Propensity scores based methods for estimating average treatment effect and average treatment effect among treated: A comparative study

Early childcare and developmental delay risk at 3.5 years: Insights from the French ELFE cohort

We tested the association between early childcare attendance in the first three years of life and child development at age 3.5 years in the French context, where early childcare is subsidized. In the ELFE (Étude Longitudinale Français depuis l'Enfance) birth cohort study set in metropolitan France, children's development was reported by parents at age 3.5 years (n = 11,033) via the Child Development Inventory (CDI) questionnaire. CDI scores were transformed into a development quotient (DQ), with a DQ < 90 corresponding to possible and a DQ < 85 corresponding to a probable developmental delay. Inverse probability weighted multivariable regression models were used to analyse whether early childcare in the first three years of life (centre-based, childminder, informal or parental care) was associated to development delay. Compared to children in exclusive parental care, those in centre-based childcare (CBC) or with a childminder prior to school entry were significantly less likely to experience possible (OR = 0.56, [95% CI = 0.51-0.61] for CBC and OR = 0.77, [95% CI = 0.72-0.83] for childminder attendance) and probable developmental delay (OR = 0.62, [0.58-0.67] for CBC and OR = 0.80 [0.76-0.83] for childminder). Informal childcare attendance was not significantly associated with children's possible nor probable developmental delay ((OR = 0.97, [0.84-1.12]) and (OR = 0.97, [0.82-1.15]), respectively). Conclusions: Overall, our findings add to the existing scientific literature, showing that in the French context, where childcare can start as early as 3 months of age, early childcare attendance can contribute to child's development. What's Known on This Subject: • Studies on early childcare attendance and child development have shown mixed results, associations with better psychomotor development mainly being observed in Nordic countries, while some studies in other countries such as the USA showed no or negative associations. What This Study Adds: • In a country with broad and subsidized access to childcare such as France, access to early childhood education can positively contribute to children's psychomotor development. However, we found that access to childcare does not appear to reduce social inequalities in children's psychomotor development.

Estimating the benefit of esketamine nasal spray versus real-world treatment on patient-reported functional remission: results from the ICEBERG study

Introduction

Treatment resistant depression (TRD) affects approximately 10-30% of patients with major depressive disorder, and most patients with TRD do not respond to real-world treatments (RWT). Treatment with esketamine nasal spray (NS) plus a selective serotonin or serotonin norepinephrine reuptake inhibitor (SSRI/SNRI) has significant long-term clinical benefit over RWT in patients with TRD. However, the impact on patient-reported function remains to be determined.

Methods

The ICEBERG analysis was an indirect treatment comparison performed using data from two studies of patients with TRD: SUSTAIN-2 (esketamine NS; NCT02497287) and the European Observational TRD Cohort (EOTC; RWT; NCT03373253; clinicaltrials.gov). Here, patient-reported functional remission, assessed using the Sheehan Disability Scale (SDS), was defined as SDS ≤6 at Month 6. Analyses were conducted using propensity score re-weighting and multivariable models based on 18 covariates.

Results

At Month 6, the probability of functional remission in esketamine NS-treated patients from SUSTAIN-2 (n=512) was 25.6% (95% confidence interval [CI] 21.8-29.4), while the adjusted probability for RWT patients from the EOTC (n=184) was 11.5% (95% CI 6.9-16.1; relative risk: 2.226 [95% CI 1.451-3.416]; p=0.0003). In the total combined population (N=696), patients who did not achieve clinical response or remission had a low probability of achieving functional remission (5.84% and 8.76%, respectively). However, for patients who did achieve clinical response or remission, the probability of achieving functional remission was greater (43.38% and 54.15%, respectively), although many still did not achieve this status.

Conclusions

For patients with TRD, esketamine NS had a significant functional benefit versus RWT after 6 months of treatment. Irrespective of treatment, achievement of clinical response or remission was insufficient to attain functional remission. Nevertheless, clinical remission increased the likelihood of achieving functional remission, further supporting an important role for clinical remission in for the path towards functional recovery.

Hair cortisol concentrations across pregnancy and maternal postpartum depressive symptoms - The ELFE cohort

Postpartum depression and depressive symptoms have a major impact on maternal and infant health and well-being, yet to date their aetiology remains unclear. One hypothesis suggests a link between these symptoms and variations in prenatal cortisol levels, but existing evidence is limited and inconclusive. This study aims to provide additional evidence to disentangle the relationship between prenatal cortisol concentrations and subsequent occurrence of postpartum depressive symptoms. Cortisol for all three trimesters of pregnancy was extracted from the hair of 775 women participating in the French ELFE cohort. Depressive symptomatology at two months postpartum was assessed through the Edinburgh Postpartum Depression Scale (EPDS). Associations between prenatal cortisol levels and EPDS scores were tested using propensity-score weighted logistic regression models to control for confounders. An increase in mean cortisol concentrations was observed from the first to the third trimester of pregnancy. No significant differences in hair cortisol concentrations were found during the first and second trimesters between women who experienced postpartum depressive symptoms and those who did not. However, an association was observed between third trimester hair cortisol concentrations and depressive symptoms at two months postpartum. Women whose cortisol concentrations fell within the second quartile had a higher risk of subsequent PPDS (aOR = 2.67, 95%CI [1.01, 7.08]). Using a large sample from the general population, we observed an association between hair cortisol levels during the third trimester of pregnancy and postpartum depressive symptoms. Nevertheless, our results suggest that future studies could benefit from investigating other biomarkers of the reactivity of the corticotropic axis.

