PubMed Health⌕ Search
Home/Archive/PubMed

PubMed · 10474144

Meta-analysis by random effect modelling in generalized linear models.

M Aitkin
View original source ↗Export metadata ↓

Abstract

The meta-analysis of multi-centre trials can be based on either fixed or random effect models. This paper argues for the general use of random effect models, and illustrates the value of non-parametric maximum likelihood (NPML) analysis of such trials. The same general approach unifies administrative 'league table' analyses in epidemiological and other studies. Several examples of the NPML analysis are given, including a 70-centre trial.

Explore related subjects

Adrenergic beta-AntagonistsData Interpretation, StatisticalFibrinolytic AgentsHumansLikelihood FunctionsLinear ModelsMeta-Analysis as TopicMulticenter Studies as TopicMyocardial InfarctionNormal DistributionStatistics, NonparametricJournal Article

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

M Aitkin. Meta-analysis by random effect modelling in generalized linear models.. https://doi.org/10.1002/(sici)1097-0258(19990915%2F30)18%3A17%2F18%3C2343%3A%3Aaid-sim260%3E3.0.co%3B2-3

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related citations

PubMed2003

Serum glucose and lipid changes during the course of clozapine treatment: the effect of concurrent beta-adrenergic antagonist treatment.

We examined the effects of long-term clozapine treatment, concurrent treatment with beta-adrenergic antagonists, and clozapine-induced weight gain on serum glucose and lipid measures. Fifty subjects met the DSM-III-R criteria for schizophrenia or schizoaffective disorder, participated in a 10-week, double-blind comparison of haloperidol and clozapine and a 1-year, open-label clozapine trial, and had available serum glucose and lipid levels. Weight and glucose, and lipid laboratory values were measured at the baseline and throughout the double-blind and year-long study. There were significant increases in serum triglyceride, total cholesterol, and glucose levels during the course of clozapine treatment. There were no significant changes in high-density lipoprotein (HDL) or low-density lipoprotein (LDL). Propranolol and atenolol had additive effects on changes in the total cholesterol and triglycerides, with propranolol having the most pronounced effects. Propranolol and atenolol had no significant effect on the serum glucose levels. There were significant correlations between the triglyceride and HDL level changes and clozapine-associated weight gain during the study. There were no significant correlations between the change in serum total cholesterol, LDL, or glucose and weight gain. Clozapine therapy has adverse effects on glucose and lipid homeostasis, with clozapine-induced changes in serum glucose likely due to the inherent pharmacological properties of clozapine. Concurrent beta-adrenergic receptor antagonist treatment may have an additive effect on serum lipids, and clozapine-associated weight gain also plays a modest role in triglyceride increases.

Adrenergic beta-Antagonists↗
PubMed2002

Comparison of actions of irbesartan versus atenolol on cardiac repolarization in hypertensive left ventricular hypertrophy: results from the Swedish Irbesartan Left Ventricular Hypertrophy Investigation Versus Atenolol (SILVHIA).

Left ventricular (LV) hypertrophy is associated with a substantial risk for malignant arrhythmias and sudden death. The effects of antihypertensive therapy on QT dispersion, which reflects cardiac repolarization heterogeneity, in relation to changes in LV mass has not been well studied. Repeat echocardiography and QT measurements (standard 12-lead electrocardiograms) were performed in hypertensive patients with LV hypertrophy, who were randomized double-blind to receive the angiotensin II type 1-receptor blocker irbesartan (n = 44) or the beta(1)-receptor blocker atenolol (n = 48) for 48 weeks, and in 37 matched hypertensive control subjects without LV hypertrophy. LV mass index was related to QT dispersion (r = 0.34, p <0.001). The reduction in LV mass was greater using irbesartan than using atenolol (-27 +/- 28 vs -15 +/- 21 g/m(2) at 48 weeks, p = 0.021), with similar reductions in blood pressure. Irbesartan decreased QT dispersion (from 56 +/- 24 ms to 45 +/- 20 ms at 48 weeks; p <0.001) and QTc dispersion (from 57 +/- 24 to 44 +/- 19 ms at 48 weeks; p <0.001). In contrast, atenolol had minor effects. The decreases in QT and QTc dispersions were greater using irbesartan than using atenolol (p = 0.001 and p = 0.011, respectively); the same results were found when changes in LV mass, blood pressure, and heart rate were also included in multivariate analyses. Thus, heterogeneity of ventricular repolarization is related to the degree of LV hypertrophy. Irbesartan, but not atenolol, reduces QT and QTc dispersions independent of changes in LV mass, blood pressure, or heart rate, and thus seems to induce structural and electrical remodeling in a direction that could decrease the risk of fatal events in hypertensive patients.

Adrenergic beta-Antagonists↗
PubMed2002

Effects of long-term beta-blocker (metoprolol or carvedilol) therapy on QT variability in subjects with chronic heart failure secondary to ischemic cardiomyopathy.

Chronic heart failure (CHF) is a risk factor for sudden death. Temporal and spatial changes in repolarization are among the most studied mechanisms for inducing fatal ventricular arrhythmias. Beta blockers effectively reduce the risk of sudden death in CHF. Our aim in this study was to investigate changes induced by metoprolol and carvedilol on the QT variability index (QTVI), a new measure reflecting the temporal heterogeneity of cardiac repolarization. A total of 82 subjects, who were in New York Heart Association functional class II or III, underwent short-term spectral analysis of RR and QT variability before and after a 1-year course of high-dose metoprolol (40 subjects) or carvedilol (42 subjects) at baseline (rest) and after sympathetic stress (head-up tilt). At rest, both drug-treated groups had lower QTVI (p <0.001) than after placebo, but during tilt patients treated with carvedilol had a lower QTVI than those treated with metoprolol (p <0.05). Although both beta-blocker treatments helped to normalize the QTVI measured in normal subjects at rest, they each differentially altered the index after tilt. Carvedilol seemed to improve the QTVI more than metoprolol.

Adrenergic beta-Antagonists↗