Reference Interval Development
CALIPER reference intervals are calculated in a systematic manner following guidelines published by the Clinical Laboratory Standards Institute (CLSI) in their report Defining, Establishing, and Verifying Reference Intervals in the Clinical Laboratory; Approved Guideline – Third Edition (C28-A3).
Testing has already been performed on several key instruments that are most commonly used in paediatric laboratories in Canada and worldwide including the Abbott Architect, Ortho Vitros, Roche Modular/Cobas, Beckman, and Siemens Immulite systems. Future studies are planned for additional tests still outstanding on these instruments, as well as for other techniques and instruments not yet assessed. All of the analytical methods are rigorously controlled according to the current standards for clinical laboratory operations. To achieve harmonization and standardization across CALIPER study centres, paediatric centres in Canada participating in the CALIPER project are required to complete a quality assurance/standardization program through CEQAL. CEQAL provides reference materials for a number of clinical chemistry assays which are being tested by all CALIPER sites to assess analytical performance and comparability of laboratory assays on various chemistry and immunoassay platforms.
Children from birth to 18 years of age are recruited to take part in CALIPER. Recruitment initiatives target schools, universities, community centers, neighborhood groups, festivals, summer camps, and arts and cultural groups in Canada. Before donating a blood sample, participants complete a short questionnaire and give written informed consent or assent, where applicable.
- Inclusion criteria: healthy children (determined based on responses to questionnaires) of both sexes between 0 and 18 years of age.
- Exclusion criteria: history of chronic illness or metabolic disease, acute illness within the past month, or prescribed medication over the past month. Demographic data collected includes previous and current health status, diet, exercise status, ethnicity, BMI parameters, and Tanner stage (where appropriate).
Sample Collection and Analysis
Depending on the age of the child, 2 to 10 mL of whole blood is collected by trained paediatric phlebotomists using standardized conditions. Serum separated tubes (BD) are used, which allow for the collection of samples following centrifugation at 5000 g for 8 minutes. Blood samples are separated less than 5 hours after collection and are serum aliquoted in 0.5 mL aliquots in cryovials and stored at – 80oC. Samples have been analyzed on several analytical platforms including the Abbott Architect c800, i2000SR, and ci4100, Ortho Vitros 5.1 FS Chemistry System, Roche Modular/Cobas system, Beckman Coulter Lx20/Dx800, and Siemens Immulite 2500. Analytical methods are controlled according to manufacturers’ instructions via preventive maintenance and function checks, calibration, and quality control. All specimens tested are subjected to automated interference analysis for hemolysis, icterus, and turbidity.
Data Analysis and Statistics
To establish all reference intervals published to date, a highly standardized approach was used. Preliminary analysis using visual graphical (scatter and distribution plots) and statistical approaches was performed to examine the data for skewness. Following this, significant outliers were removed.
The data were then examined to determine appropriate age and sex partitions. Partitions were decided on the basis of trends observed in distribution plots and were then statistically evaluated using the Harris and Boyd test.
Non-parametric or robust methods were used to determine specific reference intervals. For partitions with sample sizes greater than 120, the non-parametric rank method was used, while the robust statistical methods were applied to partitions smaller than 120. The 90 per cent confidence intervals around the lower and upper reference limits were then calculated.
All statistical analysis was performed using MS Excel, SPSS, and R software.
For a more detailed description of the methods previously applied in calculating intervals and subsequently transferring those values to other instruments, please refer to specific papers within our complete list of CALIPER publications.