Data Science Methods for Real-World Evidence Generation in Real-World Data

In the healthcare landscape, data science (DS) methods have emerged as indispensable tools to harness real-world data (RWD) from various data sources such as electronic health records, claim and registry data, and data gathered from digital health technologies. Real-world evidence (RWE) generated from RWD empowers researchers, clinicians, and policymakers with a more comprehensive understanding of real-world patient outcomes. Nevertheless, persistent challenges in RWD (e.g., messiness, voluminousness, heterogeneity, multimodality) and a growing awareness of the need for trustworthy and reliable RWE demand innovative, robust, and valid DS methods for analyzing RWD. In this article, I review some common current DS methods for extracting RWE and valuable insights from complex and diverse RWD. This article encompasses the entire RWE-generation pipeline, from study design with RWD to data preprocessing, exploratory analysis, methods for analyzing RWD, and trustworthiness and reliability guarantees, along with data ethics considerations and open-source tools. This review, tailored for an audience that may not be experts in DS, aspires to offer a systematic review of DS methods and assists readers in selecting suitable DS methods and enhancing the process of RWE generation for addressing their specific challenges.

High-dimensional mediation analysis for continuous outcome with confounders using overlap weighting method in observational epigenetic study

BACKGROUND

Mediation analysis is a powerful tool to identify factors mediating the causal pathway of exposure to health outcomes. Mediation analysis has been extended to study a large number of potential mediators in high-dimensional data settings. The presence of confounding in observational studies is inevitable. Hence, it's an essential part of high-dimensional mediation analysis (HDMA) to adjust for the potential confounders. Although the propensity score (PS) related method such as propensity score regression adjustment (PSR) and inverse probability weighting (IPW) has been proposed to tackle this problem, the characteristics with extreme propensity score distribution of the PS-based method would result in the biased estimation.

METHODS

In this article, we integrated the overlapping weighting (OW) technique into HDMA workflow and proposed a concise and powerful high-dimensional mediation analysis procedure consisting of OW confounding adjustment, sure independence screening (SIS), de-biased Lasso penalization, and joint-significance testing underlying the mixture null distribution. We compared the proposed method with the existing method consisting of PS-based confounding adjustment, SIS, minimax concave penalty (MCP) variable selection, and classical joint-significance testing.

RESULTS

Simulation studies demonstrate the proposed procedure has the best performance in mediator selection and estimation. The proposed procedure yielded the highest true positive rate, acceptable false discovery proportion level, and lower mean square error. In the empirical study based on the GSE117859 dataset in the Gene Expression Omnibus database using the proposed method, we found that smoking history may lead to the estimated natural killer (NK) cell level reduction through the mediation effect of some methylation markers, mainly including methylation sites cg13917614 in CNP gene and cg16893868 in LILRA2 gene.

CONCLUSIONS

The proposed method has higher power, sufficient false discovery rate control, and precise mediation effect estimation. Meanwhile, it is feasible to be implemented with the presence of confounders. Hence, our method is worth considering in HDMA studies.

Association between spinal manipulative therapy and lumbar spine reoperation after discectomy: a retrospective cohort study

BACKGROUND

Patients who undergo lumbar discectomy may experience ongoing lumbosacral radiculopathy (LSR) and seek spinal manipulative therapy (SMT) to manage these symptoms. We hypothesized that adults receiving SMT for LSR at least one year following lumbar discectomy would be less likely to undergo lumbar spine reoperation compared to matched controls not receiving SMT, over two years' follow-up.

METHODS

We searched a United States network of health records (TriNetX, Inc.) for adults aged ≥ 18 years with LSR and lumbar discectomy ≥ 1 year previous, without lumbar fusion or instrumentation, from 2003 to 2023. We divided patients into two cohorts: (1) chiropractic SMT, and (2) usual care without chiropractic SMT. We used propensity matching to adjust for confounding variables associated with lumbar spine reoperation (e.g., age, body mass index, nicotine dependence), calculated risk ratios (RR), with 95% confidence intervals (CIs), and explored cumulative incidence of reoperation and the number of SMT follow-up visits.

RESULTS

Following propensity matching there were 378 patients per cohort (mean age 61 years). Lumbar spine reoperation was less frequent in the SMT cohort compared to the usual care cohort (SMT: 7%; usual care: 13%), yielding an RR (95% CIs) of 0.55 (0.35-0.85; P = 0.0062). In the SMT cohort, 72% of patients had ≥ 1 follow-up SMT visit (median = 6).

