Search

A difficult decision for patients in need of kidney–pancreas transplant is whether to seek a living kidney donor or wait to receive both organs from one deceased donor. The framework of dynamic treatment regimes (DTRs) can inform this choice, but a patient‐relevant strategy such as "wait for deceased‐donor transplant" is ill‐defined because there are multiple versions of treatment (i.e., wait times, organ qualities). Existing DTR methods average over the distribution of treatment versions in the data, estimating survival under a "representative intervention." This is undesirable if transporting inferences to a target population such as patients today, who experience shorter wait times thanks to evolutions in allocation policy. We, therefore, propose the concept of a generalized representative intervention (GRI): a random DTR that assigns treatment version by drawing from the distribution among strategy compliers in the target population (e.g., patients today). We describe an inverse‐probability‐weighted product‐limit estimator of survival under a GRI that performs well in simulations and can be implemented in standard statistical software. For continuous treatments (e.g., organ quality), weights are reformulated to depend on probabilities only, not densities. We apply our method to a national database of kidney–pancreas transplant candidates from 2001–2020 to illustrate that variability in transplant rate across years and centers results in qualitative differences in the optimal strategy for patient survival. [ABSTRACT FROM AUTHOR]

In recent sequential multiple assignment randomized trials, outcomes were assessed multiple times to evaluate longer-term impacts of the dynamic treatment regimes (DTRs). Q-learning requires a scalar response to identify the optimal DTR. Inverse probability weighting may be used to estimate the optimal outcome trajectory, but it is inefficient, susceptible to model mis-specification, and unable to characterize how treatment effects manifest over time. We propose modified Q-learning with generalized estimating equations to address these limitations and apply it to the M-bridge trial, which evaluates adaptive interventions to prevent problematic drinking among college freshmen. Simulation studies demonstrate our proposed method improves efficiency and robustness. [ABSTRACT FROM AUTHOR]

Support vector regression (SVR) is particularly beneficial when the outcome and predictors are nonlinearly related. However, when many covariates are available, the method's flexibility can lead to overfitting and an overall loss in predictive accuracy. To overcome this drawback, we develop a feature selection method for SVR based on a genetic algorithm that iteratively searches across potential subsets of covariates to find those that yield the best performance according to a user-defined fitness function. We evaluate the performance of our feature selection method for SVR, comparing it to alternate methods including LASSO and random forest, in a simulation study. We find that our method yields higher predictive accuracy than SVR without feature selection. Our method outperforms LASSO when the relationship between covariates and outcome is nonlinear. Random forest performs equivalently to our method in some scenarios, but more poorly when covariates are correlated. We apply our method to predict donor kidney function 1 year after transplant using data from the United Network for Organ Sharing national registry. [ABSTRACT FROM AUTHOR]

Cluster analysis is an unsupervised learning strategy that is exceptionally useful for identifying homogeneous subgroups of observations in data sets of unknown structure. However, it is challenging to determine if the identified clusters represent truly distinct subgroups rather than noise. Existing approaches for addressing this problem tend to define clusters based on distributional assumptions, ignore the inherent correlation structure in the data, or are not suited for high‐dimension low‐sample size (HDLSS) settings. In this paper, we propose a novel method to evaluate the significance of identified clusters by comparing the explained variation due to the clustering from the original data to that produced by clustering a unimodal reference distribution that preserves the covariance structure in the data. The reference distribution is generated using kernel density estimation, and thus, does not require that the data follow a particular distribution. By utilizing sparse covariance estimation, the method is adapted for the HDLSS setting. The approach can be used to test the null hypothesis that the data cannot be partitioned into clusters and to determine the optimal number of clusters. Simulation examples, theoretical evaluations, and applications to temporomandibular disorder research and cancer microarray data illustrate the utility of the proposed method. [ABSTRACT FROM AUTHOR]

