Projection‐Based Inference in Randomised Clinical Trials

Covariate information is often available in randomised clinical trials for each subject prior to treatment assignment and is commonly utilised to make covariate adjustment for baseline characteristics predictive of the outcome in order to increase precision and improve power in the detection of a treatment effect. Motivated by a nonparametric covariance analysis, we study a projection approach to making objective covariate adjustment in randomised clinical trials on the basis of two unbiased estimating functions that decouple the outcome and covariate data. The proposed projection approach extends a weighted least‐squares procedure by projecting one of the estimating functions onto the linear subspace spanned by the other estimating function that is E‐ancillary for the average treatment effect. Compared with the weighted least‐squares method, the projection method allows for objective inference on the average treatment effect by exploiting the treatment specific covariate–outcome associations. The resulting projection‐based estimator of the average treatment effect is asymptotically efficient when the treatment‐specific working regression models are correctly specified and is asymptotically more efficient than other existing competitors when the treatment‐specific working regression models are misspecified. The proposed projection method is illustrated by an analysis of data from an HIV clinical trial. In a simulation study, we show that the proposed projection method compares favourably with its competitors in finite samples.     

The influence of population characteristics on household response to a charity book collection based on pledges and social pressure

The success of a charitable campaign will depend on many factors, including solicitation technique, population characteristics, and type of charity. We implemented a randomised controlled trial to assess the effect of pledging and social pressure on the charitable donation behaviour of households. We implemented a charitable book collection in two different geographical areas, with one being more affluent, more ethnically diverse, and with higher levels of education. We received a marked variation in response: overall, across both treatment groups and control group, the number of households who donated books was higher in one area than the other. However, we did not find any evidence of a heterogeneous response to the pledge and social pressure interventions, compared to the control group.     

Sampling-based Randomised Designs for Causal Inference under the Potential Outcomes Framework

We establish the inferential properties of the mean-difference estimator for the average treatment effect in randomised experiments where each unit in a population is randomised to one of two treatments and then units within treatment groups are randomly sampled. The properties of this estimator are well understood in the experimental design scenario where first units are randomly sampled and then treatment is randomly assigned but not for the aforementioned scenario where the sampling and treatment assignment stages are reversed. We find that the inferential properties of the mean-difference estimator under this experimental design scenario are identical to those under the more common sample-first-randomise-second design. This finding will bring some clarifications about sampling-based randomised designs for causal inference, particularly for settings where there is a finite super-population. Finally, we explore to what extent pre-treatment measurements can be used to improve upon the mean-difference estimator for this randomise-first-sample-second design. Unfortunately, we find that pre-treatment measurements are often unhelpful in improving the precision of average treatment effect estimators under this design, unless a large number of pre-treatment measurements that are highly associative with the post-treatment measurements can be obtained. We confirm these results using a simulation study based on a real experiment in nanomaterials.     

The Effectiveness of Educational and Needle Exchange Programs: A Meta-analysis of HIV Prevention Strategies for Injecting Drug Users

A meta-analysis of 16 educational interventions and 10 needle exchange programs was performed to estimate the effectiveness of reducing HIV risk behaviors in the injecting drug user population. Information on intervention, outcome, design and demographics was coded and analyzed for all educational and needle exchange program evaluation studies published between January 1984 and May 1995. The weighted mean effect size for the 6,251 study subjects of the 16 educational interventions was 0.749 (95% CI, 0.708 to 0.790), and the weighted mean effect size for the 1,675 study subjects of the 10 needle exchange programs was 0.279 (95% CI, 0.207 to 0.352), suggesting that both interventions had a positive impact on reducing HIV risk behaviors associated with injecting drug use. However, these results were dependent upon research design, outcome type and follow-up time.     

A flexible parametric approach for estimating switching regime models and treatment effect parameters

In this paper, we propose a flexible, parametric class of switching regime models allowing for both skewed and fat-tailed outcome and selection errors. Specifically, we model the joint distribution of each outcome error and the selection error via a newly constructed class of multivariate distributions which we call generalized normal mean–variance mixture distributions. We extend Heckman’s two-step estimation procedure for the Gaussian switching regime model to the new class of models. When the distributions of the outcome errors are asymmetric, we show that an additional correction term accounting for skewness in the outcome error distribution (besides the analogue of the well known inverse mill’s ratio) needs to be included in the second step regression. We use the two-step estimators of parameters in the model to construct simple estimators of average treatment effects and establish their asymptotic properties. Simulation results confirm the importance of accounting for skewness in the outcome errors in estimating both model parameters and the average treatment effect and the treatment effect for the treated.     

