Pediatric reference ranges derived by different methodologies

Medical Laboratory Observer, March, 2006 by Daniel M. Baer

Q We are a mid-size hospital with low pediatric volume due to a nearby children's hospital. Can I use a reference range from literature if I do not have access to pediatric specimens to establish my own? I am nervous about using a reference range based on another methodology.

A Identifying pediatric reference intervals is a continuing problem for all laboratories. Numerous factors need to be considered including age, sex, diet, drugs, posture, stress, and time of day. The process of transferring and validating appropriate reference ranges or reference intervals, however, is the same for both adult and pediatric patients. This process has been spelled out in the document "How to Define and Determine Reference Intervals in the Clinical Library" available from the Clinical Laboratory Standards Institute (CLSI; formerly NCCLS) in Wayne, PA.

It is acceptable to use a manufacturer's reference interval or a reference interval from another lab using a similar analytical system. If the method is markedly different, this process cannot be used. This process, called transference, needs to be assessed for acceptability (validation). Three validation processes are delineated in the CLSI reference; I will only describe two in general terms. Please see the reference for further details.

The first validation process is a subjective assessment. The laboratorian must determine that the population tested for the transferred reference interval is similar to your testing population. Next, the pre-analytical and analytical procedures must be reviewed to determine if they are consistent with the processes in your laboratory.

In the second validation process, 20 specimens that represent your sampling population are identified and tested. The 95% reference interval is validated if no more than two of 20 test values exceed the proposed interval. If three or more values are identified, then the process should be repeated with an additional 20 specimens and checked for outliers. If there are three or more outliers on the second set, then further investigation into the analytical process and reference populations need to be initiated. If there is no explanation for the observed differences, then the laboratory may need to develop its own reference interval.

--Stanley F. Lo, PhD

Technical Director

Clinical Chemistry

Children's Hospital of Wisconsin

Milwaukee, WI

Reference

1 Clinical and Laboratory Standards Institute. How to Define and Determine Reference Intervals in the Clinical Laboratory. Wayne, PA: NCCLS; 2000. Approved Guideline-2nd ed. C28-A2.

Daniel M. Baer, MD, is professor emeritus of laboratory medicine at Oregon Health and Science University in Portland, OR, and a member of MLO's editorial advisory board.

[ILLUSTRATION OMITTED]