A number of statistical approaches have been proposed for incorporating supplemental information in randomized clinical trials. Existing methods often compare the marginal treatment effects to evaluate the degree of consistency between sources. Dissimilar marginal treatment effects would either lead to increased bias or down-weighting of the supplemental data. This represents a limitation in the presence of treatment effect heterogeneity, in which case the marginal treatment effect may differ between the sources solely due to differences between the study populations. We introduce the concept of covariate-adjusted exchangeability, in which differences in the marginal treatment effect can be explained by differences in the distributions of the effect modifiers. The potential outcomes framework is used to conceptualize covariate-adjusted and marginal exchangeability. We utilize a linear model and the existing multisource exchangeability models framework to facilitate borrowing when marginal treatment effects are dissimilar but covariate-adjusted exchangeability holds. We investigate the operating characteristics of our method using simulations. We also illustrate our method using data from two clinical trials of very low nicotine content cigarettes. Our method has the ability to incorporate supplemental information in a wider variety of situations than when only marginal exchangeability is considered. [ABSTRACT FROM AUTHOR]

We propose an inverse probability of censoring weighted (IPCW) bagging (bootstrap aggregation) pre‐processing that enables the application of any machine learning procedure for classification to be used to predict the cause‐specific cumulative incidence, properly accounting for right‐censored observations and competing risks. We consider the IPCW area under the time‐dependent ROC curve (IPCW‐AUC) as a performance evaluation metric. We also suggest a procedure to optimally stack predictions from any set of IPCW bagged methods. We illustrate our proposed method in the Swedish InfCareHIV register by predicting individuals for whom treatment will not maintain an undetectable viral load for at least 2 years following initial suppression. The R package stackBagg that implements our proposed method is available on Github. [ABSTRACT FROM AUTHOR]

Recent evidence from randomized clinical trials (RCTs) of very low nicotine content (VLNC) cigarettes indicates that smokers randomized to VLNC cigarettes had significantly lower cigarette use, dependence, and biomarkers of exposure than smokers randomized to normal nicotine content control cigarettes. In these trials, a substantial number of participants did not adhere to their randomized treatment assignment, i.e., they used commercial cigarettes not provided by the trial in place of or in addition to the VLNC cigarettes provided by the trial. As with most RCTs, the analysis of these trials followed the intention-to-treat principle, where participants are analyzed according to their randomized treatment assignment regardless of adherence. Alternately, the analysis of an RCT could focus on the estimation and testing of the causal effect of the intervention, which is the treatment effect if all subjects were to adhere to their randomized treatment assignment. In this commentary, we compare these two approaches, highlighting the important role of causal estimation and inference for evaluating the regulatory effect of a nicotine standard for cigarettes. Additionally, we review the results of the secondary analyses of randomized trials of VLNC cigarettes using causal inference methodology to account for non-adherence to the assigned treatment and discuss the implications for a nicotine standard for cigarettes. [ABSTRACT FROM AUTHOR]

Non-adherence has been well recognized for years to be a common issue that significantly impacts clinical outcomes and health care costs. Medication adherence is remarkably low even in the controlled environment of clinical trials where it has potentially complex major implications. Collection of non-adherence data diverge markedly among cardiovascular randomized trials and, even where collected, is rarely incorporated in the statistical analysis to test the consistency of the primary endpoint(s). The imprecision introduced by the inconsistent assessment of non-adherence in clinical trials might confound the estimate of the calculated efficacy of the study drug. Hence, clinical trials may not accurately answer the scientific question posed by regulators, who seek an accurate estimate of the true efficacy and safety of treatment, or the question posed by payers, who want a reliable estimate of the effectiveness of treatment in the marketplace after approval. The Non-adherence Academic Research Consortium is a collaboration among leading academic research organizations, representatives from the U.S. Food and Drug Administration and physician-scientists from the USA and Europe. One in-person meeting was held in Madrid, Spain, culminating in a document describing consensus recommendations for reporting, collecting, and analysing adherence endpoints across clinical trials. The adoption of these recommendations will afford robustness and consistency in the comparative safety and effectiveness evaluation of investigational drugs from early development to post-marketing approval studies. These principles may be useful for regulatory assessment, as well as for monitoring local and regional outcomes to guide quality improvement initiatives. [ABSTRACT FROM AUTHOR]

