Transcutaneous bilirubinometer: intermeter reliability.
This study examined intermeter reliability on two transcutaneous bilirubinometers on 30 white prephototherapy infants between 36 and 42 weeks' gestation with birthweights greater than 2500 g. All infants were screened with both meters, on the forehead, within a 5-minute period. Meter readings were taken within 30 minutes following infant heelstick for serum bilirubin. The Pearson correlations between serum bilirubin and jaundice meter measurements were .75 for meter 1 and .76 for meter 2. The correlation between jaundice meter indices from the two meters was .87, leaving 13% of the variance in one meter not accounted for by the correlation with the other meter. An examination of the differences between the two meters revealed that in the upper range (greater than or equal to 16) meter 1 was always higher than meter 2 by, on the average, 2.6 points (SD = 1.3). In readings of 15 or less, the differences between the meters were inconsistent. Additionally, in 16 of 30 cases, the action levels from the two meters led to the same conclusion; in the remaining 14 cases, meter 1 gave a positive reading (indicating the need to determine serum bilirubin) while meter 2 gave an equivocal reading (indicating user discretion in deciding whether serum bilirubin should be determined). These data indicate that each meter should be individually evaluated to determine action indices, and that jaundice meters should not be used interchangeably on the same infant.