Biomedical subjects

Scott Zeger

Publications and source records attributed to Scott Zeger.

5 recordsLinked to original sources

PubMed2004

Underestimation of standard errors in multi-site time series studies.

Multi-site time series studies of the association of air pollution with mortality and morbidity have figured prominently in the literature as comprehensive approaches for estimating short-term effects of air pollution on health. Hierarchical models are generally used to combine site-specific information and to estimate pooled air pollution effects while taking into account both within-site statistical uncertainty and across-site heterogeneity. Within a site, characteristics of time series data of air pollution and health (small pollution effects, missing data, and highly correlated predictors) make the modeling of all sources of uncertainty challenging. One potential consequence is underestimation of the statistical variance of the site-specific effects to be combined.In this paper, we investigate the impact of variance underestimation on the pooled relative rate estimate. We focused on two-stage normal-normal hierarchical models and on underestimation of the statistical variance at the first stage. By mathematical considerations and simulation studies, we found that variance underestimation did not affect the pooled estimate substantially. However, the pooled estimate was somewhat sensitive to variance underestimation when the number of sites was small and underestimation was severe. These simulation results are applicable to any two-stage normal-normal hierarchical model for combining information of site-specific results (including meta-analyses), and they can easily be extended to more general hierarchical formulations. We also examined the impact of variance underestimation on the national average relative rate estimate from the National Morbidity, Mortality and Air Pollution Study. We found that variance underestimation as large as 40% had little effect on the national average.

Air Pollutants↗

PubMed2002

Methods for evaluating the performance of diagnostic tests in the absence of a gold standard: a latent class model approach.

In many areas of medical research, 'gold standard' diagnostic tests do not exist and so evaluating the performance of standardized diagnostic criteria or algorithms is problematic. In this paper we propose an approach to evaluating the operating characteristics of diagnoses using a latent class model. By defining 'true disease' as our latent variable, we are able to estimate sensitivity, specificity and negative and positive predictive values of the diagnostic test. These methods are applied to diagnostic criteria for depression using Baltimore's Epidemiologic Catchment Area Study Wave 3 data.

Baltimore↗

PubMed2002

Growth curve analyses of neuropsychological profiles in children with neurofibromatosis type 1: specific cognitive tests remain "spared" and "impaired" over time.

Cognitive deficits in neurofibromatosis Type 1 (NF-1) have been documented in both the verbal and visuospatial domains. Previous investigations from our laboratory have determined a specific pattern of "spared" (Picture Arrangement, Picture Completion, and Rapid Automatized Naming) and "impaired" (Judgment of Line Orientation, Vocabulary, and Block Design) performance on cognitive measures in this population when compared to sibling-matched controls in pairwise designs. Growth curve analyses were conducted on these repeated measures in 19 patients with NF-1 and their siblings to investigate the longitudinal course and growth pattern of these spared and impaired measures. Results indicated that over time children with NF-1 do not catch up to their siblings on impaired measures, and they continue to perform similarly to their siblings on the spared measures. With respect to growth rates, on average across the 6 cognitive measures there was no significant difference between the groups. However, the variation among families for level of performance was estimated to be larger than variation among siblings within a family for 2 out of 6 cognitive measures (i.e., providing for these 2, Vocabulary and Rapid Automatized Naming, evidence of substantial familial correlation), suggesting that there is need to consider NF-1 associated deficits within a familial context.

Adolescent↗

PubMed2002

SNOMAD (Standardization and NOrmalization of MicroArray Data): web-accessible gene expression data analysis.

SNOMAD is a collection of algorithms for the normalization and standardization of gene expression datasets derived from diverse biological and technological sources. In addition to conventional transformations and visualization tools, SNOMAD includes two non-linear transformations which correct for bias and variance which are non-uniformly distributed across the range of microarray element signal intensities: (1). Local mean normalization; and (2). Local variance correction (Z-score generation using a locally calculated standard deviation).

Algorithms↗

PubMed2002

Local mean normalization of microarray element signal intensities across an array surface: quality control and correction of spatially systematic artifacts.

Here we present a methodology for the normalization of element signal intensities to a mean intensity calculated locally across the surface of a DNA microarray. These methods allow the detection and/or correction of spatially systematic artifacts in microarray data. These include artifacts that can be introduced during the robotic printing, hybridization, washing, or imaging of microarrays. Using array element signal intensities alone, this local mean normalization process can correct for such artifacts because they vary across the surface of the array. The local mean normalization can be usedfor quality control and data correction purposes in the analysis of microarray data. These algorithms assume that array elements are not spatially ordered with regard to sequence or biological function and require that this spatial mapping is identical between the two sets of intensities to be compared. The tool described in this report was developed in the R statistical language and is freely available on the Internet as part of a larger gene expression analysis package. This Web implementation is interactive and user-friendly and allows the easy use of the local mean normalization tool described here, without programming expertise or downloading of additional software.

Algorithms↗