CONCLUSIONS

This study found that adults experiencing LSR at least one year after lumbar discectomy who received SMT were less likely to undergo lumbar spine reoperation compared to matched controls not receiving SMT. While these findings hold promise for clinical implications, they should be corroborated by a prospective study including measures of pain, disability, and safety to confirm their relevance. We cannot exclude the possibility that our results stem from a generalized effect of engaging with a non-surgical clinician, a factor that may extend to related contexts such as physical therapy or acupuncture.

REGISTRATION

Open Science Framework ( https://osf.io/vgrwz ).

Economics of widowhood mortality in adult women in India

The economic consequence of widowhood on health is well-established, demonstrating that economic factors can significantly link with health outcomes, even the risk of mortality for widows. However, empirical evidence is restricted only to developed countries. Thus, this study assesses the role of economic factors (paid work, pension and household economic status) on the mortality of widows in broad age groups in India. We used two waves of the India Human Development Survey (IHDS), a nationally representative prospective dataset in India for 42,009 women (married and widows) aged 25 years and above at IHDS wave 1 whose survival status was observed between two waves. Further, 6,953 widows were considered for sub-sample analysis in this study. Logistic regression and propensity score matching (PSM) were applied to understand the association and causality between economic factors and mortality for widows. Poor household economic status, paid regular work, and receiving a widowed pension were significantly associated with lower mortality risk for young widows. In comparison, unpaid and paid regular work was linked with lower mortality risk for old widows. The result of causal analysis suggests that receiving a widows' pension had a slight impact on mortality reduction for young widows while engaging in paid regular work significantly reduced the mortality of old widows. This research confirms that the link between economic factors and mortality among widows is age dependent in the Indian context.

Directed Acyclic Graph Assisted Method For Estimating Average Treatment Effect

Observational data, such as electronic clinical records and claims data, can prove invaluable for evaluating the Average Treatment Effect (ATE) and supporting decision-making, provided they are employed correctly. The Inverse Probability of Treatment Weighting (IPTW) method, based on propensity scores, has demonstrated remarkable efficacy in estimating ATE, assuming that the assumptions of exchangeability, consistency, and positivity are met. Directed Acyclic Graphs (DAGs) offer a practical approach to assess the exchangeability assumption, which asserts that treatment assignment and potential outcomes are independent given a set of confounding variables that block all backdoor paths from treatment assignment to potential outcomes. To ensure a consistent ATE estimator, one can adjust for a minimally sufficient adjustment set of confounding variables that block all backdoor paths from treatment assignment to the outcome. To enhance the efficiency of ATE estimators, our proposal involves incorporating both the minimally sufficient adjustment set of confounding variables and predictors into the propensity score model. Extensive simulations were conducted to evaluate the performance of propensity score-based IPTW methods in estimating ATE when different sets of covariates were included in the propensity score models. The simulation results underscored the significance of including the minimally sufficient adjustment set of confounding variables along with predictors in the propensity score models to obtain a consistent and efficient ATE estimator. We applied this proposed method to investigate whether tracheostomy was causally associated with in-hospital infant mortality, utilizing the 2016 Healthcare Cost and Utilization Project Kids' Inpatient Database. The estimated ATE was found to be approximately 2.30%-2.46% with p-value >0.05.

Dementia risk in patients with type 2 diabetes: Comparing metformin with no pharmacological treatment

INTRODUCTION

Metformin has been suggested as a therapeutic agent for dementia, but the relevant evidence has been partial and inconsistent.

METHODS

We established a national cohort of 210,237 type 2 diabetes patients in the UK Clinical Practice Research Datalink. Risks of incident dementia were compared between metformin initiators and those who were not prescribed any anti-diabetes medication during follow-up.

RESULTS

Compared with metformin initiators (n = 114,628), patients who received no anti-diabetes medication (n = 95,609) had lower HbA1c and better cardiovascular health at baseline. Both Cox regression and propensity score weighting analysis showed metformin initiators had lower risk of dementia compared to those non-users (adjusted hazard ratio = 0.88 [95% confidence interval: 0.84-0.92] and 0.90 [0.84-0.96]). Patients on long-term metformin treatment had an even lower risk of dementia.

DISCUSSION

Metformin may act beyond its glycemic effect and reduce dementia risk to an even lower level than that of patients with milder diabetes and better health profiles.

HIGHLIGHTS

Metformin initiators had a significantly lower risk of dementia compared with patients not receiving anti-diabetes medication. Compared with metformin initiators, diabetes patients not receiving pharmacological treatment had better glycemic profiles at baseline and during follow-up. Patients on long-term metformin treatment had an even lower risk of subsequent dementia incidence. Metformin may act beyond its effect on hyperglycemia and has the potential of being repurposed for dementia prevention.

