessed variation and bias when covariates with each possible combination of these factors were used to transport RDs using outcome modeling or inverse odds weighting. Inclusion of variables that differed in distribution between the populations but were non-EMMs reduced precision, regardless of whether they were associated with the outcome. However, non-EMMs associated with selection did not amplify bias resulting from omission of necessary EMMs. Including all variables associated with the outcome may result in unnecessarily imprecise estimates when estimating treatment effects in external target populations.

Human immune and metabolic biomarker levels, and stress-biomarker associations, differ by season: Implications for biomedical health research

Although seasonal changes in physiology are well documented, little is known about how human immune and metabolic markers vary across seasons, and no studies have examined how stress → health biomarker associations differ across the year. To investigate these issues, we analyzed data from 2118 participants of the Midlife in the United States (MIDUS) study to determine whether there were differences in (a) levels of 19 immune and metabolic markers, and (b) the association between perceived stress and each biomarker across the year. Results of component-wide boosted generalized additive models revealed seasonal patterning for most biomarkers, with immune proteins generally peaking when days were shorter. Moreover, whereas levels of hemoglobin A1C rose from late fall to spring, triglycerides were elevated in the summer and fall, and high-density lipoprotein decreased steadily from January to December. Urinary cortisol and cortisone exhibited opposite patterns, peaking at the beginning and end of the year, respectively. Most critically, we found that the effects of perceived stress on 18 of the 19 health biomarkers assessed varied by month of measurement. In some cases, these differences involved the magnitude of the stress → biomarker association but, in other cases, it was the direction of the effect that changed. Studies that do not account for month of biomarker assessment may thus yield misleading or unreproducible results.

Interaction analysis of subgroup effects in randomized trials: the essential methodological points

Subgroup analysis aims to identify subgroups (usually defined by baseline/demographic characteristics), who would (or not) benefit from an intervention under specific conditions. Often performed post hoc (not pre-specified in the protocol), subgroup analyses are prone to elevated type I error due to multiple testing, inadequate power, and inappropriate statistical interpretation. Aside from the well-known Bonferroni correction, subgroup treatment interaction tests can provide useful information to support the hypothesis. Using data from a previously published randomized trial where a p value of 0.015 was found for the comparison between standard and Hemopatch® groups in (the subgroup of) 135 patients who had hand-sewn pancreatic stump closure we first sought to determine whether there was interaction between the number and proportion of the dependent event of interest (POPF) among the subgroup population (patients with hand-sewn stump closure and use of Hemopatch®), Next, we calculated the relative excess risk due to interaction (RERI) and the "attributable proportion" (AP). The p value of the interaction was p = 0.034, the RERI was - 0.77 (p = 0.0204) (the probability of POPF was 0.77 because of the interaction), the RERI was 13% (patients are 13% less likely to sustain POPF because of the interaction), and the AP was - 0.616 (61.6% of patients who did not develop POPF did so because of the interaction). Although no causality can be implied, Hemopatch® may potentially decrease the POPF after distal pancreatectomy when the stump is closed hand-sewn. The hypothesis generated by our subgroup analysis requires confirmation by a specific, randomized trial, including only patients undergoing hand-sewn closure of the pancreatic stump after distal pancreatectomy.Trial registration: INS-621000-0760.

Sufficient component cause simulations: an underutilized epidemiologic teaching tool

Simulation studies are a powerful and important tool in epidemiologic teaching, especially for understanding causal inference. Simulations using the sufficient component cause framework can provide students key insights about causal mechanisms and sources of bias, but are not commonly used. To make them more accessible, we aim to provide an introduction and tutorial on developing and using these simulations, including an overview of translation from directed acyclic graphs and potential outcomes to sufficient component causal models, and a summary of the simulation approach. Using the applied question of the impact of educational attainment on dementia, we offer simple simulation examples and accompanying code to illustrate sufficient component cause-based simulations for four common causal structures (causation, confounding, selection bias, and effect modification) often introduced early in epidemiologic training. We show how sufficient component cause-based simulations illuminate both the causal processes and the mechanisms through which bias occurs, which can help enhance student understanding of these causal structures and the distinctions between them. We conclude with a discussion of considerations for using sufficient component cause-based simulations as a teaching tool.

