PubMed Health⌕ Search

SEARCH · PubMed Health

Results for “Regression Analysis”

Explore indexed PubMed citations for clinical trials, systematic reviews and public health research. Read source abstracts and follow each citation to its original PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 577 records · Page 32Linked to original sources
PubMed2004

Regression analysis of multiple source and multiple informant data from complex survey samples.

In this tutorial, we describe regression-based methods for analysing multiple source data arising from complex sample survey designs. We use the term 'multiple-source' data to encompass all cases where data are simultaneously obtained from multiple informants, or raters (e.g. self-reports, family members, health care providers, administrators) or via different/parallel instruments, indicators or methods (e.g. symptom rating scales, standardized diagnostic interviews, or clinical diagnoses). We review regression models for analysing multiple source risk factors or multiple source outcomes and show that they can be considered special cases of generalized linear models, albeit with correlated outcomes. We show how these methods can be extended to handle the common survey features of stratification, clustering, and sampling weights. We describe how to fit regression models with multiple source reports derived from complex sample surveys using general purpose statistical software. Finally, the methods are illustrated using data from two studies: the Stirling County Study and the Eastern Connecticut Child Survey.

Data Collection↗
PubMed2000

Regression analysis of correlated binary outcomes.

The purpose of this paper is to describe and illustrate a regression approach to the analysis of correlated binary outcomes (Liang & Zeger, 1986). Ignoring the correlations between repeated observations can lead to invalid inferences. This approach extends logistic regression to account for repeated observations in each of a series of individuals. In this paper, I present a nontechnical introduction to the generalized estimating equations (GEE) approach. A fictitious example is used to demonstrate that GEE regression correctly adjusts for the correlations between repeated binary observations. The approach is illustrated with an analysis of safer sex practices among high-risk teenagers.

Adolescent↗
PubMed2002

Corneal endothelial cell loss after nonmechanical penetrating keratoplasty depends on diagnosis: a regression analysis.

PURPOSE: To assess the impact of diagnosis on corneal endothelial cell loss after nonmechanical penetrating keratoplasty (PK) with regression models. METHODS: Five hundred eighty-nine eyes [273 with keratoconus (group I; PK only), 187 with Fuchs' endothelial dystrophy (group II; 77 PK only and 110 triple procedures) and 129 with bullous keratopathy (group III; 115 pseudophakic and 14 aphakic eyes)] were included in this prospective study. The time course of the endothelial cell density after PK was assessed by specular microscopy. Endothelial cell density was analyzed in cross-sectional fashion at 6 and 24 months follow-up and in longitudinal manner (follow-up 2.4+/-1.8 years) with linear and exponential regression models (minimizing the residuum between observed and predicted endothelial cell count). Donor age, post-mortem time, storage time and ratio of organ-cultured to short-term-preserved donor corneas did not differ significantly between groups of patients. RESULTS: In a cross section, endothelial cell density in group I / group II / group III decreased from 1,959+/-499 / 1,524+/-528 / 1,526+/-670 cells/mm(2)to 1,617+/-553 / 1,222+/-520 / 1,063+/-384 cells/mm(2) (P=0.06 / 0.04 / 0.005). Endothelial cell count was significantly higher in group I than in group II (P=0.048) and significantly higher in group II than in group III (P=0.01) at 24 months. In the linear regression model, cell count decreased in group I / group II / group III by 136+/-465 / 241+/-374 / 421+/-484 cells annually. In the exponential regression model, cell count decreased in group I / group II / group III by 2.9+/-28.0% / 11.2+/-21.8% / 19.3+/-29.4% annually. Cell loss was highly significantly lower in group I than in group II ( P<0.0001) and highly significantly lower in group II than in group III (P<0.0001). CONCLUSION: Endothelial cell loss seems to be least pronounced after PK in keratoconus (2.9%), followed by Fuchs' dystrophy (11.2%) and bullous keratopathy (19.3%). These differences between diagnoses indicate migration of endothelial cells along a density gradient after PK.

Adolescent↗
PubMed1988

Families of lines: random effects in linear regression analysis.

