John Van Ryzin

Race, poverty, and survival in multiple myeloma

To identify possible interracial differences in the behavior of multiple myeloma, the records of 52 black myeloma patients at Harlem Hospital Center (HHC) and 46 black and 46 white patients at Columbia‐Presbyterian Medical Center (CPMC) were reviewed. In addition to clinical variables such as tumor burden, azotemia, and hypoalbuminemia, the effect of poverty on prognosis was examined, using socioeconomic indices from the United States census block group data of each patient. The median survival of CPMC black and white patients was comparable (34 and 29 months, respectively) whereas that of the HHC group was 12 months (Breslow test, P < 0.0001). Overcrowding and hypoalbuminemia were the most significant prognostic factors by multivariate regression analysis on all 144 patients (P = 0.001); for HHC patients, overcrowding was the single significant variable affecting survival (P = 0.004). By all socioeconomic indices, HHC patients were more impoverished than CPMC patients (P < 0.001); they also presented with more advanced disease. Race is not a significant prognostic factor in myeloma, whereas the effect of socioeconomic status on survival appears to equal that of previously described clinical features.

A generalized multihit dose-response model for low-dose extrapolation.

A quantal dose-response model based on a multihit theory of toxic response is presented. When spontaneous background toxic response is included, the model involves three unknown parameters. The maximum likelihood estimators for these three parameters are given as the solution of a nonlinear iterative algorithm. The resulting three-dimensional vector of estimators is shown to be asymptotically strongly consistent, asymptotically unique with probability one, and, when suitably normalized, it has asymptotically a trivariate normal distribution. On the basis of these results, a large-sample goodness-of-fit test is given. The use of this model for low-dose extrapolation is indicated. Application of the results is illustrated using three sets of toxicity data.

Maximum entropy histograms

The derivation of a new class of nonparametric probability density estimators, maximum entropy histogram estimators (MEHE) is presented. Some of their asymptotic properties are summaried.

Large sample properties of maximum entropy histograms

The large sample properties of a new class of histogram estimators whose derivation is based on an information theory criterion--the maximum entropy principle, which preserves the observed mass and mean--are studied. The pointwise strong consistency, the point-wise asymptotic normality, and the rate of convergence to normality are investigated. The asymptotic mean square error (MSE) of these estimates is also compared relative to the histogram based on spacings, the classical k -nearest neighbor, the kernel estimator, and the generalized k -nearest neighbor density estimator.

Race, Poverty and Survival in Multiple Myeloma

To identify possible interracial differences in the behavior of multiple myeloma, we reviewed the records of 52 black myeloma patients at Harlem Hospital Center (HHC) and 46 black and 46 white controls at Columbia-Presbyterian Medical Center (CPMC). In addition to clinical variables such as tumor burden, azotemia and hypoalbuminemia, we examined the effect of poverty on prognosis using socio-economic indices from the United States census block group data of each patient. The median survival of CPMC black and white patients was comparable (34 and 29 months, respectively) whereas that of the HHC group was 12 months (Breslow test, p < 0.0001). Overcrowding and hypoalbuminemia were the most significant prognostic factors by multivariate regression analysis on all 144 patients (p = 0.001); for HHC patients, overcrowding was the single significant variable affecting survival (p = 0.004). By all socio-economic indices, HHC patients were more impoverished than CPMC patients (p < 0.001); they also presented with more advanced disease. Race is not a significant prognostic factor in myeloma whereas the effect of socio-economic status on survival appears to equal that of previously described clinical features.

An empirical Bayes approach to outliers

Abstract A formulation of the problem of detecting outliers as an empirical Bayes problem is studied. In so doing we encounter a non-standard empirical Bayes problem for which the notion of average risk asymptotic optimality (a.r.a.o.) of procedures is defined. Some general theorems giving sufficient conditions for a.r.a.o. procedures are developed. These general results are then used in various formulations of the outlier problem for underlying normal distributions to give a.r.a.o. empirical Bayes procedures. Rates of convergence results are also given using the methods of Johns and Van Ryzin (1971, 1972).

Use of Models in Low-Dose Extrapolation

A major problem faced by regulatory agencies today is establishing acceptably low exposure levels to known or potential human carcinogens in the environment. Such regulations are required of the Environmental Protection Agency for toxic substances, of the Food and Drug Administration for food contaminants, of the Occupational Safety and Health Administration for workplace environments, and of the Consumer Products Safety Commission for hazardous consumer products.

Fixed width confidence intervals for the exponential mean survival time under random censorship

Abstract This paper provides asymptotically optimal fixed width confidence intervals for the parameter θ of an exponential distribution when the i.i.d. observations from this exponential distribution are randomly right censored.

A note on a multivariate version of Rolle's theorem and uniqueness of maximum likelihood roots

This note gives a multivariate version of Rolles theorem and shows its usefulness in establishing the uniqueness of the root of the maximum likelihood equations, the so-called maximum likelihood equation estimator. The technique is used to prove uniqueness in two situations from the literature where the original proof of uniqueness was in error.

Estimation of the subsurvival function for time-to-tumor in survival/sacrifice experiments

We propose a new closed-form nonparametric estimator for the subsurvival function for time-to-tumor. This estimator combines sacrifice data and data from natural deaths of animals without assumptions concerning the lethality of the tumor and without specification of cause of death.