Biomedical subjects

M Ewell

Publications and source records attributed to M Ewell.

4 recordsLinked to original sources

PubMed1998

Inference using conditional logistic regression with missing covariates.

When there are many nuisance parameters in a logistic regression model, a popular method for eliminating these nuisance parameters is conditional logistic regression. Unfortunately, another common problem in a logistic regression analysis is missing covariate data. With many nuisance parameters to eliminate and missing covariates, many investigators exclude any subject with missing covariates and then use conditional logistic regression, often called a complete-case analysis. In this article, we derive a modified conditional logistic regression that is appropriate with covariates that are missing at random. Performing a conditional logistic regression with only the complete cases is convenient with existing statistical packages, but it may give bias if missingness is not completely at random.

Bias↗

PubMed1997

Risk factors associated with preeclampsia in healthy nulliparous women. The Calcium for Preeclampsia Prevention (CPEP) Study Group.

OBJECTIVE: Our goal was to identify risk factors for the development of preeclampsia in nulliparous women enrolled in a multicenter trial comparing calcium supplementation to a placebo. STUDY DESIGN: A total of 4589 women from five centers was studied. Analysis of risk factors for preeclampsia was performed in 4314 who carried the pregnancy to > 20 weeks. Baseline systolic and diastolic blood pressure, demographic characteristics, and findings after randomization were examined for the prediction of preeclampsia. Preeclampsia was defined as hypertension (diastolic blood pressure > or = 90 mm Hg on two occasions 4 hours to 1 week apart) and proteinuria (> or = 300 mg/24 hours, a protein/creatinine ratio > or = 0.35, one dipstick measurement > or = 2+ or two dipstick measurements > or = 1+ at an interval as specified for diastolic blood pressure). RESULTS: Preeclampsia developed in 326 women (7.6%). The first analysis treated each risk factor as a categoric variable in a univariate regression. Maternal age, blood group and Rh factor, alcohol use, previous abortion or miscarriage, private insurance, and calcium supplementation were not statistically significant. Risk factors initially found to be significant were body mass index, systolic blood pressure, diastolic blood pressure, non-white race (African-American and other), clinical center, and smoking. Adjusted odds ratios computed with a logistic regression model revealed that body mass index (odds ratio 3.22 for > or = 35 kg/m2 vs < 19.8 kg/m2), systolic blood pressure (odds ratio 2.66 for > or = 120 vs < 101 mm Hg), diastolic blood pressure (odds ratio 1.72 for > or = 61 mm Hg vs < 60 mm Hg), and clinical center (odds ratio 1.85 for Memphis vs the other clinical centers) were statistically significant predictors of preeclampsia. Results of the final model fit revealed that preeclampsia risk increases significantly (p < 0.0001) with increased body mass index at randomization, as well as with increased systolic and diastolic blood pressure at randomization. Calcium supplementation had no effect on the risks posed by body mass index and blood pressure. Among risk factors developing after randomization, an abnormal results of a glucose screen (plasma glucose > or = 140 mg/dl 1 hour after a 50 gm glucose challenge) was not found to be associated with a significant risk of preeclampsia. CONCLUSION: These risk factors should be of value in counseling women regarding preeclampsia and should aid in understanding the pathophysiologic characteristics of this syndrome.

Adolescent↗

PubMed1997

The large sample distribution of the weighted log rank statistic under general local alternatives.

We derive the large sample distribution of the weighted log rank statistic under a general class of local alternatives in which both the cure rates and the conditional distribution of time to failure among those who fail are assumed to vary in the two treatment arms. The analytic result presented here is important to data analysts who are designing clinical trials for diseases such as non-Hodgkins lymphoma, leukemia and melanoma, where a significant proportion of patients are cured. We present a numerical illustration comparing powers obtained from the analytic result to those obtained from simulations.

Clinical Trials as Topic↗

PubMedDate not supplied

Comparing methods for calculating confidence intervals for vaccine efficacy.

A method is introduced for computing a Bayesian 95 per cent posterior probability region for vaccine efficacy. This method assumes independent vague gamma prior distributions for the incidence rates on each arm of the trial, and a Poisson likelihood for the counts of incident cases of infection. The approach is similar in spirit to the Bayesian analysis of the binomial risk ratio described by Aitchison and Bacon-Shone. However, the focus of our interest is not on incorporating prior information into the design of trials for efficacy, but rather on evaluating whether or not the Bayesian approach with vague prior information produces comparable results to a frequentist approach. A review of methods for constructing exact and large sample intervals for vaccine efficacy is provided as a framework for comparison. The confidence interval methods are assessed by comparing the size and power of tests of vaccine efficacy in proposed intermediate sized randomized double blinded placebo controlled trials.

AIDS Vaccines↗