Reference Intervals (RI's):
In the clinical laboratory, test results are typically expressed in numbers with units. Reference intervals reflect the amount of analyte in a given concentration. The patient's test results are compared to a reference interval, or a range that has been approved to reflect the results expected for a healthy population.
Test results are able to be correlated with medical decision levels or reference intervals. Sometimes correlation studies are done by the lab to verify test results. Values are based on both peer data, clinical research, clinical experience, instrument data, the FDA, and each individual laboratory. This is important because each laboratory you work in may vary slightly in its reference intervals or ranges.
Not all tests have a clear association with a single disease or condition, however, and there is not always sufficient medical evidence to define a specific reference interval. For this reason, a statistical approach may be taken. In this case, the RI is usually based on the statistical distribution of values obtained when the same test is done on hundreds of healthy people in a population. It is then represented on a Gaussian Bell Curve. The bell curve shows the values that are expected along with variations (standard deviation), which is typically 1 SD, 2 SD or 3 SD in the clinical laboratory.
In clinical chemistry, typically 34 out of 100, or 66% of test results fall within 1 SD of the clinical mean (average) for an analyte. 5 out of 100, or 95%, fall within 2 SD of the mean. 1 out of 100, or 99%, fall within 3 SD of the mean. SD stands for "standard deviation".
A range of results represents the values that are typically seen within healthy populations. The reference range or interval for a test is generally chosen to represent about 95% of healthy people. It is set to be the range from -2 SD to +2 SD, or a 2 SD range. It could also be set to represent 99% of healthy people, or a -3 to +3 SD range. Some normal ranges fall within < or >. Some tests require an actual number, in which case a sample may need to be diluted and then rerun to get an exact number.
Test results are able to be correlated with medical decision levels or reference intervals. Sometimes correlation studies are done by the lab to verify test results. Values are based on both peer data, clinical research, clinical experience, instrument data, the FDA, and each individual laboratory. This is important because each laboratory you work in may vary slightly in its reference intervals or ranges.
Not all tests have a clear association with a single disease or condition, however, and there is not always sufficient medical evidence to define a specific reference interval. For this reason, a statistical approach may be taken. In this case, the RI is usually based on the statistical distribution of values obtained when the same test is done on hundreds of healthy people in a population. It is then represented on a Gaussian Bell Curve. The bell curve shows the values that are expected along with variations (standard deviation), which is typically 1 SD, 2 SD or 3 SD in the clinical laboratory.
In clinical chemistry, typically 34 out of 100, or 66% of test results fall within 1 SD of the clinical mean (average) for an analyte. 5 out of 100, or 95%, fall within 2 SD of the mean. 1 out of 100, or 99%, fall within 3 SD of the mean. SD stands for "standard deviation".
A range of results represents the values that are typically seen within healthy populations. The reference range or interval for a test is generally chosen to represent about 95% of healthy people. It is set to be the range from -2 SD to +2 SD, or a 2 SD range. It could also be set to represent 99% of healthy people, or a -3 to +3 SD range. Some normal ranges fall within < or >. Some tests require an actual number, in which case a sample may need to be diluted and then rerun to get an exact number.
Terminology:
![Picture](/uploads/4/7/4/4/47442993/351397571.png?250)
- Standard Deviation (SD): used to quantify the amount of variation present in a set of values; dispersion
- Mean: the average of a set of values or the expected value(s)
- Median: the middle number or average of the two middle numbers in a set of values
- Mode: the most frequently seen/used number in a set of values