The impact of COVID-19 on hospitalization outcomes of patients with acute myocardial infarction in the USA

Background/study objective

The effect of the COVID-19 pandemic affected health care delivery, as it led to variable outcomes in different disease states including cardiovascular diseases. In this study, we evaluated the impact of coexisting COVID-19 on Acute Myocardial Infarction (AMI).

Design/setting

We analyzed discharge records of AMI patients from the National Inpatient Sample (NIS) in the year 2020.

Main outcome measures

Using propensity score matching, we assessed the impact of COVID-19 infection on the in-hospital outcomes of patients presenting with AMI.

Results

There were 1154 patients with concomitant COVID-19 infection and AMI who were matched with 109,990 patients with AMI and without COVID-19. We found that patients with COVID-19 who had AMI were less likely to have dyslipidemia (64.6 % vs. 70.4 %, p < 0.001), peripheral vascular disease (2.4 % vs. 3.8 % p = 0.0017), smoking history (23.5 % vs. 28.2 % p < 0.0001) and hypertension (37.1 % vs. 40.1 % p = 0.004).COVID-19 was associated with higher hospital mortality rates (Adjusted odds ratio aOR: 2.72, CI: 2.23-3.30, p < 0.001), cardiac arrest (aOR: 1.65, 95 % CI: 1.26-2.15, p < 0.001), cardiogenic shock (aOR:1.36,95 % CI: 1.10-1.68, p = 0.004) and respiratory failure (aOR:1.81, 95 % CI: 1.55-2.11 p < 0.001) compared to AMI patients without COVID-19. There was also a significant association between coexisting COVID-19 and longer duration of hospital stay (Adjusted mean differences:1.40, 95 % CI: 1.31-1.59 p < 0.0001) in AMI patients.

Conclusion

COVID-19 infection is associated with worse in-hospital mortality and cardiorespiratory complications in patients with AMI.

An overview of propensity score matching methods for clustered data

Propensity score matching is commonly used in observational studies to control for confounding and estimate the causal effects of a treatment or exposure. Frequently, in observational studies data are clustered, which adds to the complexity of using propensity score techniques. In this article, we give an overview of propensity score matching methods for clustered data, and highlight how propensity score matching can be used to account for not just measured confounders, but also unmeasured cluster level confounders. We also consider using machine learning methods such as generalized boosted models to estimate the propensity score and show that accounting for clustering when using these methods can greatly reduce the performance, particularly when there are a large number of clusters and a small number of subjects per cluster. In order to get around this we highlight scenarios where it may be possible to control for measured covariates using propensity score matching, while using fixed effects regression in the outcome model to control for cluster level covariates. Using simulation studies we compare the performance of different propensity score matching methods for clustered data across a number of different settings. Finally, as an illustrative example we apply propensity score matching methods for clustered data to study the causal effect of aspirin on hearing deterioration using data from the conservation of hearing study.

Causal inference in case of near-violation of positivity: comparison of methods

In causal studies, the near-violation of the positivity may occur by chance, because of sample-to-sample fluctuation despite the theoretical veracity of the positivity assumption in the population. It may mostly happen when the exposure prevalence is low or when the sample size is small. We aimed to compare the robustness of g-computation (GC), inverse probability weighting (IPW), truncated IPW, targeted maximum likelihood estimation (TMLE), and truncated TMLE in this situation, using simulations and one real application. We also tested different extrapolation situations for the sub-group with a positivity violation. The results illustrated that the near-violation of the positivity impacted all methods. We demonstrated the robustness of GC and TMLE-based methods. Truncation helped in limiting the bias in near-violation situations, but at the cost of bias in normal conditions. The application illustrated the variability of the results between the methods and the importance of choosing the most appropriate one. In conclusion, compared to propensity score-based methods, methods based on outcome regression should be preferred when suspecting near-violation of the positivity assumption.

Maize productivity amidst northern rural growth credit programme in Ghana

The need to improve maize production and develop agriculture led to the design and implementation of many flagship programmes in Ghana. Among these programmes is the rural growth credit programme. This paper used current data (2021) from credit-constrained maize farmers in the rural growth credit programme to extend the propensity score matching method to the analysis of credit impacts on farm productivity. The study used a sample of 130 farmers, comprising 65 farmers as a treatment group and 65 farmers as a control group. The findings of this paper indicate that, credit-constrained farmers who have access to the rural growth credit relatively have more productivity than credit-constrained farmers who did not have access to the credit. The paper therefore conclude that, the rural growth credit intervention program did achieved its intended purpose in respect of improving farm productivity in Ghana. It could therefore be deduced that credit interventions programs do have a positive impact on farm productivity in Ghana.

Effects of Social Participation by Middle-Aged and Elderly Residents on the Utilization of Medical Services: Evidence From China

Objectives

Aim to evaluate the effect of social participation on utilization of medical services among middle-aged and elderly residents in China.

