The number of patients required in a clinical trial

Patients with early breast cancer benefit from effective axillary treatment

“…This leaves 26 patients who could possibly have benefited from initial axillary treatment, or 7% of the total group. It is important to note that in order to have a 90% chance of detecting a 7% difference between 2 treatment arms of a clinical trial at a statistically significant level of p = 0.05, 2000 patients are required (10). We conclude from this analysis that the negative result obtained in the B-04 study does not prove that axillary treatment is of no value, but rather, that the number of patients who might benefit is small.…”

Repeated assessment of results in clinical trials of cancer treatment

“…The latter can be determined from the charts published by Boag et al (1971), provided that one is able to specify the minimum difference one is hoping to detect and the risk one is prepared to take of not detecting such a difference when it does in fact exist. As pointed out by Schneiderman (1964), it is often very difficult for a clinician to specify his requirements in this way.…”

“…In practice therefore he is more likely to be able to state *T (and, taking into account the difficulties of maintaining consistent documentation and enthusiasm over a long period, *T is usually likely to be about five years and rarely more than ten) and also give an estimate of the total number of patients likely to enter the trial if it continues for its full length. The charts of Boag et al (1971) can then be used to show what chance there will be with this number of demonstrating a given difference between the groups. With this information the clinician may decide either to commence his trial, or that it is not worth proceeding unless he can collaborate with others to increase the rate of patient entry.…”

“…For example, when the difference in cs is 0-15, the two-year rates are 0-536 and 0-420 having a difference of 0-116. The charts given by Boag et al (1971) predict that a total of about 500 patients in a trial would be required to stand a three in four chance of detecting a difference of 0-116. Table II shows that 751 of the 1,000 trials produced a final significant difference when the difference in cs was 0-15.…”

“…These preliminary assessments may be used to stop the trial if a new unproven treatment method is seen to be unexpectedly much worse than the usual technique, or if the results from the new treatment are already found to be significantly better. While such surveys are ethically desirable, since their aim is to minimize the number of patients treated by an inferior method, it is not always appreciated that, as pointed out by Boag, Haybittle, Fowler and Emery (1971), the procedure may seriously invalidate the use of a conventional test of significance in comparing the results. Armitage, McPherson and Rowe (1969) have examined this effect in data having a binomial, normal or exponential distribution, and it is of equal interest to study the situation when survival rates are compared in a typical clinical trial of cancer treatment.…”