Factors Contributing to the Link between Physical Well-Being and Chronic Pain in Young People from Galicia, Northwest Spain

Introduction: The relation between physical well-being and chronic pain is complex and involves several subjective and objective covariates. We aimed to assess the role of mediator, confounder, or interactor played by covariates, including sleep quality, physical activity, perceived stress, smoking, and alcohol drinking in the relation between physical well-being and chronic pain. Method: We used Poisson regression to obtain incidence rate ratios (IRR) of the association between physical well-being and chronic pain in a cohort study carried out among university students. We applied General Structural Equation Modeling (GSEM) to assess mediation and stratum-specific analyses to distinguish confounding from interaction. We computed Relative Excess Risks due to Interaction (RERI), Attributable Proportion (AP), and the Synergy index (S) to measure additive interaction. Results: High physical well-being is related to a large decrease in the risk of chronic pain (IRR^{Total Effect} = 0.58; 95% CI: 0.50-0.81). Perceived stress mediates 12.5% of the total effect of physical well-being on chronic pain. The stratum-specific IRRs of current smokers and non-current smokers were different from each other and were larger than the crude IRR (IRR = 1.49; 95% CI: 1.24-1.80), which indicates that smoking could be both confounder and interactor. Interaction analyses showed that physical activity could act as a potential interactor (RERI = 0.25; 95% CI: 0.13, 0.60). Conclusions: Perceived stress is an important mediator of the relation between physical well-being and chronic pain, while smoking is both a confounder and an interactor. Our findings may prove useful in distinguishing high-risk groups from low-risk groups, in the interventions aimed at reducing chronic pain.

Improving trial generalizability using observational studies

Complementary features of randomized controlled trials (RCTs) and observational studies (OSs) can be used jointly to estimate the average treatment effect of a target population. We propose a calibration weighting estimator that enforces the covariate balance between the RCT and OS, therefore improving the trial-based estimator's generalizability. Exploiting semiparametric efficiency theory, we propose a doubly robust augmented calibration weighting estimator that achieves the efficiency bound derived under the identification assumptions. A nonparametric sieve method is provided as an alternative to the parametric approach, which enables the robust approximation of the nuisance functions and data-adaptive selection of outcome predictors for calibration. We establish asymptotic results and confirm the finite sample performances of the proposed estimators by simulation experiments and an application on the estimation of the treatment effect of adjuvant chemotherapy for early-stage non-small-cell lung patients after surgery.

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.

Neutralizing Monoclonal Antibody Use and COVID-19 Infection Outcomes

Importance

Evidence on the effectiveness and safety of COVID-19 therapies across a diverse population with varied risk factors is needed to inform clinical practice.

Objective

To assess the safety of neutralizing monoclonal antibodies (nMAbs) for the treatment of COVID-19 and their association with adverse outcomes.

Design, Setting, and Participants

This retrospective cohort study included 167 183 patients from a consortium of 4 health care systems based in California, Minnesota, Texas, and Utah. The study included nonhospitalized patients 12 years and older with a positive COVID-19 laboratory test collected between November 9, 2020, and January 31, 2022, who met at least 1 emergency use authorization criterion for risk of a poor outcome.

Exposure

Four nMAb products (bamlanivimab, bamlanivimab-etesevimab, casirivimab-imdevimab, and sotrovimab) administered in the outpatient setting.

Main Outcomes and Measures

Clinical and SARS-CoV-2 genomic sequence data and propensity-adjusted marginal structural models were used to assess the association between treatment with nMAbs and 4 outcomes: all-cause emergency department (ED) visits, hospitalization, death, and a composite of hospitalization or death within 14 days and 30 days of the index date (defined as the date of the first positive COVID-19 test or the date of referral). Patient index dates were categorized into 4 variant epochs: pre-Delta (November 9, 2020, to June 30, 2021), Delta (July 1 to November 30, 2021), Delta and Omicron BA.1 (December 1 to 31, 2021), and Omicron BA.1 (January 1 to 31, 2022).

Results

Among 167 183 patients, the mean (SD) age was 47.0 (18.5) years; 95 669 patients (57.2%) were female at birth, 139 379 (83.4%) were White, and 138 900 (83.1%) were non-Hispanic. A total of 25 241 patients received treatment with nMAbs. Treatment with nMAbs was associated with lower odds of ED visits within 14 days (odds ratio [OR], 0.76; 95% CI, 0.68-0.85), hospitalization within 14 days (OR, 0.52; 95% CI, 0.45-0.59), and death within 30 days (OR, 0.14; 95% CI, 0.10-0.20). The association between nMAbs and reduced risk of hospitalization was stronger in unvaccinated patients (14-day hospitalization: OR, 0.51; 95% CI, 0.44-0.59), and the associations with hospitalization and death were stronger in immunocompromised patients (hospitalization within 14 days: OR, 0.31 [95% CI, 0.24-0.41]; death within 30 days: OR, 0.13 [95% CI, 0.06-0.27]). The strength of associations of nMAbs increased incrementally among patients with a greater probability of poor outcomes; for example, the ORs for hospitalization within 14 days were 0.58 (95% CI, 0.48-0.72) among those in the third (moderate) risk stratum and 0.41 (95% CI, 0.32-0.53) among those in the fifth (highest) risk stratum. The association of nMAb treatment with reduced risk of hospitalizations within 14 days was strongest during the Delta variant epoch (OR, 0.37; 95% CI, 0.31-0.43) but not during the Omicron BA.1 epoch (OR, 1.29; 95% CI, 0.68-2.47). These findings were corroborated in the subset of patients with viral genomic data. Treatment with nMAbs was associated with a significant mortality benefit in all variant epochs (pre-Delta: OR, 0.16 [95% CI, 0.08-0.33]; Delta: OR, 0.14 [95% CI, 0.09-0.22]; Delta and Omicron BA.1: OR, 0.10 [95% CI, 0.03-0.35]; and Omicron BA.1: OR, 0.13 [95% CI, 0.02-0.93]). Potential adverse drug events were identified in 38 treated patients (0.2%).

