Introduction to computational causal inference using reproducible Stata, R, and Python code: A tutorial

Going beyond randomised controlled trials to assess treatment effect heterogeneity across target populations

Methods have been developed for transporting evidence from randomised controlled trials (RCTs) to target populations. However, these approaches allow only for differences in characteristics observed in the RCT and real-world data (overt heterogeneity). These approaches do not recognise heterogeneity of treatment effects (HTE) according to unmeasured characteristics (essential heterogeneity). We use a target trial design and apply a local instrumental variable (LIV) approach to electronic health records from the Clinical Practice Research Datalink, and examine both forms of heterogeneity in assessing the comparative effectiveness of two second-line treatments for type 2 diabetes mellitus. We first estimate individualised estimates of HTE across the entire target population defined by applying eligibility criteria from national guidelines (n = 13,240) within an overall target trial framework. We define a subpopulation who meet a published RCT's eligibility criteria ('RCT-eligible', n = 6497), and a subpopulation who do not ('RCT-ineligible', n = 6743). We compare average treatment effects for pre-specified subgroups within the RCT-eligible subpopulation, the RCT-ineligible subpopulation, and within the overall target population. We find differences across these subpopulations in the magnitude of subgroup-level treatment effects, but that the direction of estimated effects is stable. Our results highlight that LIV methods can provide useful evidence about treatment effect heterogeneity including for those subpopulations excluded from RCTs.

Machine Learning Methods to Estimate Individualized Treatment Effects for Use in Health Technology Assessment

BACKGROUND

Recent developments in causal inference and machine learning (ML) allow for the estimation of individualized treatment effects (ITEs), which reveal whether treatment effectiveness varies according to patients' observed covariates. ITEs can be used to stratify health policy decisions according to individual characteristics and potentially achieve greater population health. Little is known about the appropriateness of available ML methods for use in health technology assessment.

METHODS

In this scoping review, we evaluate ML methods available for estimating ITEs, aiming to help practitioners assess their suitability in health technology assessment. We present a taxonomy of ML approaches, categorized by key challenges in health technology assessment using observational data, including handling time-varying confounding and time-to event data and quantifying uncertainty.

RESULTS

We found a wide range of algorithms for simpler settings with baseline confounding and continuous or binary outcomes. Not many ML algorithms can handle time-varying or unobserved confounding, and at the time of writing, no ML algorithm was capable of estimating ITEs for time-to-event outcomes while accounting for time-varying confounding. Many of the ML algorithms that estimate ITEs in longitudinal settings do not formally quantify uncertainty around the point estimates.

LIMITATIONS

This scoping review may not cover all relevant ML methods and algorithms as they are continuously evolving.

CONCLUSIONS

Existing ML methods available for ITE estimation are limited in handling important challenges posed by observational data when used for cost-effectiveness analysis, such as time-to-event outcomes, time-varying and hidden confounding, or the need to estimate sampling uncertainty around the estimates.

IMPLICATIONS

ML methods are promising but need further development before they can be used to estimate ITEs for health technology assessments.

HIGHLIGHTS

Estimating individualized treatment effects (ITEs) using observational data and machine learning (ML) can support personalized treatment advice and help deliver more customized information on the effectiveness and cost-effectiveness of health technologies.ML methods for ITE estimation are mostly designed for handling confounding at baseline but not time-varying or unobserved confounding. The few models that account for time-varying confounding are designed for continuous or binary outcomes, not time-to-event outcomes.Not all ML methods for estimating ITEs can quantify the uncertainty of their predictions.Future work on developing ML that addresses the concerns summarized in this review is needed before these methods can be widely used in clinical and health technology assessment-like decision making.

A practical guide to estimating treatment effects in patients with rheumatic diseases using real-world data

OBJECTIVE

Randomized controlled trials are considered the gold standard in study methodology. However, due to their study design and inclusion criteria, these studies may not capture the heterogeneity of real-world patient populations. In contrast, the lack of randomization and the presence of both measured and unmeasured confounding factors could bias the estimated treatment effect when using observational data. While causal inference methods allow for the estimation of treatment effects, their mathematical complexity may hinder their application in clinical research.

METHODS

We present a practical, nontechnical guide using a common statistical package (Stata) and a motivational simulated dataset that mirrors real-world observational data from patients with rheumatic diseases. We demonstrate regression analysis, regression adjustment, inverse-probability weighting, propensity score (PS) matching and two robust estimation methods.

RESULTS

Although the methods applied to control for confounding factors produced similar results, the commonly used one-to-one PS matching method could yield biased results if not thoroughly assessed.

CONCLUSION

The guide we propose aims to facilitate the use of readily available methods in a common statistical package. It may contribute to robust and transparent epidemiological and statistical methods, thereby enhancing effectiveness research using observational data in rheumatology.

Reducing Stimulant Prescribing Error: A Quality Improvement Initiative in Pediatric Outpatient Setting

OBJECTIVE

To evaluate the impact of the Songklanagarind ADHD Multidisciplinary Assessment and Care Team for Quality Improvement (SAMATI) initiative on reducing stimulant prescribing errors in a pediatric outpatient setting.

