Subject-based reference values.

Medical decisions based on comparison with group-based reference value have often less than desired sensitivity to significant changes in the biochemical or physiological state of an individual under investigation. Subgrouping of the reference values according to sex, age, or other criteria does frequently not solve this problem. The better solution is to compare new values with old results from the same individual, i.e., with subject-based reference values. Two types of criteria may be used: those based on critical differences between two results (the reference change limit) and those based on time-series models.