Conclusions and Relevance

In this study, nMAb treatment for COVID-19 was safe and associated with reductions in ED visits, hospitalization, and death, although it was not associated with reduced risk of hospitalization during the Omicron BA.1 epoch. These findings suggest that targeted risk stratification strategies may help optimize future nMAb treatment decisions.

Toward a Clearer Definition of Selection Bias When Estimating Causal Effects

Selection bias remains a subject of controversy. Existing definitions of selection bias are ambiguous. To improve communication and the conduct of epidemiologic research focused on estimating causal effects, we propose to unify the various existing definitions of selection bias in the literature by considering any bias away from the true causal effect in the referent population (the population before the selection process), due to selecting the sample from the referent population, as selection bias. Given this unified definition, selection bias can be further categorized into two broad types: type 1 selection bias owing to restricting to one or more level(s) of a collider (or a descendant of a collider) and type 2 selection bias owing to restricting to one or more level(s) of an effect measure modifier. To aid in explaining these two types-which can co-occur-we start by reviewing the concepts of the target population, the study sample, and the analytic sample. Then, we illustrate both types of selection bias using causal diagrams. In addition, we explore the differences between these two types of selection bias, and describe methods to minimize selection bias. Finally, we use an example of "M-bias" to demonstrate the advantage of classifying selection bias into these two types.

A Causal Framework for Making Individualized Treatment Decisions in Oncology

Simple Summary

Physicians routinely make individualized treatment decisions by accounting for the joint effects of patient prognostic covariates and treatments on clinical outcomes. Ideally, this is performed using historical randomized clinical trial (RCT) data. Randomization ensures that unbiased estimates of causal treatment effect parameters can be obtained from the historical RCT data and used to predict each new patient’s outcome based on the joint effect of their baseline covariates and each treatment being considered. However, this process becomes problematic if a patient seen in the clinic is very different from the patients who were enrolled in the RCT. That is, if a new patient does not satisfy the entry criteria of the RCT, then the patient does not belong to the population represented by the patients who were studied in the RCT. In such settings, it still may be possible to utilize the RCT data to help choose a new patient’s treatment. This may be achieved by combining the RCT data with data from other clinical trials, or possibly preclinical experiments, and using the combined dataset to predict the patient’s expected outcome for each treatment being considered. In such settings, combining data from multiple sources in a way that is statistically reliable is not entirely straightforward, and correctly identifying and estimating the effects of treatments and patient covariates on clinical outcomes can be complex. Causal diagrams provide a rational basis to guide this process. The first step is to construct a causal diagram that reflects the plausible relationships between treatment variables, patient covariates, and clinical outcomes. If the diagram is correct, it can be used to determine what additional data may be needed, how to combine data from multiple sources, how to formulate a statistical model for clinical outcomes as a function of treatment and covariates, and how to compute an unbiased treatment effect estimate for each new patient. We use adjuvant therapy of renal cell carcinoma to illustrate how causal diagrams may be used to guide these steps.

Abstract

We discuss how causal diagrams can be used by clinicians to make better individualized treatment decisions. Causal diagrams can distinguish between settings where clinical decisions can rely on a conventional additive regression model fit to data from a historical randomized clinical trial (RCT) to estimate treatment effects and settings where a different approach is needed. This may be because a new patient does not meet the RCT’s entry criteria, or a treatment’s effect is modified by biomarkers or other variables that act as mediators between treatment and outcome. In some settings, the problem can be addressed simply by including treatment–covariate interaction terms in the statistical regression model used to analyze the RCT dataset. However, if the RCT entry criteria exclude a new patient seen in the clinic, it may be necessary to combine the RCT data with external data from other RCTs, single-arm trials, or preclinical experiments evaluating biological treatment effects. For example, external data may show that treatment effects differ between histological subgroups not recorded in an RCT. A causal diagram may be used to decide whether external observational or experimental data should be obtained and combined with RCT data to compute statistical estimates for making individualized treatment decisions. We use adjuvant treatment of renal cell carcinoma as our motivating example to illustrate how to construct causal diagrams and apply them to guide clinical decisions.

