Adjusting for regression effect in uncontrolled studies

When clinical studies require enrolled patients to have abnormal assays, the natural tendency of repeat measurements to regress toward the mean can lead to a false assessment of effectiveness of therapy. We propose a method to more accurately estimate the true effect of therapy by adjusting for a component of improvement that can be attributed to regression effect. The model we use allows for a combination of additive and/or multiplicative effects of the therapy.     

Bayesian inferences in the Cox model for order-restricted hypotheses

In studying the relationship between an ordered categorical predictor and an event time, it is standard practice to include dichotomous indicators of the different levels of the predictor in a Cox model. One can then use a multiple degree-of-freedom score or partial likelihood ratio test for hypothesis testing. Often, interest focuses on comparing the null hypothesis of no difference to an order-restricted alternative, such as a monotone increase across levels of a predictor. This article proposes a Bayesian approach for addressing hypotheses of this type. We reparameterize the Cox model in terms of a cumulative product of parameters having conjugate prior densities, consisting of mixtures of point masses at one, and truncated gamma densities. Due to the structure of the model, posterior computation can proceed via a simple and efficient Gibbs sampling algorithm. Posterior probabilities for the global null hypothesis and subhypotheses, comparing the hazards for specific groups, can be calculated directly from the output of a single Gibbs chain. The approach allows for level sets across which a predictor has no effect. Generalizations to multiple predictors are described, and the method is applied to a study of emergency medical treatment for stroke.     

A diagnostic for Cox regression with discrete failure-time models

Changes in maximum likelihood parameter estimates due to deletion of individual observations are useful statistics, both for regression diagnostics and for computing robust estimates of covariance. For many likelihoods, including those in the exponential family, these delete-one statistics can be approximated analytically from a one-step Newton-Raphson iteration on the full maximum likelihood solution. But for general conditional likelihoods and the related Cox partial likelihood, the one-step method does not reduce to an analytic solution. For these likelihoods, an alternative analytic approximation that relies on an appropriately augmented design matrix has been proposed. In this paper, we extend the augmentation approach to explicitly deal with discrete failure-time models. In these models, an individual subject may contribute information at several time points, thereby appearing in multiple risk sets before eventually experiencing a failure or being censored. Our extension also allows the covariates to be time dependent. The new augmentation requires no additional computational resources while improving results.     

Residuals for proportional hazards models with interval-censored survival data

We develop diagnostic tools for use with proportional hazards models for interval-censored survival data. We propose counterparts to the Cox-Snell, Lagakos (or martingale), deviance, and Schoenfeld residuals. Many of the properties of these residuals carry over to the interval-censored case. In particular, the interval-censored versions of the Lagakos and Schoenfeld residuals may be derived as components of suitable score statistics. The Lagakos residuals may be used to check regression relationships, while the Schoenfeld residuals can help to detect nonproportional hazards in semiparametric models. The methods apply to parametric models and to the semiparametric model with discrete observation times.     

A Bayesian Semiparametric Survival Model with Longitudinal Markers

Summary We consider inference for data from a clinical trial of treatments for metastatic prostate cancer. Patients joined the trial with diverse prior treatment histories. The resulting heterogeneous patient population gives rise to challenging statistical inference problems when trying to predict time to progression on different treatment arms. Inference is further complicated by the need to include a longitudinal marker as a covariate. To address these challenges, we develop a semiparametric model for joint inference of longitudinal data and an event time. The proposed approach includes the possibility of cure for some patients. The event time distribution is based on a nonparametric Pólya tree prior. For the longitudinal data we assume a mixed effects model. Incorporating a regression on covariates in a nonparametric event time model in general, and for a Pólya tree model in particular, is a challenging problem. We exploit the fact that the covariate itself is a random variable. We achieve an implementation of the desired regression by factoring the joint model for the event time and the longitudinal outcome into a marginal model for the event time and a regression of the longitudinal outcomes on the event time, i.e., we implicitly model the desired regression by modeling the reverse conditional distribution.     

