Each laboratory must determine the referrence intervals according to their own patient population. However, the reference interval studies are both costly and laborious. In recent years, the reference intervals was easily and indirectly calculated from patient results registered in the hospital information systems. We aimed to determine reference intervals for some hematological test parameters from patient results in the hospital information systems non-parametric percent estimation method and compared with the reference intervals that we use in our laboratory and recommended by the existing manufacturers. In this study, whole blood test results of 39,723 patients in the Hospital Information System were used. For detecting the high and low reference intervals MedCalc non-parametric percent estimation statistical program was used. Reference intervals were calculated separately in gender and age subgroups. Our results were indicated separately using of reference intervals for all of the sub-parameters for both gender and in the 0-12 age group. In addition, for hematocrit and hemoglobine tests in men over the age of 65 must be used different reference intervals than other groups. The reference interval values that we have obtained at the end of our study were shown significant differences from the values recommended by the manufacturer company and we use. Furthermore, Our reference interval values are more suitable for use in our laboratory.
Keywords: Reference interval, non-parametric method, indirect method, hospital information system, whole blood test.
