Confidence Limits in Evaluating Controlled Therapeutic Trials

Confidence Intervals and Survival Estimates

“…[1][2][3][4][5][6][7] In Lancet in 1973, Wulff commented on the need for CI so that the magnitude of differences could be better understood between groups. 7 He noted that clinicians are not just interested in knowing whether 2 treatments are equally effective but rather "by how much the new treatment is better (or worse) than the old or placebo." Others have commented on the importance of reporting CI, including Shakespeare and Holecek who observed that only 3 of 21 published studies in a single issue of the IJROBP used CI.…”

“…Conversely, one would expect the true 1-year survival rate to lie outside the CI on 5 occasions. [1][2][3][4][5][6][7] In the oncology literature the relevance of data is often defined by whether the P value has reached a level of "statistical significance". The P value helps the clinician to answer the question of whether or not a hypothesis has been rejected, however, the P value itself says nothing about the magnitude or direction of a treatment effect.…”

From Statistical Significance to Effect Estimation: Statistical Reform in Psychology, Medicine and Ecology

“…In the 1970s criticisms of NHST became increasingly common (Band & Boen, 1972;Shulman, Kupst & Suran, 1976). Some critics of this era began advocating CI reporting instead of p values (Green, 1972;Wulff, 1973). To facilitate rapid reform, others provided guides to calculating CIs for relevant bio-medical effect sizes, such as odds ratios and relative risk values (Rothman, 1975(Rothman, , 1978a.…”

Confidence Intervals in Procedural Dermatology: An Intuitive Approach to Interpreting Data

“…Investigation of the efficacy of a new procedure can be used to illustrate the concept of confidence intervals. Assuming that we do not know whether a new surgical technique will be successful, the true success rate can be said to between 0 and 100% 4–8 . To test the therapy, we can (hypothetically) conduct sequential clinical trials.…”