METHODS

A retrospective study examined attention deficit hyperactivity disorder (ADHD) registry data from January 2017 to June 2023 to assess the impact of the SAMATI initiative, implemented in early 2020. This initiative, integrating multiple components such as audit and feedback, clinical pharmacist involvement, and Electronic Medical Record utilization, aimed to enhance ADHD medication management. Using interrupted time series and control chart analyses, the study evaluated the initiative's effect on reducing stimulant prescribing errors. Additionally, parental satisfaction was measured to gauge the initiative's overall success.

RESULTS

Out of 282 patients enrolled, 267 were included in the final analysis after exclusions. Post-intervention analysis showed significant reductions in various prescribing errors per thousand prescriptions: prescribing without concern drug-condition interaction (443 to 145, p < 0.001), prescribing without adequate monitoring (115 to 14, p < 0.001), lack of regular office visits (98 to 21, p = 0.007), and inappropriate dosage (66 to 14, p = 0.05). Medication errors severity classification also showed significant changes, with notable decreases in classes C and D errors. Parental satisfaction improved from 84% to 95%.

CONCLUSION

The SAMATI initiative significantly reduced stimulant prescribing errors and enhanced parental satisfaction in ADHD care management. This study demonstrates the potential of comprehensive quality improvement strategies in improving medication management in pediatric healthcare. Further research in diverse settings is warranted to confirm these findings.

On variance estimation of the inverse probability-of-treatment weighting estimator: A tutorial for different types of propensity score weights

Propensity score methods, such as inverse probability-of-treatment weighting (IPTW), have been increasingly used for covariate balancing in both observational studies and randomized trials, allowing the control of both systematic and chance imbalances. Approaches using IPTW are based on two steps: (i) estimation of the individual propensity scores (PS), and (ii) estimation of the treatment effect by applying PS weights. Thus, a variance estimator that accounts for both steps is crucial for correct inference. Using a variance estimator which ignores the first step leads to overestimated variance when the estimand is the average treatment effect (ATE), and to under or overestimated estimates when targeting the average treatment effect on the treated (ATT). In this article, we emphasize the importance of using an IPTW variance estimator that correctly considers the uncertainty in PS estimation. We present a comprehensive tutorial to obtain unbiased variance estimates, by proposing and applying a unifying formula for different types of PS weights (ATE, ATT, matching and overlap weights). This can be derived either via the linearization approach or M-estimation. Extensive R code is provided along with the corresponding large-sample theory. We perform simulation studies to illustrate the behavior of the estimators under different treatment and outcome prevalences and demonstrate appropriate behavior of the analytical variance estimator. We also use a reproducible analysis of observational lung cancer data as an illustrative example, estimating the effect of receiving a PET-CT scan on the receipt of surgery.

Use and reporting of inverse-probability-of-treatment weighting for multicategory treatments in medical research: a systematic review

OBJECTIVES

Causal inference methods for observational data represent an alternative to randomised controlled trials when they are not feasible or when real-world evidence is sought. Inverse-probability-of-treatment weighting (IPTW) is one of the most popular approaches to account for confounding in observational studies. In medical research, IPTW is mainly applied to estimate the causal effect of a binary treatment, even when the treatment has in fact multiple categories, despite the availability of IPTW estimators for multiple treatment categories. This raises questions about the appropriateness of the use of IPTW in this context. Therefore, we conducted a systematic review of medical publications reporting the use of IPTW in the presence of a multi-category treatment. Our objectives were to investigate the frequency of use and the implementation of these methods in practice, and to assess the quality of their reporting.

STUDY DESIGN AND SETTING

Using Pubmed, Embase and Web of Science, we screened 5660 articles and retained 106 articles in the final analysis that were from 17 different medical areas. This systematic review is registered on PROSPERO (CRD42022352669).

RESULTS

The number of treatment groups varied between 3 and 9, with a large majority of articles (90 [84.9%]) including 3 or 4 groups. The most commonly used method for estimating the weights was multinomial regression (51 [48.1%]) and generalized boosted models (48 [45.3%]). The covariates of the weight model were reported in 91 articles (85.9 %). Twenty-six articles (24.5 %) did not discuss the balance of covariates after weighting, and only 16 articles (15.1 %) referred to the assumptions needed to obtain correct inferences.

CONCLUSION

The results of this systematic review illustrate that medical publications scarcely use IPTW methods for more than two treatment categories. Among the publications that did, the quality of reporting was suboptimal, in particular in regard to the assumptions and model building. IPTW for multi-category treatments could be applied more broadly in medical research, and the application of the proposed guidelines in this context will help researchers to report their results and to ensure reproducibility of their research.

Geospatial Distribution of Racial Disparities in Influenza Vaccination in Nursing Homes

OBJECTIVES

This study aimed to assess the distribution of racial disparities in influenza vaccination between White and Black short-stay and long-stay nursing home residents among states and hospital referral regions (HRRs).

DESIGN

Retrospective cohort study.

SETTING AND PARTICIPANTS

We included short-stay and long-stay older adults residing in US nursing homes during influenza seasons between 2011 and 2018. Included residents were aged ≥65 years and enrolled in Traditional Medicare. Analyses were conducted using resident-seasons, whereby residents could contribute to one or more influenza seasons if they resided in a nursing home across multiple seasons.

