Estimate Treatment Effects Research Articles

Comparative safety and efficacy of cognitive enhancers for Alzheimer’s dementia: a systematic review with individual patient data network meta-analysis

ObjectiveTo examine the comparative efficacy and safety of cognitive enhancers by patient characteristics for managing Alzheimer’s dementia (AD).DesignSystematic review and individual patient data (IPD) network meta-analysis (NMA) based on our...

Does the Adoption of Climate-Smart Agricultural Practices Impact Farmers' Income? Evidence from Ghana.

People’s lives, particularly farmers’, have been affected by extreme weather conditions that have reduced the yield of numerous crops due to climate change. Climate-smart agriculture practices can reduce or eliminate greenhouse gas emissions and have the propensity to increase farm income and productivity. Therefore, the purpose of this study is to ascertain whether CSA practices impact farmers’ income. This study includes all cocoa farmers in the selected districts in the Ashanti Region. The population includes those who live in the six cocoa production villages. The multistage sampling procedure was considered based on the dominants of literature. The study used an endogenous switching regression framework to examine the effects of the adoption of climate-smart agricultural practices (CSAPs) on farmers’ income. While estimating treatment effects, telasso uses lasso techniques to select the appropriate variable sets. The results revealed that gender, farm experience, age, household size, and farm size do not significantly influence the adoption of irrigation and crop insurance. The study revealed a significant positive impact of access to credit on adopting irrigation and crop insurance. The adoption of climate-smart practices has a positive coefficient. This indicates that if all respondents in each region adopts these practices, their income would increase significantly. This study shows that adopting irrigation practices leads to an increase in household income of 8.6% and 11.1%, respectively, for cocoa farmers. Crop insurance has a positive coefficient and is statistically significant on household income, on-farm, and off-farm. This paper shows that climate-smart practices such as crop insurance can positively influence farmers’ income in Ghana. We also conjecture that crop insurance is the most effective and efficient climate-smart practice among the various agricultural practices. The study suggests that access to credit and mass awareness should be compulsory modules coupled with the consistent training of farmers on new technologies for effective policy implementation. Expanding access to extension officers could enhance farmers’ adaptive capacity and warrant the efficiency of implemented practices.

Personalization of renal replacement therapy initiation: a secondary analysis of the AKIKI and IDEAL-ICU trials

BackgroundTrials comparing early and delayed strategies of renal replacement therapy in patients with severe acute kidney injury may have missed differences in survival as a result of mixing together patients at heterogeneous levels of risks. Our aim was to evaluate the heterogeneity of treatment effect on 60-day mortality from an early vs a delayed strategy across levels of risk for renal replacement therapy initiation under a delayed strategy.MethodsWe used data from the AKIKI, and IDEAL-ICU randomized controlled trials to develop a multivariable logistic regression model for renal replacement therapy initiation within 48 h after allocation to a delayed strategy. We then used an interaction with spline terms in a Cox model to estimate treatment effects across the predicted risks of RRT initiation.ResultsWe analyzed data from 1107 patients (619 and 488 in the AKIKI and IDEAL-ICU trial respectively). In the pooled sample, we found evidence for heterogeneous treatment effects (P = 0.023). Patients at an intermediate-high risk of renal replacement therapy initiation within 48 h may have benefited from an early strategy (absolute risk difference, − 14%; 95% confidence interval, − 27% to − 1%). For other patients, we found no evidence of benefit from an early strategy of renal replacement therapy initiation but a trend for harm (absolute risk difference, 8%; 95% confidence interval, − 5% to 21% in patients at intermediate-low risk).ConclusionsWe have identified a clinically sound heterogeneity of treatment effect of an early vs a delayed strategy of renal replacement therapy initiation that may reflect varying degrees of kidney demand-capacity mismatch.

Being Transparent About Brilliant Failures: An Attempt to Use Real-World Data in a Disease Model for Patients with Castration-Resistant Prostate Cancer.