Background Acute kidney injury (AKI) has been extensively studied in hospital settings. Limited data exist regarding outcomes for patients with outpatient AKI who are not subsequently admitted. We investigated whether outpatient AKI, defined by a 50% increase in creatinine (Cr), is associated with increased mortality and renal events. Methods In this retrospective study, outpatient serum Cr values from adults receiving primary care at a health system during an 18-month exposure period were used to categorize patients into one of five groups (no outpatient AKI, outpatient AKI with recovery, outpatient AKI without recovery, outpatient AKI without repeat Cr and no Cr). Principal outcomes of all-cause mortality and renal events (50% decline in estimated glomerular filtration rate to <30 mL/min/1.73 m2) were examined using Cox proportional hazards models. Results Among 384 869 eligible patients, 51% had at least one Cr measured during the exposure period. Outpatient AKI occurred in 1.4% of patients while hospital AKI occurred in only 0.3% of patients. The average follow-up was 5.3 years. Outpatient AKI was associated with an increased risk of all-cause mortality {adjusted hazard ratio [aHR] 1.90 [95% confidence interval (CI) 1.76–2.06]} and results were consistent across all AKI groups. Outpatient AKI was also associated with an increased risk of renal events [aHR 1.33 (95% CI 1.11–1.59)], even among those who recovered. Conclusions Outpatient AKI is more prevalent than inpatient AKI and is a risk factor for all-cause mortality and renal events, even among those who recover kidney function. Further research is necessary to determine risk factors and identify strategies for preventing outpatient AKI. [ABSTRACT FROM AUTHOR]

Background Nonadherence to posttransplant immunosuppressive medication is associated with increased rates of rejection and graft loss, yet it is unknown to what degree ideal adherence is associated with the sequelae of overimmunosuppression. Specifically, we questioned whether excellent adherence increased the posttransplant cancer risk. Methods Between August 1998 and August 2006, 195 consenting kidney transplant recipients had electronic monitoring of theirimmunosuppressive medication adherence. Results Based on their average quantitative adherence to a single immunosuppressant drug over the first 6 months posttransplant, recipients were grouped into adherence tertiles (highest, >97.9% adherence; middle, 91–97.8%; lowest, <91%). The cumulative incidence of cancer was calculated for patients in each tertile, treating death as a competing risk. The association between adherence and cancer rate was calculated after adjusting for recipient risk factors, using a competing risk proportional hazards model. The median duration of follow-up was 10.1 years. The 10-year estimated cumulative cancer incidence was 59.4% in the most adherent, 36.1% in the middle group and 38.1% in the least adherent group (P = 0.006). Excluding nonmelanocytic skin cancers, cancer incidence remained significantly higher in the highest adherence group (P = 0.002). Conclusions These data provide additional support for the need to individualize immunosuppression to minimize both rejection and immunosuppressive drug–related complications including cancer. [ABSTRACT FROM AUTHOR]

Summary: Patients awaiting cadaveric organ transplantation face a difficult decision if offered a low‐quality organ: accept the organ or remain on the waiting list and hope a better organ is offered in the future. A dynamic treatment regime (DTR) for transplantation is a rule that determines whether a patient should decline an offered organ. Existing methods can estimate the effect of DTRs on survival outcomes, but these were developed for applications where treatment is abundantly available. For transplantation, organ availability is limited, and existing methods can only estimate the effect of a DTR assuming a single patient follows the DTR. We show for transplantation that the effect of a DTR depends on whether other patients follow the DTR. To estimate the anticipated survival if the entire population awaiting transplantation were to adopt a DTR, we develop a novel inverse probability weighted estimator (IPCW) which re‐weights patients based on the probability of following their transplant history in the counterfactual world in which all patients follow the DTR of interest. We estimate this counterfactual probability using hot deck imputation to fill in data that is not observed for patients who are artificially censored by IPCW once they no longer follow the DTR of interest. We show via simulation that our proposed method has good finite‐sample properties, and we apply our method to a lung transplantation observational registry. [ABSTRACT FROM AUTHOR]