METHODS

Our comparison of interest was marginalized vs privileged racial group membership measured as Black vs White race. We obtained influenza vaccination documentation from resident Minimum Data Set assessments from October 1 through June 30 of a particular influenza season. Nonparametric g-formula was used to estimate age- and sex-standardized disparities in vaccination, measured as the percentage point (pp) difference in the proportions of individuals vaccinated between Black and White nursing home residents within states and HRRs.

RESULTS

The study included 7,807,187 short-stay resident-seasons (89.7% White and 10.3% Black) in 14,889 nursing homes and 7,308,111 long-stay resident-seasons (86.7% White and 13.3% Black) in 14,885 nursing homes. Among states, the median age- and sex-standardized disparity between Black and White residents was 10.1 percentage points (pps) among short-stay residents and 5.3 pps among long-stay residents across seasons. Among HRRs, the median disparity was 8.6 pps among short-stay residents and 5.0 pps among long-stay residents across seasons.

CONCLUSIONS AND IMPLICATIONS

Our analysis revealed that the magnitudes of vaccination disparities varied substantially across states and HRRs, from no disparity in vaccination to disparities in excess of 25 pps. Local interventions and policies should be targeted to high-disparity geographic areas to increase vaccine uptake and promote health equity.

Comparison of outcomes between off-pump and on-pump coronary artery bypass graft surgery using collaborative targeted maximum likelihood estimation

There are some discrepancies about the superiority of the off-pump coronary artery bypass grafting (CABG) surgery over the conventional cardiopulmonary bypass (on-pump). The aim of this study was estimating risk ratio of mortality in the off-pump coronary bypass compared with the on-pump using a causal model known as collaborative targeted maximum likelihood estimation (C-TMLE). The data of the Tehran Heart Cohort study from 2007 to 2020 was used. A collaborative targeted maximum likelihood estimation and targeted maximum likelihood estimation, and propensity score (PS) adjustment methods were used to estimate causal risk ratio adjusting for the minimum sufficient set of confounders, and the results were compared. Among 24,883 participants (73.6% male), 5566 patients died during an average of 8.2 years of follow-up. The risk ratio estimates (95% confidence intervals) by unadjusted log-binomial regression model, PS adjustment, TMLE, and C-TMLE methods were 0.86 (0.78-0.95), 0.88 (0.80-0.97), 0.88 (0.80-0.97), and 0.87(0.85-0.89), respectively. This study provides evidence for a protective effect of off-pump surgery on mortality risk for up to 8 years in diabetic and non-diabetic patients.

Age at first drink and its influence on alcohol use behaviours in young adulthood: Evidence from an Australian household-based panel study

BACKGROUND

Public health guidelines recommend delaying the initiation age for alcohol. However, the causal link between age-at-first-drink (AFD) and future alcohol use in young adulthood is uncertain. This study examined the association between AFD and alcohol-related outcomes at age 20 years using an Australian sample.

METHODS

Data were obtained from Waves 1-19 (years 2001-2019) of the Household, Income and Labour Dynamics in Australia Survey on 20-year-olds with responses across ≥3 consecutive waves (n = 2278). The AFD for each respondent (between 15 and 20 years) was analysed relative to Australian legal drinking age (18 years). Inverse probability treatment weighting was used to evaluate associations between AFD and four outcomes at age 20 years: risk of current alcohol use; quantity of weekly alcohol consumption; risk of binge drinking; and frequency of binge drinking. Adjustments were made for confounders (e.g., heavy drinking by parents). Robustness of study findings was evaluated using several diagnostic tests/sensitivity analyses.

RESULTS

Among 20-year-olds, those with an AFD of 15-16 years consumed significantly more alcohol per week compared to an AFD of 18 years. Additionally, 20-year-old drinkers with an AFD of 16 years were significantly more likely to binge drink (though this association was likely confounded). An inverse dose-response relationship was observed between AFD and weekly alcohol consumption at 20 years, where a higher AFD led to lower alcohol consumption.

CONCLUSION

Study findings indicate an association between a higher AFD and consuming less alcohol in young adulthood, which could potentially support the scale-up of prevention programs to delay AFD among Australian adolescents.

Methodological biases in observational hospital studies of COVID-19 treatment effectiveness: pitfalls and potential

Introduction

This study aims to discuss and assess the impact of three prevalent methodological biases: competing risks, immortal-time bias, and confounding bias in real-world observational studies evaluating treatment effectiveness. We use a demonstrative observational data example of COVID-19 patients to assess the impact of these biases and propose potential solutions.

Methods

We describe competing risks, immortal-time bias, and time-fixed confounding bias by evaluating treatment effectiveness in hospitalized patients with COVID-19. For our demonstrative analysis, we use observational data from the registry of patients with COVID-19 who were admitted to the Bellvitge University Hospital in Spain from March 2020 to February 2021 and met our predefined inclusion criteria. We compare estimates of a single-dose, time-dependent treatment with the standard of care. We analyze the treatment effectiveness using common statistical approaches, either by ignoring or only partially accounting for the methodological biases. To address these challenges, we emulate a target trial through the clone-censor-weight approach.

