When a new test is introduced the expected values for the "normal" population need to be determined. It is important to both include a representative cross-section of the local community and exclude those who might bias the distribution.
Inclusion |
Exclusion |
healthy |
disease or comorbid condition that might affect the analyte |
afebrile |
febrile |
normal hydration |
dehydrated |
not exposed to excessive stress |
moderate to severe stress or exertion |
no recent surgery or illness |
recovering from recent surgery or illness |
if female, nonpregnant, nonlactating and not on oral contraceptives |
if female, pregnant. lactating or on oral contraceptives |
representative of local population where the test will be used |
notably different from the local population |
no occupational or environmental chemical or toxin exposures |
occupational or environmental chemical or toxin exposure |
not recently transfused |
recently transfused |
no medications, or drug unlikely to affect the analyte |
drugs that might affect the analyte |
no recent use of vitamin supplements, or use not likely to affect the analyte |
recent use of vitamin supplements likely to affect the analyte |
no recent use of stimulants |
recent use of alcohol, tobacco or stimulants |
proper pretest preparation (fasting, etc.) |
not properly prepared for the test |
sample properly collected, handled and stored |
collection, handling or storage artifact |
Ideally the persons used to determine the normal reference range will be evaluated prospectively. However, sometimes it is necessary to check a reference range retrospectively, selecting values from those persons at low risk for disease (health fairs, pre-employment screening, elective surgery, etc.). This has to be done carefully to prevent bias.
Reference ranges can also be developed for persons affected by a specific disease, provided greater care is taken to exclude potentially confounding variables.
Specialty: Clinical Laboratory