Search PubMed⌕ Search

Biomedical subjects

M R Segal

Publications and source records attributed to M R Segal.

At least 19 recordsLinked to original sources
PubMed1997

Radiological evaluation of lymph node metastases in patients with cervical cancer. A meta-analysis.

OBJECTIVE: To apply meta-analysis to compare the utility of lymphangiography (LAG), computed tomography (CT), and magnetic resonance (MR) imaging for the diagnosis of lymph node metastasis in patients with cervical cancer. DATA SOURCES: MEDLINE literature search and manual reviews of article bibliographies. STUDY SELECTION: Studies selected included at least 20 patients with imaging-histologic correlation, described diagnostic criteria for lymph node metastasis, and presented data to allow calculation of contingency tables. DATA EXTRACTION: Independently by 2 investigators, stratified for stage of disease (early vs late) and for lymph node location (pelvic vs para-aortic). DATA SYNTHESIS: Seventeen studies met the inclusion criteria for LAG, 17 for CT, and 10 for MR imaging. Summary receiver operator characteristic analysis showed no significant differences in the overall performance of LAG, CT, and MR imaging. There was, however, a trend toward better performance for MR imaging than for LAG or CT, both globally and when stratified for stage of disease or for lymph node location. Bayesian analysis of clinical utility showed only moderate increases in positive posttest probability of lymph node metastasis for all methods. Negative test results had a greater impact and, depending on the clinical setting, decreased the probability of lymph node metastasis from 15% to 44% (pretest) to 3% to 18% (posttest). CONCLUSIONS: The LAG, CT, and MR imaging perform similarly in the detection of lymph node metastasis from cervical cancer. As CT and MR imaging are less invasive than LAG and also assess local tumor extent, they should be considered the preferred adjuncts to clinical evaluation of invasive cervical cancer.

Female↗
PubMed1997

Maternal smoking during pregnancy and birth outcomes with weight gain adjustments via varying-coefficient models.

There is considerable interest in the impact of maternal exposures during pregnancy on birth outcomes. Clearly, exposures associated with poor birth outcomes need modification or avoidance. However, arriving at such estimates of association is made challenging by a number of features characteristic of the relevant data. First, exposures may be time varying (for example, cigarette and alcohol consumption) so that, to relate them to birth outcomes, one needs to model them and then extract derived parameters. Secondly, there are likely to be unequal numbers and spacings of exposure determinations during pregnancy. Thirdly, one needs to account for a variety of additional covariates. Finally, the variability and non-linearities inherent in birth outcomes mandate flexible modelling approaches. Here we use data from a cohort of East Boston mothers to assess the impact of smoking during pregnancy on birth weights. We emphasize modelling of, and then adjusting for, maternal weight gain during pregnancy and a proxy measure for pre-pregnancy weight, so as to obtain better estimates of the smoking effect. Throughout, our analysis is guided by appropriate graphics. The adjustment features an interesting application of varying-coefficient models. Results indicate that smoking related deficits in birth weights depend on the mode of adjustment, and that previously observed deficits of approximately 200 g are best recaptured with use of varying-coefficient models.

Adult↗
PubMed1997

Dependence estimation for marginal models of multivariate survival data.

We have previously (Segal and Neuhaus, 1993) devised methods for obtaining marginal regression coefficients and associated variance estimates for multivariate survival data, using a synthesis of the Poisson regression formulation for univariate censored survival analysis and generalized estimating equations (GEE's). The method is parametric in that a baseline survival distribution is specified. Analogous semiparametric models, with unspecified baseline survival, have also been developed (Wei, Lin and Weissfeld, 1989; Lin, 1994). Common to both these approaches is the provision of robust variances for the regression parameters. However, none of this work has addressed the more difficult area of dependence estimation. While GEE approaches ostensibly provide such estimates, we show that there are problems adopting these with multivariate survival data. Further, we demonstrate that these problems can affect estimation of the regression coefficients themselves. An alternate, ad hoc approach to dependence estimation, based on design effects, is proposed and evaluated via simulation and illustrative examples.

Algorithms↗
PubMed1997

An empirical comparison of multivariable methods for estimating risk of death from coronary heart disease.

