Clinical trials statistical significance

A recent study showed abnormal niacin sensitivity in schizophrenia patients as evidenced by attenuation of the flush response to niacin in schizophrenia. However, there is still an ongoing debate whether this response is due to altered pharmacological sensitivity to niacin or an inadequate cutaneous vasodilatory response to the stimulus (Messamore et al. 2003). Early studies even attempted to use niacin as an augmenting agent for the treatment of schizophrenia with mixed results (Ananth et al. 1973 Petrie et al. 1981). In a placebo-controlled comparative study by Ramsay et al. (1970), it was found that, while no significant differences were seen in total Brief Psychiatric Rating Scale (BPRS) scores prior to commencement of the clinical trial, statistically significant... 

The justification for randomized trials is not that no worthwhile observations can be made without them, but that important biases can occur in non-randomized comparisons which are particularly problematic if the benefits of treatment are, in reality, small or absent. For example, a non-randomized comparison of the effect of aspirin dosage on the operative risk of carotid endarterectomy (Table 18.2) reported a clinically and statistically significant lower operative risk in patients on high-dose aspirin (1300 mg) than taking low-dose aspirin (325 mg or less) (Barnett et al. 1998) however, a subsequent randomized trial (Taylor et al. 1999), performed to confirm this observation, showed that high-dose aspirin was, in fact, harmful (Table 18.1). It is likely that the non-randomized comparison had been biased by unmeasured differences between the patients in the low-dose and high-dose aspirin groups. 

One might conclude from reading the published account of a clinical trial that significance testing was the, be all and end all of statistical analysis. In fact, a more useful application of statistics is in estimating the magnitude of effects. In providing an estimate it is also usual to include an estimate of its reliability. Two common approaches are to quote standard errors and confidence limits. 

Clinical trials for r-IEN-y in RA indicated that the dmg is well tolerated (52). Consistent improvement in tender and swollen joint scores was observed, but a large number of patients were needed in the trial to show statistical significance for r-IEN-y treatment. In certain individuals, responses were remarkable. An additive effect between r-IEN-y and penicillamine was detected. Efficacy was lower when r-IEN-y was combined with gold therapy. Research is continuing. 

A key element in planning and conducting clinical trials is to ensure that they have scientific validity and objectivity. This is particularly relevant with respect to Phase II and III studies, where it is desired to demonstrate a positive benefit to risk outcome. Responses to a drug among a patient population are rarely homogeneous and clear-cut. Thus, sound statistical principles must be applied in order to be able to distinguish significant effects from random events. 

Compound 34 (BCZ-1812, RWJ-270201, peramivir) showed selective inhibition of influenza virus sialidases over bacterial and mammalian sialidases (Babu et al. 2000 Bantia et al. 2001 Sidwell and Smee 2002). Successful inhibition of influenza virus infectivity in vitro (Smee et al. 2001) and upon oral administration in vivo [mice (Bantia et al. 2001) and ferrets, reviewed in Sidwell and Smee 2002] led to human clinical trials of orally administered peramivir (Barroso et al. 2005). While orally administrated peramivir successfully completed animal studies and Phase I and Phase II clinical trials, in which the compound was showing neither major side effects nor toxicity (Sidwell and Smee 2002), preliminary results of the Phase III trials (June 2002) demonstrated no statistically significant difference in the primary efficacy endpoint, possibly due to low bioavailability (Barroso et al. 2005). 

A very common analysis in clinical trials involves the analysis of two binomial variables to see if there is a statistically significant association between them. A binomial variable is one that can have only one of two values. For example, let s assume that we have a variable called treatment whose value is either a 1 to indicate active drug therapy or a 0 to indicate placebo. We also have a variable called headache whose value is a 1 if the patient experiences headache after therapy and a 0 if not. What we want to know is whether a change in the level of therapy is significantly associated with a change in the level of headache. The 2x2 table looks like this ... 

For continuous variables you may be required to provide inferential statistics along with the descriptive statistics that you generate from PROC UNIVARIATE. The inferential statistics discussed here are all focused on two-sided tests of mean values and whether they differ significantly in either direction from a specified value or another population mean. Many of these tests of the mean are parametric tests that assume the variable being tested is normally distributed. Because this is often not the case with clinical trial data, we discuss substitute nonparametric tests of the population means as well. Here are some common continuous variable inferential tests and how to get the inferential statistics you need out of SAS. 

But do the clinical-trial data submitted to the FDA even establish proof of principle Recall that the rather small differences found between drug and placebo in the trials submitted to the FDA could have been due to the breaking of blind on the basis of perceived side effects. It may simply be evidence of an enhanced placebo effect, rather than a true drug effect. As I noted in Chapter i, once side effects are taken into account, the difference between SSRI and placebo is not even statistically significant.30... 

It is remarkable that most of the data collected from the available SERMs are unanimous in reproducing an estrogen agonistic profile in venous thrombogenesis. The vast clinical experience acquired with tamoxifen confirms an augmented risk for both deep venous thrombosis and pulmonary embolism. This increase, however, did not presuppose increased mortality in the overview of randomized trials of adjuvant tamoxifen for early breast cancer, where the one extra death per 5000 woman-years of tamoxifen attributed to pulmonary embolus was not statistically significant (Early Breast Cancer Trialists Collaborative Group 1998). 

A number of studies in humans show that PUFAs can generate significant immunomodulatory effects. Generally, these studies have utilized considerably lower amounts of fish oil to treat subjects than found in most animal studies. Numerous clinical trials have examined the effects of fish oil on rheumatoid arthritis and many have reported statistically significant benefits such as decreased morning stiffness and numbers of tender joints [57]. Several other studies have reported that PUFAs can provide therapeutic benefits for patients with IgA nephropathy, the most common primary human glomerulonephritis... 

---