Biostatistics, Dept. ofhttp://jhir.library.jhu.edu/handle/1774.2/328282018-01-22T14:32:28Z2018-01-22T14:32:28ZRelationship of somatosensory evoked potentials and cerebral oxygen consumption during hypoxic hypoxia in dogsTraystman, R.J.Zeger, Scott L.McPherson, R.W.http://jhir.library.jhu.edu/handle/1774.2/328652017-07-03T19:52:58Z1986-01-01T00:00:00ZRelationship of somatosensory evoked potentials and cerebral oxygen consumption during hypoxic hypoxia in dogs
Traystman, R.J.; Zeger, Scott L.; McPherson, R.W.
The effects of hypoxic hypoxia on cerebral hemodynamics and somatosensory evoked potential (SEP) were studied in 10 pentobarbital anestheteized dogs. Cerebral blood flow (CBF) was measured using the venous outflow technique and cerebral oxygen consumption (CMRO2) was calculated from the arterio-cerebro-venous oxygen difference times CBF. SEP was evaluated by percutaneous stimulation of an upper extremity nerve and was recorded over the contralateral somatosensory cortex. The latencies of the initial negative wave (N1), second positive wave (P2) and the amplitude of the primary complex (P1N1) were measured. Animals were breathed sequentially with oxygen concentrations of 21, 10, 6, 5, and 4.5% for five minutes each. Animals were returned to room air breathing when the amplitude of the SEP decreased to less than 20% of control and were observed for 30 minutes following reoxygenation. Severe hypoxia (4.5% O2) increased CBF to 200% of control, decreased CMRO2 to 45% of control, decreased amplitude and increased latency of SEP. Following reoxygenation, as CMRO2 increased toward control, latency of SEP decreased and amplitude increased and CBF returned to baseline within 30 min. During hypoxia and reoxygenation, the latencies of N1 and P2 and the amplitude of P1N1 were correlated with CMRO2 in individual animals. We conclude that changes in SEP amplitude and latency reflect changes in CMRO2 despite high CBF during rapidly progressive hypoxic hypoxia and following reoxygenation.
1986-01-01T00:00:00ZInference Based on Estimating Functions in the Presence of Nuisance ParametersZeger, Scott L.Liang, Kung-Yeehttp://jhir.library.jhu.edu/handle/1774.2/328642017-07-03T19:52:58Z1995-01-01T00:00:00ZInference Based on Estimating Functions in the Presence of Nuisance Parameters
Zeger, Scott L.; Liang, Kung-Yee
In many studies, the scientific objective can be formulated in terms of a statistical model indexed by parameters, only some of which are of scientific interest. The other "nuisance parameters" are required to complete the specification of the probability mechanism but are not of intrinsic value in themselves. It is well known that nuisance parameters can have a profound impact on inference. Many approaches have been proposed to eliminate or reduce their impact. In this paper, we consider two situations: where the likelihood is completely specified; and where only a part of the random mechanism can be reasonably assumed. In either case, we examine methods for dealing with nuisance parameters from the vantage point of parameter estimating functions. To establish a context, we begin with a review of the basic concepts and limitations of optimal estimating functions. We introduce a hierarchy of orthogonality conditions for estimating functions that helps to characterize the sensitivity of inferences to nuisance parameters. It applies to both the fully and partly parametric cases. Throughout the paper, we rely on examples to illustrate the main ideas.
1995-01-01T00:00:00ZMarginalized Multilevel Models and Likelihood InferenceZeger, Scott L.Heagerty, Patrick J.http://jhir.library.jhu.edu/handle/1774.2/328632017-07-03T19:52:58Z2000-01-01T00:00:00ZMarginalized Multilevel Models and Likelihood Inference
Zeger, Scott L.; Heagerty, Patrick J.
Hierarchical or "multilevel" regression models typically parameterize the mean response conditional on unobserved latent variables or "random" effects and then make simple assumptions regarding their distribution. The interpretation of a regression parameter in such a model is the change in possibly transformed mean response per unit change in a particular predictor having controlled for all conditioning variables including the random effects. An often overlooked limitation of the conditional formulation for nonlinear models is that the interpretation of regression coefficients and their estimates can be highly sensitive to difficult-to-verify assumptions about the distribution of random effects, particularly the dependence of the latent variable distribution on covariates. In this article, we present an alternative parameterization for the multilevel model in which the marginal mean, rather than the conditional mean given random effects, is regressed on covariates. The impact of random effects model violations on the marginal and more traditional conditional parameters is compared through calculation of asymptotic relative biases. A simple two-level example from a study of teratogenicity is presented where the binomial overdispersion depends on the binary treatment assignment and greatly influences likelihood-based estimates of the treatment effect in the conditional model. A second example considers a three-level structure where attitudes toward abortion over time are correlated with person and district level covariates. We observe that regression parameters in conditionally specified models are more sensitive to random effects assumptions than their counterparts in the marginal formulation.
2000-01-01T00:00:00ZFeedback Models for Discrete and Continuous Time SeriesLiang, Kung-YeeZeger, Scott L.http://jhir.library.jhu.edu/handle/1774.2/328622017-07-03T19:52:58Z1991-01-01T00:00:00ZFeedback Models for Discrete and Continuous Time Series
Liang, Kung-Yee; Zeger, Scott L.
In public health research, it is common to follow a cohort of subjects over time, observing a vector of health indicators and a set of covariates at each of many visits. An objective of analysis is to characterize the inter-dependencies, in particular, the feedback of one response upon another while accounting for the covariates. With Gaussian responses, multivariate autoregressive models that incorporate feedback are commonly used. This paper discusses analogous Markov models for multivariate discrete and mixed discrete/continuous response variables. One special case is an extension of seemingly unrelated regressions to discrete and continuous outcomes. A generalized estimating equations approach that requires correct specification of only conditional means and variances is discussed. The methods are illustrated by a study of infectious diseases and vitamin A deficiency in Indonesian children.
1991-01-01T00:00:00Z