BACKGROUND: Logistic regression and, more recently, Cox regression have been the predominant methods for identifying risk factors and developing risk estimation equations for coronary heart disease (CHD). Software for the regression tree method is now available for binary and survival outcomes and thus offers an alternative methodology. This paper compares these four methods for identifying significant risk factors from among a set of candidate factors and for estimating the risk of death from CHD using baseline and mortality follow-up data on 1,701 men participating in the Busselton Health Study. The candidate risk factors were age, body mass index, systolic and diastolic blood pressure, treatment for hypertension, cholesterol and smoking. METHODS: Logistic regression, Cox proportional hazards regression, binary regression tree, and survival regression tree analyses have been applied to data obtained from the same cohort of men for CHD death risk estimation and prediction. The four methods are compared in terms of the variables selected, goodness-of-fit of models, similarity of cross-validated estimated risks for individuals, and ability to discriminate between those who died from CHD and those who did not die from CHD during the follow-up period, including the comparison of Receiver Operating Characteristic (ROC) curves. RESULTS: Although age and a blood pressure variable were selected by all four methods, body mass index was also selected by the regression tree methods and smoking was also selected by Cox regression. There was good, but not excellent, agreement between methods in estimates of risk for individuals, the areas under the ROC curves were 0.66 for the binary tree, 0.72 for logistic regression, 0.71 for the survival tree method and 0.78 for Cox regression. The average differences in estimated risk between those who died from CHD and those who did not die from CHD during the follow-up period were 0.051 for logistic regression, 0.070 for the binary tree method, 0.073 for the survival tree method and 0.088 for Cox regression. CONCLUSION: For a moderately sized cohort typical of many applications of these methods in the literature, the two methods which used the survival outcome performed better than the methods using a binary outcome. Despite selecting some different variables and showing moderate differences in risk estimates for individuals, the two binary approaches were similar in performance. Cox regression appeared to be superior to the survival tree method, but further larger studies of completely separate samples for model development and evaluation of prediction performance are required to confirm this finding.

Age Factors↗
PubMed1995

Relation of FEV1 and peripheral blood leukocyte count to total mortality. The Normative Aging Study.

Increased peripheral blood leukocyte count and decreased level of pulmonary function have both been implicated as causes of increased total mortality in population-based studies. The extent to which these factors are independent of cigarette smoking is controversial. The authors explored the relation of leukocyte count and the level of forced expiratory volume in 1 second to total mortality in the Normative Aging Study population in the Boston, Massachusetts, area. Other covariates examined included forced vital capacity, height, body mass index, systolic and diastolic blood pressure, and total cholesterol. The sample for the current analysis consisted of 1,956 men who underwent the baseline Normative Aging Study examination during 1961-1969. Subjects ranged in age from 21 to 80 years of age at the time of entry. A total of 170 deaths occurred over the 30 years of follow-up. Statistical analysis was conducted utilizing Cox proportional hazards modeling and regression trees for censored survival data. The Cox proportional hazards model suggested that age, forced expiratory volume in 1 second, and peripheral blood leukocyte count were the three most important predictors of increased mortality in this cohort. A regression tree analysis in general confirmed these results. Both methods of analysis suggest that forced expiratory volume in 1 second and peripheral blood leukocyte count were predictors of mortality, independent of cigarette smoking.

Adult↗
PubMed1995

Survival trees with time-dependent covariates: application to estimating changes in the incubation period of AIDS.

Tree-structured methods for exploratory data analysis have previously been extended to right-censored survival data. We further extend these methods to allow for truncation and time-dependent covariates. We apply the new methods to a data set on incubation times of acquired immunodeficiency syndrome (AIDS), using calendar time as a time-dependent covariate. Contrary to expectation, we find that rates of progression to AIDS appear to be faster after August 1989 than before.

Adult↗
PubMed1995

Extending the elements of tree-structured regression.

The usage of tree-structured or recursive partitioning methods has grown steadily in the decade since the appearance of the definitive monograph 'Classification and Regression Trees'. Accompanying this growth have been many methodologic and software extensions that have served to give the tree-structured approach even wider applicability. This overview highlights some of these developments, emphasizing the regression setting. An illustrative example describes how tree-structured regression, modified to handle right-censored, left-truncated survival data with time-dependent covariates, can be used to assess whether rates of progression from HIV to AIDS have changed over time.

Acquired Immunodeficiency Syndrome↗
PubMed1993

Trees and tracking.

Epidemiologists have used the term 'tracking' to connote an individual's maintenance of relative rank of some longitudinally measured characteristic over a given time span. To assess the extent to which an attribute tracks we have first to summarize individual growth curves, and second to quantify the notion of maintenance of relative rank, both in the face of random error. A sequence of papers appearing in 1981 provided differing methodologies for appraising tracking. Here we take a different approach to tracking by using regression trees for longitudinal data. The above two concerns are simultaneously addressed in that the procedure identifies subgroups, defined in terms of covariates, within which the collection of growth curves is homogeneous. After reviewing the existing approaches to tracking we describe the tree-structured methodology, and present an illustrative example pertaining to lung function growth in children.

