(1289) Excess Daily Blood Count Sampling in Children: A PICU Data Collaborative Study

Introduction: Excessive laboratory testing is harmful and multiple medical societies have recommended against routine testing in stable patients. Here we identify practice patterns in pediatric intensive care units (ICUs) for obtaining daily complete blood counts (CBCs) and describe the information gained for hemoglobin (HGB) and platelet count (PLT) values. We hypothesized that (1) daily CBC testing practices vary across sites, (2) daily testing reveals newly actionable results < 10% of the time, and (3) >50% of children who have daily CBC testing never have newly actionable HGB or PLT results.

Methods: Data were extracted from the PICU Data Collaborative (PDC) [pedsdata.org] database, which includes granular ICU data. A “daily CBC” was defined as one taken 20-28 hours from an immediately prior or subsequent CBC. The daily CBCs per ICU-day were compared across sites with ANOVA. Actionable results were defined as either (1) HGB < 7.0 g/dL when the most recent HGB was >= 7.0 g/dL, (2) HGB decrease by at least 25% from prior, or (3) PLT < 50 K/uL when the most recent PLT was >= 50 K/uL.

Results: 147,787 CBCs were identified during an ICU stay from 4 PDC sites and 54,967 hospital encounters. There were 59,563 (40.3%) “daily CBCs.” 13,354 (24.3%) encounters had at least one daily CBC while in an ICU. The mean rate of daily CBCs for the 4 sites ranged from 0.111/patient-day to 0.323/patient-day (p < 0.001). There were 1,123 (1.9%) daily CBCs where the HGB was < 7.0 g/dL when the most recent value was above that, 1,102 (1.8%) instances of a HGB drop of > 25% from most recent, and 1,268 (2.1%) instances of PLT < 50 K/uL when the most recent value was above that. 56,435 (94.7%) of the daily CBCs did not meet any actionable criteria. Of the encounters that had at least one daily CBC in the ICU, 11,204 (83.9%) never had a daily CBC that met any actionable criteria.

Conclusions: Daily CBCs are commonly performed in critically ill children. There is significant practice variability across sites. These routine tests reveal newly actionable changes in HGB or PLT values < 10% of the time, and most children did not have an actionable result during their entire ICU stay. These results provide further support for the recommendation that routine daily testing in critically ill children should be reserved for selected patients.