Regression Calibration in Semiparametric Accelerated Failure Time Models

Summary In large cohort studies, it often happens that some covariates are expensive to measure and hence only measured on a validation set. On the other hand, relatively cheap but error‐prone measurements of the covariates are available for all subjects. Regression calibration (RC) estimation method ( Prentice, 1982 , Biometrika 69 , 331–342) is a popular method for analyzing such data and has been applied to the Cox model by Wang et al. (1997, Biometrics 53 , 131–145) under normal measurement error and rare disease assumptions. In this article, we consider the RC estimation method for the semiparametric accelerated failure time model with covariates subject to measurement error. Asymptotic properties of the proposed method are investigated under a two‐phase sampling scheme for validation data that are selected via stratified random sampling, resulting in neither independent nor identically distributed observations. We show that the estimates converge to some well‐defined parameters. In particular, unbiased estimation is feasible under additive normal measurement error models for normal covariates and under Berkson error models. The proposed method performs well in finite‐sample simulation studies. We also apply the proposed method to a depression mortality study.     

Biophysical stratification of the Amazon basin

In field measurement programmes, stratified sampling can optimize sampling efficiency, but stratification is often undertaken subjectively, and is frequently based on a priori classification schemes such as those used for vegetation maps. In order to avoid the problems associated with a priori subjective schemes, we explore here an objective procedure, Regression Tree Analysis (RTA). RTA has previously been used in local-scale studies, but here we apply it to a very large study domain, namely the entire humid tropical zone of South America. The aim of the study was to develop an optimal sampling design in preparation for the Large Scale Biosphere-Atmosphere Experiment in Amazonia (LBA). Co-registered spatially continuous fields of rainfall, temperature, photosynthetically active radiation (PAR), the normalized difference index (NDVI), an index of surface moisture, and other independent variables were used to predict three dependent variables, annual net radiation (Rn), latent heat (LE) and net primary production (NPP). Rather than simply dividing the study area based on differing levels of the three dependent variables, empirical models were developed using RTA to indicate how the relationships between these and possible forcing variables vary across the study area. For each variable long-term seasonal indices such as annual average, monthly minimum and amplitude were used to exclude effects of temporal phase differences between the hemispheres. The resulting hierarchical models revealed variations in the interdependence of the forcing variables throughout the study area and therefore provided a basis for a stratified sampling and identifying the most important variables to be collected in LBA for the Amazon basin as a whole as well as optimizing the sampling scheme for scaling up findings from the field scale to larger areas.     

Uncertainty in allometric exponent estimation: a case study in scaling metabolic rate with body mass

Many factors could influence the allometric scaling exponent β estimation, but have not been explored systematically. We investigated the influences of three factors on the estimate of β based on a data set of 626 species of basal metabolic rate and mass in mammals. The influence of sampling error was tested by re-sampling with different sample sizes using a Monte Carlo method. Small random errors were introduced to measured data to examine their influence on parameter estimations. The influence of analysis method was also evaluated by applying nonlinear and linear regressions to the original data. Results showed that a relative large sample size was required to lower statistical inference errors. When sample size n was 10% of the base population size (n=63), 35% of the samples supported β=2/3, 39% supported β=3/4, and 15% rejected β=0.711, even though the base population had a β=0.711. The controversy surrounding the estimation of β in the literature could be partially attributable to such small sample sizes in many studies. Measurement errors in body mass and base metabolic rate, especially in body mass, could largely increase alpha and beta errors. Analysis methods also affected parameter estimations. Nonlinear regressions provided better estimates of the scaling exponent that were significantly higher than these commonly estimated by linear regressions. This study demonstrated the importance of the quantity and quality of data as well as analysis method in power law analysis, raising caution in interpreting power law results. Meta-data synthesis using data from independent studies seems to be a proper approach in the future, but caution should be taken to make sure that such measurements are made using similar protocols.