Results

Overlooking competing risk bias and employing the naïve Kaplan-Meier estimator led to increased in-hospital death probabilities in patients with COVID-19. Specifically, in the treatment effectiveness analysis, the Kaplan-Meier estimator resulted in an in-hospital mortality of 45.6% for treated patients and 59.0% for untreated patients. In contrast, employing an emulated trial framework with the weighted Aalen-Johansen estimator, we observed that in-hospital death probabilities were reduced to 27.9% in the "X"-treated arm and 40.1% in the non-"X"-treated arm. Immortal-time bias led to an underestimated hazard ratio of treatment.

Conclusion

Overlooking competing risks, immortal-time bias, and confounding bias leads to shifted estimates of treatment effects. Applying the naïve Kaplan-Meier method resulted in the most biased results and overestimated probabilities for the primary outcome in analyses of hospital data from COVID-19 patients. This overestimation could mislead clinical decision-making. Both immortal-time bias and confounding bias must be addressed in assessments of treatment effectiveness. The trial emulation framework offers a potential solution to address all three methodological biases.

Unanchored Population-Adjusted Indirect Comparison Methods for Time-to-Event Outcomes Using Inverse Odds Weighting, Regression Adjustment, and Doubly Robust Methods With Either Individual Patient or Aggregate Data

OBJECTIVES

Several methods for unanchored population-adjusted indirect comparisons (PAICs) are available. Exploring alternative adjustment methods, depending on the available individual patient data (IPD) and the aggregate data (AD) in the external study, may help minimize bias in unanchored indirect comparisons. However, methods for time-to-event outcomes are not well understood. This study provides an overview and comparison of methods using a case study to increase familiarity. A recent method is applied to marginalize conditional hazard ratios, which allows for the comparisons of methods, and a doubly robust method is proposed.

METHODS

The following PAIC methods were compared through a case study in third-line small cell lung cancer, comparing nivolumab with standard of care based on a single-arm phase II trial (CheckMate 032) and real-world study (Flatiron) in terms of overall survival: IPD-IPD analyses using inverse odds weighting, regression adjustment, and a doubly robust method; IPD-AD analyses using matching-adjusted indirect comparison, simulated treatment comparison, and a doubly robust method.

RESULTS

Nivolumab extended survival versus standard of care with hazard ratios ranging from 0.63 (95% CI 0.44-0.90) in naive comparisons (identical estimates for IPD-IPD and IPD-AD analyses) to 0.69 (95% CI 0.44-0.98) in the IPD-IPD analyses using regression adjustment. Regression-based and doubly robust estimates yielded slightly wider confidence intervals versus the propensity score-based analyses.

CONCLUSIONS

The proposed doubly robust approach for time-to-event outcomes may help to minimize bias due to model misspecification. However, all methods for unanchored PAIC rely on the strong assumption that all prognostic covariates have been included.

Differentially Private Outcome-Weighted Learning for Optimal Dynamic Treatment Regime Estimation

Precision medicine is a framework for developing evidence-based medical recommendations that seeks to determine the optimal sequence of treatments tailored to all of the relevant patient-level characteristics which are observable. Because precision medicine relies on highly sensitive, patient-level data, ensuring the privacy of participants is of great importance. Dynamic treatment regimes (DTRs) provide one formalization of precision medicine in a longitudinal setting. Outcome-Weighted Learning (OWL) is a family of techniques for estimating optimal DTRs based on observational data. OWL techniques leverage support vector machine (SVM) classifiers in order to perform estimation. SVMs perform classification based on a set of influential points in the data known as support vectors. The classification rule produced by SVMs often requires direct access to the support vectors. Thus, releasing a treatment policy estimated with OWL requires the release of patient data for a subset of patients in the sample. As a result, the classification rules from SVMs constitute a severe privacy violation for those individuals whose data comprise the support vectors. This privacy violation is a major concern, particularly in light of the potentially highly sensitive medical data which are used in DTR estimation. Differential privacy has emerged as a mathematical framework for ensuring the privacy of individual-level data, with provable guarantees on the likelihood that individual characteristics can be determined by an adversary. We provide the first investigation of differential privacy in the context of DTRs and provide a differentially private OWL estimator, with theoretical results allowing us to quantify the cost of privacy in terms of the accuracy of the private estimators.

Cognitive Performance of Brazilian Patients With Favorable Outcomes After Severe Traumatic Brain Injury: A Prospective Study

OBJECTIVE

The aim of this study was to investigate the cognitive performance of patients with favorable outcomes, determined by the Glasgow Outcome Scale, 1 yr after hospital discharge due to severe traumatic brain injury.

DESIGN

This was a prospective case-control study. From 163 consecutive adult patients with severe traumatic brain injury included in the study, 73 patients had a favorable outcome (Glasgow Outcome Scale score of 4 or 5) 1 yr after hospital discharge and were eligible for the cognitive evaluation, of which 28 completed the evaluations. The latter were compared with 44 healthy controls.

