Honing in on the Range: Using the Electronic Medical Record to Establish Normal Reference Ranges for Pediatric Coagulation Testing

Laboratories typically validate testing for prothrombin time (PT) and activated partial thromboplastin time (aPTT) by collecting samples from 21 to 50 healthy individuals and using the mean value, plus or minus two standard deviations, to establish a normal reference range. It is more difficult to establish pediatric reference intervals because parents are often unwilling to volunteer their children for unnecessary blood draws. At our institution, a large, tertiary care pediatric academic center, it has been our practice to establish normal coagulation test ranges by collecting extra blood from approximately 50 children getting blood drawn for other reasons.  Since well children typically require blood draws at very limited times, such as the 9-12 month well visit screen, our ranges have been based on a disproportionately large percentage of very young children. We worried that this may have resulted in an inappropriately narrow reference range for the PT in our lab as we noted a large number of otherwise healthy children referred for evaluation of a mildly prolonged PT who were found to have normal factor levels, reassuring bleeding histories and unremarkable coagulation workups.  

 We sought to address this issue by interrogating our current reference ranges for PT and aPTT by using our electronic medical record (EMR). On reviewing the literature, we have found no published reports of another institution using the EMR to validate normal laboratory reference intervals.  We believe that this method has several advantages. It has permitted us to collect data from a larger and more diverse cohort of pediatric patients and allowed us to investigate appropriate reference ranges for age and gender specific groups of patients. To carry out this study, two pediatric hematologists reviewed de-identified laboratory data obtained during routine testing from 2012-2014, along with associated medical diagnoses, on pediatric outpatients, ages 2 to 23 years, at The Children’s Hospital of Philadelphia.  265 patient samples were identified as appropriate for study inclusion. Patients were excluded if they had an underlying medical condition or were taking medications that might lead to alterations in coagulation testing.  

The previously established normal reference ranges for PT and aPTT were 11.6-13.8 and 22-36 seconds respectively. Using data obtained from the EMR, we found similar reference ranges for the aPTT at 25-38 seconds, and a narrowed ranged for the PT at 12.9 to 13.9 seconds. We were able to stratify by age and gender and while we found that there was no significant difference between the normal ranges of PT and aPTT for boys and girls, we did find that there was a significant difference in the ranges of the PT when patients were divided by age.  The range for PT in children age of 2-11 years was 12.5-13.6 seconds while the range for children 12-23 years was 13.05 -13.9 seconds (p<0.01). The PT in the younger children was narrower and not normally distributed, while in the older children, it was normally distributed but significantly wider with a much higher upper limit. Interestingly, there was no significant difference in the range of aPTTs for children of different ages.   

We anticipate that in the future, it will be possible to use the increasingly vast quantity of data stored in EMRs to establish more accurate laboratory reference ranges for unique subgroups of patients.  We hope that this information will eventually allow us to precisely identify abnormal lab results and minimize additional testing that is often medically unnecessary, expensive and anxiety provoking.  Using the EMR may be a way to validate normal ranges in a laboratory performing routine testing on pediatric specimens.  Further studies are needed to confirm these findings.