Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “Statistical Regression”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

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

At least 793 records · Page 44Linked to original sources
PubMed1980

Quantitative aspects of electrocardiogram in a Jordanian population.

Twelve lead electrocardiographic tracings were obtained from 297 subjects aged 15-64 years, selected randomly from the Jordan University Hospital staff members, student doctors and student nurses. The values of selected measurements are reported including intervals, amplitudes and the mean QRS electrical axis in the frontal plane. The relation between certain electrocardiographic characteristics and age, body built, arterial pressure and heart rate was investigated by multiple regression analysis. Statistical analysis of the difference of the data obtained in age groups was made by the "Student's" t-distribution.

Adolescent↗
PubMed2005

Asteroid hyalosis in an autopsy population: The University of California at Los Angeles (UCLA) experience.

OBJECTIVES: To study the prevalence and associations of asteroid hyalosis (AH) in a series of autopsy eyes. METHODS: Retrospective review of the University of California at Los Angeles (UCLA) autopsy eye database from 1965 to 2000 yielded 10,801 patients. The patients' medical histories were reviewed for evidence of diabetes mellitus, hypertension, hyperlipidemia, alcohol abuse, hypercalcemia, hypothyroidism, and chronic renal failure. Autopsy records were searched for evidence of optic atrophy, macular degeneration, posterior vitreous detachment, atherosclerosis, and chronic renal failure. Asteroid hyalosis was diagnosed by examination of the autopsy eyes. Univariate and multivariate statistical methods were used to analyze our data. RESULTS: The prevalence of AH was 1.96% in this autopsy population. By chi(2) analysis, AH was significantly correlated with age (P<.001), male sex (P = .006), age-related macular degeneration (P = .02), hypertension (P = .03), atherosclerosis (P<.001), and posterior vitreous attachment (P<.001). After adjusting for age in a multivariate logistic regression analysis, statistical significance was found only for posterior vitreous attachment (P = .002) and male sex (P = .046). No statistically significant association was found with diabetes mellitus or alcohol abuse by univariate or multivariate analysis. Analysis of the odds ratio showed a strong age effect that increased from 5.0 (95% confidence interval, 2.2-11.3) in age group 41 to 50 years, compared with 25.4 (95% Wald confidence interval, 8.2-77.9) in the age group of patients older than 90 years. CONCLUSIONS: A unique epidemiological autopsy cohort study of AH and its systemic associations yielded a higher prevalence of AH than previous studies. Asteroid hyalosis was strongly correlated with age and inversely correlated with posterior vitreous detachment. Unlike some previous reports, we found no statistically significant correlation between AH and diabetes mellitus.

Academic Medical Centers↗
PubMed1992

The association between wasting and stunting: an international perspective.

Wasting and stunting are widely used indices for describing the prevalence of malnutrition in childhood. Studies from Latin America, however, consistently show low prevalences of wasting in the presence of other indicators of poor health. A data base assembled by the World Health Organization, including 175 studies with information on wasting and stunting among 1-y-old children from four regions (Africa, Americas, Asia and Eastern Mediterranean), was used for examining the prevalences of the two indices and their associations. Whereas there was little variation in the median prevalences of stunting among the four regions, those of wasting showed a sevenfold differential. In Africa and Latin America, the correlations between the two prevalences were low; nevertheless, wasting was about three times higher in Africa, for any given level of stunting, than in Latin America. In Asia and the Eastern Mediterranean, there were strong positive correlations between stunting and wasting, but wasting was about twice as common in Asia for any given level of stunting. To interpret low prevalences of wasting as being indicative of a low intensity of malnutrition may be misleading; weight-for-height standards and cut-offs may have to be re-examined for use in different populations.

Africa↗
PubMed1991

In utero exposure to steroid contraceptives and survival during infancy.