RESULTS

The average loss of cognitive performance among participants with traumatic brain injury varied between 13.35% and 43.49% compared with the control group. Between 21.4% and 32% of the patients performed below the 10th percentile on three language tests and two verbal memory tests, whereas 39% to 50% performed below this threshold on one language test and three memory tests. Longer hospital stay, older age, and lower education were the most important predictors of worse cognitive performance.

CONCLUSION

One year after a severe traumatic brain injury, a significant proportion of Brazilian patients with the favorable outcome determined by Glasgow Outcome Scale still showed significant cognitive impairment in verbal memory and language domains.

Impact of special educational needs provision on hospital utilisation, school attainment and absences for children in English primary schools stratified by gestational age at birth: A target trial emulation study protocol

Introduction

One third of children in English primary schools have additional learning support called special educational needs (SEN) provision, but children born preterm are more likely to have SEN than those born at term. We aim to assess the impact of SEN provision on health and education outcomes in children grouped by gestational age at birth.

Methods

We will analyse linked administrative data for England using the Education and Child Health Insights from Linked Data (ECHILD) database. A target trial emulation approach will be used to specify data extraction from ECHILD, comparisons of interest and our analysis plan. Our target population is all children enrolled in year one of state-funded primary school in England who were born in an NHS hospital in England between 2003 and 2008, grouped by gestational age at birth (extremely preterm (24-<28 weeks), very preterm (28-<32 weeks), moderately preterm (32-<34 weeks), late preterm (34-<37 weeks) and full term (37-<42 weeks). The intervention of interest will comprise categories of SEN provision (including none) during year one (age five/six). The outcomes of interest are rates of unplanned hospital utilisation, educational attainment, and absences by the end of primary school education (year six, age 11). We will triangulate results from complementary estimation methods including the naïve estimator, multivariable regression, g-formula, inverse probability weighting, inverse probability weighting with regression adjustment and instrumental variables, along with a variety for a variety of causal contrasts (average treatment effect, overall, and on the treated/not treated).

Ethics and dissemination

We have existing research ethics approval for analyses of the ECHILD database described in this protocol. We will disseminate our findings to diverse audiences (academics, relevant government departments, service users and providers) through seminars, peer-reviewed publications, short briefing reports and infographics for non-academics (published on the study website).

Early special educational needs provision and its impact on unplanned hospital utilisation and school absences in children with isolated cleft lip and/or palate: a demonstration target trial emulation study protocol using ECHILD

Background

Special educational needs (SEN) provision is designed to help pupils with additional educational, behavioural or health needs; for example, pupils with cleft lip and/or palate may be offered SEN provision to improve their speech and language skills. Our aim is to contribute to the literature and assess the impact of SEN provision on health and educational outcomes for a well-defined population.

Methods

We will use the ECHILD database, which links educational and health records across England. Our target population consists of children identified within ECHILD to have a specific congenital anomaly: isolated cleft lip and/or palate. We will apply a trial emulation framework to reduce biases in design and analysis of observational data to investigate the causal impact of SEN provision (including none) by the start of compulsory education (Year One - age five year on entry) on the number of unplanned hospital utilisation and school absences by the end of primary education (Year Six - age ten/eleven). We will use propensity score-based estimators (inverse probability weighting (IPW) and IPW regression adjustment IPW) to compare categories of SEN provision in terms of these outcomes and to triangulate results obtained using complementary estimation methods (Naïve estimator, multivariable regression, parametric g-formula, and if possible, instrumental variables), targeting a variety of causal contrasts (average treatment effect/in the treated/in the not treated) of SEN provision.

Conclusions

This study will evaluate the impact of reasonable adjustments at the start of compulsory education on health and educational outcomes in the isolated cleft lip and palate population by triangulating complementary methods under a target-trial framework.

The effect of smoking on latent hazard classes of metabolic syndrome using latent class causal analysis method in the Iranian population

BACKGROUND

The prevalence of metabolic syndrome is increasing worldwide. Clinical guidelines consider metabolic syndrome as an all or none medical condition. One proposed method for classifying metabolic syndrome is latent class analysis (LCA). One approach to causal inference in LCA is using propensity score (PS) methods. The aim of this study was to investigate the causal effect of smoking on latent hazard classes of metabolic syndrome using the method of latent class causal analysis.

METHODS

In this study, we used data from the Tehran Lipid and Glucose Cohort Study (TLGS). 4857 participants aged over 20 years with complete information on exposure (smoking) and confounders in the third phase (2005-2008) were included. Metabolic syndrome was evaluated as outcome and latent variable in LCA in the data of the fifth phase (2014-2015). The step-by-step procedure for conducting causal inference in LCA included: (1) PS estimation and evaluation of overlap, (2) calculation of inverse probability-of-treatment weighting (IPTW), (3) PS matching, (4) evaluating balance of confounding variables between exposure groups, and (5) conducting LCA using the weighted or matched data set.