The applicability of the Tampa Scale of Kinesiophobia for patients with sub-acute neck pain: a qualitative study

The purpose of this study was to qualitatively evaluate patients understanding and interpretation of the wording used in test items of the Tampa Scale of Kinesiophobia (TSK). The TSK was developed to measure fear of movement in patients suffering from low back pain. The TSK is being increasingly used for other pain conditions. Patients with sub-acute neck pain experience problems while completing this questionnaire. The aim of this study was to elicit these problems. The study was conducted within the framework of a randomised controlled trial. The Three-Step Test Interview (TSTI) was used to collect data on the thoughts or considerations of respondents while completing the TSK. In the analysis, each transcribed interview was divided into three segments. The thoughts and considerations were then analysed and categorised per item. During the TSTI two problems were identified. One concerned the meaning of specific words used, like “dangerous” and “injury”. The other problem was that several implicit assumptions within some items make it difficult for respondents to answer these items. It was concluded that in the development and validation of questionnaires like the TSK, not only quantitative psychometric properties are important, but also qualitative research has an important contribution to enhance applicability.     

Analysing the risks of adopting circular economy initiatives in manufacturing supply chains

The concept of circular economy (CE) has proven its worth due to the scarcity of natural resources and huge amounts of wastage which impacts the environment. Thus, the adoption of the CE concept in the supply chain becomes critical. However, due to the complex nature of processes/activities in the circular supply chain (CSC), managing risk has become a priority to avoid disruption. In current literature, no discussion has been conducted on how to analyse the risks in the context of CSC. Therefore, to fill this literature gap, this study concentrates on identifying and analysing the risks to promote effective circular initiatives in supply chains in the context of the manufacturing industry, thus minimising the negative environmental impact. A total of 31 risks were identified through an extensive literature review and discussions with experts. A grey-based decision-making trial and evaluation laboratory (DEMATEL) method is applied by incorporating the experts' knowledge to compute prominence and cause/effect scores to develop an interrelationship map. Finally, a vulnerability matrix for risk categories is developed using the average of prominence and cause/effect scores of risks. The results show that transparent process is the most prominent risk and branding is the least significant risk. By using the average prominence and cause/effect score, a risk category, namely, financial risk, is identified as most vulnerable to CSC. These findings will help industry managers not only to prepare business strategies in the adoption of CE initiatives in supply chains by eliminating risks but also in minimising negative environmental impact.     

Estimating Mann–Whitney‐type Causal Effects

Mann–Whitney‐type causal effects are generally applicable to outcome variables with a natural ordering, have been recommended for clinical trials because of their clinical relevance and interpretability and are particularly useful in analysing an ordinal composite outcome that combines an original primary outcome with death and possibly treatment discontinuation. In this article, we consider robust and efficient estimation of such causal effects in observational studies and clinical trials. For observational studies, we propose and compare several estimators: regression estimators based on an outcome regression (OR) model or a generalised probabilistic index (GPI) model, an inverse probability weighted estimator based on a propensity score model and two doubly robust (DR), locally efficient estimators. One of the DR estimators involves a propensity score model and an OR model, is consistent and asymptotically normal under the union of the two models and attains the semiparametric information bound when both models are correct. The other DR estimator has the same properties with the OR model replaced by a GPI model. For clinical trials, we extend an existing augmented estimator based on a GPI model and propose a new one based on an OR model. The methods are evaluated and compared in simulation experiments and applied to a clinical trial in cardiology and an observational study in obstetrics.     

IDENTIFYING CAUSAL MECHANISMS (PRIMARILY) BASED ON INVERSE PROBABILITY WEIGHTING

This paper demonstrates the identification of causal mechanisms of a binary treatment under selection on observables, (primarily) based on inverse probability weighting; i.e. we consider the average indirect effect of the treatment, which operates through an intermediate variable (or mediator) that is situated on the causal path between the treatment and the outcome, as well as the (unmediated) direct effect. Even under random treatment assignment, subsequent selection into the mediator is generally non‐random such that causal mechanisms are only identified when controlling for confounders of the mediator and the outcome. To tackle this issue, units are weighted by the inverse of their conditional treatment propensity given the mediator and observed confounders. We show that the form and applicability of weighting depend on whether some confounders are themselves influenced by the treatment or not. A simulation study gives the intuition for these results and an empirical application to the direct and indirect health effects (through employment) of the US Job Corps program is also provided. Copyright © 2013 John Wiley & Sons, Ltd.     