A cohort study was conducted in Chiang Mai, northern Thailand, in 1,431 children of women who had used the injectable contraceptive Depo-Provera (The Upjohn Company, Kalamazoo, Michigan), 565 children of women who had used oral contraceptives during pregnancy, and a group of 2,307 control infants with no hormonal contraceptive exposures. In follow-up interviews, information was obtained on stillbirths and deaths. Cause of death was ascertained by interview, death certificate, or medical record, and underlying causes of death were ascribed by a panel. The children exposed in utero to Depo-Provera had higher neonatal and infant mortality rates (44.3 and 62.9 per 1,000 live births, respectively) than did the controls (19.8 and 29.1 per 1,000 live births). Mortality in infants exposed in utero to oral contraceptives was intermediate between that in the other two groups. Adjustment by logistic regression showed no significantly increased risk of mortality among infants exposed to oral contraceptives, but the odds ratio for death was significantly increased with Depo-Provera exposures due to accidental pregnancy (odds ratio (OR) = 1.8 (95% confidence interval (Cl) 1.1-3.0) for neonatal deaths; OR = 2.0 (95% Cl 1.3-3.2) for infant deaths). Adjustment for low birth weight reduced the risks, suggesting that low birth weight may act as an intermediate determinant of Depo-Provera-associated mortality. Among the accidental pregnancies with Depo-Provera, there was a relation between shorter injection-to-conception intervals, when maternal blood levels of the drug are high, and an increased risk of mortality. The odds ratios for neonatal mortality were 2.5 (95% Cl 1.1-5.7), 2.1 (95% Cl 1.0-4.6), and 0.9 (95% Cl 0.4-2.4) for injection-to-conception intervals of less than or equal to 4, 5-8, and greater than 9 weeks, respectively. Adjustment for low birth weight reduced these risks. Chi-square tests for trend were highly significant. Similar associations were also observed between Depo-Provera accidental pregnancies and risks of low birth weight. Thus, infants from accidental pregnancies that occur 1-2 months after a 150-mg Depo-Provera injection may be at increased risk for low birth weight and death. However, the attributable risk is low, because such pregnancies are uncommon.

Cause of Death↗
PubMedDate not supplied

Prediction of post-operative survival time by multivariate analysis in patients with advanced cancer of the stomach.

In order to give a sound basis to a surgeon for the decision that a cancer should or should not be removed, an attempt was made to produce a regression equation to estimate postoperative survival. The relation between postoperative survival and the intraoperative gross-finding of the cancer, the age and sex of the patients, operative curability, histological types of the cancer and the need for combined resections of the adjacent organs were analyzed by the method of multivariate analysis (Quantification Method) in 111 patients with advanced cancer of the stomach. A statistically significant regression equation was obtained, by which the survival months were predicted for a prospective patient. This equation appears to be useful for surgeons as one way of deciding on an appropriate method of operation. The difference between the actual and the predicted survival months was less than twelve months in seventy-six percent of the internal samples.

Adult↗
PubMed2003

An overview of the variables selection methods for the minimum sum of absolute errors regression.

The minimum sum of absolute errors regression is a robust alternative to the least squares regression whenever the errors follow a distribution for which the sample median is a more efficient estimator of location parameter than the sample mean, the errors follow a long tailed distribution, there are outliers in the values of the response variable in the data or the absolute error loss function is more appropriate than the quadratic loss function. Often an initial model may contain a large number of variables. However, in many situations, it is neither necessary nor important to include all the variables in the model. The methods for variable selection for the minimum sum of absolute errors regression are not as well documented and known as for the least squares regression. Our objective is to present an overview of the procedures to fit models with fewer variables and some criteria for selecting a model.

Algorithms↗
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↗
PubMed1981

Predicting weight loss success among obese clients in a hospital nutrition clinic.

A retrospective study of 63 obese patients seen in a hospital outpatient nutrition clinic was reported. The majority suffered from one or more other medical problems. The relationship of amount of and rate of weight loss to various factors was examined. The multivariate analysis revealed that either the amount or rate of weight loss was greater in the patients who were more overweight initially, white, male, young, single, older at age of onset of obesity, visited the clinic for a longer period of time but less frequently. Conversely, the amount or rate of weight loss was less in those patients receiving conjugated estrogens, oral contraceptives, propranolol, thioridazine HCl, and those who were receiving behavior modification as treatment.