Background: The Tobacco Longitudinal Care study was a randomized controlled trial for smoking cessation. It demonstrated that longitudinal care for smoking cessation, in which telephone-based counseling and nicotine replacement therapy were offered for 12 months, was more effective than the standard 8-week treatment.Purpose: This study aims to identify for whom and how longitudinal care increased the likelihood of abstinence.Methods: Mediated moderation analyses were utilized across three time points.Results: There was a trend towards smokers who did not respond to treatment (i.e., were still smoking) by 21 days being more likely to be abstinent at 6 months if they received longitudinal care rather than usual care. Similarly, those who did not respond to treatment by 3 months were more likely to be abstinent at 12 months if they received longitudinal care. At both time points, the likelihood of abstinence did not differ across treatment conditions among participants who responded to treatment (i.e., quit smoking). The effect on 6-month outcomes was mediated by satisfaction and readiness to quit. Cessation self-efficacy, satisfaction, and readiness to quit mediated the effect on 12-month outcomes. The effect of treatment condition on the likelihood of abstinence at 18 months was not moderated by response to treatment at 6 months.Conclusions: Smokers who did not respond to initial treatment benefited from longitudinal care. Differential effects of treatment condition were not observed among those who responded to early treatment. Conditional assignment to longitudinal care may be useful. Determining for whom and how interventions work over time will advance theory and practice. [ABSTRACT FROM AUTHOR]

The net reclassification improvement (NRI) is a widely used metric used to assess the relative ability of 2 risk models to distinguish between low- and high-risk individuals. However, the validity and usefulness of the NRI have been questioned. Criticism of the NRI focuses on its use comparing nested risk models, whereas in practice it is often used to compare nonnested risk models derived from distinct data sources. In this study, we evaluated the performance of the NRI in a nonnested context by using it to compare competing cardiovascular risk-prediction models. We explored the NRI's sensitivity to variations in risk categories and to the calibration of the compared models. We found that the NRI was very sensitive to changes in the definition of risk categories, especially when at least 1 model was miscalibrated. To address these shortcomings, we describe a novel alternative to the usual NRI that uses percentiles of risk instead of cutoffs based on absolute risk. This percentile-based NRI demonstrates the relative ability of 2 models to rank patient risk. It displays more stable behavior, and we recommend its use when there are no established risk categories or when models are miscalibrated. [ABSTRACT FROM AUTHOR]

The etiology of childhood cancers remains largely unknown, especially regarding environmental and behavioral risk factors. Unpacking the association between socioeconomic status (SES) and incidence may offer insight into such etiology. We tested associations between SES and childhood cancer incidence in a population-based case-cohort study (source cohort: Minnesota birth registry, 1989-2014). Cases, ages 0-14 years, were linked from the Minnesota Cancer Surveillance System to birth records through probabilistic record linkage. Controls were 4:1 frequency matched on birth year (2,947 cases and 11,907 controls). We tested associations of individual-level (maternal education) and neighborhood-level (census tract composite index) SES using logistic mixed models. In crude models, maternal education was positively associated with incidence of acute lymphoblastic leukemia (odds ratio (OR) = 1.10, 95% confidence interval (CI): 1.02, 1.19), central nervous system tumors (OR = 1.12, 95% CI: 1.04, 1.21), and neuroblastoma (OR = 1.15, 95% CI: 1.02,1.30). Adjustment for established risk factors--including race/ ethnicity, maternal age, and birth weight--substantially attenuated these positive associations. Similar patterns were observed for neighborhood-level SES. Conversely, higher maternal education was inversely associated with hepatoblastoma incidence (adjusted OR = 0.70, 95% CI: 0.51, 0.98). Overall, beyond the social patterning of established demographic and pregnancy-related exposures, SES is not strongly associated with childhood cancer incidence. [ABSTRACT FROM AUTHOR]