Bias↗
PubMed1993

Design effects for binary regression models fitted to dependent data.

Dependent data, such as arise with cluster sampling, typically yield variances of parameter estimates which are larger than would be provided by a simple random sample of the same size. This variance inflation factor is called the design effect of the estimator. Design effects have been derived for cluster sampling designs using simple estimators such as means and proportions, and also for linear regression coefficient estimators. In this paper, we show that a method to derive design effects for linear regression estimators extends to generalized linear models for binary responses. In particular, some simple expressions for design effects in the linear regression model provide accurate approximations for binary regression models such as those based on the logistic, probit and complementary log-log links. We corroborate our findings with two examples and some simulation studies.

Acquired Immunodeficiency Syndrome↗
PubMed1993

Robust inference for multivariate survival data.

Multivariate survival data arise when an individual records multiple survival events or when individuals recording single survival events are grouped into clusters. In this paper we propose a new method for the analysis of multivariate survival data. The technique is a synthesis of the Poisson regression formulation for univariate censored survival analysis and the generalized estimating equation approach for obtaining valid variance estimates for generalized linear models in the presence of clustering. When the survival data are clustered, combining the methods provides not only valid estimates for the variances of regression parameters but also estimates of the dependence between survival times. The approach entails specifying parametric models for the marginal hazards and a dependence structure, but does not require specification of the joint multivariate survival distribution. Properties of the methodology are investigated by simulation and through an illustrative example.

Humans↗
PubMed1993

Different AIDS incubation periods and their impacts on reconstructing human immunodeficiency virus epidemics and projecting AIDS incidence.

The objective of this study was to investigate heterogeneity in incubation distributions in different cohorts and to assess the sensitivity of back-calculated infection rates to different assumptions about incubation times from human immunodeficiency virus (HIV) infection to AIDS diagnosis. Incubation distributions were estimated by using data from three different cohort studies. These and one other published incubation model were used as inputs for a back-calculation procedure that reconstructed smooth HIV-infection rates from AIDS incidence among adults in the United States, allowing for changes over time in incubation. Incubation estimates from the different cohorts differed substantially. The cumulative HIV incidence estimates that result from using the different incubations are very different, but the back-calculated models all produce good fits to the observed diagnosis counts. We conclude that systematic differences in incubation times of different groups add substantially to the uncertainty inherent in using the back-calculation method to reconstruct HIV epidemics and project future numbers of AIDS cases.

Acquired Immunodeficiency Syndrome↗
PubMed1992

An annotated bibliography of methods for analysing correlated categorical data.

This paper provides an annotated bibliography of over 100 articles concerning methods for analysing correlated categorical response data. Most of the papers listed here concern categorical regression models and estimation, with particular emphasis on binary responses. The papers are classified by several characteristics which group them according to common themes. The bibliography serves as a reference of methods for analysts of correlated categorical data, as well as for persons interested in methodologic work in this active area of statistical research.

Clinical Trials as Topic↗
PubMed1992

The relationship of nasal disorders to lower respiratory tract symptoms and illness in a random sample of children.

We examined the relationship of nasal disorders, defined as frequent colds and sinus trouble, to lower respiratory tract symptoms in a random population of 718 children aged 4 to 11 years in East Boston, Massachusetts. Frequent colds were significantly associated with maternal smoking (odds ratio (OR) = 3.00; 95% confidence interval (CI) = 1.97, 4.58), and so was sinus trouble (OR = 4.73; 95% CI = 1.78, 12.51). After adjustment for maternal smoking, age and sex, frequent colds (OR = 2.88; 95% CI = 1.87, 4.42) and sinus trouble (OR = 4.95, 95% CI = 1.83, 13.39) remained significant predictors of lower respiratory tract symptoms in separate logistic regressions. If one restricted the cohort to the 513 children who also had personal smoking information and adjusted for this variable as well, the results for colds were unchanged (OR = 2.94; 95% CI = 1.78, 4.84) but the results for sinus trouble were now not statistically significant (OR = 2.30, 95% CI = 0.67, 7.94). We conclude that nasal disorders are associated with lower respiratory tract symptoms in children.

Child↗
PubMed1992

Effects of asthma on pulmonary function in children. A longitudinal population-based study.