RESULTS

Based on the results of IPTW which compared the low, medium and high risk classes of metabolic syndrome (compared to a class without metabolic syndrome), no association was found between smoking and the metabolic syndrome latent classes. PS matching which compared low and moderate risk classes compared to class without metabolic syndrome, showed that smoking increases the probability of being in the low-risk class of metabolic syndrome (OR: 2.19; 95% CI: 1.32, 3.63). In the unadjusted analysis, smoking increased the chances of being in the low-risk (OR: 1.45; 95% CI: 1.01, 2.08) and moderate-risk (OR: 1.68; 95% CI: 1.18, 2.40) classes of metabolic syndrome compared to the class without metabolic syndrome.

CONCLUSIONS

Based on the results, the causal effect of smoking on latent hazard classes of metabolic syndrome can be different based on the type of PS method. In adjusted analysis, no relationship was observed between smoking and moderate-risk and high-risk classes of metabolic syndrome.

Application of targeted maximum likelihood estimation in public health and epidemiological studies: a systematic review

PURPOSE

The targeted maximum likelihood estimation (TMLE) statistical data analysis framework integrates machine learning, statistical theory, and statistical inference to provide a least biased, efficient, and robust strategy for estimation and inference of a variety of statistical and causal parameters. We describe and evaluate the epidemiological applications that have benefited from recent methodological developments.

METHODS

We conducted a systematic literature review in PubMed for articles that applied any form of TMLE in observational studies. We summarized the epidemiological discipline, geographical location, expertize of the authors, and TMLE methods over time. We used the Roadmap of Targeted Learning and Causal Inference to extract key methodological aspects of the publications. We showcase the contributions to the literature of these TMLE results.

RESULTS

Of the 89 publications included, 33% originated from the University of California at Berkeley, where the framework was first developed by Professor Mark van der Laan. By 2022, 59% of the publications originated from outside the United States and explored up to seven different epidemiological disciplines in 2021-2022. Double-robustness, bias reduction, and model misspecification were the main motivations that drew researchers toward the TMLE framework. Through time, a wide variety of methodological, tutorial, and software-specific articles were cited, owing to the constant growth of methodological developments around TMLE.

CONCLUSIONS

There is a clear dissemination trend of the TMLE framework to various epidemiological disciplines and to increasing numbers of geographical areas. The availability of R packages, publication of tutorial papers, and involvement of methodological experts in applied publications have contributed to an exponential increase in the number of studies that understood the benefits and adoption of TMLE.

Calculating risk and prevalence ratios and differences in R: developing intuition with a hands-on tutorial and code

Epidemiologic research questions often focus on evaluating binary outcomes, yet curricula and scientific literature do not always provide clear guidance or examples on selecting and calculating an appropriate measure of association in these scenarios. Reporting inappropriate measures may lead to misleading statistical conclusions. We present a hands-on tutorial that includes annotated code written in an open-source statistical programming language (R) showing readers how to apply, compare, and understand four methods used to estimate a risk or prevalence ratio (or difference), rather than presenting an odds ratio. We will provide guidance on when to use each method, discussing the strengths and limitations of each approach, and compare the results obtained across them. Ultimately, we aim to help trainees, public health researchers, and interdisciplinary professionals develop an intuition for these methods and empower them to implement and interpret these methods in their own research.

Compliance Trajectory and Patterns of COVID-19 Preventive Measures, Japan, 2020-2022

COVID-19 remains a global health threat. Compliance with nonpharmaceutical interventions is essential because of limited effectiveness of COVID-19 vaccines, emergence of highly contagious variants, and declining COVID-19 antibody titers over time. We evaluated compliance with 14 nonpharmaceutical intervention-related COVID-19 preventive behaviors, including mask wearing, ventilation, and surface sanitation, in a longitudinal study in Japan using 4 waves of Internet survey data obtained during 2020-2022. Compliance with most preventive behaviors increased or remained stable during the 2-year period, except for surface sanitation and going out behaviors; compliance with ventilation behavior substantially decreased in winter. Compliance patterns identified from latent class analysis showed that the number of persons in the low compliance class decreased, whereas those in the personal hygiene class increased. Our findings reflect the relaxation of mobility restriction policy in Japan, where the COVID-19 pandemic continues. Policymakers should consider behavioral changes caused by new policies to improve COVID-19 prevention strategies.

Post-Acute Care Setting After Hip Fracture Hospitalization and Subsequent Opioid Use in Older Adults

OBJECTIVE

Pain management in post-acute care (PAC) requires careful balance, with both opioid use and inadequate pain treatment linked to poor outcomes. We describe opioid use among older adults following discharge from PAC for hip fracture in skilled nursing facilities (SNFs) and inpatient rehabilitation facilities (IRFs).

DESIGN

Retrospective cohort.

SETTING AND PARTICIPANTS

Medicare beneficiaries with Medicare Provider Analysis (MedPAR) claims, aged 66 years and older with a hip fracture hospitalization between 2012 and 2018 followed by PAC in SNFs or IRFs and then discharge to the community.

METHODS

Individuals were followed from PAC discharge for up to 1 year to assess opioid use. Covariate-standardized risk ratios (RR) and risk differences (RD) for opioid use within 7 days of PAC discharge were estimated via parametric g-formula with modified Poisson regression, and hazard ratios (HRs) for any post-PAC opioid use and long-term opioid use via Fine-Gray sub-distribution hazards regression.