Modelling international sovereign risk information spillovers: A multilayer network approach

Applying the TVP-VAR model, we creatively construct multilayer information spillover networks containing return spillover layer, volatility spillover layer and extreme risk spillover layer among 23 countries in the G20 to explore international sovereign risk spillovers. From the perspective of system-level and country-level measures, this article explores the topological structures of static and dynamic multilayer networks. We observe that (i) at the system-level, multilayer measures containing uniqueness edge ratio and average edge overlap show each layer has unique network structures and spillover evolution behavior, especially for dynamic networks. Average connectedness strength shows volatility and extreme risk spillover layers are more sensitive to extreme events. Meanwhile, three layers have highly intertwined and interrelated relations. Notably, their spillovers all show a great upsurge during the crisis (financial and European debt crisis) and the COVID-19 pandemic period. (ii) At the country-level, average overlapping net-strength shows that countries’ roles are different during distinct periods. Multiplex participation coefficient on out-strength indicates we’ll focus on countries with highly heterogeneous connectedness among three layers during the stable period since their underestimated spillovers soar in extreme events or crises. Multilayer networks supply comprehensive information that cannot obtain by single-layer.     

Improving social harm indices with a modulated Hawkes process

Communities are affected adversely by a range of social harm events, such as crime, traffic crashes, medical emergencies, and drug use. The police, fire, health and social service departments are tasked with mitigating such social harm through various types of interventions. While various different social harm indices have been proposed for allocating resources to spatially fixed hotspots, the risk of social harm events is dynamic, and new algorithms and software systems that are capable of quickly identifying risks and triggering appropriate public safety responses are needed. We propose a novel modulated Hawkes process for this purpose that offers flexible approaches to both (i) the incorporation of spatial covariates and leading indicators for variance reduction in the case of rarer event categories, and (ii) the capture of dynamic hotspot formation through self-excitation. We present an efficient l1-penalized EM algorithm for estimating the model that performs feature selection for the spatial covariates of each incident type simultaneously. We provide simulation results using data from the Indianapolis Metropolitan Police Department in order to illustrate the advantages of the modulated Hawkes process model of social harm over various recently introduced social harm indices and property crime Hawkes processes.     

Channels through which Public Employment Services and Small Business Assistance Programmes Work*

Empirical evidence has found that public employment services (PES) and small business assistance (SBA) programmes are successful at getting the unemployed back to work. Policywise it is important to know which of these two programmes is more effective, for whom, and when. Using unusually rich survey data and matching methods, this study evaluates the relative effectiveness of PES and SBA for different subgroups in Romania in the late 1990s, where the outcome variables involve earnings, employment and unemployment in 2000–1 and early 2002. It finds that heterogeneity matters and that these programmes need to be tailored to the problem at hand.     

Semiparametric bounds on treatment effects

We present a variety of semiparametric models that produce bounds on the average causal effect of a binary treatment on a binary outcome. The semiparametric assumptions exploit variation in observable covariates to narrow the bounds. In our main model, the outcome is determined by a generalized linear model, but the treatment may be arbitrarily endogenous. Our bounding strategy does not require the existence of an instrument, but incorporating an instrument narrows the bounds. The bounds are further improved by combining the semiparametric model with the joint threshold-crossing assumption of Shaikh and Vytlacil (2005).     

Assessment of meaningful change in routine outcome measurement (ROM) with a combination of a longitudinal and a ‘classify and count’ approach

To assess significant changes of health status in people receiving health care, distribution-based and anchor-based methods have been proposed. However, there is no real consensus on what method is the best for evaluating clinically meaningful change. To maximize the internal and external validity of outcome assessment, we propose combining two approaches as recommended by recent practical guidelines on this field. Specifically, we suggest applying longitudinal hierarchical linear models on subgroups of patients showing reliable change and reliable and clinically significant change. This combined approach improved the model’s ability (1) to quantify the magnitude of changes to be reliable and clinically meaningful and (2) to select significant predictors of changes. An empirical application on a prevalence sample of Italian outpatients attending four community mental health services was done. A cross-sectional model and three longitudinal models were applied on the entire study sample and reliable and clinically meaningful change subsamples to investigate the magnitude of change and the predictive effect on outcomes of clinical, socio-demographic and process variables on different patients’ subgroups. Differences were found suggesting that both the statistical method and the sample used to calculate individual changes affect the estimates. The main conclusion is that ignoring the longitudinal data structure or including patients with unreliable change at the follow-up might result in misleading inferences that can alter the real magnitude of changes and the contributions of predictors. The approach proposed provides robust feedback to clinicians on clinically significant change and can be recommended in outcome studies and research.     