Data from a longitudinal study of childhood factors influencing the development of chronic obstructive lung disease were used to assess the effects of asthma on lung function development in male and female children. A population-based cohort of 602 white children, initially aged 5 to 9 yr, was observed prospectively for 13 yr. Spirometry was performed and a standardized respiratory and illness questionnaire was administered by trained interviewers on a yearly basis. Forced vital capacity (FVC), forced expiratory volume in one second (FEV1), and forced expiratory flow between 25 and 75% of vital capacity (FEF25-75) were used as measures of lung function. The total number of children reporting asthma over the course of the study was 67. Male asthmatic subjects (n = 42) had larger average percentage of predicted FVC than nonasthmatic males (n = 277). Female asthmatic subjects (n = 23) had a lower average percentage of predicted FEV1 than nonasthmatic females (n = 260). In a multivariate analysis of the individual lung function measures, adjusting for previous level of pulmonary function, age, height, change in height, and personal and maternal smoking, males reporting active asthma had a significantly larger FVC than males with no history of asthma. In contrast, females with active asthma had a significantly smaller FEV1 than females with no history of asthma. Both males and females with active asthma had decreased FEF25-75. From our analysis, we would predict that a female who develops asthma at age 7 would experience a 5% reduction in FEV1 by age 10 and a 7% deficit by age 15.(ABSTRACT TRUNCATED AT 250 WORDS)

Asthma↗
PubMed1992

The effect of maternal smoking during pregnancy on early infant lung function.

We studied the effect of prenatal maternal cigarette smoking on the pulmonary function (PF) of 80 healthy infants tested shortly after birth (mean, 4.2 +/- 1.9 wk). Mothers' prenatal smoking was measured by: (1) questionnaire reports at each prenatal visit of the number of cigarettes smoked per day, and (2) urine cotinine concentrations (corrected for creatinine) obtained at each visit. Infant PF was assessed by partial expiratory flow-volume curves and helium-dilution measurement of FRC. Forced expiratory flow rates were significantly lower in infants born to smoking mothers, both when unadjusted and after controlling for infant size, age, sex, and passive exposure to environmental tobacco smoke (ETS) between birth and the time of PF testing. Flow at functional residual capacity (VFRC) in infants born to smoking mothers was lower than that found in infants whose mothers did not smoke during pregnancy (74.3 +/- 15.9 versus 150.4 +/- 8.9 ml/s; p = 0.0007). Differences remained significant when flow was corrected for lung size (VFRC/FRC: 0.87 +/- 0.26 versus 1.77 +/- 0.12 s-1; p = 0.013). No differences in pulmonary function were evident among infants exposed and unexposed to ETS in the home after stratifying by prenatal exposure status. We conclude that maternal smoking during pregnancy is associated with significant reductions in forced expiratory flow rates in young infants. The results suggest that maternal smoking during pregnancy may impair in utero airway development and/or alter lung elastic properties. We speculate that these effects of maternal prenatal smoking on early levels of forced expiratory flow may be an important factor predisposing infants to the occurrence of wheezing illness later in childhood.

Adult↗
PubMed1992

Living arrangements and survival among middle-aged and older adults in the NHANES I epidemiologic follow-up study.

BACKGROUND: There is concern about but little information on how living alone affects the health and survival of older adults. METHODS: We examined the association between living arrangements (living alone, with a spouse, or with someone other than a spouse) and survival among 7651 adults, aged 45 to 74 years in the National Health and Nutrition Examination Survey (NHANES I) (1971-1975) and traced at the NHANES I Follow-up Study (1982-1984), to see whether certain sociodemographic factors (race, education, income, and employment), health behaviors (alcohol, smoking, physical activity, and obesity), or chronic medical conditions were influential in the association. RESULTS: We found a stronger association of living arrangements with survival for men than for women, and for middle-aged men than for older men. For men, those living alone and those living with someone other than a spouse were equally disadvantaged in terms of survival. Income, race, employment, and physical activity influenced the association of living arrangements and survival, but their impact varied by age, gender, and living arrangement. CONCLUSION: Living arrangements had a weak impact on survival among men, but had no effect among women.

Age Factors↗
PubMed1992

Correlates of knee pain among US adults with and without radiographic knee osteoarthritis.

We examined the associations of sociodemographic variables, health behaviors, health status and psychological well being with radiographic knee osteoarthritis (OA) and self-reported knee pain for 4056 US adults aged 45-74 years. Among persons with or without knee OA known correlates of radiographic knee OA (age, sex, race, obesity) were generally not associated with knee pain. Radiographic severity, psychological well being and health status were associated with knee pain, both among persons with and without radiographic knee OA, suggesting that nonradiographic correlates of self-reported knee pain are independent of whether a person has radiographic knee OA.

Age Factors↗