Laboratory experiments often involve two groups of subjects, with a linear phenomenon observed in each subject. Simple linear regression as propounded in standard textbooks is inadequate to treat this experimental design, particularly when it comes to dealing with random variation of slopes and intercepts among subjects. The author describes several techniques that can be used to compare two independent families of lines and illustrates their use with laboratory data. The methods are described tutorially, compared, and discussed in the context of more sophisticated and more naive approaches to this common data-analytic problem. Technical details are supplied in APPENDIX A.

Adolescent↗
PubMed1991

Multiple regression analysis of differential response to treatment in randomized controlled clinical trials.

A multiple regression model is presented for the analysis of the components of individual change in clinical trials. Of primary interest is the condition where treatment effects vary according to patient baseline level. The model differentiates the average effects of treatment from baseline-dependent treatment effects, as well as identifying regression toward the mean and temporal effects. This model is applied in the analysis of data from a clinical trial of nutritional supplementation.

Calcium↗
PubMed1976

Replicating psychiatric ratings through multiple regression analysis: The Midtown Manhattan Restudy.

This paper examines three related questions: First, can psychiatrists' judgments be successfully predicted by multiple regression techniques? Second, assuming that they can, are such ratings a valid measure of mental health for the same sample at a later point in time? Third, what is the relation between mental health ratings made in 1954 and such subsequently reported behavioral outcomes as nervous breakdown, mental hospitalization, or seeking professional help for emotional problems? The evidence presented warrants two conclusion. (1) The computer-derived mental health ratings are an adequate substitute for the original ratings. The regression equation accounts for 69 percent of the variance in those ratings; and the computer-derived ratings behave in the same way as the psychiatrists' ratings in relation to other variables. (2) However, neither the psychiatrists' ratings nor the computer-derived ratings are very accurate in predicting subsequent self-reported behavior indicative of mental impairment.

Adult↗
PubMed1992

Linear regression analysis based on Buckley-James estimating equation.

In this note we consider the problem of drawing inference about the regression parameters in a linear model with survival data. A simple procedure based on the Buckley-James (1979, Biometrika 66, 429-436) estimating equation is proposed and illustrated with an example.

Heart Transplantation↗
PubMed1997

Hepatitis C virus infection among institutionalised psychiatric patients: a regression analysis of indicators of risk.

BACKGROUND/AIMS: Institutionalised psychiatric patients are at increased risk of developing chronic infection with hepatitis B virus (HBV). However, little information is available on transmission and epidemiology of hepatitis C virus (HCV) in this setting. The aim of this study was to identify potential risk factors of acquiring HCV infection in two large psychiatric institutions in northern Italy. METHODS: We designed a case-control study using randomly selected controls from the same study database, consisting of a total of 1180 patients, in order to satisfy the principle that both cases and controls should be representative of the same base experience. A multiple regression logistic analysis was used to identify features that could predict exposure to HCV as evidenced by the presence of circulating anti-HCV antibodies. RESULTS: Anti-HCV was detected in 79 patients (6.7%). The prevalence of viraemia and the distribution of genotypes were very similar to those found in subjects with chronic HCV infection drawn from the same geographical area. Multivariate analysis indicated that a diagnosis of psychosis and a history of trauma were statistically significant independent risk factors associated with a positive anti-HCV result (OR 2.615, 1.273-5.373 95% CI and OR 2.096, 1.133-3.877 95% CI, respectively). CONCLUSIONS: The findings of this large epidemiological study show for the first time that prolonged residence in psychiatric institutions does not entail per se a significant risk of acquiring HCV infection. Since transmission of HCV in this setting appears to occur predominantly via classical parenteral routes, simple prophylactic measures appear to be adequate to prevent infection.

Adult↗
PubMed1985

Applications of microcomputer spreadsheet packages as adjuncts to multiple logistic regression analysis.

This paper illustrates how a microcomputer spreadsheet package can be used by epidemiologists to facilitate the computation of multiple logistic regression (MLR) probabilities, as well as odds ratios and associated confidence intervals, given the coefficients of the MLR model. By formatting a spreadsheet, data entry is greatly simplified, and computations are accomplished without any arithmetic manipulations on the part of the user. This approach makes it feasible for clerical support staff to assist in the computation of seemingly complex expressions. The increasing availability of microcomputers in clinical and research settings suggests that numerous analytic applications are amenable to this approach, thereby decreasing reliance on mainframe computers and desk-top calculators.

Computers↗