BackgroundReal-world disease models spanning multiple treatment lines can provide insight into the (cost) effectiveness of treatment sequences in clinical practice.ObjectiveOur objective was to explore whether a disease model based solely on real-world data (RWD) could be used to estimate the effectiveness of treatments for patients with castration-resistant prostate cancer (CRPC) that could then be suitably used in a cost-effectiveness analysis.MethodsWe developed a patient-level simulation model using patient-level data from the Dutch CAPRI registry as input parameters. Time to event (TTE) and overall survival (OS) were estimated with multivariate regression models, and type of event (i.e., next treatment or death) was estimated with multivariate logistic regression models. To test internal validity, TTE and OS from the simulation model were compared with the observed outcomes in the registry.ResultsAlthough patient characteristics and survival outcomes of the simulated data were comparable to those in the observed data (median OS 20.6 vs. 19.8 months, respectively), the disease model was less accurate in estimating differences between treatments (median OS simulated vs. observed population: 18.6 vs. 17.9 [abiraterone acetate plus prednisone], 24.0 vs. 25.0 [enzalutamide], 20.2 vs. 18.7 [docetaxel], and 20.0 vs. 23.8 months [radium-223]).ConclusionsOverall, the disease model accurately approximated the observed data in the total CRPC population. However, the disease model was unable to predict differences in survival between treatments due to unobserved differences. Therefore, the model is not suitable for cost-effectiveness analysis of CRPC treatment. Using a combination of RWD and data from randomised controlled trials to estimate treatment effectiveness may improve the model.Supplementary InformationThe online version contains supplementary material available at 10.1007/s40801-022-00294-7.

Effects of Motherwort Injection Versus Intramuscular Oxytocin for Preventing Postpartum Hemorrhage Among Women Who Underwent Cesarean Section.

Objectives: Subject to ethical constraints, real-world data are an important resource for evaluating treatment effects of medication use during pregnancy and the postpartum period. This study investigated whether motherwort injection, a traditional Chinese medicine preparation, was more effective than intramuscular (IM) oxytocin for preventing postpartum hemorrhage (PPH) in a real-world setting when intravenous (IV) oxytocin is administered. Methods: We conducted an active-controlled, propensity-score matched cohort study using an established pregnancy registry database. Women who underwent cesarean section and received IV oxytocin at the third stage of labor were included. We used an active-comparator design to minimize indication bias, in which we compared IM motherwort injection in the uterus versus IM oxytocin, both on top of IV oxytocin use. We applied 1:1 propensity-score matching (PSM) to balance patient baseline characteristics and used a logistic regression model to estimate treatment effect (i.e., risk difference (RD) and odds ratio (OR)) by using the counterfactual framework. The outcomes of interest were blood loss over 500 ml within 2 h after delivery (PPH, primary) and blood loss over 1,000 ml (severe PPH, secondary). We conducted four sensitivity analyses to examine the robustness of the results. Results: A total of 22,519 pregnant women underwent cesarean sections, among which 4,081 (18.12%) PPH and 480 (2.13%) severe PPH occurred. Among included women, 586 (2.60%) were administrated with IM motherwort injection, and 21,933 (97.40%) used IM oxytocin. After PSM, patient baseline characteristics were well balanced. Compared with IM oxytocin, the use of IM motherwort injection was associated with significantly lower risk of PPH (RD −25.26%, 95% CI −30.04% to −20.47%, p < 0.001; OR 0.25, 95% CI 0.18 to 0.32, p < 0.001) and severe PPH (RD −3.58%, 95% CI −5.87% to −1.30%, p < 0.001; OR 0.39, 95% CI 0.20 to 0.71, p < 0.002). Sensitivity analyses showed that the results were similar. Conclusion: With the use of data from a real-world setting, the findings consistently showed that among women undergoing cesarean section who had received IV oxytocin, the additional use of IM motherwort injection could achieve a lower risk of PPH as compared to the additional use of IM oxytocin. Our study suggested a paradigm for investigating the treatment effect of Chinese herbal medicine in the real-world practice setting.

Random effect misspecification in stepped wedge designs.

