Statistical Methods in Medical Research

Most existing dose-ranging study designs focus on assessing the dose–efficacy relationship and identifying the minimum effective dose. There is an increasing interest in optimizing the dose based on the benefit–risk tradeoff. We propose a Bayesian quasi-likelihood dose-ranging design that jointly considers safety and efficacy to simultaneously identify the minimum effective dose and the maximum utility dose to optimize the benefit–risk tradeoff. The binary toxicity endpoint is modeled using a beta-binomial model. The efficacy endpoint is modeled using the quasi-likelihood approach to accommodate various types of data (e.g. binary, ordinal or continuous) without imposing any parametric assumptions on the dose–response curve. Our design utilizes a utility function as a measure of benefit–risk tradeoff and adaptively assign patients to doses based on the doses’ likelihood of being the minimum effective dose and maximum utility dose. The design takes a group-sequential approach. At each interim, the doses that are deemed overly toxic or futile are dropped. At the end of the trial, we use posterior probability criteria to assess the strength of the dose–response relationship for establishing the proof-of-concept. If the proof-of-concept is established, we identify the minimum effective dose and maximum utility dose. Our simulation study shows that compared with some existing designs, the Bayesian quasi-likelihood dose-ranging design is robust and yields competitive performance in establishing proof-of-concept and selecting the minimum effective dose. Moreover, it includes an additional feature for further maximum utility dose selection.

In single-arm trials with a predefined subgroup based on baseline biomarkers, it is often assumed that a biomarker defined subgroup, the biomarker positive subgroup, has the same or higher response to treatment compared to its complement, the biomarker negative subgroup. The goal is to determine if the treatment is effective in each of the subgroups or in the biomarker positive subgroup only or not effective at all. We propose the isotonic stratified design for this problem. The design has a joint set of decision rules for biomarker positive and negative subjects and utilizes joint estimation of response probabilities using assumed monotonicity of response between the biomarker negative and positive subgroups. The new design reduces the sample size requirement when compared to running two Simon's designs in each biomarker positive and negative. For example, the new design requires 23%–35% fewer patients than running two Simon's designs for scenarios we considered. Alternatively, the new design allows evaluating the response probability in both biomarker negative and biomarker positive subgroups using only 40% more patients needed for running Simon's design in the biomarker positive subgroup only.

In psychophysics and psychometrics, an integral method to the discipline involves charting how a person’s response pattern changes according to a continuum of stimuli. For instance, in hearing science, Visual Analog Scaling tasks are experiments in which listeners hear sounds across a speech continuum and give a numeric rating between 0 and 100 conveying whether the sound they heard was more like word “a” or more like word “b” (i.e. each participant is giving a continuous categorization response). By taking all the continuous categorization responses across the speech continuum, a parametric curve model can be fit to the data and used to analyze any individual’s response pattern by speech continuum. Standard statistical modeling techniques are not able to accommodate all of the specific requirements needed to analyze these data. Thus, Bayesian hierarchical modeling techniques are employed to accommodate group-level non-linear curves, individual-specific non-linear curves, continuum-level random effects, and a subject-specific variance that is predicted by other model parameters. In this paper, a Bayesian hierarchical model is constructed to model the data from a Visual Analog Scaling task study of mono-lingual and bi-lingual participants. Any nonlinear curve function could be used and we demonstrate the technique using the 4-parameter logistic function. Overall, the model was found to fit particularly well to the data from the study and results suggested that the magnitude of the slope was what most defined the differences in response patterns between continua.

The Fellegi-Sunter model is a latent class model widely used in probabilistic linkage to identify records that belong to the same entity. Record linkage practitioners typically employ all available matching fields in the model with the premise that more fields convey greater information about the true match status and hence result in improved match performance. In the context of model-based clustering, it is well known that such a premise is incorrect and the inclusion of noisy variables could compromise the clustering. Variable selection procedures have therefore been developed to remove noisy variables. Although these procedures have the potential to improve record matching, they cannot be applied directly due to the ubiquity of the missing data in record linkage applications. In this paper, we modify the stepwise variable selection procedure proposed by Fop, Smart, and Murphy and extend it to account for missing data common in record linkage. Through simulation studies, our proposed method is shown to select the correct set of matching fields across various settings, leading to better-performing algorithms. The improved match performance is also seen in a real-world application. We therefore recommend the use of our proposed selection procedure to identify informative matching fields for probabilistic record linkage algorithms.

The focus of precision medicine is on decision support, often in the form of dynamic treatment regimes, which are sequences of decision rules. At each decision point, the decision rules determine the next treatment according to the patient’s baseline characteristics, the information on treatments and responses accrued by that point, and the patient’s current health status, including symptom severity and other measures. However, dynamic treatment regime estimation with ordinal outcomes is rarely studied, and rarer still in the context of interference – where one patient’s treatment may affect another’s outcome. In this paper, we introduce the weighted proportional odds model: a regression based, approximate doubly-robust approach to single-stage dynamic treatment regime estimation for ordinal outcomes. This method also accounts for the possibility of interference between individuals sharing a household through the use of covariate balancing weights derived from joint propensity scores. Examining different types of balancing weights, we verify the approximate double robustness of weighted proportional odds model with our adjusted weights via simulation studies. We further extend weighted proportional odds model to multi-stage dynamic treatment regime estimation with household interference, namely dynamic weighted proportional odds model. Lastly, we demonstrate our proposed methodology in the analysis of longitudinal survey data from the Population Assessment of Tobacco and Health study, which motivates this work. Furthermore, considering interference, we provide optimal treatment strategies for households to achieve smoking cessation of the pair in the household.

