Statistics blinded trials

All patients had a 3-month single-blind baseline amphetamine titration period and all 62 patients improved significantly during this time. They were then randomized to amphetamine or placebo. During the 12-month double-blind phase, 71% of the children in the placebo group and 29% in the amphetamine group stopped treatment or were switched to open treatment. Most of these dropouts occurred in the first 3 months of the double-blind trial. Thirty-two children (8 on placebo and 24 on amphetamine) completed the study as planned. Early dropouts were considered in a separate statistical analysis. 

An international, multicenter, double-blind trial addressed the acute efficacy and safety of a single-dose range of olanzapine (5 to 20 mg/day) compared with a single-dose range of haloperidol (5 to 20 mg/day) (11.6). A total of 1996 patients with a DSM-lll-R diagnosis of schizophrenia (83.1%), schizophreniform disorder (1.9%), or schizoaffective disorder (15%) participated in this study. The primary overall efficacy analysis (i.e., the difference in baseline to endpoint (last observation carried forward [LOCF]) mean change on the BPRS) found olanzapine to be statistically superior to haloperidol (HPDL) (i.e., -10.98 -7.93 p < 0.015). 

The results of a 1-year open extension of two randomized, double-blind studies of zaleplon have been reported (9). In 316 older patients who took zaleplon nightly from 6 to 12 months and were then followed through a 7-day singleblind, placebo-controlled, run-out period, the safety profile was similar to that observed in a short-term trial in an equivalent population. The data also suggested that therapy for up to 12 months produced and maintained statistically significant improvement in time to persistent sleep onset, duration of sleep, and the number of nocturnal wakenings. Withdrawal was not associated with rebound insomnia. The authors concluded that placebo-controlled, double-blind trials are needed to confirm these results. 

Clinically, 160 mg of Permixon twice daily was superior to placebo in a double-blind trial in 110 men with BPH (Champault et al., 1984). A statistically significant (p < 0.001) benefit compared to placebo was seen in nocturia, flow rate, post-void residual, self-rating, physician rating, and dysuria. Compared to baseline, both placebo and saw palmetto were beneficial in improving nocturia (p < 0.001), but only saw palmetto improved flow rate and post-void residual compared to baseline (p < 0.001). Headache was the only adverse effect. 

Its principal use is against superficial herpes virus infections of humans in which it produces a moderate cure rate without harming the tissue to which the drug is applied. Since dermal herpes simplex is usually self-limiting and of short duration, the assessment of clinical effect is not easy when double-blind trials are carried out. Some authors claim that a statistically positive result is obtained and an almost equal number conclude that the difference between drug and placebo treated patients is not significant. Some patients, however, consider the treatment to be beneficial [190]. 

The two other available double-blind trials on Echinaceae purpureae herba [14] and on a combination of Echinacea angustifolia with Eupatorium perfoliatum, Bap-tisia tinctoria and Arnica montana [15] yielded very similar rates of relative risk reduction 12% (95% Cl -13 to 32%), and 14% (95% Cl -2 to 27%) respectively. In the second trial [15] which included more than 600 volunteers the difference just missed statistical significance. 

A randomized, double-blind, placebo-controlled trial evaluating the use of a monophasic OC containing 30 meg ethinyl estradiol and 3 mg drospirenone, a progesterone with anti-androgenic effects, showed improvement in the treatment arm compared with placebo.31 In particular, appetite, food cravings, and acne improved. However, active treatment was not associated with a statistically significant improvement in the overall outcome measure, the Calendar of Premenstrual Experiences (COPE) scale, perhaps because of the small sample size (n = 82). 

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... 

Although there is some evidence for the efficacy of long-term treatment with rifaximin for symptomatic relief in patients with uncomplicated diverticular disease, an unresolved issue is whether rifaximin can prevent major inflammatory complications of diverticular disease. In the two prospective open trials discussed above, the occurrence rate of complications in 12 months was lower in patients treated with glucomannan plus rifaximin compared to patients treated with glucomannan only 2.7 versus 0.9% [43] and 3.2 versus 1.3% [44], This observation was not confirmed in the double-blind placebo-controlled trial [45] in which no difference in the 1-year complication rate was observed between the rifaximin and placebo groups. However, in all the studies, the number of patients suffering complications in a 12-month period was too small to detect any statistically significant... 

Refer to Section 6.4.7 and Exhibit 6.12 to explain randomization and the techniques used. Randomization and double-blinding are necessary to prevent bias in data collection so that statistical analysis based on normal distribution can be used to evaluate the trial results. 

In the third randomized trial of an MMPI (British Biotech Study 145) (12), patients with locally advanced or metastatic gastric cancer were randomized in a double-blind fashion to a low dose (10 mg bid) of marimastat or matching placebo. Marimastat 10 mg po bid, which had inferior survival compared to the higher dose of marimastat, was presumably selected as the active control arm for this trial based on its superior tolerability compared to high-dose marimastat. Patients randomized to marimastat had a trend to better overall survival (167 d vs 135 d p = 0.07) at the protocol stipulated endpoint of the study, and this statistical trend strengthened with an additional 6 mo of follow up (p = 0.048). Nevertheless, patients randomized to marimastat 10 mg po bid had statistically superior progression-free survival compared to patients randomized to placebo (p = 0.027). Marimastat 25 mg po bid, which was the most efficacious dose in British Biotech Study 128, was not tested in Study 145. 

Statistical testing for baseline imbalance has no role in a trial where the handling of randomisation and blinding has been fully satisfactory. ... 

This was a multi-centre, pan-European, randomised double-blind placebo-controlled clinical trial in acute stroke to evaluate the effect of ancrod, a natural defribrinogenating agent (Hennerici et al. (2006)). The primary endpoint was based on the Barthel Index a favourable score of 95 or 100 or a return to the pre-stroke level at three months was viewed as a success. The primary method of statistical analysis was based on a logistic model including terms for treatment, age category, baseline Scandinavian Stroke Scale and centre. 

There is one final opportunity to revisit the proposed methods of statistical analysis prior to the breaking of the blind, or in an unblinded trial, before the statistics group have seen study data. This so-called blind review usually takes place around the time of database lock and the following lists some of the aspects of analysis that would generally be considered ... 

---