Stepped wedge cluster randomized trials are often analysed using linear mixed effects models that may include random effects for cluster, time and/or treatment. We investigate the impact of misspecification of the random effects structure of the model. Specifically, we considered two cases of misspecification of the random effects in a cross-sectional stepped wedge cluster randomized trials model - fit a linear mixed effects model with random time effects but the true model includes random treatment effects (case 1) or fit a linear mixed effects model with random treatment effect but the true model includes random time effects (case 2) - and derived the variance of the estimated treatment effect under misspecification. We defined two measures of the effect of misspecification: validity and efficiency. Validity is the ratio of the model-based variance of the treatment effect from the mis-specified model divided by the true variance of the treatment effect from the mis-specified model (based on a sandwich estimate of the variance). Efficiency is the ratio of the model-based variance of the treatment effect from the correctly specified model divided by the true variance of the treatment effect from the mis-specified model. We found that validity is less than 1.0 (anti-conservative) in almost all situations investigated with the exception of case 1 with two sequences, when validity could be greater than 1.0. Efficiency is less than 1 in all cases and depends on the intracluster correlation coefficient, the relative magnitude of the variance of the misclassified variance component, and the number of sequences. In general, there is no universal recommendation as to the most robust approach except for the case of a classic stepped wedge cluster randomized trial with only 2 sequences, where fitting a random time model is less likely to lead to anti-conservative inference compared with fitting a random intervention model.

Potential Effects of Elimination of the Black Race Coefficient in eGFR Calculations in the CREDENCE Trial.

The effect of including race in the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation on screening, recruitment, and outcomes of clinical trials is unclear. The inclusion and outcomes of participants in the Canagliflozin and Renal Events in Diabetes with Established Nephropathy Clinical Evaluation (CREDENCE) trial, which randomized individuals with type 2 diabetes and CKD to canagliflozin or placebo, were evaluated after calculating eGFR using the 2009 CKD-EPI creatinine equation with and without a race-specific coefficient or the 2021 CKD-EPI creatinine equation. Treatment effects were estimated using proportional hazards models and piecewise linear mixed effects models for eGFR slope. Of 4401 randomized participants, 2931 (67%) were White participants, 224 (5%) were Black participants, 877 (20%) were Asian participants, and 369 (8%) participants were other race. Among randomized participants, recalculation of screening eGFR using the 2009 equation without a race-specific coefficient had no effect on the likelihood of non-Black participants meeting inclusion criteria but would have excluded 22 (10%) randomized Black participants for eGFR<30 ml/min per 1.73 m2. Recalculation with the 2021 equation would have excluded eight (4%) Black participants for low eGFR and one (0.4%) Black participant for eGFR≥90 ml/min per 1.73 m2, whereas 30 (0.7%) and 300 (7%) non-Black participants would have been excluded for low and high eGFR, respectively. A high proportion (eight of 22; 36%) of end points in Black participants occurred in individuals who would have been excluded following recalculation using the race-free 2009 equation but not when recalculated with the 2021 equation (one of eight; 13%). Cardiovascular and kidney treatment effects remained consistent across eGFR categories following recalculation with either equation. Changes in estimated treatment effects on eGFR slope were modest but were qualitatively larger following recalculation using the 2021 equation. However, the effect of canagliflozin on chronic change in eGFR was attenuated by 7% among Black participants and increased 6% in non-Black participants. In the CREDENCE trial, eGFR recalculation without the race-specific coefficient had small but potentially important effects on event rates and the relative proportion of Black participants without substantially changing efficacy estimates. Canagliflozin and Renal Events in Diabetes with Established Nephropathy Clinical Evaluation (CREDENCE), NCT02065791.

Effects of ACE inhibitors and angiotensin receptor blockers: protocol for a UK cohort study using routinely collected electronic health records with validation against the ONTARGET trial

IntroductionCardiovascular disease is a leading cause of death globally, responsible for nearly 18 million deaths worldwide in 2017. Medications to reduce the risk of cardiovascular events are prescribed based on...