Missing data is a common challenge when analyzing epidemiological data, and imputation is often used to address this issue. Here, we investigate the scenario where a covariate used in an analysis has missingness and will be imputed. There are recommendations to include the outcome from the analysis model in the imputation model for missing covariates, but it is not necessarily clear if this recommendation always holds and why this is sometimes true. We examine deterministic imputation (i.e. single imputation with fixed values) and stochastic imputation (i.e. single or multiple imputation with random values) methods and their implications for estimating the relationship between the imputed covariate and the outcome. We mathematically demonstrate that including the outcome variable in imputation models is not just a recommendation but a requirement to achieve unbiased results when using stochastic imputation methods. Moreover, we dispel common misconceptions about deterministic imputation models and demonstrate why the outcome should not be included in these models. This article aims to bridge the gap between imputation in theory and in practice, providing mathematical derivations to explain common statistical recommendations. We offer a better understanding of the considerations involved in imputing missing covariates and emphasize when it is necessary to include the outcome variable in the imputation model.

We propose a novel framework based on the RuleFit method to estimate heterogeneous treatment effect in randomized clinical trials. The proposed method estimates a rule ensemble comprising a set of prognostic rules, a set of prescriptive rules, as well as the linear effects of the original predictor variables. The prescriptive rules provide an interpretable description of the heterogeneous treatment effect. By including a prognostic term in the proposed model, the selected rule is represented as an heterogeneous treatment effect that excludes other effects. We confirmed that the performance of the proposed method was equivalent to that of other ensemble learning methods through numerical simulations and demonstrated the interpretation of the proposed method using a real data application.

Ordinal response is commonly found in medicine, biology, and other fields. In many situations, the predictors for this ordinal response are compositional, which means that the sum of predictors for each sample is fixed. Examples of compositional data include the relative abundance of species in microbiome data and the relative frequency of nutrition concentrations. Moreover, the predictors that are strongly correlated tend to have similar influence on the response outcome. Conventional cumulative logistic regression models for ordinal responses ignore the fixed-sum constraint on predictors and their associated interrelationships, and thus are not appropriate for analyzing compositional predictors.To solve this problem, we proposed Bayesian Compositional Models for Ordinal Response to analyze the relationship between compositional data and an ordinal response with a structured regularized horseshoe prior for the compositional coefficients and a soft sum-to-zero restriction on coefficients through the prior distribution. The method was implemented with R package rstan using efficient Hamiltonian Monte Carlo algorithm. We performed simulations to compare the proposed approach and existing methods for ordinal responses. Results revealed that our proposed method outperformed the existing methods in terms of parameter estimation and prediction. We also applied the proposed method to a microbiome study HMP2Data, to find microorganisms linked to ordinal inflammatory bowel disease levels. To make this work reproducible, the code and data used in this paper are available at https://github.com/Li-Zhang28/BCO.

We used Monte Carlo simulations to compare the performance of marginal structural models (MSMs) based on weighted univariate generalized linear models (GLMs) to estimate risk differences and relative risks for binary outcomes in observational studies. We considered four different sets of weights based on the propensity score: inverse probability of treatment weights with the average treatment effect as the target estimand, weights for estimating the average treatment effect in the treated, matching weights and overlap weights. We considered sample sizes ranging from 500 to 10,000 and allowed the prevalence of treatment to range from 0.1 to 0.9. We examined both the robust variance estimator when using generalized estimating equations with an independent working correlation matrix and a bootstrap variance estimator for estimating the standard error of the risk difference and the log-relative risk. The performance of these methods was compared with that of direct weighting. Both the direct weighting approach and MSMs based on weighted univariate GLMs resulted in the identical estimates of risk differences and relative risks. When sample sizes were small to moderate, the use of an MSM with a bootstrap variance estimator tended to result in the most accurate estimates of standard errors. When sample sizes were large, the direct weighting approach and an MSM with a bootstrap variance estimator tended to produce estimates of standard error with similar accuracy. When using a MSM to estimate risk differences and relative risks, in general it is preferable to use a bootstrap variance estimator than the robust variance estimator. We illustrate the application of the different methods for estimating risks differences and relative risks using an observational study on the effect on mortality of discharge prescribing of a beta-blocker in patients hospitalized with acute myocardial infarction.

For the analysis of time-to-event data, frequently used methods such as the log-rank test or the Cox proportional hazards model are based on the proportional hazards assumption, which is often debatable. Although a wide range of parametric and non-parametric methods for non-proportional hazards has been proposed, there is no consensus on the best approaches. To close this gap, we conducted a systematic literature search to identify statistical methods and software appropriate under non-proportional hazard. Our literature search identified 907 abstracts, out of which we included 211 articles, mostly methodological ones. Review articles and applications were less frequently identified. The articles discuss effect measures, effect estimation and regression approaches, hypothesis tests, and sample size calculation approaches, which are often tailored to specific non-proportional hazard situations. Using a unified notation, we provide an overview of methods available. Furthermore, we derive some guidance from the identified articles.

This paper aims to extend the Besag model, a widely used Bayesian spatial model in disease mapping, to a non-stationary spatial model for irregular lattice-type data. The goal is to improve the model’s ability to capture complex spatial dependence patterns and increase interpretability. The proposed model uses multiple precision parameters, accounting for different intensities of spatial dependence in different sub-regions. We derive a joint penalized complexity prior to the flexible local precision parameters to prevent overfitting and ensure contraction to the stationary model at a user-defined rate. The proposed methodology can be used as a basis for the development of various other non-stationary effects over other domains such as time. An accompanying R package fbesag equips the reader with the necessary tools for immediate use and application. We illustrate the novelty of the proposal by modeling the risk of dengue in Brazil, where the stationary spatial assumption fails and interesting risk profiles are estimated when accounting for spatial non-stationary. Additionally, we model different causes of death in Brazil, where we use the new model to investigate the spatial stationarity of these causes.