Evaluation of transcutaneous bilirubinometer (DRAEGER JM 103) use in Zimbabwean newborn babies

Abstract

Background: Acute Bilirubin Encephalopathy in the neonatal period is a major cause of permanent disability. Effective screening and surveillance are essential in the newborn period to enable timely management. Noninvasive transcutaneous bilirubin devices have been successfully used for screening in many settings. We evaluated the accuracy of the Draeger JM 103 (Medical Systems, USA) for estimating serum bilirubin in Zimbabwean newborns.

Methods: Paired transcutaneous (forehead and sternum) and serum bilirubin measurements were compared on 283 infants consecutively recruited between 01 August and 30 November 2015 at Harare Hospital Neonatal Unit. Using serum bilirubin as gold standard, Pearson Correlation Coefficient (r) was calculated for the two transcutaneous measurement sites. Linear regression plots of transcutaneous versus serum estimates were performed. Comparison was made between preterm and term babies. Specificity, sensitivity, positive predictive value and negative predictive value of the JM103 were calculated including ROC curves to assess the accuracy of the diagnostic tests.

Results: Fifty-five percent of the babies were male. Median gestational age was 38 weeks (range 28-42). One hundred and fifteen (41%) were preterm. Median postnatal age was 3 days (range 0-10). Serum bilirubin ranged 85-408 μmol/l, transcutaneous bilirubin sternum; 170-544 μmol/l and forehead; 119-510 μmol/l. Correlation between serum and transcutaneous bilirubin (sternum) was 0.77 and between serum and transcutaneous (forehead) was 0.72. Preterm babies correlation for sternum was 0.77 and forehead was 0.75. Term babies correlation for sternum was 0.76 and forehead was 0.70. The sensitivity for the sternum site was 76%, specificity 90%, Positive Predictive Value of 70 and Negative Predictive Value 92. Sensitivity for forehead site was 62%, specificity 95% with a Positive Predictive Value of 80 and Negative Predictive Value of 90. Bland-Altman plot of serum versus transcutaneous measurements showed agreement between the tests. The ROC curves showed that the accuracy of the two diagnostic tests were good with no significant difference between the two, p = 0.2954.

Conclusion: The study demonstrated a strong positive correlation for both sternum and forehead sites with serum bilirubin in this Zimbabwean population of African origin. However, the sternum is a better site for identifying babies with jaundice compared to forehead. The Draeger JM-103 can be used to screening for neonatal jaundice in this population.

Keywords: Bilirubin; Correlation; JM-103; Jaundice; Neonate; Zimbabwe.

Fig. 1

Linear Regression Plot for Forehead TcB Versus TSB. Linear regression plot showing the relationship between Forehead TcB and TSB. As Forehead TcB concentration increase TSB concentration also increase showing a positive linear relationship with a coefficient of determination, R² = 0.52

Fig. 2

Linear Regression Plot of Sternum TcB versus TSB. Linear regression plot showing the relationship between Sternum TcB and TSB. As Sternum TcB concentration increase TSB concentration also increase showing a positive linear relationship with a coefficient of determination, R² = 0.59

Fig. 3

Bland Altman Plot showing difference between TSB and TcB forehead. The Bland Altman plot shows the agreement between Forehead TcB and TSB measurements. The plot shows that only 3.89% of the points lie outside the limits of agreement

Fig. 4

Receiver Operator Characteristic Curves for Sternum and Forehead TcB. The ROC curve shows the accuracy of Forehead TcB and Sternum TcB. The ROC curves show good accuracy in the two methods in separating babies needing phototherapy and those not in need with no significant difference in the two methods