Intestinal permeability before and after albendazole treatment in low and high socioeconomic status schoolchildren in Makassar, Indonesia

Intestinal helminths are highly prevalent in low-SES children and could contribute to poor health outcomes either directly or via alteration of the gut microbiome and gut barrier function. We analysed parasitic infections and gut microbiota composition in 325 children attending high- and low-SES schools in Makassar, Indonesia before and after albendazole treatment. Lactulose/Mannitol Ratio (LMR, a marker of gut permeability); I-FABP (a surrogate marker of intestinal damage) as well as inflammatory markers (LBP) were measured. Helminth infections were highly prevalent (65.6%) in low-SES children. LMR and I-FABP levels were higher in low-SES children (geomean (95%CI): 4.03 (3.67–4.42) vs. 3.22 (2.91–3.57); p. adj < 0.001; and 1.57 (1.42–1.74) vs. 1.25 (1.13–1.38); p. adj = 0.02, respectively) while LBP levels were lower compared to the high-SES (19.39 (17.09–22.01) vs. 22.74 (20.07–26.12); p.adj = 0.01). Albendazole reduced helminth infections in low-SES and also decreased LMR with 11% reduction but only in helminth-uninfected children (estimated treatment effect: 0.89; p.adj = 0.01). Following treatment, I-FABP decreased in high- (0.91, p.adj < 0.001) but increased (1.12, p.adj = 0.004) in low-SES children. Albendazole did not alter the levels of LBP. Microbiota analysis showed no contribution from specific bacterial-taxa to the changes observed. Intestinal permeability and epithelial damage are higher while peripheral blood inflammatory marker is lower in children of low-SES in Indonesia. Furthermore, treatment decreased LMR in helminth-uninfected only.

Distributional regression in clinical trials: treatment effects on parameters other than the mean

BackgroundThe classical linear model is widely used in the analysis of clinical trials with continuous outcomes. However, required model assumptions are frequently not met, resulting in estimates of treatment effect that can be inefficient and biased. In addition, traditional models assess treatment effect only on the mean response, and not on other aspects of the response, such as the variance. Distributional regression modelling overcomes these limitations. The purpose of this paper is to demonstrate its usefulness for the analysis of clinical trials, and superior performance to that of traditional models.MethodsDistributional regression models are demonstrated, and contrasted with normal linear models, on data from the LIPID randomized controlled trial, which compared the effects of pravastatin with placebo in patients with coronary heart disease. Systolic blood pressure (SBP) and the biomarker midregional pro-adrenomedullin (MR-proADM) were analysed. Treatment effect was estimated in models that used response distributions more appropriate than the normal (Box-Cox-t and Johnson’s Su for MR-proADM and SBP, respectively), applied censoring below the detection limit of MR-proADM, estimated treatment effect on distributional parameters other than the mean, and included random effects for longitudinal observations. A simulation study was conducted to compare the performance of distributional regression models with normal linear regression, under conditions mimicking the LIPID study. The R package gamlss (Generalized Additive Models for Location, Scale and Shape), which implements maximum likelihood estimation for distributional regression modelling, was used throughout.ResultsIn all cases the distributional regression models fit the data well, in contrast to poor fits obtained for traditional models; for MR-proADM a small but significant treatment effect on the mean was detected by the distributional regression model and not the normal model; and for SBP a beneficial treatment effect on the variance was demonstrated. In the simulation study distributional models strongly outperformed normal models when the response variable was non-normal and heterogeneous; and there was no disadvantage introduced by the use of distributional regression modelling when the response satisfied the normal linear model assumptions.ConclusionsDistributional regression models are a rich framework, largely untapped in the clinical trials world. We have demonstrated a sample of the capabilities of these models for the analysis of trials. If interest lies in accurate estimation of treatment effect on the mean, or other distributional features such as variance, the use of distributional regression modelling will yield superior estimates to traditional normal models, and is strongly recommended.Trial registrationThe LIPID trial was retrospectively registered on ANZCTR on 27/04/2016, registration number ACTRN12616000535471.

