Introduction

Neonatal hyperbilirubinemia is defined as a total serum bilirubin above 5 mg/dL (86 umol/L) during the newborn period [1], with 60% of newborns experiencing this condition [2]. Complications, such as kernicterus, are extremely rare [2], and universal bilirubin screening among hospitals has decreased the prevalence of severe hyperbilirubinemia and potential complications [3].

Hemolytic disease of the newborn (HDN), a pathologic etiology of neonatal hyperbilirubinemia, is frequently due to blood antigen incompatibility between the pregnant person and fetus (e.g., ABO or Rh). ABO incompatibility (ABOi) occurs in ~20% of pregnancies, with only about 1% of these newborns developing HDN [2]. This low rate may be due to decreased anti-ABO antibody transport across the placenta due to these antibodies being predominantly IgM class [4]. It may also be due to decreased development of A and B antigens on fetal red blood cells [4]. Parameters used to identify and diagnose HDN vary, but the diagnosis is suggested with rapidly progressive hyperbilirubinemia and/or neonatal anemia [4, 5]. A proposed method of detecting those who are ABOi and at increased risk of HDN has been the direct antiglobulin (Coomb’s) test (DAT), a screening immunoassay used to determine presence of maternal antibodies on the neonate’s red blood cells [6]. Data on the utility of DAT screening varies, which may lead to overuse of phototherapy for patients at lower risk of HDN [6,7,8,9].

Despite the guidelines via the American Academy of Pediatrics (AAP) in 2004 and 2009 [2, 10], there has been wide variation in hospital practice related to hyperbilirubinemia [11]. Hospital wide clinical practice guidelines for hyperbilirubinemia have been shown to improve outcomes related to unnecessary phototherapy and hospital costs [12, 13]. The 2004 AAP guidelines presented risk factors associated with increased neurotoxicity risk related to hyperbilirubinemia [2], which was further clarified in the 2009 update [10]. Isoimmune hemolytic disease is considered a neurotoxicity risk factor, but as previously stated, definitions may vary, and not all ABOi neonates exhibit hemolytic disease [2, 10]. In 2022, the AAP published another update to prior guidelines for the management and prevention of hyperbilirubinemia in the newborn [14]. In this update, the authors clearly define parameters that may indicate hemolysis, including a rapid rate (>0.3 mg/dL/h) of increase in the serum or transcutaneous bilirubin (TcB) in the first 24 h or >0.2 mg/dL/h thereafter [14]. They also propose that the DAT helps identify infants at risk for hyperbilirubinemia attributable to hemolysis and identify a positive direct antiglobulin test as a neurotoxicity risk factor [14]. Infants that are DAT-negative may be managed with usual care per these updated guidelines [14].

In our hospital, a large community-based academic hospital in the northeastern United States, a Hyperbilirubinemia Clinical Pathway (HCP) was implemented by physicians in the Newborn Nursery in January 2021, aimed at standardizing hyperbilirubinemia care among ABOi newborns. Specifically within this guideline, in attempt to decrease unnecessary phototherapy, full-term newborns that were ABOi and DAT negative were placed on the low risk curve (LRC) on the phototherapy nomogram [2, 10, 15] given likely lower risk of HDN [16]. We performed a retrospective single-center cohort study for ABO-incompatible newborns and compared rates of phototherapy before and after implementation of this guideline in the newborn nursery. We then compared rates of phototherapy among DAT negative newborns.

Methods

Context

Within our community hospital, there are ~2800–2900 newborns each year with phototherapy rates approximating 4% among all newborns. Prior to the implementation of our HCP, our department’s practice was based on the AAP guidelines [2, 10] as well as guidelines from surrounding hospitals. Providers were placing all newborns with ABOi, regardless of DAT status, on the medium risk curve (MRC) on the phototherapy nomogram. Of note, in our hospital, we obtain blood type and DAT status on all newborns born to mothers with blood type O or Rh-negative status. The new clinical guideline, started in January 2021 (Supplementary Fig. 1), implemented the following interventions: newborns that were ABOi and DAT negative will be placed on the LRC (previously placed on MRC) on the phototherapy nomogram if full term (≥38 weeks gestational age) [2, 10, 15], newborns that undergo phototherapy will be assessed immediately for need for supplementation, phototherapy will be discontinued when the total bilirubin is at least 2–3 mg/dL below the light level [17,18,19], and rebound bilirubin level will be obtained only if the newborn is deemed at high risk for continued phototherapy or exhibiting a poor response to phototherapy [20].

