Dear Editor,

In the context of neonatal jaundice, clinicians discharge the patient or make appropriate clinical decisions according to current guidelines [1, 2] based on total bilirubin (TBil) concentrations. Point-of-care testing (POCT) devices are commonly used for this purpose, and only in selected cases, clinicians request TBil measurement by the central laboratory. The Department of Laboratory Medicine of La Paz University Hospital, one of Spain’s largest tertiary hospitals, has led a multiparameter, multisite POCT network accredited by ISO 22870 over the last 24 years. TBil POCT is conducted exclusively at the Neonatology and Pediatric Emergency Departments.

Considering the clinical impact of interference by nucleated red blood cells (NRBC) in TBil measurements via POCT, as reported by Colombo, et al. [3], we aimed to evaluate the impact of this interference in our patients.

This retrospective analysis was conducted between January 2017 and September 2022. All newborns with NRBC results>150/100 white blood cells (WBC) and TBil measurements by both POCT and the laboratory performed on the same day were included (Fig. 1). The differences in concentration and hour between the POCT and laboratory TBil results were calculated. NRBC manual counts were obtained from peripheral blood smears. TBil was measured by POCT in the clinical setting using whole blood subjected to spectrophotometry on an ABL90 Flex blood gas analyzer (Radiometer, Brønshøj, Denmark) and using serum/plasma samples (BD Vacutainer, Mexico City, Mexico) exposed to chemical oxidation (vanadate method) on an Atellica Solution-CH instrument (Siemens Healthineers, Erlangen, Germany) in the central laboratory. The data were obtained from the laboratory information system (LabTrak; Intersystems, Cambridge, MA, USA).

During the study period, 263 of 2522 samples had NRBC>150/WBC (range, 151–1,615). Eighty-six samples from 59 patients had TBil results from POCT and laboratory measurement performed on the same day. None of the samples required dilution for TBil measurement.

Both methods showed an acceptable concordance, with y= 0.87 (95% confidence interval [CI], –3.38–3.85)+0.9 (95% CI, 0.72–1.15) x. The predicted values at the clinical decision levels showed a difference of 5%, which is below the laboratory specification (optimum allowable measurement uncertainty based on biological variation: 10%).

In 96.5% (83/86) of samples, the TBil concentration difference was <2 mg/dL (range of TBil measured by the laboratory: 0.7–15.4 mg/dL; SI units, μmol/L; conversion factor, 17.1), and 77.1% of these samples were measured in a <1-hour interval. Given this time interval, we assumed that no intervention had occurred in these samples between the two measurements. In the remaining three cases (3.5%), the discrepancy between the POCT and laboratory TBil results was >2 mg/dL. These patients were evaluated individually.

Patient 1: We observed a 256 NRBC/100 WBC count, an increase in the TBil concentration from 10 mg/dL (POCT) to 12.5 mg/dL (laboratory), and a 7.5-hour difference between the blood collections. The patient was diagnosed as having rhesus (Rh) isoimmunization.

Patient 2: We observed a 291 NRBC/100 WBC count, an increase in the TBil concentration from 9.4 mg/dL (laboratory) to 12.8 mg/dL (POCT), and a 2.3-hour difference between the blood collections. The diagnosis here as well was Rh isoimmunization.

Patient 3: We observed a 229 NRBC/100 WBC count, a decrease in the TBil concentration from 23.6 mg/dL (POCT) to 14.3 mg/dL (laboratory), and a 6.5-hour difference between the blood collections. This patient had a complex heart disease with a secondary hemophagocytic syndrome, and daily plasmapheresis was prescribed.

The medication of the three patients was reviewed, and none received drugs that could interfere with TBil measurement by POCT. Discrepancies in these patients were attributed to a difference of >1 hour between the blood collections. As these were different blood draws, a change in the patients’ clinical situation or an intercurrent treatment may explain the observed changes in the TBil concentration. Although we did not review the clinical course in these three patients, they displayed a similar pattern, with a decrease in the TBil concentration, measured by POCT and in the laboratory, over time.

In our experience across 6 years of this study, based on the analysis of a large number of samples with a wide range of NRBC count, we have not observed the interference described by Colombo, et al. [3] of NRBC in the measurement of neonatal TBil by POCT, despite using the same model of blood gas analyzer [3]. Our results suggest that this type of interference is very rare or that other factors (i.e., interference of another substance cleared with the exchange transfusion) may be involved. Nevertheless, we agree with comments in previous clinical case reports [4, 5] regarding the need for effective communication with clinicians and POCT users about the importance of continuous education, to ensure a good understanding of the measurement methods and their potential interferences and limitations, especially in environments where knowledge of analytical issues may be limited.