Health technology assessments and real-world evidence: tell us what you want, what you really, really want.

Health technology assessments and real-world evidence: tell us what you want, what you really, really want.

Locally Optimal Design for A/B Tests in the Presence of Covariates and Network Dependence

A/B test, a simple type of controlled experiment, refers to the statistical procedure of experimenting to compare two treatments applied to test subjects. For example, many IT companies frequently conduct A/B tests on their users who are connected and form social networks. Often, the users’ responses could be related to the network connection. In this paper, we assume that the users, or the test subjects of the experiments, are connected on an undirected network, and the responses of two connected users are correlated. We include the treatment assignment, covariate features, and network connection in a conditional autoregressive model. Based on this model, we propose a design criterion that measures the variance of the estimated treatment effect and allocate the treatment settings to the test subjects by minimizing the criterion. Since the design criterion depends on an unknown network correlation parameter, we adopt the locally optimal design method and develop a hybrid optimization approach to obtain the optimal design. Through synthetic and real social network examples, we demonstrate the value of including network dependence in designing A/B experiments and validate that the proposed locally optimal design is robust to the choices of parameters.

Challenges encountered during the systematic review of newer and enhanced influenza vaccines and recommendations for the future.

There are a variety of challenges in the conduct of systematic reviews of influenza vaccines. We describe our experience of completing four systematic reviews of newer and enhanced inactivated seasonal influenza vaccines. The reporting of the included studies created significant challenges for study identification, data extraction and analysis. Those challenges have implications for the resources required to conduct reviews and, more significantly, for the accuracy of the estimated treatment effect. There is a substantial burden of morbidity and mortality associated with seasonal influenza, and the evidence used to support vaccination strategies requires regular review. An improved review process will facilitate robust decision-making both nationally and internationally. We recommend the development of reporting guidelines, increased engagement between researchers and decision makers, a database of identified trials, and research into search optimisation.

Analysing outcome variables with floor effects due to censoring: a simulation study with longitudinal trial data

ackground: Randomised controlled trials (RCTs) are the gold standard to estimate treatment effects. When patients receive effective treatment over time they may reach the limit of a certain measurement scale. This phenomenon is known as censoring and lead to skewed distributions of the outcome variable with an excess of either low (floor effect) or high values (ceiling effect). Applying traditional methods such as linear mixed models to analyse this kind of longitudinal RCT data may result in bias of the regression parameters. To deal with floor effects due to censoring, a tobit mixed model can be used. The objective of this study was to compare the results of longitudinal linear mixed model analyses with longitudinal tobit mixed model analyses.Methods: First, a simulation study was performed in which several situations of RCTs with floor effects were simulated. Second, data from an empirical RCT was analysed with both methods.Results: Although all analyses underestimated the intervention effects, the tobit mixed model performed much better than the linear mixed model in handling floor effects. However, with an increasing number of follow-up measurements in combination with a strong floor effect estimates from the tobit mixed model were also not accurate.Conclusion: tobit mixed model analysis should be used to estimate treatments effects in longitudinal RCTs with floor effects due to censoring.

Treatment of COVID-19-associated ARDS with mesenchymal stromal cells: a multicenter randomized double-blind trial