Adolescent↗
PubMed1981

Modelling determinants of health.

The linear regression procedure of stepwise elimination has been used to examine the association between a range of environmental variables and infant mortality rates (IMR) in Sri Lanka. The nature of the water supply was found to be significantly associated with IMR. Latrines were significantly associated with IMR once qualitative differences between districts were allowed for. Other 'environmental determinants of health', 3 indices of housing quality and the presence of electricity supplies, were not significant. The cost of effecting changes in IMR by extending subsidies to water supply improvements is estimated.

Developing Countries↗
PubMed1991

Evaluation of how age modifies the risk for Parkinson's disease, based on stratified comparisons of descriptive data.

This is a study of selected information on incidence and prevalence of Parkinson's disease (PD), mortality from Parkinsonism, and levodopa use (LDU). The concept of effect modification (EM) by age (EMA), a variation across age of incidence or odds ratios, was here extended to comparisons of prevalences, mortalities and drug use, and mathematically defined following a reported model. Age-specific EM from comparisons of empirical data was measured and combined with figures for calendar time and the annual infant mortality rates (IMRs) in the population of each observation, using linear regression analysis. Statistically significant associations of EM at ages 70-79 and 80+ were found: with IMR from prevalence comparisons; and with IMR and calendar time in comparisons of proportionate mortalities in longitudinal observations. A similar age-specific pattern of associations was observed with time from incidence in Rochester. The observations for LDU and PD incidence in Iceland and Turku, Finland, fitted well the age-specific regression lines of the IMRs with the age-specific EM components from prevalence data. We concluded that there is a universal age-specific pattern of variation of epidemiological characteristics of PD related to calendar time, and to local factors represented by the IMR. Moreover, this change can be perceived from different parameters under certain conditions. Comparisons of survey, death-records, and LDU data when used for generation of causal hypotheses for PD require that EMA be taken into account.

Age Factors↗
PubMed1994

[An industrial toxicological study on workers who synthesize aminophenol and anisidine from chloronitrobenzene].

Aromatic nitro- and amino-compounds (ANA) are notorious methemoglobin forming substances. From 1987 to 1993, the author surveyed the exposure level to ANA and health status of workers in a chemical industry where they synthetize nitrophenol and anisidine from chloro-nitrobenzene. The ANA exposure indicator was the diazo-reaction positive substance in the worker's urine (Dz) and the biological effect of ANA was monitored through the methemoglobin concentration in the blood (MHb). The following results were obtained from the survey. 1. Both the Dz and MHb values tended to become lower between 1987 and 1993. 2. The MHb levels of the workers were a little higher than the control levels, and the difference between the values was statistically significant (p < 0.01); both levels were, however, within the normal range. 3. The MHb value was estimable from a mathematical model: MHb = (k1-k2/a) x DU x CH, where k1 and k2 were constants and a, DU and CH indicated the activity of glutathione S transferase in erythrocytes (GST), DZ in urine and blood hemoglobin concentration, respectively. The correlation coefficient of the multiple regression was statistically significant (R2 = 0.263, p < 0.01). This model suggests that the biotransformed fraction of ANA in the pulmonary uptake site, the lung, could not penetrate into erythrocytes or act as an MHb former. The model seemed to be consistent and valid when urinary DZ values were less than 1.0 (mg/mg creatinine) and MHb values were less than 2.0 (mg/ml blood). Thus, the upper limit of normal levels for MHb can be considered to be 2.0 (mg/ml blood), or 12-1.6% of total hemoglobin in blood. 4. The significance of the uptake sites, i.e., lung, skin etc., was discussed with reference to the evidence obtained from the longshoremen cases, who were percutaneously exposed to chloronitrobenzene in 1984. 5. Hemolysate solutions from the blood of workers were applied to isoelectric focusing electrophoresis (IEF), and Hb species of different valencies were separated; intermediate MHbs, where the heme iron in one of the components of Hb, i.e., in the alpha or beta chain was oxidized and MHb where both the alpha and beta chain heme iron oxidized and Hb X and Y, were separated and easily discernible. The correlation matrix was calculated using the data relevant to IEF and other indicators; MHb estimated spectrophotometrically, GST activity of erythrocytes, Hb and urinary DZ vale. The statistical procedure of principal component analysis was applied to the matrix, and two major principal components were extracted from the calculation.(ABSTRACT TRUNCATED AT 400 WORDS)

Adult↗
PubMed2005

A robustified modeling approach to analyze pediatric length of stay.

