Laboratory tests provide physicians with invaluable information to aid in clinical decision-making, including prognosis, diagnosis, and treatment. Population-based reference intervals are widely-used benchmarks for laboratory test result interpretation. Reference intervals are most commonly defined as the central 95% of the distribution of laboratory test results expected in a healthy reference population. Establishing reference intervals in the pediatric population poses challenges including recruitment difficulties and sample volume size. Furthermore, profound stages of growth and development throughout childhood and adolescence is reflected in the concentration of several circulating biomarkers. To ensure reference intervals reflect these changes, they have traditionally been separated into groups based on characteristics, such as age, sex, and pubertal status. However, this static partitioning oversimplifies the dynamic and gradual changes true to pediatric physiology.
The Canadian Laboratory Initiative on Pediatric Reference (CALIPER) project develops pediatric reference intervals based on thousands of healthy children and adolescents (www.caliperproject.ca). CALIPER reference intervals are currently available for over 170 laboratory tests and are partitioned into discrete reference intervals by age and/or sex, when appropriate. To better reflect the actual influence of age on pediatric laboratory values, CALIPER has developed continuous reference intervals based on a non-parametric quantile regression method.
Several studies have utilized reference curves in the pediatric population. For example, the World Health Organization (WHO)1 and the Centers for Disease Control and Prevention (CDC)2 developed reference curves for body measurements to assess pediatric growth. These methods are also valuable in laboratory medicine, as several biochemical markers similarly exhibit gradual age-related trends. For example, Zierk and colleagues established continuous reference intervals for hematological and biochemical analytes using an indirect statistical approach on retrospective laboratory data to estimate the healthy distribution from a population containing both healthy and unhealthy subjects3,4. Indirect approaches are undoubtedly advantageous for their ability to bypass obstacles involved in recruiting healthy subjects and obtain large amounts of laboratory data. However, directly sampling a healthy population remains the recommended approach. Direct sampling provides a well-defined reference population with the application of pre-defined inclusion and exclusion criteria and greatly reduces pre-analytical and analytical variation in laboratory test results. CALIPER continuous reference intervals combine both direct sampling and continuous modeling of laboratory values with age.
Some biochemical markers that exhibit minimal change with age (e.g. electrolytes), will likely not benefit from continuous reference curves. However, for biochemical markers that demonstrate large age-related changes, continuous reference intervals will likely improve the accuracy of laboratory test interpretation compared to discrete reference intervals, which simplify the effect of age. However, objective comparisons between discrete and continuous reference intervals for laboratory test result interpretation are warranted, including comparing flagging rates as well as concordance and discordance ratios.
The implementation of continuous reference intervals into current laboratory information systems poses an obstacle, as laboratories have limited capacity to accommodate mathematical functions for laboratory test interpretation based on continuous reference curves specific for age, sex, and potentially additional covariates. Therefore, reducing the complexity of continuous reference intervals into narrow age-groups is likely still required for implementation into clinical laboratories. Third-party web-based platforms can be a helpful way for laboratory professionals and physicians to input their patient’s information and laboratory test result and obtain interpretation based on continuous reference intervals.
Continuous reference intervals reflect the true dynamic relationship between age and laboratory values and will thus lead to more accurate laboratory test result interpretation in populations with dynamic age-related trends, including pediatrics.
- de Onis M, Onyango AW, Borghi E, et al. Development of a WHO growth reference for school-aged children and adolescents. Bull World Health Organ 2007;85:660-7.
- Kuczmarski RJ, Ogden CL, Guo SS, et al. 2000 CDC Growth Charts for the United States: methods and development. Vital Health Stat 11 2002;(246):1-190.
- Zierk J, Arzideh F, Rechenauer T, et al. Age- and sexspecific dynamics in 22 hematologic and biochemical analytes from birth to adolescence. Clin Chem 2015;61:964-73.
- Zierk J, Arzideh F, Haeckel R, et al. Pediatric reference intervals for alkaline phosphatase. Clin Chem Lab Med. 2017;55:102-10.