Methods

We used data from the 2018 wave of the China Health and Retirement Longitudinal Study. Social participation is classified into three types. Furthermore, to control for confounding factors, our study computed propensity score matching (PSM) to evaluate the effect of social participation on the utilization of medical services.

Result

The result of PSM indicates that social participation significantly positively affects the utilization of outpatient services, the average treatment effect on the treated (ATT = 0.038***) and the utilization of inpatient services (ATT = 0.015**) by middle-aged and elderly residents. Furthermore, the utilization of outpatient health care services was significantly positively associated with leisure activities (ATT = 0.035***), social activities to help others (ATT = 0.031***), and learning activities to gain new knowledge (ATT = 0.034***) among middle-aged and elderly residents. The utilization of inpatient health care was significantly positively associated with leisure activities (ATT = 0.015***) but had no significant association with social deeds that help others and increased new knowledge among middle-aged and elderly residents.

Conclusion

Thus, social participation significantly positively affects healthcare utilization by middle-aged and elderly residents. Hence, the government and society should provide more conveniences and promote social participation among middle-aged and elderly residents.

A tutorial comparing different covariate balancing methods with an application evaluating the causal effects of substance use treatment programs for adolescents

Randomized controlled trials are the gold standard for measuring causal effects. However, they are often not always feasible, and causal treatment effects must be estimated from observational data. Observational studies do not allow robust conclusions about causal relationships unless statistical techniques account for the imbalance of pretreatment confounders across groups and key assumptions hold. Propensity score and balance weighting (PSBW) are useful techniques that aim to reduce the observed imbalances between treatment groups by weighting the groups to look alike on the observed confounders. Notably, there are many methods available to estimate PSBW. However, it is unclear a priori which will achieve the best trade-off between covariate balance and effective sample size for a given application. Moreover, it is critical to assess the validity of key assumptions required for robust estimation of the needed treatment effects, including the overlap and no unmeasured confounding assumptions. We present a step-by-step guide to the use of PSBW for estimation of causal treatment effects that includes steps on how to evaluate overlap before the analysis, obtain estimates of PSBW using multiple methods and select the optimal one, check for covariate balance on multiple metrics, and assess sensitivity of findings (both the estimated treatment effect and statistical significance) to unobserved confounding. We illustrate the key steps using a case study examining the relative effectiveness of substance use treatment programs and provide a user-friendly Shiny application that can implement the proposed steps for any application with binary treatments.

Personalized treatment selection using observational data

Estimating the optimal treatment regime based on individual patient characteristics has been a topic of discussion in many forums. Advanced computational power has added momentum to this discussion over the last two decades and practitioners have been advocating the use of new methods in determining the best treatment. Treatments that are geared toward the 'best' outcome for a patient based on his/her genetic markers and characteristics are of high importance. In this article, we develop an approach to predict the optimal personalized treatment based on observational data. We have used inverse probability of treatment weighted machine learning methods to obtain score functions to predict the optimal treatment. Extensive simulation studies showed that our proposed method has desirable performance in selecting the optimal treatment. We provided a case study to examine the Statin use on cognitive function to illustrate the use of our proposed method.

Adolescent cannabis experimentation and unemployment in young to mid-adulthood: Results from the French TEMPO Cohort study

BACKGROUND

France accounts for one of the highest levels of recreational cannabis use, with almost 40% of youth aged 17 reporting having experimented with cannabis. We investigated the impact of early cannabis experimentation (defined as first-time use ≤ 16 years) on future probability of unemployment in young to mid-adulthood using a longitudinal, community sample over the span of 9 years.

METHODS

Data were obtained from the French TEMPO Cohort study, set up in 2009 among young adults aged 22-25 years old. Participants who reported information on age of cannabis experimentation and employment status in at least one study wave (2009, 2011, 2015 and 2018) were included in the statistical analyses (N = 1487, 61.2% female).

RESULTS

In A-IPW-adjusted analyses, early cannabis experimenters (≤ 16 years) had 1.71 (95% CI: 1.46-2.02) times higher odds of experiencing unemployment compared to late cannabis experimenters (> 16 years) and 2.40 (95% CI: 2.00 - 2.88) times higher odds of experiencing unemployment compared to non-experimenters. Late cannabis experimenters experienced 1.39 (95% CI: 1.17-1.68) times higher odds of being unemployed compared to non-experimenters, and early cannabis experimenters experienced 3.84 (95%CI: 2.73-5.42) times higher odds of experiencing long-term unemployment (defined as unemployed at least twice) compared to non-experimenters.

CONCLUSIONS

Participants who ever used cannabis, especially at or before the age of 16, had higher odds of experiencing unemployment, even when accounting for many psychological, academic and family characteristics which preceded cannabis initiation.