Nonparametric identification and estimation of sample selection models under symmetry

Under a conditional mean restriction Das et al. (2003) considered nonparametric estimation of sample selection models. However, their method can only identify the outcome regression function up to a constant. In this paper we strengthen the conditional mean restriction to a symmetry restriction under which selection biases due to selection on unobservables can be eliminated through proper matching of propensity scores; consequently we are able to identify and obtain consistent estimators for the average treatment effects and the structural regression functions. The results from a simulation study suggest that our estimators perform satisfactorily.     

Why homeowners’ documentation went missing under the Home Affordable Mortgage Program (HAMP)?: An analysis of strategic behavior of homeowners and servicers

This study investigated why documentation of certain homeowners who participated in the trial period plan (TPP) of HAMP went missing using a model of strategic behavior of homeowners and servicers. I found that the likelihood of cancellation of the TPP for missing documentation was higher for those who were current on their mortgage payments, compared to those 30 days delinquent, prior to entering the plan—both types of homeowners entered the program because they were in danger of imminent default. This finding, which is consistent with servicers “steering” homeowners with low credit risk away from HAMP to their own (proprietary) modifications, is more likely for loans owned or guaranteed by the government-sponsored enterprises (GSE) than non-GSE loans—the GSE loans have relatively more homeowners with low credit risks. The outcome, not widespread across servicers, is possibly related to the lack of standardization across servicers in identifying borrowers in danger of imminent default. Missing documentation is also associated with severely delinquent homeowners who had less incentive or were unable to submit complete documentation. This cohort of borrowers could benefit from financial education and Treasury’s proposal that servicers provide a single point of contact for borrowers would help reduce the problem of missing documentation.     

Non‐parametric Analysis of Gap Times for Multiple Event Data: An Overview

Multiple event data are frequently encountered in medical follow‐up, engineering and other applications when the multiple events are considered as the major outcomes. They may be repetitions of the same event (recurrent events) or may be events of different nature. Times between successive events (gap times) are often of direct interest in these applications. The stochastic‐ordering structure and within‐subject dependence of multiple events generate statistical challenges for analysing such data, including induced dependent censoring and non‐identifiability of marginal distributions. This paper provides an overview of a class of existing non‐parametric estimation methods for gap time distributions for various types of multiple event data, where sampling bias from induced dependent censoring is effectively adjusted. We discuss the statistical issues in gap time analysis, describe the estimation procedures and illustrate the methods with a comparative simulation study and a real application to an AIDS clinical trial. A comprehensive understanding of challenges and available methods for non‐parametric analysis can be useful because there is no existing standard approach to identifying an appropriate gap time method that can be used to address research question of interest. The methods discussed in this review would allow practitioners to effectively handle a variety of real‐world multiple event data.     

Impact of treatment heterogeneity on drug resistance and supply chain costs

The efficacy of scarce drugs for many infectious diseases is threatened by the emergence and spread of resistance. Multiple studies show that available drugs should be used in a socially optimal way to contain drug resistance. This paper studies the tradeoff between risk of drug resistance and operational costs when using multiple drugs for a specific disease. Using a model for disease transmission and resistance spread, we show that treatment with multiple drugs, on a population level, results in better resistance-related health outcomes, but more interestingly, the marginal benefit decreases as the number of drugs used increases. We compare this benefit with the corresponding change in procurement and safety stock holding costs that result from higher drug variety in the supply chain. Using a large-scale simulation based on malaria transmission dynamics, we show that disease prevalence seems to be a less important factor when deciding the optimal width of drug assortment, compared to the duration of one episode of the disease and the price of the drug(s) used. Our analysis shows that under a wide variety of scenarios for disease prevalence and drug cost, it is optimal to simultaneously deploy multiple drugs in the population. If the drug price is high, large volume purchasing discounts are available, and disease prevalence is high, it may be optimal to use only one drug. Our model lends insights to policy makers into the socially optimal size of drug assortment for a given context.     

我国商业体育赛事的经营风险及其规避措施探讨

本文从界定商业体育赛事以及商业体育赛事经营风险的概念出发,分析了我国商业体育赛事经营风险产生的原因及种类,在此基础上从风险控制和风险转嫁的角度提出了商业体育赛事经营风险的规避措施。