PURPOSE: Length of stay (LOS) is an important measure of the cost of pediatric hospitalizations, but the guidelines developed so far are not rigorously evidence-based. This study demonstrates a robust gamma mixed regression approach to analyze the positively skewed LOS variable, which has implications for future studies of pediatric health care management. METHODS: The robustified approach is applied to analyze hospital discharge data on childhood gastroenteritis in Western Australia (n=514). The model accounts for demographic characteristics and co-morbidities of the patients, as well as the dependency of LOS outcomes nested within the 58 hospitals in the State. The method is compared with the standard linear mixed regression with trimming of extreme observations. RESULTS: For the empirical application, the linear mixed regression results are sensitive to the magnitude of trimming. The identified significant factors from the robust regression model, namely infection, failure to thrive, and iron deficiency anemia are resistant to high-LOS outliers. CONCLUSIONS: Robust gamma mixed regression appears to be a suitable alternative to analyze the clustered and positively skewed pediatric LOS, without transforming and trimming the data arbitrarily.

Child, Hospitalized↗
PubMed2001

Hospital readmission among long-term ventilator patients.

STUDY OBJECTIVES: Patients experiencing prolonged periods of in-hospital mechanical ventilation have been described as long-term ventilator (LTV) patients. The purpose of this study was to document the incidence of hospital readmission and to identify risk factors for readmission for LTV patients up to 6 months after hospital discharge. DESIGN: This study was part of a larger prospective longitudinal descriptive study of posthospital outcomes for LTV patients. SETTING AND PARTICIPANTS: One hundred ninety-nine ICU patients admitted to a university medical center, Veterans Administration hospital, or small community hospital who required > 96 h of continuous in-hospital mechanical ventilation were enrolled. MEASUREMENTS AND RESULTS: Descriptive statistics, logistic regression, and survival analytic techniques were used. The 6-month hospital readmission rate was 38%. Readmission occurred most often within days 1 to 60 days (mean, 39.2 days) posthospital discharge. Predictive variables for readmission were the following: length of the index hospital stay; length of the index mechanical ventilation; and the need for oxygen at hospital discharge. Using survival analysis, the age category of 66 to 71 years was statistically significant for the relative risk of readmission within the first 30 days of the index hospital discharge. CONCLUSIONS: LTV patients should be considered at risk for hospital readmission. Further study examining the impact of closer follow-up in the first 60 days posthospital discharge is necessary in order to determine whether there is a more effective way of reducing the risk of readmission for LTV patients.

Aged↗
PubMed2004

Constructing covariance functions for random regression models for growth in Gelbvieh beef cattle.

Genetic parameters for a random regression model of growth in Gelbvieh beef cattle were constructed using existing estimates. Information for variances along ages was provided by parameters used for routine Gelbvieh multiple-trait evaluation, and information on correlations among different ages was provided by random regression model estimates from literature studies involving Nellore cattle. Both sources of information were combined into multiple-trait estimates; corrected for continuity, smoothness, and general agreement with literature estimates; and extrapolated to 730 d. Covariance functions using standardized Legendre polynomials were fit for the following effects: additive genetic (direct and maternal), and animal and maternal permanent environment. Residual variances at different ages were fitted using linear splines with three knots. Fit was by least squares. The order of polynomials was varied from third to sixth. Increasing the fit beyond cubic provided small improvements in R2 and increased the number of small eigenvalues of covariance matrices, especially for the additive effect. Parameters for a random regression model in beef cattle can be constructed with negligible artifacts from literature estimates. Formulas can easily be modified for other types of polynomials and splines.

Algorithms↗