Performance of propensity score matching to estimate causal effects in small samples

Propensity score (PS) matching is a very popular causal estimator usually used to estimate the average treatment effect on the treated (ATT) from observational data. However, opting for this estimator may raise some efficiency issues when the sample size is limited. Therefore, we aimed to evaluate the performance of propensity score matching in this context. We started with a motivating example based on a cohort of 66 children with sickle cell anemia who received either allogeneic bone-marrow transplant or chronic transfusion. We found substantial differences in the ATT estimate according to the model selected for propensity score estimation and subsequent matching. Then, we assessed the performance of the different propensity score matching methods and post-matching analyses to estimate the ATT using a simulation study. Although all selected propensity score matching methods were based of previous recommendations, we found important discrepancies in the estimation of treatment effect between them, underlining the importance of thorough sensitivity analyses when using propensity score matching in the context of small sample sizes.

Influence of chronic hepatitis C infection on the monocyte-to-platelet ratio: data analysis from the National Health and Nutrition Examination Survey (2009-2016)

BACKGROUND

Hepatitis C virus (HCV) causes life-threatening chronic infections. Implementation of novel, economical or widely available screening tools can help detect unidentified cases and facilitate their linkage to care. We investigated the relationship between chronic HCV infection and a potential complete blood count biomarker (the monocyte-to-platelet ratio) in the United States.

METHODS

The analytic dataset was selected from cycle years 2009-2016 of the National Health and Nutrition Examination Survey. Complete case data- with no missingness- was available for n = 5281 observations, one-hundred and twenty-two (n = 122) of which were exposed to chronic HCV. The primary analysis used survey-weighted logistic regression to model the effect of chronic HCV on the monocyte-to-platelet ratio adjusting for demographic and biological confounders in a causal inference framework. Missing data and propensity score methods were respectively performed as a secondary and sensitivity analysis.

RESULTS

In the analytic dataset, outcome data was available for n = 5281 (n = 64,245,530 in the weighted sample) observations of which n = 122 (n = 1,067,882 in the weighted sample) tested nucleic acid positive for HCV. Those exposed to chronic HCV infection in the United States have 3.10 times the odds of a high monocyte-to-platelet ratio than those not exposed (OR = 3.10, [95% CI: 1.55-6.18]).

CONCLUSION

A relationship exists between chronic HCV infection and the monocyte-to-platelet ratio in the general population of the United States. Reversing the direction of this association to predict chronic HCV infection from complete blood counts, could provide an economically feasible and universal screening tool, which would help link patients with care.

Independent action models and prediction of combination treatment effects for response rate, duration of response and tumor size change in oncology drug development

An unprecedented number of new cancer targets are in development, and most are being developed in combination therapies. Early oncology development is strategically challenged in choosing the best combinations to move forward to late stage development. The most common early endpoints to be assessed in such decision-making include objective response rate, duration of response and tumor size change. In this paper, using independent-drug-action and Bliss-drug-independence concepts as a foundation, we introduce simple models to predict combination therapy efficacy for duration of response and tumor size change. These models complement previous publications using the independent action models (Palmer 2017, Schmidt 2020) to predict progression-free survival and objective response rate and serve as new predictive models to understand drug combinations for early endpoints. The models can be applied to predict the combination treatment effect for early endpoints given monotherapy data, or to estimate the possible effect of one monotherapy in the combination if data are available from the combination therapy and the other monotherapy. Such quantitative work facilitates strategic planning and decision making in early stage oncology drug development.

Use of dipeptidyl peptidase-4 inhibitors and prognosis of COVID-19 in hospitalized patients with type 2 diabetes: A propensity score analysis from the CORONADO study

AIM

To investigate the association between routine use of dipeptidyl peptidase-4 (DPP-4) inhibitors and the severity of coronavirus disease 2019 (COVID-19) infection in patient with type 2 diabetes in a large multicentric study.

MATERIALS AND METHODS

This study was a secondary analysis of the CORONADO study on 2449 patients with type 2 diabetes (T2D) hospitalized for COVID-19 in 68 French centres. The composite primary endpoint combined tracheal intubation for mechanical ventilation and death within 7 days of admission. Stabilized weights were computed for patients based on propensity score (DPP-4 inhibitors users vs. non-users) and were used in multivariable logistic regression models to estimate the average treatment effect in the treated as inverse probability of treatment weighting (IPTW).

RESULTS

Five hundred and ninety-six participants were under DPP-4 inhibitors before admission to hospital (24.3%). The primary outcome occurred at similar rates in users and non-users of DPP-4 inhibitors (27.7% vs. 28.6%; p = .68). In propensity analysis, the IPTW-adjusted models showed no significant association between the use of DPP-4 inhibitors and the primary outcome by Day 7 (OR [95% CI]: 0.95 [0.77-1.17]) or Day 28 (OR [95% CI]: 0.96 [0.78-1.17]). Similar neutral findings were found between use of DPP-4 inhibitors and the risk of tracheal intubation and death.

CONCLUSIONS

These data support the safety of DPP-4 inhibitors for diabetes management during the COVID-19 pandemic and they should not be discontinued.