Is the Effect Larger in Group A or B? It Depends: Understanding Results From Nonlinear Probability Models

Demographers and other social scientists often study effect heterogeneity (defined here as differences in outcome-predictor associations across groups defined by the values of a third variable) to understand how inequalities evolve between groups or how groups differentially benefit from treatments. Yet answering the question "Is the effect larger in group A or group B?" is surprisingly difficult. In fact, the answer sometimes reverses across scales. For example, researchers might conclude that the effect of education on mortality is larger among women than among men if they quantify education's effect on an odds-ratio scale, but their conclusion might flip (to indicate a larger effect among men) if they instead quantify education's effect on a percentage-point scale. We illuminate this flipped-signs phenomenon in the context of nonlinear probability models, which were used in about one third of articles published in Demography in 2018-2019. Although methodologists are aware that flipped signs can occur, applied researchers have not integrated this insight into their work. We provide formal inequalities that researchers can use to easily determine if flipped signs are a problem in their own applications. We also share practical tips to help researchers handle flipped signs and, thus, generate clear and substantively correct descriptions of effect heterogeneity. Our findings advance researchers' ability to accurately characterize population variation.

Exchangeability of measures of association before and after exposure status is flipped: its relationship with confounding in the counterfactual model

BACKGROUND

The counterfactual definition of confounding is often explained in the context of exchangeability between the exposed and unexposed groups. One recent approach is to examine whether the measures of association (e.g., associational risk difference) are exchangeable when exposure status is flipped in the population of interest. We discuss the meaning and utility of this approach, showing their relationships with the concept of confounding in the counterfactual framework.

METHODS

Three hypothetical cohort studies are used, in which the target population is the total population. After providing an overview of the notions of confounding in distribution and in measure, we discuss the approach from the perspective of exchangeability of measures of association (e.g., factual associational risk difference vs. counterfactual associational risk difference).

RESULTS

In general, if the measures of association are non-exchangeable when exposure status is flipped, confounding in distribution is always present, although confounding in measure may or may not be present. Even if the measures of association are exchangeable when exposure status is flipped, there could be confounding both in distribution and in measure. When we use risk difference or risk ratio as a measure of interest and the exposure prevalence in the population is 0.5, testing the exchangeability of measures of association is equivalent to testing the absence of confounding in the corresponding measures.

CONCLUSIONS

The approach based on exchangeability of measures of association essentially does not provide a definition of confounding in the counterfactual framework. Subtly differing notions of confounding should be distinguished carefully.

Association between Patellofemoral and medial Tibiofemoral compartment osteoarthritis progression: exploring the effect of body weight using longitudinal data from osteoarthritis initiative (OAI)

OBJECTIVES

To investigate the associations of medial and lateral patellofemoral osteoarthritis (PF-OA) at baseline with symptomatic and radiographic OA outcomes in the medial tibiofemoral compartment (MTFC) over 4 years, according to baseline overweight status.

METHODS

Data and MRI images of 600 subjects in the FNIH-OA biomarkers consortium were used. Symptomatic worsening and radiographic progression of MTFC-OA were defined using Western Ontario and McMaster Universities Arthritis Index (WOMAC) pain scores and MTFC joint space narrowing (JSN) from baseline to 4-year follow-up. Baseline MRIs were read to establish PF-OA diagnosis. The association between baseline regional PF-OA pattern and odds for MTFC-OA progression was evaluated using regression models (adjusted for relevant confounding covariates including body mass index (BMI), age, sex, PF alignment measurements, KL grade, and knee alignment). To evaluate the effect modifying role for overweight status, stratification analysis was performed (BMI ≥ 25 vs. < 25 kg/m²).

RESULTS

At baseline, 340 (56.7%), 255 (42.5%), and 199 (33.2%) subjects had OA in the medial, lateral, and both PF compartments. Baseline medial PF-OA was associated with WOMAC pain score and MTFC JSN progression at 4 years (Adjusted OR:1.56[95%CI:1.09-2.23] and 1.59[1.11-2.28], respectively) but not lateral PF-OA. In stratification analysis, overweight status was found to be an effect modifier for medial PF-OA and WOMAC pain (OR in overweight vs. non-overweight subjects:1.65[1.13-2.42] vs. 0.50[0.12-1.82]) as well as MTFC-JSN progression (1.63[1.12-2.4] vs. 0.75[0.19-2.81]).

CONCLUSIONS

In addition to the known confounding effect of BMI for PF-OA and MTFC-OA, the overweight status may also play an effect modifier role in the association between baseline medial PF-OA and MTFC-OA progression, which is amenable to secondary prevention.

Assessing causality in epidemiology: revisiting Bradford Hill to incorporate developments in causal thinking