Summary: The efficiency of doubly robust estimators of the average causal effect (ACE) of a treatment can be improved by including in the treatment and outcome models only those covariates which are related to both treatment and outcome (i.e., confounders) or related only to the outcome. However, it is often challenging to identify such covariates among the large number that may be measured in a given study. In this article, we propose GLiDeR (Group Lasso and Doubly Robust Estimation), a novel variable selection technique for identifying confounders and predictors of outcome using an adaptive group lasso approach that simultaneously performs coefficient selection, regularization, and estimation across the treatment and outcome models. The selected variables and corresponding coefficient estimates are used in a standard doubly robust ACE estimator. We provide asymptotic results showing that, for a broad class of data generating mechanisms, GLiDeR yields a consistent estimator of the ACE when either the outcome or treatment model is correctly specified. A comprehensive simulation study shows that GLiDeR is more efficient than doubly robust methods using standard variable selection techniques and has substantial computational advantages over a recently proposed doubly robust Bayesian model averaging method. We illustrate our method by estimating the causal treatment effect of bilateral versus single‐lung transplant on forced expiratory volume in one year after transplant using an observational registry. [ABSTRACT FROM AUTHOR]

Noncompliance or non-adherence to randomized treatment is a common challenge when interpreting data from randomized clinical trials. The effect of an intervention if all participants were forced to comply with the assigned treatment (i.e., the causal effect) is often of primary scientific interest. For example, in trials of very low nicotine content (VLNC) cigarettes, policymakers are interested in their effect on smoking behavior if their use were to be compelled by regulation. A variety of statistical methods to estimate the causal effect of an intervention have been proposed, but these methods, including inverse probability of compliance weighted (IPCW) estimators, assume that participants' compliance statuses are reported without error. This is an untenable assumption when compliance is based on self-report. Biomarkers (e.g., nicotine levels in the urine) may provide more reliable indicators of compliance but cannot perfectly discriminate between compliers and non-compliers. However, by modeling the distribution of the biomarker as a mixture distribution and writing the probability of compliance as a function of the mixture components, we show how the probability of compliance can be directly estimated from the data even when compliance status is unknown. To estimate the causal effect, we develop a new approach which re-weights participants by the product of their probability of compliance given the observed data and the inverse probability of compliance given confounders. We show that our proposed estimator is consistent and asymptotically normal and show that in some situations the proposed approach is more efficient than standard IPCW estimators. We demonstrate via simulation that the proposed estimator achieves smaller bias and greater efficiency than ad hoc approaches to estimating the causal effect when compliance is measured with error. We apply our method to data from a recently completed randomized trial of VLNC cigarettes. [ABSTRACT FROM AUTHOR]

Because the number of patients waiting for organ transplants exceeds the number of organs available, a better understanding of how transplantation affects the distribution of residual lifetime is needed to improve organ allocation. However, there has been little work to assess the survival benefit of transplantation from a causal perspective. Previous methods developed to estimate the causal effects of treatment in the presence of time-varying confounders have assumed that treatment assignment was independent across patients, which is not true for organ transplantation. We develop a version of G-estimation that accounts for the fact that treatment assignment is not independent across individuals to estimate the parameters of a structural nested failure time model. We derive the asymptotic properties of our estimator and confirm through simulation studies that our method leads to valid inference of the effect of transplantation on the distribution of residual lifetime. We demonstrate our method on the survival benefit of lung transplantation using data from the United Network for Organ Sharing. [ABSTRACT FROM AUTHOR]

Mixed models are commonly used to represent longitudinal or repeated measures data. An additional complication arises when the response is censored, for example, due to limits of quantification of the assay used. While Gaussian random effects are routinely assumed, little work has characterized the consequences of misspecifying the random-effects distribution nor has a more flexible distribution been studied for censored longitudinal data. We show that, in general, maximum likelihood estimators will not be consistent when the random-effects density is misspecified, and the effect of misspecification is likely to be greatest when the true random-effects density deviates substantially from normality and the number of noncensored observations on each subject is small. We develop a mixed model framework for censored longitudinal data in which the random effects are represented by the flexible seminonparametric density and show how to obtain estimates in SAS procedure NLMIXED. Simulations show that this approach can lead to reduction in bias and increase in efficiency relative to assuming Gaussian random effects. The methods are demonstrated on data from a study of hepatitis C virus. [ABSTRACT FROM PUBLISHER]