The confidence interval is a powerful tool for evaluating evidence that aids in making decisions about the care of individual patients. In this brief report, I will define the confidence interval and describe its usefulness in the care of individual patients.
The results of more than 230 000 controlled trials have been published in the medical literature, many of which involve the treatment of diseases relevant to the practice of dermatology. One of the problems facing practicing dermatologists is to figure out whether the results of controlled trials are valid and useful in making decisions about the care of their patients. The magnitude of the treatment effect and its precision are important concepts that must be understood to determine whether the results of a study are valid and useful. In evaluating a clinical trial, look for clinical outcome measures that are clear cut and clinically meaningful to you and your patients,1 and determine the magnitude of the difference in achieving these meaningful outcomes of the treatments studied. In addition, determine the precision of the estimate of the differences among treatments. The most useful measures of the magnitude of the treatment effect are the differences between response rates and its reciprocal, the number needed to treat (NNT).2- 4 The NNT represents the number of patients one would need to treat to achieve 1 additional cure.
Bigby M. Confidence Intervals. Arch Dermatol. 2000;136(3):386-387. doi:10.1001/archderm.136.3.386