The nine Bradford Hill (BH) viewpoints (sometimes referred to as criteria) are commonly used to assess causality within epidemiology. However, causal thinking has since developed, with three of the most prominent approaches implicitly or explicitly building on the potential outcomes framework: directed acyclic graphs (DAGs), sufficient-component cause models (SCC models, also referred to as 'causal pies') and the grading of recommendations, assessment, development and evaluation (GRADE) methodology. This paper explores how these approaches relate to BH's viewpoints and considers implications for improving causal assessment. We mapped the three approaches above against each BH viewpoint. We found overlap across the approaches and BH viewpoints, underscoring BH viewpoints' enduring importance. Mapping the approaches helped elucidate the theoretical underpinning of each viewpoint and articulate the conditions when the viewpoint would be relevant. Our comparisons identified commonality on four viewpoints: strength of association (including analysis of plausible confounding); temporality; plausibility (encoded by DAGs or SCC models to articulate mediation and interaction, respectively); and experiments (including implications of study design on exchangeability). Consistency may be more usefully operationalised by considering an effect size's transportability to a different population or unexplained inconsistency in effect sizes (statistical heterogeneity). Because specificity rarely occurs, falsification exposures or outcomes (i.e., negative controls) may be more useful. The presence of a dose-response relationship may be less than widely perceived as it can easily arise from confounding. We found limited utility for coherence and analogy. This study highlights a need for greater clarity on BH viewpoints to improve causal assessment.

Assessing causality in epidemiology: revisiting Bradford Hill to incorporate developments in causal thinking

The nine Bradford Hill (BH) viewpoints (sometimes referred to as criteria) are commonly used to assess causality within epidemiology. However, causal thinking has since developed, with three of the most prominent approaches implicitly or explicitly building on the potential outcomes framework: directed acyclic graphs (DAGs), sufficient-component cause models (SCC models, also referred to as 'causal pies') and the grading of recommendations, assessment, development and evaluation (GRADE) methodology. This paper explores how these approaches relate to BH's viewpoints and considers implications for improving causal assessment. We mapped the three approaches above against each BH viewpoint. We found overlap across the approaches and BH viewpoints, underscoring BH viewpoints' enduring importance. Mapping the approaches helped elucidate the theoretical underpinning of each viewpoint and articulate the conditions when the viewpoint would be relevant. Our comparisons identified commonality on four viewpoints: strength of association (including analysis of plausible confounding); temporality; plausibility (encoded by DAGs or SCC models to articulate mediation and interaction, respectively); and experiments (including implications of study design on exchangeability). Consistency may be more usefully operationalised by considering an effect size's transportability to a different population or unexplained inconsistency in effect sizes (statistical heterogeneity). Because specificity rarely occurs, falsification exposures or outcomes (i.e., negative controls) may be more useful. The presence of a dose-response relationship may be less than widely perceived as it can easily arise from confounding. We found limited utility for coherence and analogy. This study highlights a need for greater clarity on BH viewpoints to improve causal assessment.

A generalizability score for aggregate causal effect

Scientists frequently generalize population level causal quantities such as average treatment effect from a source population to a target population. When the causal effects are heterogeneous, differences in subject characteristics between the source and target populations may make such a generalization difficult and unreliable. Reweighting or regression can be used to adjust for such differences when generalizing. However, these methods typically suffer from large variance if there is limited covariate distribution overlap between the two populations. We propose a generalizability score to address this issue. The score can be used as a yardstick to select target subpopulations for generalization. A simplified version of the score avoids using any outcome information and thus can prevent deliberate biases associated with inadvertent access to such information. Both simulation studies and real data analysis demonstrate convincing results for such selection.

Tutorial: A nontechnical explanation of the counterfactual definition of effect modification and interaction

Effect modification and interaction are important concepts for answering causal questions about interdependent effects of two (or more) exposures on some outcome of interest. Although conceptually alike and often mistakenly regarded as synonymous, effect modification and interaction actually refer to slightly different concepts when considered from a causal perspective. Their subtle yet relevant distinction lies in how the interplay between exposures is defined and the causal roles attributed to the exposures involved in the effect modification and interaction. To gain more insight into similarities and differences between the concepts of effect modification and interaction, the counterfactual theory of causation, albeit complicated, can be very helpful. Therefore, this article presents a nontechnical explanation of the counterfactual definition of effect modification and interaction. Essentially, effect modification and interaction are reflections of the reality and complexity of multicausality. The causal effect of an exposure of interest often depends on the levels of other exposures (effect modification) or causal effects of other exposures (interaction). Consequently, exposure effects should not be regarded in isolation but in combination. Understanding the underlying principles of effect modification and interaction on a conceptual level enables researchers to better anticipate, detect, and interpret these causal phenomena when setting up, analyzing, and reporting findings of (clinical) epidemiological studies. Effect modification and interaction are not biases to be avoided but properties of causal effects that ought to be unveiled. Hence, evidence for effect modification and interaction needs to be shown in order to delineate in whom and which instances causal effects occur.

The association between waiting time and multidisciplinary pain treatment outcomes in patients with rheumatic conditions

BACKGROUND

Access to multidisciplinary pain treatment facilities (MPTF) is limited by extensive waiting time in many countries. However, there is a lack of knowledge about the impact of waiting time on clinical outcomes, particularly for patients with rheumatic conditions. This study examined the association between waiting time for MPTF and clinical outcomes in patients with rheumatic conditions.

METHODS

Data were extracted from the Quebec Pain Registry, a large database of patients who received services in MPTF. The associations between waiting time (classified as < 2 months, 2-6 months and >  6 months) and change in pain interference, pain intensity and health-related quality of life, from the initial visit at the MPTF to the 6-month follow-up, were tested using generalized estimating equations.

