Plotting Bilirubin Levels in Newborns: How Labs Are Wrestling with IT Barriers to Target Severe Neonatal Hyperbilirubinemia
Many healthy newborns develop jaundice during their first few days of life, but a few are at risk for developing severe hyperbilirubinemia, a condition that can lead to kernicterus. Among the complications of this lifelong brain syndrome are cerebral palsy, mental retardation, and hearing loss.
Last July, the American Academy of Pediatrics (AAP) released updated clinical guidelines for risk assessment and management of severe hyperbilirubinemia in newborns. These guidelines indicate measuring total serum bilirubin (TSB) prior to discharging a newborn as the best documented method of assessing the baby’s risk of developing hyperbilirubinemia, and that any TSB measurement—regardless when taken—must be interpreted in terms of the infant’s age in hours.
Incorporating this guideline in the laboratory has been a challenge for lab directors, however, because most lab information systems (LIS) are not equipped to capture an infants’ time of birth, making it impossible to analyze the bilirubin levels by hour. In this issue, Strategies looks at how one medical director has developed a “workaround” algorithm with help from LIS technicians to comply with these AAP guidelines.
While severe hyperbilirubinemia is rare, it is treatable if detected early enough. The key to early detection is identifying abnormal bilirubin levels, which is done by interpreting these levels by plotting the hour specific measurements on a Bhutani graph—a percentile-based predictive bilirubin nomogram that was developed by the University of Pennsylvania’s Vinod K. Bhutani, MD, FAAP, and associates and published in Pediatrics in Jan. 1999 (see Figure 1 below).
Figure 1: Bhutani Nomograph. Reproduced by Permission of Pediatrics 2004; 114(1): 297-316.
Plotting an infant’s bilirubin levels on an hourly graph was normal practice when pediatricians were treating cases involving incompatibility of Rh antigens. This can lead to erythroblastosis fetalis, although Rh-associated incompatibility is now prevented by immunizing mothers with Rhogam, explained M. Jeffery Maisels, MD, chair of the AAP Subcommittee on Hyperbilirubinemia and with the Department of Pediatrics, William Beaumont Hospital in Royal Oak, Mich. But since Rh-associated erythroblastosis fetalis is rarely seen anymore, pediatricians have forgotten that it’s critical to look at infant bilirubin levels in hours, rather than days, he added. “As babies nowadays leave the hospital before they are 36-48 hours old, talking about a bilirubin level on day one or day two is completely irrelevant,” Maisels said. “A doctor needs to understand that a bilirubin level of 8 at 24 hours means there is some problem with the baby because it is above the 95th percentile. That same bilirubin level of 8 at 47.9 hours is below the 40th percentile in the Bhutani nomogram.”
A group of physicians and laboratorians with Salt Lake City’s Intermountain Health Care (IHC) system believed in the importance of hourly bilirubin reporting even before the AAP guidelines came out, but it required some extra effort and technological maneuvering to bring their LIS component in line. “Our major difficulty in doing this has been that our LIS does not store birth time, just birthdate,” said Phil Bach, PhD, of IHC’s Primary Children’s Medical Center. “Since the bilirubin is interpreted based on the number of hours since birth, our LIS can’t provide an interpretation.”
In many of IHC’s pediatric units, nurses were plotting out the results by hand on paper charts, which was problematic for many reasons. “It took up their time and they weren’t very accurate,” said Larry D. Eggert, MD, System Medical Director of IHC’s Newborn Clinical Program. In 2002, Eggert and his associates spearheaded an initiative to work with IHC’s computer team to find a way to store an infant’s birth time, draw the Bhutani graph, and then plot the bilirubin levels.
“We were able to go to an alternative source in our data repository, which is the time of birth in the labor and delivery information, which we record in real time,” explained Eggert. “We were able to use the patient identification number and match the birth time from that secondary source. Once we did that, it was relatively trivial to print those values out on the chart.” The IT team also implemented transcutaneous bilirubin monitoring with point-of-care testing. If clinicians are doing transcutaneous bilirubin testing using a point-of-care device off site, they can use a Web-based program to bring the same plotting tool up on their computer, and then just put in the date and time that the sample was obtained or the transcutaneous bilirubin test was done.
With a bit of trial and error, it took Eggert and the IT team about six weeks to develop a system of using this secondary data source, and then three months to refine and troubleshoot. “It seems to be working well so far in the facilities that we’ve rolled out,” he added. “We did a six month trial at two facilities, and now we’re rolling it out facility-by-facility through the rest of our 18-hospital system.”
Even though he is satisfied with this system, Eggert warned that lab directors and clinicians need to be persistent and stubborn to push this initiative forward, because they’re likely to encounter resistance from the IT department. “Don’t take ‘no’ for an answer,” he advised. “When the IT people tell you it can’t be done, it really means they don’t know how to do it. But other people do, and they just need to go out and find the right resources.”
For more information:
- The updated clinical guidelines for assessing hyperbilirubinemia:
Management of Hyperbilirubinemia in the Newborn Infant 35 or More Weeks of Gestation Subcommittee on Hyperbilirubinemia, Pediatrics 2004; 114; 297-316, www.pediatrics.org/cgi/content/full/114/1/297
The Bhutani chart was introduced in a January 1999 Pediatrics article:
Bhutani VK, Johnson l, Sivieri EM, Predictive ability of a predischarge hour-specific serum bilirubin for sunsequent significant hyperbilirubinemia in healthy term and near-term newborns, Pediatrics 1999 Jan; 103(1):6-14.