BackgroundSevere acute respiratory syndrome coronavirus-2 (SARS–CoV-2)-induced acute respiratory distress syndrome (ARDS) causes high mortality. Umbilical cord-derived mesenchymal stromal cells (UC-MSCs) have potentially relevant immune-modulatory properties, whose place in ARDS treatment is not established. This phase 2b trial was undertaken to assess the efficacy of UC-MSCs in patients with SARS–CoV-2-induced ARDS.MethodsThis multicentre, double-blind, randomized, placebo-controlled trial (STROMA–CoV-2) recruited adults (≥ 18 years) with SARS–CoV-2-induced early (< 96 h) mild-to-severe ARDS in 10 French centres. Patients were randomly assigned to receive three intravenous infusions of 106 UC-MSCs/kg or placebo (0.9% NaCl) over 5 days after recruitment. For the modified intention-to-treat population, the primary endpoint was the partial pressure of oxygen to fractional inspired oxygen (PaO2/FiO2)-ratio change between baseline (day (D) 0) and D7.ResultsAmong the 107 patients screened for eligibility from April 6, 2020, to October 29, 2020, 45 were enrolled, randomized and analyzed. PaO2/FiO2 changes between D0 and D7 did not differ significantly between the UC-MSCs and placebo groups (medians [IQR] 54.3 [− 15.5 to 93.3] vs 25.3 [− 33.3 to 104.6], respectively; ANCOVA estimated treatment effect 7.4, 95% CI − 44.7 to 59.7; P = 0.77). Six (28.6%) of the 21 UC-MSCs recipients and six of 24 (25%) placebo-group patients experienced serious adverse events, none of which were related to UC-MSCs treatment.ConclusionsD0-to-D7 PaO2/FiO2 changes for intravenous UC-MSCs-versus placebo-treated adults with SARS–CoV-2-induced ARDS did not differ significantly. Repeated UC-MSCs infusions were not associated with any serious adverse events during treatment or thereafter (until D28). Larger trials enrolling patients earlier during the course of their ARDS are needed to further assess UC-MSCs efficacy in this context.Trial registration: NCT04333368. Registered 01 April 2020, https://clinicaltrials.gov/ct2/history/NCT04333368.

Deep treatment-adaptive network for causal inference

Causal inference is capable of estimating the treatment effect (i.e., the causal effect of treatment on the outcome) to benefit the decision making in various domains. One fundamental challenge in this research is that the treatment assignment bias in observational data. To increase the validity of observational studies on causal inference, representation-based methods as the state-of-the-art have demonstrated the superior performance of treatment effect estimation. Most representation-based methods assume all observed covariates are pre-treatment (i.e., not affected by the treatment) and learn a balanced representation from these observed covariates for estimating treatment effect. Unfortunately, this assumption is often too strict a requirement in practice, as some covariates are changed by doing an intervention on treatment (i.e., post-treatment). By contrast, the balanced representation learned from unchanged covariates thus biases the treatment effect estimation. In light of this, we propose a deep treatment-adaptive architecture (DTANet) that can address the post-treatment covariates and provide a unbiased treatment effect estimation. Generally speaking, the contributions of this work are threefold. First, our theoretical results guarantee DTANet can identify treatment effect from observations. Second, we introduce a novel regularization of orthogonality projection to ensure that the learned confounding representation is invariant and not being contaminated by the treatment, meanwhile mediate variable representation is informative and discriminative for predicting the outcome. Finally, we build on the optimal transport and learn a treatment-invariant representation for the unobserved confounders to alleviate the confounding bias.

Pricing for Heterogeneous Products: Analytics for Ticket Reselling

Problem definition: We present a data-driven study of the secondary ticket market. In particular, we are primarily concerned with accurately estimating price sensitivity for listed tickets. In this setting, there are many issues including endogeneity, heterogeneity in price sensitivity for different tickets, binary outcomes, and nonlinear interactions between ticket features. Our secondary goal is to highlight how this estimation can be integrated into a prescriptive trading strategy for buying and selling tickets in an active marketplace. Academic/practical relevance: We present a novel method for demand estimation with heterogeneous treatment effect in the presence of confounding. In practice, we embed this method within an optimization framework for ticket reselling, providing the ticket reselling platform with a new framework for pricing tickets on its platform. Methodology: We introduce a general double/orthogonalized machine learning method for classification problems. This method allows us to isolate the causal effects of price on the outcome by removing the conditional effects of the ticket and market features. Furthermore, we introduce a novel loss function that can be easily incorporated into powerful, off-the-shelf machine learning algorithms, including gradient boosted trees. We show how, in the presence of hidden confounding variables, instrumental variables can be incorporated. Results: Using a wide range of synthetic data sets, we show this approach beats state-of-the-art machine learning and causal inference approaches for estimating treatment effects in the classification setting. Furthermore, using National Basketball Association ticket listings from the 2014–2015 season, we show that probit models with instrumental variables, previously used for price estimation of tickets in the resale market, are significantly less accurate and potentially misspecified relative to our proposed approach. Through pricing simulations, we show our proposed method can achieve an 11% return on investment by buying and selling tickets, whereas existing techniques are not profitable. Managerial implications: The knowledge of how to price tickets on its platform offers a range of potential opportunities for our collaborator, both in terms of understanding sellers on their platform and in developing new products to offer them.

