Clinical decision support for preterm hyperbilirubinemia

What are clinical decision support (CDS) tools?

CDS tools are digital solutions designed to assist healthcare providers to improve patient outcomes, reduce medical errors, and increase compliance with standards of care. Common applications include preventing the overuse of certain tests to support laboratory stewardship and reminding healthcare providers of newly implemented workflows to ensure adherence.

How can CDS tools help evaluate preterm infants for hyperbilirubinemia?

CDS tools can automate many manual evaluation steps that may otherwise lead to inconsistent and delayed evaluations of hyperbilirubinemia.

Accurate and timely evaluation of hyperbilirubinemia is critical for preterm infants since they are at higher risk of developing kernicterus and irreversible damage to the brain. With prompt evaluation and intervention, most cases of kernicterus can be prevented. Unfortunately, unlike newborn infants 35 or more weeks of gestation, there is limited consensus on how to evaluate preterm infants for hyperbilirubinemia.

Most clinicians use medical decision levels (MDLs) based on serum total bilirubin results and a combination of factors including the patient’s gestational age, age, and weight to guide the initiation of phototherapy or exchange transfusion.

Although the evaluation workflows vary among medical institutions, multiple manual steps are generally followed: 1) identify preterm infants at risk of developing hyperbilirubinemia; 2) retrieve information scattered across different databases within the laboratory information system (LIS) and electronic health record (EHR), such as the most recent bilirubin result, specimen collection date/time, gestational age, birth date/time, and the most recent weight; 3) perform the necessary calculations, such as a preterm infant’s age in hours; 4) identify the corresponding MDLs to initiate phototherapy or exchange transfusion from a worksheet, nomogram, or chart, as indicated by each medical institution’s policy; 5) compare the identified MDLs with the most recent bilirubin result; and 6) order phototherapy or exchange transfusion, if not already in place and not contraindicated.

This multistep manual evaluation workflow is laborious, and a 4% error rate was reported for manual hyperbilirubinemia risk stratification. Children’s Hospital Los Angeles implemented a CDS tool and has significantly reduced delayed phototherapy orders. Nonetheless, CDS tools are only meant to assist clinical decision-making, not to replace clinical judgment and dictate clinical practices, as each case is unique.

What are some challenges to implementing a CDS tool to manage neonatal hyperbilirubinemia?

Early engagement of the hospital IT team is critical to implementing CDS tools with complex algorithms. Conventional LIS tools are limited, as some patient information required to evaluate hyperbilirubinemia — including gestational age, birth time, and weight — is stored in other modules of the EHR and cannot be readily used by the LIS. Support from healthcare providers is also essential.

In addition, laboratory medicine professionals, healthcare providers, and the hospital IT team must discuss many nuances. What are the manual steps to be automated and replaced? What happens when critical information is later documented or corrected? Is the automated evaluation process triggered by each new bilirubin result, or as indicated by healthcare provider? How should evaluation results be communicated? Are there other capabilities of the CDS tool that may be beneficial? Thoroughly discussing each decision is essential to ensure that healthcare providers' needs are fully addressed, and limitations are clearly understood.

Finally, teamwork is required to thoroughly evaluate the CDS tool, including multiple iterations with collaborative discussions. It would also be beneficial to track relevant quality indicators, such as the number of delayed phototherapy orders and length of stay for preterm infants with hyperbilirubinemia, in order to monitor the outcome of this quality improvement project.

Yi Xiao, PhD, DABCC, FADLM, is an assistant professor of clinical pathology at the Keck School of Medicine of University of Southern California and the assistant director of the core laboratory at Children’s Hospital Los Angeles. +EMAIL: [email protected]