RESULTS

A total of 3230 patients with rheumatic conditions (mean age: 55.8 ± 14.0 years; 66% were women) were included in the analysis. Small significant differences in improvement between waiting time groups were revealed, with patients waiting less than 2 months having a larger improvement in all clinical outcomes compared to patients who waited 2-6 months or over 6 months before their initial visit (adjusted time X group effect p ≤ 0.001). Only patients waiting less than 2 months reached a clinically important improvement in pain interference (1.12/10), pain intensity (1.3/10) and physical and mental quality of life (3.9 and 3.7/100).

CONCLUSIONS

Longer delays experienced by patients before receiving services in MPTF were associated with statistically significant smaller improvements in pain interference, pain intensity and health-related quality of life; these differences were, however, not clinically significant. Based on these results, we advise that strategies are developed not only to reduce waiting times and mitigate their impacts on patients with rheumatic conditions, but also to improve treatment effectiveness in MPTF.

Heterogeneity of antidiabetic treatment effect on the risk of major adverse cardiovascular events in type 2 diabetes: a systematic review and meta-analysis

BACKGROUND

We explored whether clinically relevant baseline characteristics of patients with type 2 diabetes can modify the effect of glucagon-like peptide-1 receptor agonists (GLP-1 RA) or sodium-glucose cotransporter-2 inhibitors (SGLT-2i) on the risk of major adverse cardiovascular events (MACE).

METHODS

We investigated Medline and EMBASE through June 2019. We included randomized clinical trials reporting the effect of GLP-1 RA or SGLT-2i on MACE in subgroups of patients with type 2 diabetes, identified through key baseline factors: established cardiovascular disease; heart failure; chronic kidney disease; uncontrolled diabetes; duration of diabetes; hypertension; obesity; age; gender and race. Hazard ratios (HRs) and 95% confidence intervals (CIs) from trials were meta-analyzed using random-effects models.

RESULTS

Ten trials enrolling 89,790 patients were included in the analyses. Subgroup meta-analyses showed a 14% risk reduction of MACE in patients with established cardiovascular disease [GLP1-RA: HR, 0.86 (95% CI, 0.80-0.93); SGLT-2i: 0.86 (0.80-0.93)], and no effect in at-risk patients without history of cardiovascular events [GLP1-RA: 0.94 (0.82-1.07); SGLT-2i: 1.00 (0.87-1.16)]. We observed a trend toward larger treatment benefits with SGLT-2i among patients with chronic kidney disease [0.82 (0.69-0.97)], and patients with uncontrolled diabetes for both GLP1-RA or SGLT-2i [GLP1-RA: 0.82 (0.71-0.95); SGLT-2i: 0.84 (0.75-0.95)]. Uncontrolled hypertension, obesity, gender, age and race did not appear to modify the effect of these drugs.

CONCLUSIONS

In this exploratory analysis, history of cardiovascular disease appeared to modify the treatment effect of SGLT2i or GLP1-RA on MACE. Chronic kidney disease and uncontrolled diabetes should be further investigated as potential effect modifiers.

The Interaction Continuum

A common reason given for assessing interaction is to evaluate “whether the effect is larger in one group versus another”. It has long been known that the answer to this question is scale dependent: the “effect” may be larger for one subgroup on the difference scale, but smaller on the ratio scale. In this article, we show that if the relative magnitude of effects across subgroups is of interest then there exists an “interaction continuum” that characterizes the nature of these relations. When both main effects are positive then the placement on the continuum depends on the relative magnitude of the probability of the outcome in the doubly exposed group. For high probabilities of the outcome in the doubly exposed group, the interaction may be positive-multiplicative positive-additive, the strongest form of positive interaction on the “interaction continuum”. As the probability of the outcome in the doubly exposed group goes down, the form of interaction descends through ranks, of what we will refer to as the following: positive-multiplicative positive-additive, no-multiplicative positive-additive, negative-multiplicative positive-additive, negative-multiplicative zero-additive, negative-multiplicative negative-additive, single pure interaction, single qualitative interaction, single-qualitative single-pure interaction, double qualitative interaction, perfect antagonism, inverted interaction. One can thus place a particular set of outcome probabilities into one of these eleven states on the interaction continuum. Analogous results are also given when both exposures are protective, or when one is protective and one causative. The “interaction continuum” can allow for inquiries as to relative effects sizes, while also acknowledging the scale dependence of the notion of interaction itself.

