Calculation of reference intervals for complete blood count in elderly population using patient data

Conventional method of calculation of reference interval (RI) for complete blood count (CBC) parameters for elderly population is challenging as finding healthy elderly volunteers is difficult. This study was carried out to assess the possibility of use of stored patient data of CBC reports of elderly patients in calculation of RI for CBC parameters. A retrospective, cross sectional study was carried out in a hospital laboratory using stored data of patients >65 years of age (n=3094). Relevant clinical criteria and statistical tools were used to refine data to obtain representative samples for each CBC parameter. RI were calculated using standard statistical tools. Calculated RI were compared with RI provided by the manufacturer (RI for >18years old) and with RI defined in five other countries. All data sets showed normal distribution thus, parametric approach was adopted to calculate RI. The RI of all CBC parameters were similar to those of other countries. The RI obtained in this study for red cell indices, platelet count and white blood cells are similar to the RI provided by the manufacturer for adults. The RI obtained for RBC, Hb, PCV were clinically significantly lower compared to the RI provided by the manufacturer for adults. When appropriate selection criteria and relevant statistical tools are applied, patient data can be effectively used to calculate RI for CBC parameters in elderly.