Study design

We performed a retrospective chart review for newborns between January 2020 and October 2021. The study population included newborns in the nursery that had ABOi. Patients who were transferred or immediately admitted to the NICU within 24 h of life, unless for severe hyperbilirubinemia, were excluded, given difference in hyperbilirubinemia guidelines. Patients who are born <36 weeks gestation are admitted directly to the NICU within our hospital, and were excluded. Patients were identified via EPIC queries for maternal blood type, extracting data for mothers with blood type O. Infant blood type was then manually verified and a manual chart review was then performed to extract data if ABOi was identified. Risk curve assignment, assigned by providers for each patient on the phototherapy nomogram, was obtained by review of admission and progress notes within the electronic medical record. Patient data was then de-identified and recorded in Excel. This study was determined to not be human subjects research and exempt by our Institutional Review Board under the policies and procedures of the Human Research Protection Program.

Measures

Patient demographic and baseline information included: sex, gestational age at birth, delivery type, birth weight, and infant blood type, including DAT status. The DAT is routinely performed in cord blood samples for all ABOi newborns at our hospital. DAT status was obtained via review of infant laboratory results.

The primary outcomes were: (1) risk curve assignments on the phototherapy nomogram used in the 2004 AAP guidelines [2] and (2) rates of phototherapy across intervention groups and among DAT negative newborns. If risk curve assignments were changed by providers during hospitalization due to high risk bilirubins and concern for isoimmune hemolytic disease, the final assignment was recorded.

Secondary outcome measures included length of stay, 7-day re-admission rate related to hyperbilirubinemia, nadir weight decrease during hospitalization, feeding method at discharge, number of bilirubin laboratory tests, bilirubin at 12 and 24 h of life. Bilirubin levels were initially measured as TcB at 12 and 24 h of life and confirmed with serum bilirubin (TsB) level if above light level, which is standard practice within our hospital for all ABOi newborns. If both TcB and TsB were reported at 12 or 24 h of life, TsB was used in analysis. The length of stay was calculated as the difference between admission electronic time stamp and discharge time stamp from the hospital.

Statistical analysis

Patient baseline/demographic and treatment characteristics distributions were summarized using counts and frequencies for categorical data and median (min–max) for continuous data. Differences in patient characteristics were assessed using the Fisher’s Exact test (categorical) or the Mann-Whitney-Wilcoxon test (continuous) where appropriate. All p values are based on a two-sided hypothesis with significance testing at a 0.05 level. R version 4.0 was used for all analyses.

Results

In our initial EPIC query, 573 newborns were found to have ABOi at 36 weeks gestational age or above between January 2020 and October 2021 (Fig. 1). Of these, 31 newborns were excluded due to immediate admission or transfer to the NICU within 24 h of life for reasons other than hyperbilirubinemia. Our final sample size included 542 total ABOi newborns born during our analyzed time period. Of these, 279 newborns were pre-guideline implementation (January 2020–December 2020) and 263 newborns were post-guideline implementation (January 2021–October 2021). There were no significant differences among baseline characteristics between all newborns across intervention groups, except for birth weight (pre-intervention median: 3365 g; post-intervention median 3285 g; p = 0.043) (Table 1). There were no significant differences among baseline characteristics among ABOi DAT negative newborns across intervention groups (Table 2).

Fig. 1: Outline for inclusion and exclusion criteria for retrospective study.
figure 1

Pre-intervention includes the timeframe from January 1, 2020 to December 31, 2020. Post-intervention includes the timeframe from January 1, 2021 to October 31, 2021. ABOi ABO incompatibility, NICU neonatal intensive care unit.

Full size image
Table 1 Baseline characteristics among ABOi newborns within pre- and post-intervention groups.
Full size table
Table 2 Baseline characteristics among ABOi DAT negative newborns within pre- and post-intervention groups.
Full size table

Primary outcomes

A significantly higher proportion of newborns were assigned to the LRC on the phototherapy nomogram after the HCP implementation (pre/post-intervention non-LRC: 91.4% vs 35.4% respectively, p < 0.0001; pre/post-intervention LRC: 8.6% vs 64.6% respectively, p < 0.0001) (Table 1). There were also a significantly higher proportion of DAT negative newborns assigned to the LRC after the HCP implementation (pre/post intervention non-LRC: 89.9% vs 28.2%, respectively, p < 0.0001; pre/post-intervention LRC: 10.1% vs 71.8%, respectively, p < 0.0001) (Table 2). There were no significant differences among the intervention groups for rates of phototherapy (24.0 % vs 19.4%, respectively; p = 0.212), including among DAT negative newborns (pre/post-intervention: 22.3% vs 15.8%; p = 0.08) (Fig. 2).