Evaluation of propensity score used in cardiovascular research: a cross-sectional survey and guidance document

BACKGROUND

Propensity score (PS) methods are frequently used in cardiovascular clinical research. Previous evaluations revealed poor reporting of PS methods, however a comprehensive and current evaluation of PS use and reporting is lacking. The objectives of the present survey were to (1) evaluate the quality of PS methods in cardiovascular publications, (2) summarise PS methods and (3) propose key reporting elements for PS publications.

METHODS

A PubMed search for cardiovascular PS articles published between 2010 and 2017 in high-impact general medical (top five by impact factor) and cardiovascular (top three by impact factor) journals was performed. Articles were evaluated for the reporting of PS techniques and methods. Data extraction elements were identified from the PS literature and extraction forms were pilot tested.

RESULTS

Of the 306 PS articles identified, most were published in Journal of the American College of Cardiology (29%; n=88), and Circulation (27%, n=81), followed by European Heart Journal (15%; n=47). PS matching was performed most often, followed by direct adjustment, inverse probability of treatment weighting and stratification. Most studies (77%; n=193) selected variables to include in the PS model a priori. A total of 38% (n=116) of studies did not report standardised mean differences, but instead relied on hypothesis testing. For matching, 92% (n=193) of articles presented the balance of covariates. Overall, interpretations of the effect estimates corresponded to the PS method conducted or described in 49% (n=150) of the reviewed articles.

DISCUSSION

Although PS methods are frequently used in high-impact medical journals, reporting of methodological details has been inconsistent. Improved reporting of PS results is warranted and these proposals should aid both researchers and consumers in the presentation and interpretation of PS methods.

Propensity score specification for optimal estimation of average treatment effect with binary response

Propensity score methods are commonly used in statistical analyses of observational data to reduce the impact of confounding bias in estimations of average treatment effect. While the propensity score is defined as the conditional probability of a subject being in the treatment group given that subject's covariates, the most precise estimation of average treatment effect results from specifying the propensity score as a function of true confounders and predictors only. This property has been demonstrated via simulation in multiple prior research articles. However, we have seen no theoretical explanation as to why this should be so. This paper provides that theoretical proof. Furthermore, this paper presents a method for performing the necessary variable selection by means of elastic net regression, and then estimating the propensity scores so as to obtain optimal estimates of average treatment effect. The proposed method is compared against two other recently introduced methods, outcome-adaptive lasso and covariate balancing propensity score. Extensive simulation analyses are employed to determine the circumstances under which each method appears most effective. We applied the proposed methods to examine the effect of pre-cardiac surgery coagulation indicator on mortality based on a linked dataset from a retrospective review of 1390 patient medical records at Jewish Hospital (Louisville, KY) with the Society of Thoracic Surgeons database.

A quantitative analysis of the effect of continuity of care on 30-day readmission and in-hospital mortality among patients with acute ischemic stroke

BACKGROUND

Continuity of care is a core element of high-quality patient care in a primary care setting and one of a national priority.

OBJECTIVE

To assess and quantify the impact of continuity of care on 30-day readmissions, 30-day inpatient mortality, and hospital length of stay (LOS), among hospitalized patients with acute ischemic stroke disease.

DESIGN AND SUBJECTS

Observational retrospective cohort (n = 356,134) using a 2.75% random sample (n=1,036,753) from the State of Florida Agency for Health Care Administration (AHCA) database from 2006 to 2016.

MEASURES

We assessed continuity of care using an integrated continuity of care CoC score, calculated by merging three standard indices of continuity of care - Bice-Boxerman Continuity of Care Index (COCI), Herfindahl Index (HI), and Usual Provider of Care (UPC) Index via a Principal Component Analysis (PCA). We measured 30-day hospital readmissions, 30-day inpatient mortality, and LOS.

RESULTS

Our analysis revealed that hospital LOS was significantly affected by CoC. The statistically significant average treatment effect (ATEs), expressed in risk difference (RD), ranged between 0.27 [95%CI: (0.07, 0.48)] and 1.0 day [95%CI: (0.57, 1.43)]. A similar trend was observed for 30-day readmission (ATEs ranging from 0.0067 [95%CI: (0.0002, 0.0132) to 0.0071 [95%CI: (0.0005, 0.0136)]), and inpatient mortality (ATEs ranging from 0.0006 [95% confidence interval (CI): (0.0001, 0.0012)] to 0.0007 [95%CI: (0.0001, 0.0012)]).

CONCLUSIONS

Our findings suggest a strong association between continuity of care and clinical outcomes. Continuity of care leads to a reduction in mortality, rehospitalization, and hospital length of stay.