Causal Diagrams: Pitfalls and Tips

Graphical models are useful tools in causal inference, and causal directed acyclic graphs (DAGs) are used extensively to determine the variables for which it is sufficient to control for confounding to estimate causal effects. We discuss the following ten pitfalls and tips that are easily overlooked when using DAGs: 1) Each node on DAGs corresponds to a random variable and not its realized values; 2) The presence or absence of arrows in DAGs corresponds to the presence or absence of individual causal effect in the population; 3) “Non-manipulable” variables and their arrows should be drawn with care; 4) It is preferable to draw DAGs for the total population, rather than for the exposed or unexposed groups; 5) DAGs are primarily useful to examine the presence of confounding in distribution in the notion of confounding in expectation; 6) Although DAGs provide qualitative differences of causal structures, they cannot describe details of how to adjust for confounding; 7) DAGs can be used to illustrate the consequences of matching and the appropriate handling of matched variables in cohort and case-control studies; 8) When explicitly accounting for temporal order in DAGs, it is necessary to use separate nodes for each timing; 9) In certain cases, DAGs with signed edges can be used in drawing conclusions about the direction of bias; and 10) DAGs can be (and should be) used to describe not only confounding bias but also other forms of bias. We also discuss recent developments of graphical models and their future directions.

Effect heterogeneity and variable selection for standardizing causal effects to a target population

The participants in randomized trials and other studies used for causal inference are often not representative of the populations seen by clinical decision-makers. To account for differences between populations, researchers may consider standardizing results to a target population. We discuss several different types of homogeneity conditions that are relevant for standardization: Homogeneity of effect measures, homogeneity of counterfactual outcome state transition parameters, and homogeneity of counterfactual distributions. Each of these conditions can be used to show that a particular standardization procedure will result in an unbiased estimate of the effect in the target population, given assumptions about the relevant scientific context. We compare and contrast the homogeneity conditions, in particular their implications for selection of covariates for standardization and their implications for how to compute the standardized causal effect in the target population. While some of the recently developed counterfactual approaches to generalizability rely upon homogeneity conditions that avoid many of the problems associated with traditional approaches, they often require adjustment for a large (and possibly unfeasible) set of covariates.

Religiosity and health: A global comparative study

The objective of this paper is to understand global connections between indicators of religiosity and health and how these differ cross-nationally. Data are from World Values Surveys (93 countries, N=121,770). Health is based on a self-assessed question about overall health. First, country-specific regressions are examined to determine the association separately in each country. Next, country-level variables and cross-level interactions are added to multilevel models to assess whether and how context affects health and religiosity slopes. Results indicate enormous variation in associations between religiosity and health across countries and religiosity indicators. Significant positive associations between all religiosity measures and health exist in only three countries (Georgia, South Africa, and USA); negative associations in only two (Slovenia and Tunisia). Macro-level variables explain some of this divergence. Greater participation in religious activity relates to better health in countries characterized as being religiously diverse. The importance in god and pondering life's meaning is more likely associated with better health in countries with low levels of the Human Development Index. Pondering life's meaning more likely associates with better health in countries that place more stringent restrictions on religious practice. Religiosity is less likely to be related to good health in communist and former communist countries of Asia and Eastern Europe. In conclusion, the association between religiosity and health is complex, being partly shaped by geopolitical and macro psychosocial contexts.

The choice of effect measure for binary outcomes: Introducing counterfactual outcome state transition parameters

Standard measures of effect, including the risk ratio, the odds ratio, and the risk difference, are associated with a number of well-described shortcomings, and no consensus exists about the conditions under which investigators should choose one effect measure over another. In this paper, we introduce a new framework for reasoning about choice of effect measure by linking two separate versions of the risk ratio to a counterfactual causal model. In our approach, effects are defined in terms of "counterfactual outcome state transition parameters", that is, the proportion of those individuals who would not have been a case by the end of follow-up if untreated, who would have responded to treatment by becoming a case; and the proportion of those individuals who would have become a case by the end of follow-up if untreated who would have responded to treatment by not becoming a case. Although counterfactual outcome state transition parameters are generally not identified from the data without strong monotonicity assumptions, we show that when they stay constant between populations, there are important implications for model specification, meta-analysis, and research generalization.

Understanding and misunderstanding randomized controlled trials

Randomized Controlled Trials (RCTs) are increasingly popular in the social sciences, not only in medicine. We argue that the lay public, and sometimes researchers, put too much trust in RCTs over other methods of investigation. Contrary to frequent claims in the applied literature, randomization does not equalize everything other than the treatment in the treatment and control groups, it does not automatically deliver a precise estimate of the average treatment effect (ATE), and it does not relieve us of the need to think about (observed or unobserved) covariates. Finding out whether an estimate was generated by chance is more difficult than commonly believed. At best, an RCT yields an unbiased estimate, but this property is of limited practical value. Even then, estimates apply only to the sample selected for the trial, often no more than a convenience sample, and justification is required to extend the results to other groups, including any population to which the trial sample belongs, or to any individual, including an individual in the trial. Demanding 'external validity' is unhelpful because it expects too much of an RCT while undervaluing its potential contribution. RCTs do indeed require minimal assumptions and can operate with little prior knowledge. This is an advantage when persuading distrustful audiences, but it is a disadvantage for cumulative scientific progress, where prior knowledge should be built upon, not discarded. RCTs can play a role in building scientific knowledge and useful predictions but they can only do so as part of a cumulative program, combining with other methods, including conceptual and theoretical development, to discover not 'what works', but 'why things work'.

