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 649 records · Page 36Linked to original sources
PubMed1997

A meta-analysis of 61 sperm count studies revisited.

OBJECTIVE: To re-examine data on sperm counts over time from 61 studies from around the world. DESIGN: Parametric analyses and flexible nonlinear models of the relation between sperm counts and time. MAIN OUTCOME MEASURE(S): Mean sperm concentrations per milliliter and regression coefficients for possible trends of concentrations over time. RESULT(S): A significant decline was found only in U.S. studies. CONCLUSION(S): Studies from specific sites have found declines in sperm counts, but a world-wide decline has not been demonstrated. Rigorous assessment of statistical models should be done before conclusions are drawn. Flexible smoothing models are a useful addition to currently available analytic methods.

Global Health↗
PubMed2003

Predicting mass rapid transit noise levels on an elevated station.

This study developed a noise prediction model for elevated mass rapid transit (MRT) platforms. Relevant physical and operational parameters (e.g. cruise speed, acceleration and deceleration rates for trains, building façade setbacks and so on) were collected from the Bangkok mass transit system (BTS), the first elevated MRT system operated in Bangkok, Thailand. The equivalent continuous sound pressure levels (L(Aeq)) were collected from both sides of the MRT stations at the center of each platform. The relevant parameters were collected on both platforms and ground level, on both sides of MRT stations. These parameters were statistically tested to determine their correlation with MRT noise. The final model was built from highly correlated parameters using multiple regression analysis with a stepwise regression technique. Statistical evaluation showed a high degree of goodness-of-fit test for the model to the observed data. Therefore, it can be efficiently used for the projection of MRT noise in the affected areas.

Engineering↗
PubMed2004

The perils of adding up: a cautionary tale about modeling and regression analysis.

To turn incomplete or imperfectly reliable information into estimates useful for policy, some modeling is often necessary or helpful. It does not follow that every statistical relation constitutes a model. There has to be an underlying theory, and the numerical estimates must respect any definitional or accounting identities that constrain the results. Ad hoc relations that do not meet these criteria can lead to uninterpretable or meaningless values, especially when different partial values, separately estimated, are added together. The imputed values of "responsiveness" in the WHO World Health Report 2000 illustrate this danger.

Global Health↗
PubMed1976

Regression of injury-induced atheromatous lesions in rabbits.

For four consecutive weeks, 61 rabbits received weekly injections of lymphocytotoxic-positive human serum into the left carotid artery and of autologous serum into the right carotid artery as a control. Serum cholesterol and serum triglyceride levels were measured before the study, in the second and fourth weeks of the study, and weekly thereafter. The results show that repeated intimal injury caused raised, lipid-containing thromboatherosclerotic lesions and that there was a consistent regression to lipid-free fibromusculoelastic plaques from the first week after completion of the injection regimen to the fourth week. Apparently, regeneration of an intact covering cell layer resulted in the elimination of lipid deposits from raised lesions, resulting in lipid-free fibromusculoelastic plaques. In addition, fatty streaks were observed to occur during regression. A statistically significant rise in serum cholesterol level during the phase of progression of lesions and a subsequent fall during regression were observed.

Animals↗
PubMed2000

Identifying demand for health resources using waiting times information.

In this paper the differences in average waiting times are utilized to identify the determinants of demand for health services. The equilibrium waiting time framework is used, but the full equilibrium assumption is relaxed by selecting areas with low waiting times and by estimating a (semi-)parametric selection model. Determinants of supply are used as instruments for the endogeneity of waiting times. A model for the demand for acute services at the ward level in the UK is estimated. The model estimates, and their implications for health service allocations in the UK, are contrasted against more standard allocation models. The present results show that it is critically important to account for rationing by waiting times when identifying needs from care utilization data.

Health Care Rationing↗
PubMed2007

LongCriSP: a test for bump hunting in longitudinal data.

