Statistical methods for establishing and validating reference intervals

Reference intervals are an essential part of laboratory medicine, and accreditation standards require that every laboratory result is accompanied by an appropriate reference interval to provide guidance in the interpretation of the test. Furthermore, the Clinical Laboratory Improvement Act of 1988 (CLIA '88) requires laboratories to verify that the reference interval accompanying a laboratory result is appropriate for the patient population the laboratory serves. These 2 tasks are fundamental to providing quality laboratory services. In this review, we will consider the statistical methods that can be applied to establish and validate reference intervals.