Using group data to treat individuals: understanding heterogeneous treatment effects in the age of precision medicine and patient-centred evidence

Although often conflated, determining the best treatment for an individual (the task of a doctor) is fundamentally different from determining the average effect of treatment in a population (the purpose of a trial). In this paper, we review concepts of heterogeneity of treatment effects (HTE) essential in providing the evidence base for precision medicine and patient-centred care, and explore some inherent limitations of using group data (e.g. from a randomized trial) to guide treatment decisions for individuals. We distinguish between person-level HTE (i.e. that individuals experience different effects from a treatment) and group-level HTE (i.e. that subgroups have different average treatment effects), and discuss the reference class problem, engendered by the large number of potentially informative subgroupings of a study population (each of which may lead to applying a different estimated effect to the same patient), and the scale dependence of group-level HTE. We also review the limitations of conventional 'one-variable-at-a-time' subgroup analyses and discuss the potential benefits of using more comprehensive subgrouping schemes that incorporate information on multiple variables, such as those based on predicted outcome risk. Understanding the conceptual underpinnings of HTE is critical for understanding how studies can be designed, analysed, and interpreted to better inform individualized clinical decisions.

A Bayesian semiparametric latent variable approach to causal mediation

In assessing causal mediation effects in randomized studies, a challenge is that the direct and indirect effects can vary across participants due to different measured and unmeasured characteristics. In that case, the population effect estimated from standard approaches implicitly averages over and does not estimate the heterogeneous direct and indirect effects. We propose a Bayesian semiparametric method to estimate heterogeneous direct and indirect effects via clusters, where the clusters are formed by both individual covariate profiles and individual effects due to unmeasured characteristics. These cluster-specific direct and indirect effects can be estimated through a set of regression models where specific coefficients are clustered by a stick-breaking prior. To let clustering be appropriately informed by individual direct and indirect effects, we specify a data-dependent prior. We conduct simulation studies to assess performance of the proposed method compared to other methods. We use this approach to estimate heterogeneous causal direct and indirect effects of an expressive writing intervention for patients with renal cell carcinoma.

A typology of four notions of confounding in epidemiology

Confounding is a major concern in epidemiology. Despite its significance, the different notions of confounding have not been fully appreciated in the literature, leading to confusion of causal concepts in epidemiology. In this article, we aim to highlight the importance of differentiating between the subtly different notions of confounding from the perspective of counterfactual reasoning. By using a simple example, we illustrate the significance of considering the distribution of response types to distinguish causation from association, highlighting that confounding depends not only on the population chosen as the target of inference, but also on the notions of confounding in distribution and confounding in measure. This point has been relatively underappreciated, partly because some literature on the concept of confounding has only used the exposed and unexposed groups as the target populations, while it would be helpful to use the total population as the target population. Moreover, to clarify a further distinction between confounding “in expectation” and “realized” confounding, we illustrate the usefulness of examining the distribution of exposure status in the target population. To grasp the explicit distinction between confounding in expectation and realized confounding, we need to understand the mechanism that generates exposure events, not the product of that mechanism. Finally, we graphically illustrate this point, highlighting the usefulness of directed acyclic graphs in examining the presence of confounding in distribution, in the notion of confounding in expectation.

All your data are always missing: incorporating bias due to measurement error into the potential outcomes framework

Epidemiologists often use the potential outcomes framework to cast causal inference as a missing data problem. Here, we demonstrate how bias due to measurement error can be described in terms of potential outcomes and considered in concert with bias from other sources. In addition, we illustrate how acknowledging the uncertainty that arises due to measurement error increases the amount of missing information in causal inference. We use a simple example to show that estimating the average treatment effect requires the investigator to perform a series of hidden imputations based on strong assumptions.

On the definition of a confounder

The causal inference literature has provided a clear formal definition of confounding expressed in terms of counterfactual independence. The causal inference literature has not, however, produced a clear formal definition of a confounder, as it has given priority to the concept of confounding over that of a confounder. We consider a number of candidate definitions arising from various more informal statements made in the literature. We consider the properties satisfied by each candidate definition, principally focusing on (i) whether under the candidate definition control for all "confounders" suffices to control for "confounding" and (ii) whether each confounder in some context helps eliminate or reduce confounding bias. Several of the candidate definitions do not have these two properties. Only one candidate definition of those considered satisfies both properties. We propose that a "confounder" be defined as a pre-exposure covariate C for which there exists a set of other covariates X such that effect of the exposure on the outcome is unconfounded conditional on (X, C) but such that for no proper subset of (X, C) is the effect of the exposure on the outcome unconfounded given the subset. A variable that helps reduce bias but not eliminate bias we propose referring to as a "surrogate confounder."