Bias Reduction through Analysis of Competing Events (BRACE) Correction to Address Cancer Treatment Selection Bias in Observational Data.

Cancer treatments can paradoxically appear to reduce the risk of noncancer mortality in observational studies, due to residual confounding. Here we introduce a method, Bias Reduction through Analysis of Competing Events (BRACE), to reduce bias in the presence of residual confounding. BRACE is a novel method for adjusting for bias from residual confounding in proportional hazards models. Using standard simulation methods, we compared BRACE with Cox proportional hazards regression in the presence of an unmeasured confounder. We examined estimator distributions, bias, mean squared error (MSE), and coverage probability. We then estimated treatment effects of high versus low intensity treatments in 36,630 prostate cancer, 4,069 lung cancer, and 7,117 head/neck cancer patients, using the Veterans Affairs database. We analyzed treatment effects on cancer-specific mortality (CSM), noncancer mortality (NCM), and overall survival (OS), using conventional multivariable Cox and propensity score (adjusted using inverse probability weighting) models, versus BRACE-adjusted estimates. In simulations with residual confounding, BRACE uniformly reduced both bias and MSE. In the absence of bias, BRACE introduced bias toward the null, albeit with lower MSE. BRACE markedly improved coverage probability, but with a tendency toward overcorrection for effective but nontoxic treatments. For each clinical cohort, more intensive treatments were associated with significantly reduced hazards for CSM, NCM, and OS. BRACE attenuated OS estimates, yielding results more consistent with findings from randomized trials and meta-analyses. BRACE reduces bias and MSE when residual confounding is present and represents a novel approach to improve treatment effect estimation in nonrandomized studies.

Statistical Power for Estimating Treatment Effects Using Difference-in-Differences and Comparative Interrupted Time Series Estimators With Variation in Treatment Timing

This article develops new closed-form variance expressions for power analyses for commonly used difference-in-differences (DID) and comparative interrupted time series (CITS) panel data estimators. The main contribution is to incorporate variation in treatment timing into the analysis. The power formulas also account for other key design features that arise in practice: autocorrelated errors, unequal measurement intervals, and clustering due to the unit of treatment assignment. We consider power formulas for both cross-sectional and longitudinal models and allow for covariates. An illustrative power analysis provides guidance on appropriate sample sizes. The key finding is that accounting for treatment timing increases required sample sizes. Further, DID estimators have considerably more power than standard CITS and ITS estimators. An available Shiny R dashboard performs the sample size calculations for the considered estimators.

Recurrent individual treatment assignment: a treatment policy approach to account for heterogeneous treatment effects

Longitudinal randomized controlled trials generally assign individuals randomly to interventions at baseline and then evaluate how differential average treatment effects evolve over time. This study shows that longitudinal settings could benefit from Recurrent Individual Treatment Assignment (RITA) instead, particularly in the face of (dynamic) heterogeneous treatment effects. Focusing on the optimization of treatment assignment, rather than on estimating treatment effects, acknowledges the presence of unobserved heterogeneous treatment effects and improves overall intervention response when compared to intervention policies in longitudinal settings based on Randomized Controlled Trials (RCTs)-derived average treatment effects. This study develops a RITA-algorithm and evaluates its performance in a multi-period simulation setting, considering two alternative interventions and varying the extent of unobserved heterogeneity in individual treatment response. The results show that RITA learns quickly, and adapts individual assignments effectively. If treatment heterogeneity exists, the inherent focus on both exploit and explore enables RITA to outperform a conventional assignment strategy that relies on RCT-derived average treatment effects.