A basic introduction to fixed‐effect and random‐effects models for meta‐analysis

Research Synthesis Methods 11/2010; 1(2):97 - 111. DOI: 10.1002/jrsm.12

ABSTRACT There are two popular statistical models for meta-analysis, the fixed-effect model and the random-effects model. The fact that these two models employ similar sets of formulas to compute statistics, and sometimes yield similar estimates for the various parameters, may lead people to believe that the models are interchangeable. In fact, though, the models represent fundamentally different assumptions about the data. The selection of the appropriate model is important to ensure that the various statistics are estimated correctly. Additionally, and more fundamentally, the model serves to place the analysis in context. It provides a framework for the goals of the analysis as well as for the interpretation of the statistics.In this paper we explain the key assumptions of each model, and then outline the differences between the models. We conclude with a discussion of factors to consider when choosing between the two models. Copyright © 2010 John Wiley & Sons, Ltd.

  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Network meta-analysis (NMA) has emerged as a useful analytical tool allowing comparison of multiple treatments based on direct and indirect evidence. Commonly, a hierarchical Bayesian NMA model is used, which allows rank probabilities (the probability that each treatment is best, second best, and so on) to be calculated for decision making. However, the statistical properties of rank probabilities are not well understood. This study investigates how rank probabilities are affected by various factors such as unequal number of studies per comparison in the network, the sample size of individual studies, the network configuration, and effect sizes between treatments. In order to explore these factors, a simulation study of four treatments (three equally effective treatments and one less effective reference) was conducted. The simulation illustrated that estimates of rank probabilities are highly sensitive to both the number of studies per comparison and the overall network configuration. An unequal number of studies per comparison resulted in biased estimates of treatment rank probabilities for every network considered. The rank probability for the treatment that was included in the fewest number of studies was biased upward. Conversely, the rank of the treatment included in the most number of studies was consistently underestimated. When the simulation was altered to include three equally effective treatments and one superior treatment, the hierarchical Bayesian NMA model correctly identified the most effective treatment, regardless of all factors varied. The results of this study offer important insight into the ability of NMA models to rank treatments accurately under several scenarios. The authors recommend that health researchers use rank probabilities cautiously in making important decisions.
    Clinical epidemiology. 01/2014; 6:451-60.
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Potential cardiovascular (CV) risks of testosterone replacement therapy (TRT) are currently a topic of intense interest. However, no studies have addressed CV risk as a function of the route of administration of TRT. Two meta-analyses were conducted, one of CV adverse events (AEs) in 35 randomized controlled trials (RCTs) of TRT lasting 12 weeks or more, and one of 32 studies reporting the effect of TRT on serum testosterone and dihydrotestosterone (DHT). CV risks of TRT: Of 2,313 studies identified, 35 were eligible and included 3,703 mostly older men who experienced 218 CV-related AEs. No significant risk for CV AEs was present when all TRT administration routes were grouped (relative risk (RR) = 1.28, 95% confidence interval (CI): 0.76 to 2.13, P = 0.34). When analyzed separately, oral TRT produced significant CV risk (RR = 2.20, 95% CI: 1.45 to 3.55, P = 0.015), while neither intramuscular (RR = 0.66, 95% CI: 0.28 to 1.56, P = 0.32) nor transdermal (gel or patch) TRT (RR = 1.27, 95% CI: 0.62 to 2.62, P = 0.48) significantly altered CV risk. Serum testosterone/DHT following TRT: Of 419 studies identified, 32 were eligible which included 1,152 men receiving TRT. No significant difference in the elevation of serum testosterone was present between intramuscular or transdermal TRT. However, transdermal TRT elevated serum DHT (5.46-fold, 95% CI: 4.51 to 6.60) to a greater magnitude than intramuscular TRT (2.20-fold, 95% CI: 1.74 to 2.77). Oral TRT produces significant CV risk. While no significant effects on CV risk were observed with either injected or transdermal TRT, the point estimates suggest that further research is needed to establish whether administration by these routes is protective or detrimental, respectively. Differences in the degree to which serum DHT is elevated may underlie the varying CV risk by TRT administration route, as elevated serum dihydrotestosterone has been shown to be associated with CV risk in observational studies.
    BMC medicine. 01/2014; 12:211.
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Associating liver partition and portal vein ligation for staged hepatectomy (ALPPS) is a novel strategy to resect liver tumors despite the small size of the liver remnant. It is an hepatectomy in two stages, with PVL and parenchymal transection during the first stage, which induces rapid growth of the remnant liver exceeding any other technique. Despite high postoperative morbidity and mortality in most reports, the technique was adopted by a number of surgeons. This systematic review explores current data regarding the feasibility, safety, and oncologic efficacy of ALPPS; the search strategy has been published online. A meta-analysis of hypertrophy, feasibility (ALPPS stage 2 performed), mortality, complications, and R0 (complete) resection was performed. A literature search revealed a total of 13 publications that met the search criteria, reporting data from 295 patients. Evidence levels were low, with the highest Oxford evidence level being 2c. The most common indication was colorectal liver metastasis in 203 patients. Hypertrophy in the meta-analysis was 84 %, feasibility (ALPPS stage 2 performed) 97 % (CI 94-99 %), 90-day mortality 11 % (CI 8-16 %), and complications grade IIIa or higher occured in 44 % (CI 38-50 %) of patients. A standardized reporting format for complications is lacking despite the widespread use of the Clavien-Dindo classification. Oncological outcome is not well-documented. The most common topics in the selected studies published were technical feasibility and indications for the procedures. Publication bias due to case-series and single-center reports is common. A systematic exploration of this novel operation with a rigid methodology, such as registry analyses and a randomized controlled trial, is highly advised.
    Annals of Surgical Oncology 12/2014; · 4.12 Impact Factor


Available from
May 21, 2014