RESUMO
BACKGROUND: Venepuncture-related blood loss is a common cause of neonatal anemia. Currently, this is the only way to obtain hemoglobin levels. This causes distress for the infant but can also lead to the need for blood transfusions. Recently, a new technique for measuring hemoglobin levels non-invasively has been developed to reduce iatrogenic blood loss and pain. OBJECTIVE: To compare hemoglobin levels obtained using a transcutaneous spectroscopic device (Mediscan 2000, MBR Optical Systems, Wuppertal, Germany) with venous or capillary blood samples in neonates. METHODS: Single-center prospective cohort study of term and preterm infants. The white light spectroscopic device was placed on the forearm for 60 s to measure hemoglobin content within 4 h of venous or capillary blood sampling. Pain reactions of the infants were assessed by using a neonatal pain assessment tool. Results were analyzed by Bland-Altman comparison and Wilcoxon signed-rank test. RESULTS: 80 infants (mean gestational age 29.8 +/- 3.8 weeks, mean birth weight 1,300 +/- 690 g) were enrolled into the study. A total of 313 spectroscopic recordings within 2 h of a clinically indicated blood sample (181 capillary, 142 venous) were taken. The correlation coefficient R(2) was 0.96 for capillary/spectroscopic and 0.99 for venous/spectroscopic pairs. Pain scores were significantly less for the spectroscopic measurements (p < 0.01). CONCLUSION: The results show good correlation between the hemoglobin blood levels and spectroscopic measurements. The slightly lower correlation coefficient for the capillary samples demonstrates a naturally higher variance in these values due to the laboratory method.