We propose an extension of the Harezlak and Heckman (J. Comput. Graph. Statist. 2001; 10(4): 713-729) test for detecting local extrema to the longitudinal data setting. We use penalized spline regression techniques (Statist. Sci. 1996; 11:89-102) to provide a computationally efficient method of testing for relatively large data sets. We estimate the p-values of our test, LongCriSP, with a smoothed bootstrap. Our simulation studies indicate that the test is generally conservative and has power exceeding 70 per cent at the alpha = 0.1 nominal level in most considered settings. Finally, we apply our testing procedure to the longitudinal measurements of body mass index of former prisoners of war in Vietnam and conclude that the mean population curve exhibits non-monotone behaviour.

Body Mass Index↗
PubMed2004

QSAR study of anticonvulsant negative allosteric modulators of the AMPA receptor.

A quantitative structure-activity relationship (QSAR) study was performed on a set of 49 negative allosteric modulators of AMPA receptor, acting as anticonvulsant agents, using multiple linear regression. The predictive ability of the resulting model was evaluated against a set of 12 compounds; the results showed good statistics in regression and revealed high correlation between anticonvulsant activity and some electrotopological descriptors.

Allosteric Regulation↗
PubMed1990

Regression of left ventricular hypertrophy in patients with essential hypertension: outcome of 12 years antihypertensive treatment.

To assess the regression of cardiac hypertrophy during long-term (12 years) antihypertensive treatment, the following parameters were determined in 93 patients with essential hypertension: SV1 + RV5 by electrocardiography (ECG), and septal wall (SW) and posterior wall (PW) thickness by echocardiography (UCG). The patients were treated with a thiazide diuretic alone (group 1), thiazide + beta-blocker (group 2), thiazide + methyldopa or nifedipine (group 3) or nifedipine or methyldopa alone (group 4). The blood pressure decreased gradually within 6 months of treatment. According to ECG, regression of left ventricular hypertrophy occurred during the initial 7 years in all groups, whereas in the subsequent 5 years, statistically significant regression was found only in the patients treated with thiazide + other drugs (group 3). By UCG, which was taken only at the 7th and 12th year, regression was detectable during the last 5 years in all groups. The apparent incidence of regression of hypertrophy was lower in the thiazide-alone group (group 1) than in the thiazide + beta-blocker group (group 2), most likely due to mild hypertension in group 1. A cardiovascular accident (nonfatal myocardial infarction) occurred only in 1 patient. We conclude that during long-term antihypertensive treatment, persistent, progressive reversal of cardiac hypertrophy takes place.

Adrenergic beta-Antagonists↗
PubMedDate not supplied

[Variations in the postoperative length of stay according to patient characteristics in a general surgery service].

With the purpose of identifying factors that explain variations in length of postoperative stay in a general surgery unit, information from 666 clinical records of discharged alive patients was collected. Factors related to the surgical procedure, the severity of the disease and the medical team together with sociodemographic characteristics of the patients were analyzed. Statistical analysis was based on analysis of variance and multiple linear regression. Variables with statistically significant regression coefficients (p < 0.001) were in this order: severity of complications, type of intervention, need for reintervention and severity of diagnosis. Sixty-seven percent of length of stay variation was explained by the included variables. Results allow to suggest an evaluation procedure which takes into account patient and operation characteristics.

Adolescent↗
PubMed1990

Obscure phenomena in statistical analysis of quantitative structure-activity relationships. Part 1: Multicollinearity of physicochemical descriptors.

Multicollinearity of physicochemical descriptors leads to serious consequences in quantitative structure-activity relationship (QSAR) analysis, such as incorrect estimators and test statistics of regression coefficients of the ordinary least-squares (OLS) model applied usually to QSARs. Beside the diagnosis of the known simple collinearity, principal component regression analysis (PCRA) also allows the diagnosis of various types of multicollinearity. Only if the absolute values of PCRA estimators are order statistics that decrease monotonically, the effects of multicollinearity can be circumvented. Otherwise, obscure phenomena may be observed, such as good data recognition but low predictive model power of a QSAR model.

Chemical Phenomena↗
PubMedDate not supplied

What you see may not be what you get: a brief, nontechnical introduction to overfitting in regression-type models.

