Clinical trials in urology: how many patients are required to achieve statistically significant results?

• Published in: BJU international 2005-04, Vol.95 (5), p.717-722, Article 717
• Main Authors: Parazzini, Fabio, Colli, Enrico, Chatenoud, Liliane, Malvezzi, Matteo, Olivieri, Lorenza, Montorsi, Francesco, La Vecchia, Carlo
• Format: Article
• Language: English
• Subjects: Biological and medical sciences; Clinical Trials as Topic - methods; Humans; Medical sciences; Nephrology. Urinary tract diseases; Patient Selection; Pharmacology. Drug treatments; Random Allocation; Randomized Controlled Trials as Topic - methods; Reproducibility of Results; Sample Size; Urinary system; Urologic Diseases - therapy
• ISSN: 1464-4096; 1464-410X
• DOI: 10.1111/j.1464-410X.2005.05436.x

SUMMARY During the clinical validation of a new drug there are several clinical phases. Once phase II studies have defined the efficacy of a new drug, clinical research is used to evaluate its significance in clinical practice, comparing it with other drugs or treatments in use for similar clinical conditions. The group of patients undergoing standard treatment (either untreated or treated with placebo) is thus used as a control; these phase III studies are termed ‘controlled clinical studies’. The general condition for comparing patients treated with the new drug is that they do not have characteristics (relevant to the study) that are systematically different from those in the control group. Randomization guarantees comparability between treated and untreated (or otherwise treated) patients. The comparability of the observations of the studied events are guaranteed by blinding and placebo. The fundamental question when designing a controlled clinical study to evaluate whether there are differences between two or more treatments is how many patients are needed. Generally, the smaller the clinically relevant differences in efficacy between treatments, the more patients are required, to provide sufficient statistical power and meaningful clinical results. A group of randomized patients represents the final point of sequential steps. Also of importance is to what kind of ‘population’ the results from the studied sample can be applied (qualitatively, not necessarily quantitatively), i.e. the general applicability of a study, or whether the findings can be used to treat future patients with the same or similar characteristics as those randomized, or to all patients with the same pathology. Answers to these questions depend on many aspects of the randomized selection mechanisms, the disease characteristics, and knowledge of the biological effects of the drug to be tested.