RESULTS

Of 101,021 individuals, 80% (n = 80,495) were discharged from SNFs and 20% (n = 20,526) from IRFs. Opioids were dispensed to 50,433 patients (50%) overall and the 1-year cumulative incidence was notably higher in IRF (68%) than SNF (46%) patients. The adjusted risk of discharge from PAC with an opioid was 41% lower after SNFs versus IRFs [RR: 0.59; 95% confidence limits (CLs): 0.57-0.61; and RD: -0.16; 95% CLs: -0.17 to -0.15]. The adjusted rate of any opioid use in the year after PAC discharge was 44% lower (HR: 0.56; 95% CLs: 0.54-0.57) and of long-term opioid use was 17% lower (HR: 0.83; 95% CLs: 0.80-0.87) after SNFs versus IRFs.

CONCLUSIONS AND IMPLICATIONS

Opioid use is highly prevalent upon discharge from PAC after hip fracture, with lower use after SNF versus IRF care. Future research should assess the benefits and harms of post-PAC opioid prescribing and whether care practices during PAC can be improved to optimize long-term opioid use.

Protective effect of intensive glucose lowering therapy on all-cause mortality, adjusted for treatment switching using G-estimation method, the ACCORD trial

Previous analysis of the action to control cardiovascular risk in diabetes showed an increased risk of mortality among patients receiving intensive glucose lowering therapy using conventional regression method with intention to treat approach. This method is biased when time-varying confounder is affected by the previous treatment. We used 15 follow-up visits of ACCORD trial to compare the effect of time-varying intensive vs. standard treatment of glucose lowering drugs on cardiovascular and mortality outcomes in diabetic patients. The treatment effect was estimated using G-estimation and compared with accelerated failure time model using two modeling strategies. The first model adjusted for baseline confounders and the second adjusted for both baseline and time-varying confounders. While the hazard ratio of all-cause mortality for intensive compared to standard therapy in AFT model adjusted for baseline confounders was 1.17 (95% CI 1.01-1.36), the result of time-dependent AFT model  was compatible with both protective and risk effects. However, the hazard ratio estimated by G-estimation was 0.64 (95% CI 0.39-0.92). The results of this study revealed a protective effect of intensive therapy on all-cause mortality compared with standard therapy in ACCORD trial.

A practical guide to handling competing events in etiologic time-to-event studies

Competing events are events that preclude the occurrence of the primary outcome. Much has been written on mainly the statistics behind competing events analyses. However, many of these publications and tutorials have a strong statistical tone and might fall short in providing a practical guide to clinician researchers as to when to use a competing event analysis and more importantly which method to use and why. Here we discuss the different target effects in the Fine-Gray and cause-specific methods using simple causal diagrams and provide strengths and limitations of both approaches for addressing etiologic questions. We argue why the Fine-Gray method might not be the best approach for handling competing events in etiological time-to-event studies.

Geographic Variation in Influenza Vaccination Disparities Between Hispanic and Non-Hispanic White US Nursing Home Residents

BACKGROUND

Disparities in influenza vaccination exist between Hispanic and non-Hispanic White US nursing home (NH) residents, but the geographic areas with the largest disparities remain unknown. We examined how these racial/ethnic disparities differ across states and hospital referral regions (HRRs).

METHODS

This retrospective cohort study included >14 million short-stay and long-stay US NH resident-seasons over 7 influenza seasons from October 1, 2011, to March 31, 2018, where residents could contribute to 1 or more seasons. Residents were aged ≥65 years and enrolled in Medicare fee-for-service. We used the Medicare Beneficiary Summary File to ascertain race/ethnicity and Minimum Data Set assessments for influenza vaccination. We calculated age- and sex-standardized percentage point (pp) differences in the proportions vaccinated between non-Hispanic White and Hispanic (any race) resident-seasons. Positive pp differences were considered disparities, where the proportion of non-Hispanic White residents vaccinated was greater than the proportion of Hispanic residents vaccinated. States and HRRs with ≥100 resident-seasons per age-sex stratum per racial/ethnic group were included in analyses.

RESULTS

Among 7 442 241 short-stay resident-seasons (94.1% non-Hispanic White, 5.9% Hispanic), the median standardized disparities in influenza vaccination were 4.3 pp (minimum, maximum: 0.3, 19.2; n = 22 states) and 2.8 pp (minimum, maximum: -3.6, 10.3; n = 49 HRRs). Among 6 758 616 long-stay resident-seasons (93.7% non-Hispanic White, 6.5% Hispanic), the median standardized differences were -0.1 pp (minimum, maximum: -4.1, 11.4; n = 18 states) and -1.8 pp (minimum, maximum: -6.5, 7.6; n = 34 HRRs).

CONCLUSIONS

Wide geographic variation in influenza vaccination disparities existed across US states and HRRs. Localized interventions targeted toward areas with high disparities may be a more effective strategy to promote health equity than one-size-fits-all national interventions.

Longitudinal effects of lipid indices on incident cardiovascular diseases adjusting for time-varying confounding using marginal structural models: 25 years follow-up of two US cohort studies

Background

This study assesses the effect of blood lipid indices and lipid ratios on cardiovascular diseases (CVDs) using inverse probability-of-exposure weighted estimation of marginal structural models (MSMs).