OBJECTIVE: Statistical models, such as linear or logistic regression or survival analysis, are frequently used as a means to answer scientific questions in psychosomatic research. Many who use these techniques, however, apparently fail to appreciate fully the problem of overfitting, ie, capitalizing on the idiosyncrasies of the sample at hand. Overfitted models will fail to replicate in future samples, thus creating considerable uncertainty about the scientific merit of the finding. The present article is a nontechnical discussion of the concept of overfitting and is intended to be accessible to readers with varying levels of statistical expertise. The notion of overfitting is presented in terms of asking too much from the available data. Given a certain number of observations in a data set, there is an upper limit to the complexity of the model that can be derived with any acceptable degree of uncertainty. Complexity arises as a function of the number of degrees of freedom expended (the number of predictors including complex terms such as interactions and nonlinear terms) against the same data set during any stage of the data analysis. Theoretical and empirical evidence--with a special focus on the results of computer simulation studies--is presented to demonstrate the practical consequences of overfitting with respect to scientific inference. Three common practices--automated variable selection, pretesting of candidate predictors, and dichotomization of continuous variables--are shown to pose a considerable risk for spurious findings in models. The dilemma between overfitting and exploring candidate confounders is also discussed. Alternative means of guarding against overfitting are discussed, including variable aggregation and the fixing of coefficients a priori. Techniques that account and correct for complexity, including shrinkage and penalization, also are introduced.

Computer Simulation↗
PubMed1998

Determinants of survival following hepatic resection for metastatic colorectal cancer.

Hepatic resection remains the only potentially curative treatment for metastatic colorectal cancer. This retrospective review study was undertaken in an attempt to identify factors that influence patient survival following hepatic resection for metastatic colorectal cancer. From January 1978 to December 1993, a total of 301 patients underwent a total of 345 planned hepatic resections for metastatic colorectal cancer. Of those, 245 patients had one resection, 44 had two resections, and 12 had three resections. For all patients the overall median survival was 20.6 months, operative mortality was 1.1%, and overall morbidity was 17.2%. Average hospital stay was 9 days. Statistical analysis included univariate analysis using log rank comparisons, Kaplan-Meier survival curves, and multivariate analysis using Cox proportional hazards regression. The statistically significant factors that influenced survival were distribution of liver metastases, unilobar versus bilobar (p = 0.0001), resected versus nonresected (p < 0.0001), and tumor-free surgical margins versus positive margins (p = 0.001). Surprisingly, the disease-free interval and the original stage of the primary tumor did not predict survival (p = not significant). Other factors that had no influence on survival were type of resection, size and number of liver metastases, ABO blood group, and the number of perioperative blood transfusions. For those patients who underwent resection of unilobar metastases with tumor-free margins, the 5-year survival rate was 29% with a median survival of 35 months and eight survivors > 7 years. In addition, one patient with bilobar disease had survival > 7 years and five patients who had resection of hepatic metastases and extrahepatic cancer simultaneously had survival > 3 years. Our data support the concept that patients with unilobar metastatic disease who undergo surgical resection with tumor-free surgical margins can be afforded a significant opportunity at long-term survival with acceptable morbidity, mortality, and hospital stay. Also, certain patients with bilobar or extrahepatic disease (or both) who undergo complete resection can enjoy a long-term survival. In these subgroups of patients resection should be considered on an individual basis.

Adenocarcinoma↗
PubMed1990

Cross-national models of fertility, family planning, and development: testing for reciprocal effects.

Cross-national models of fertility, family planning, and development commonly assume that there are no reciprocal effect between fertility and other variables in the model and that when path models are used, there are no reciprocal or nonrecursive effects among any set of variables in the model. The present study tests for nonrecursiveness using 2-wave panel data, and finds that nonrecursive effects are present among variables commonly used in model of fertility, family planning, and development. In addition, the pattern of relationships found has implications for the explanation of the relationship between mortality and fertility in demographic transition theory.

Birth Rate↗
PubMed1981

Suicide in England and Wales: an analysis of 100 years, 1876-1975.

The variations in suicide mortality in England and Wales over a 100-year period (1876-1975) are explained by the development of a descriptive model. The model used in a stepwise linear regression using 5 variables including toxic gas production. The variables give a good description of the observed changes in suicide mortality. The limitations of the method and the data are discussed. Despite the limitations of the method, it can be concluded that toxic gas production affected total suicide mortality.

England↗