G-computation, propensity score-based methods, and targeted maximum likelihood estimator for causal inference with different covariates sets: a comparative simulation study

Controlling for confounding bias is crucial in causal inference. Distinct methods are currently employed to mitigate the effects of confounding bias. Each requires the introduction of a set of covariates, which remains difficult to choose, especially regarding the different methods. We conduct a simulation study to compare the relative performance results obtained by using four different sets of covariates (those causing the outcome, those causing the treatment allocation, those causing both the outcome and the treatment allocation, and all the covariates) and four methods: g-computation, inverse probability of treatment weighting, full matching and targeted maximum likelihood estimator. Our simulations are in the context of a binary treatment, a binary outcome and baseline confounders. The simulations suggest that considering all the covariates causing the outcome led to the lowest bias and variance, particularly for g-computation. The consideration of all the covariates did not decrease the bias but significantly reduced the power. We apply these methods to two real-world examples that have clinical relevance, thereby illustrating the real-world importance of using these methods. We propose an R package RISCA to encourage the use of g-computation in causal inference.

Cholesterol-lowering effect of statin therapy in a clinical HIV cohort: an application of double propensity score adjustment

PURPOSE

Propensity score matching (PSM) is often used to estimate the average treatment effect among the treated (ATT) using observational data. We demonstrate how the use of "double propensity score adjustment" can reduce residual confounding and avoid bias due to incomplete matching compared with traditional PSM methods.

METHODS

The DC Cohort is an observational clinical HIV cohort in Washington, DC. We compared the mean percent change in non-high-density lipoprotein cholesterol (non-HDL-C) concentration after 3-12 months between participants treated and participants not treated with statin therapy between 2011 and 2018. We conducted traditional PSM procedures (optimal, nearest neighbor, and nearest neighbor caliper matching) and double propensity score adjustment.

RESULTS

Among 202 treated and 1252 untreated participants, the ATT was -14.5% (95% CI: -18.4, -10.6) after optimal matching (202 matched pairs; 15/22 covariates balanced), -14.9% (-18.9, -11.0) after nearest neighbor matching (202 matched pairs; 17/22 covariates balanced), and -12.0% (-16.5, -7.5) after nearest neighbor caliper matching (153 matched pairs; 21/22 covariates balanced). After double propensity score adjustment, the ATT was -13.0% (-16.0, -10.1).

CONCLUSIONS

In PSM analyses, double propensity score adjustment is a readily accessible alternative approach for estimating ATTs when sufficient covariate balance between treatment groups cannot be achieved without excluding treated participants.

Estimation of average treatment effects among multiple treatment groups by using an ensemble approach

In observational studies, generalized propensity score (GPS)-based statistical methods, such as inverse probability weighting (IPW) and doubly robust (DR) method, have been proposed to estimate the average treatment effect (ATE) among multiple treatment groups. In this article, we investigate the GPS-based statistical methods to estimate treatment effects from two aspects. The first aspect of our investigation is to obtain an optimal GPS estimation method among four competing GPS estimation methods by using a rank aggregation approach. We further examine whether the optimal GPS-based IPW and DR methods would improve the performance for estimating ATE. It is well known that the DR method is consistent if either the GPS or the outcome models are correctly specified. The second aspect of our investigation is to examine whether the DR method could be improved if we ensemble outcome models. To that end, bootstrap method and rank aggregation method are used to obtain the ensemble optimal outcome model from several competing outcome models, and the resulting outcome model is incorporated into the DR method, resulting in an ensemble DR (enDR) method. Extensive simulation results indicate that the enDR method provides the best performance in estimating the ATE regardless of the method used for estimating GPS. We illustrate our methods using the MarketScan healthcare insurance claims database to examine the treatment effects among three different bones and substitutes used for spinal fusion surgeries. We draw conclusions based on the estimates from the enDR method coupled with the optimal GPS estimation method.

Hypothermic Oxygenated Machine Perfusion of Liver Grafts from Brain-Dead Donors

Hypothermic oxygenated machine perfusion (HOPE) was introduced in liver transplantation (LT) to mitigate ischemia-reperfusion injury. Available clinical data mainly concern LT with donors after circulatory-determined death, whereas data on brain-dead donors (DBD) are scarce. To assess the impact of end-ischemic HOPE in DBD LT, data on primary adult LTs performed between March 2016 and June 2018 were analyzed. HOPE was used in selected cases of donor age >80 years, apparent severe graft steatosis, or ischemia time ≥10 hours. Outcomes of HOPE-treated cases were compared with those after static cold storage. Propensity score matching (1:2) and Bayesian model averaging were used to overcome selection bias. During the study period, 25 (8.5%) out of 294 grafts were treated with HOPE. After matching, HOPE was associated with a lower severe post-reperfusion syndrome (PRS) rate (4% versus 20%, p = 0.13) and stage 2–3 acute kidney injury (AKI) (16% versus 42%, p = 0.046). Furthermore, Bayesian model averaging showed lower transaminases peak and a lower early allograft dysfunction (EAD) rate after HOPE. A steeper decline in arterial graft resistance throughout perfusion was associated with lower EAD rate. HOPE determines a significant reduction of ischemia reperfusion injury in DBD LT.