Hypotheses meta-analysis

Meta-analysis is a method often used to determine the effectiveness of a drug but to date it has rarely been used to assess safety. One case illustrates how this technique can help. Six studies examining the use of intravenous lidocaine for acute myocardial infarction did not, on an individual basis, give strong enough evidence to support the hypothesis that this technique could cause excess mortality. The meta-analysis, however, was able to demonstrate this. ... 

Therefore, meta-analysis abstracts results from each study and expresses them in a common unit, so one can easily compare, which allows us to focus on the hypothesis under examination rather than be distracted by the myriad differences among studies. 

A meta-analysis for continuous data cannot be calculated unless the pertinent standard deviations are known. Unfortunately, clinical reports often give the sample size and mean ratings for the various groups but do not report the standard deviations (or standard error of the mean), which are necessary for effect size calculations. Thus, investigators should always report the indices of variability (e.g., confidence intervals, SDs) for the critical variables related to their primary hypothesis. 

Abstract G72 and G30 constitute a pair of primate-specific genes on complementary strands of human chromosomal 13q33. G72 is proposed to encode a protein that can activate D-amino acid oxidase (DAAO), therefore, named as D-aminoacid oxidase activator (DAOA) (Chumakov et al., 2002). This hypothesis about G72 and DAAO requires substantial further proof (see discussion later), and the neutral name G72 will be used here, although DAOA is widely used in the literature now. The G72/ G30 gene is, at this time, one of the best supported loci for both schizophrenia (SZ) and bipolar disorder (BD) by independent datasets (Craddock et al., 2006 Detera-Wadleigh and McMahon, 2006). Here, we will describe the discovery of gene complex G72/G30, association studies with SZ and BD with a meta-analysis, as well as brief information about the biology of G72/G30. 

In my view, the meta-analysis conducted by Juni et al. (2004) validly tested the null hypothesis rofecoxib is no different from any of the comparators included in this metaanalysis as regards effect on cardiovascular events. If this hypothesis is rejected (and given the direction of rejection) then its logical alternative can be asserted rofecoxib is worse than at least one comparator as regards effect on cardiovascular mortality. This alternative hypothesis is not at all the same as this one rofecoxib is worse than all the comparators as regards its effect on cardiovascular mortality. However, the cumulative meta-analysis... 

Thus, the assertion that rofecoxib could have been known to be worse than placebo by 2000 is based entirely on an auxiliary hypothesis that naproxen and placebo are identical. This hypothesis may or may not have been reasonable at the time, but the meta-analysis as carried out by Juni et cd. (2004) cannot prove it. To be fair to these authors, they did carry out some independent investigation of this hypothesis by looking at studies that had compared naproxen with placebo but here they assumed that absence of evidence was evidence of absence and did not use methods that have been developed for formally incorporating evidence regarding previous trials of comparators. (See the discussion of the approaches of Hasselblad and Kong (2001) and Hlrotsu and Yamada (1999) in Chapter 15). 

Meta-analysis may make small studies meaningful by providing a means to combine the results with those of other similar studies to enable estimates of an intervention s efficacy. Small trials may not be able to test a hypothesis, but they may provide valuable information of treatment effects using confidence intervals (Edwards et ai, 1997). Similarly, others argue that a sample size that results in a p value of 0.1 can be informative and decisions have to be made even where there is no trial evidence, a little unbiased evidence is better than none. A study might have only limited ability to detect an effect, but participants should be allowed to make an autonomous decision. 

Meta-analysis is a detailed, cumulative analysis of data from several similar studies, which is usually based on scientific publications. One of its many advantages is that the assessment or testing of any hypothesis can be carried out on a much larger sample set than in any single tests. It is widely used both in chnical and epidemiological studies. 

Whether or not to include trials with no cardiovascular events in a meta-analysis has been a highly debated topic. If a trial had a sufficiently long period of follow-up and no cardiovascular events were reported, these data would appear to support the hypothesis of noninferiority however, no formal statistical method exists for estimating the cardiovascular risk from such data, and they are typically excluded from the analysis. Tian and colleagues (2009) discussed a method of obtaining an exact confidence interval for the difference in event rates at a fixed time point that permits combining data from trials that have zero events with data from other trials. This method could be utilized as a supportive sensitivity analysis. 

Large-scale trials and meta-analysis of such trials are required to assess moderate treatment benefits. However, within trials or between trials within meta-analysis, sub-group analysis may result in spurious results due to the play of chance. Hence, it is important to avoid sub-group analyses to reduce the risk of spurious results except those sub-groups that have been pre-specified on the basis of some prior hypothesis. 

Koricheva J, Larsson S, HaukiojaE, Keinanen M (1998) Regulation of woody plant secondary metabolism by resource availability hypothesis testing by means of meta-analysis. Oikos 83 212-226... 

---