Fig. 2: Phototherapy outcomes among all ABOi newborns and ABOi DAT negative newborns pre- and post-intervention.
figure 2

DAT direct antiglobulin test, ABOi ABO incompatible. 1Fisher’s exact test; Wilcoxon rank sum test.

Full size image

Secondary outcomes

There were no significant differences among secondary outcomes between intervention groups for all newborns (Supplementary Table 1), as well as for DAT negative newborns (Supplementary Table 2). Within the entire pre-intervention group there were 10 newborns with missing bilirubin levels either 12 or 24 h of life. Within the entire post-intervention group, there were 11 newborns with missing bilirubin levels at either 12 or 24 h of life. These newborns were excluded from calculations involving bilirubin levels.

Discussion

After implementation of the HCP, providers generally adhered to the new guidelines as shown by the significant increase in LRC assignment on the phototherapy nomogram in the post-intervention group, for all ABOi newborns and for ABOi DAT negative newborns (Tables 1 and 2, respectively). With the increased LRC assignment among ABOi DAT negative newborns, there was a decreased rate of phototherapy after the intervention that did not reach significance (Fig. 2). There were reassuringly no changes in length of stay and re-admission rates for phototherapy among DAT negative newborns after the intervention (Supplementary Table 2). This supports prior findings that ABOi DAT negative newborns are likely at a low risk of HDN due to isoimmune hemolytic disease [16, 21]. Moreover, although risk curve assignment depends on gestational age (e.g., newborns with gestational age <38 weeks and no risk factors are placed on the MRC), there were no significant changes between the pre- and post-intervention groups among DAT negative newborns in regards to gestational age.

While our overall rates of phototherapy among ABOi DAT negative newborns were higher than baseline rates of phototherapy in all newborns at our hospital (~4% and 5.8% during our pre- and post-intervention period, respectively), there was no increased incidence of adverse outcomes among our patient population after placing these newborns on the LRC. Interestingly, there are highly variable rates of phototherapy among ABOi DAT negative newborns in the literature [22,23,24]. Our phototherapy rates among ABOi DAT negative newborns may be elevated due to subthreshold phototherapy initiation among providers [25], providers not adhering to the HCP, or other causes of hemolysis [16] (e.g., G6PD deficiency, in which testing may be negative during acute hemolytic episode) [26]. For example, in the post-intervention group, there were 6 ABOi DAT negative newborns that were incorrectly placed on the MRC and received phototherapy. Moreover, there were 7 ABOi DAT negative newborns with low risk 12 and 24 h of life bilirubins that received phototherapy. While this would still not account for higher than baseline rates, our study indicates the safety of treating ABOi DAT negative newborns as low risk for hyperbilirubinemia requiring phototherapy without an increase in adverse events.

With the release of the new hyperbilirubinemia guidelines by the AAP in 2022 [10], close attention to neurotoxicity risk factors is crucial. Our study supports the guidance put forth by the AAP in that ABOi DAT negative newborns may be viewed as low risk and managed with routine care. If hemolysis is suspected due to phototherapy initiation within the first 24 h of life or rapid rate of increase of bilirubin, only then should the patient be treated as having hemolytic disease and considered higher risk. Providers should then undertake further work up.

There are several limitations to our study. Our study is a single-center cohort study, which limits generalizability. We also excluded newborns that were initially admitted to the NICU for reasons other than hyperbilirubinemia, including <36 weeks gestational age. These newborns may have needed phototherapy during their hospitalization and would not be captured. While a decrease in phototherapy was observed among pre- and post-intervention ABOi DAT negative newborns, our study may have been underpowered to detect a statistically significant difference. Lastly, the time period when this study took place was during the COVID-19 pandemic, which may have impacted factors not accounted for in our analyses.

Our study shows that newborns with ABOi and a negative DAT can be safely viewed as low risk without any subsequent increase in rates of phototherapy, re-admission rates, or lengths of stay. Our findings support the new AAP guidelines recently released stating that if DAT is obtained and negative, these patients can be managed with routine care. Larger follow up studies are needed and should address further identification of risk factors contributing to hyperbilirubinemia requiring phototherapy.