Methods

A pooled dataset of two US representative cohort studies, including 16736 participants aged 42-84 years with complete information at baseline, was used. The effect of each lipid index, including low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), triglyceride (TG), ratios of TC/HDL-C, LDL-C/HDL-C, and TG/HDL-C on coronary heart disease (CHD) and stroke were estimated using weighted Cox regression.

Results

There were 1638 cases of CHD and 1017 cases of stroke during a median follow-up of 17.1 years (interquartile range: 8.5 to 25.7). Compared to optimal levels, the risk of CVD outcomes increased substantially in high levels of TC, LDL-C, TC/HDL-C, and LDL-C/HDL-C. If everyone had always had high levels of TC (≥240 mg/dL), risk of CHD would have been 2.15 times higher, and risk of stroke 1.35 times higher than if they had always had optimal levels (<200 mg/dL). Moreover, if all participants had been kept at very high (≥190 mg/dL) levels of LDL-C, risk of CHD would have been 2.62 times higher and risk of stroke would have been 1.92 times higher than if all participants had been kept at optimal levels, respectively. Our results suggest that high levels of HDL-C may be protective for CHD, but not for stroke. There was also no evidence of an adverse effect of high triglyceride levels on stroke.

Conclusions

Using MSM, this study highlights the effect of TC and LDL-C on CVD, with a stronger effect on CHD than on stroke. There was no evidence for a protective effect of high levels of HDL-C on stroke. Besides, triglyceride was not found to affect stroke.

Bridging disconnected networks of first and second lines of biologic therapies in rheumatoid arthritis with registry data: bayesian evidence synthesis with target trial emulation

OBJECTIVES

We aim to use real-world data in evidence synthesis to optimize an evidence base for the effectiveness of biologic therapies in rheumatoid arthritis to allow for evidence on first-line therapies to inform second-line effectiveness estimates.

STUDY DESIGN AND SETTING

We use data from the British Society for Rheumatology Biologics Register for Rheumatoid Arthritis to supplement randomized controlled trials evidence obtained from the literature, by emulating target trials of treatment sequences to estimate treatment effects in each line of therapy. Treatment effects estimates from the target trials inform a bivariate network meta-analysis (NMA) of first-line and second-line treatments.

RESULTS

Summary data were obtained from 21 trials of biologic therapies including two for second-line treatment and results from six emulated target trials of both treatment lines. Bivariate NMA resulted in a decrease in uncertainty around the effectiveness estimates of the second-line therapies, when compared to the results of univariate NMA, and allowed for predictions of treatment effects not evaluated in second-line randomized controlled trials.

CONCLUSION

Bivariate NMA provides effectiveness estimates for all treatments in first and second line, including predicted effects in second line where these estimates did not exist in the data. This novel methodology may have further applications; for example, for bridging networks of trials in children and adults.

Interaction Contrasts and Collider Bias

Previous papers have mentioned that conditioning on a binary collider would introduce an association between its causes in at least 1 stratum. In this paper, we prove this statement and, along with intuitions, formally examine the direction and magnitude of the associations between 2 risk factors of a binary collider using interaction contrasts. Among level one of the collider, 2 variables are independent, positively associated, and negatively associated if multiplicative risk interaction contrast is equal to, more than, and less than 0, respectively; the same results hold for the other level of the collider if the multiplicative survival interaction contrast, equal to multiplicative risk interaction contrast minus the additive risk interaction contrast, is compared with 0. The strength of the association depends on the magnitude of the interaction contrast: The stronger the interaction is, the larger the magnitude of the association will be. However, the common conditional odds ratio under the homogeneity assumption will be bounded. A figure is presented that succinctly illustrates our results and helps researchers to better visualize the associations introduced upon conditioning on a collider.

Longitudinal causal effect of modified creatinine index on all-cause mortality in patients with end-stage renal disease: Accounting for time-varying confounders using G-estimation

BACKGROUND

Standard regression modeling may cause biased effect estimates in the presence of time-varying confounders affected by prior exposure. This study aimed to quantify the relationship between declining in modified creatinine index (MCI), as a surrogate marker of lean body mass, and mortality among end stage renal disease (ESRD) patients using G-estimation accounting appropriately for time-varying confounders.

METHODS

A retrospective cohort of all registered ESRD patients (n = 553) was constructed over 8 years from 2011 to 2019, from 3 hemodialysis centers at Kerman, southeast of Iran. According to changes in MCI, patients were dichotomized to either the decline group or no-decline group. Subsequently the effect of interest was estimated using G-estimation and compared with accelerated failure time (AFT) Weibull models using two modelling strategies.

RESULTS

Standard models demonstrated survival time ratios of 0.91 (95% confidence interval [95% CI]: 0.64 to 1.28) and 0.84 (95% CI: 0.58 to 1.23) in patients in the decline MCI group compared to those in no-decline MCI group. This effect was demonstrated to be 0.57 (-95% CI: 0.21 to 0.81) using G-estimation.

CONCLUSION

Declining in MCI increases mortality in patients with ESRD using G-estimation, while the AFT standard models yield biased effect estimate toward the null.