Objective: The goal of this study was to improve compliance with published guidelines regarding management of neonatal hyperbilirubinemia in infants admitted to a general pediatric hospital ward and to improve support for their breastfeeding mothers.
Methods: This quality improvement project was conducted by using Plan-Do-Study-Act cycles and statistical process control methods. Study subjects were infants >35 weeks’ gestation admitted for hyperbilirubinemia to the general inpatient ward of a large, freestanding pediatric hospital. We developed and implemented a guideline for the inpatient management of jaundiced neonates, with ongoing feedback given to the faculty on group performance. Outcome measures included monthly compliance scores based on American Academy of Pediatrics (AAP) guidelines for management of neonates >35 weeks’ gestation and the percentage of admitted jaundiced, breastfeeding infants whose mothers received lactation consultation during hospitalization. To determine the AAP compliance score, we reviewed and assigned points to each patient admission for completion of a standard evaluation, avoidance of unnecessary intravenous (IV) fluids and peripheral IV line placement, avoidance of rebound bilirubin checks while in the hospital, and the bilirubin level at discharge.
Results: Mean monthly AAP compliance scores increased from 60.5% of total possible points during the baseline period (January 2010–December 2010) to 90.4% during the intervention period (January 2011–December 2011). Lactation consultations increased from 48% during our baseline period to 63% during our early intervention period and to 90% during the last 5 months of our intervention. Length of stay was unchanged during the baseline and intervention periods.
Conclusions: Interprofessional collaboration between nurses and physicians combined with a thoughtful campaign to increase awareness of published guidelines were successful in improving the care of infants admitted with unconjugated hyperbilirubinemia.
Neonatal jaundice is a common condition, with ∼60% of infants developing visible jaundice.1,2 Rarely, neonatal jaundice can progress to acute bilirubin encephalopathy and kernicterus, devastating conditions that are preventable with early detection and treatment of unconjugated hyperbilirubinemia.3
In 2004, the American Academy of Pediatrics (AAP) published guidelines to advise practitioners on the early detection and management of unconjugated hyperbilirubinemia.4 One of the aims of the current guideline was to detect jaundice at an intervenable stage while promoting breastfeeding and minimizing unnecessary cost and treatment. In an article that calls for improved detection of hyperbilirubinemia after discharge from the newborn nursery, Palmer et al5 noted that the mere publication of a guideline is not enough to change the practice of individual physicians. Published quality improvement efforts regarding jaundice have focused on the detection of hyperbilirubinemia and the transition from the newborn nursery to the primary care physician.6–9 Recently, a single institution published their experience with using a clinical pathway to improve the timeliness of phototherapy in the emergency department.10 To our knowledge, no studies have concentrated on improving the management of infants who are readmitted to the hospital for jaundice.
The purpose of the current project was to determine if we could improve our inpatient general pediatric staffs’ compliance with the AAP guidelines on management of jaundiced neonates admitted to a general pediatric ward. We also hoped to improve the support of mothers whose jaundiced infants were breastfed. The specific aim of this project was to determine baseline compliance with AAP guidelines for inpatient management of jaundiced neonates before project initiation and to improve compliance to >80% over the next 6 months and sustain that compliance for 6 months. A secondary aim was to improve the percentage of breastfeeding infants whose mothers received a lactation consultation during the hospitalization.
This quality improvement work involved implementation of an established standard of care. No interventions involved comparison of multiple devices or therapies, and patients were not subjected to randomization. Access to medical records by project staff and approval of the projected by exempt review were granted by the hospital institutional review board.
Nationwide Children’s Hospital is a nonprofit, freestanding children’s hospital located in Columbus, Ohio. Among children’s hospitals, it is the second largest in the country, with >450 licensed beds, >20 000 inpatient discharges, 20 000 surgeries, and nearly 1 million outpatient visits per year. Virtually all children living in central Ohio who require inpatient medical or surgical care are admitted to Nationwide Children’s Hospital. Annually, >80 noncritically ill, jaundiced infants are admitted to the general pediatric wards for inpatient management of this condition. These patients are attended by general pediatricians or pediatric hospitalists. Critically ill neonates or those requiring exchange transfusions for jaundice are admitted directly to our NICU. Nationwide Children’s Hospital does not offer a newborn service at its main campus facility. Of note, few jaundiced neonates are readmitted to maternity hospitals in central Ohio.
Planning the Intervention
This study was a quality project designed to improve compliance with established AAP guidelines for management of neonates >35 weeks’ gestation admitted to the general pediatric service with unconjugated hyperbilirubinemia as their primary diagnosis. Jaundiced infants were admitted to the hospital either by direct referral from a primary care provider or after presentation to the emergency department (ED). Because noncritically ill, jaundiced infants in our hospital were not geographically localized and were cared for by faculty-led resident teams who are at various levels of training, we set out to form an interprofessional team led by a physician (K.M.T.) and a nurse champion (J.C.).
A needs assessment completed by our inpatient nurses in December 2010 (n = 76) revealed that >50% did not feel comfortable setting up phototherapy for jaundiced infants. Furthermore, only 60% of nurses felt they could provide parents with education and support regarding jaundice and only 38% of nurses felt comfortable asking questions or providing support regarding breastfeeding. During a November 2010 survey of our pediatric resident physicians (n = 40), 55% of residents did not recognize gestational age as a risk factor for jaundice in an infant born at 37 weeks. In addition, 23% of residents stated they would obtain a blood culture on a well-appearing, term infant with unconjugated hyperbilirubinemia and no other risk factors for sepsis.
Our working group convened in October 2010 and included nurses, physicians, case managers, and a medical student. Initially, we had monthly committee meetings to identify key drivers for each of our aims, review current data, and divide up tasks. Based on this information, our initial interventions targeted attending physician, resident, and nursing education. Our goal was to standardize practice among attending physicians and the various nursing units.
For physician education, we developed a physician guideline that distilled the most important information from the 2004 AAP guidelines4,11 into a 1-page document. This document was approved by the hospital pediatrics physician group in December 2010 and served 2 purposes: (1) to gain buy-in from the attending physicians to standardize our care; and (2) to provide an educational tool for residents and students when rotating on the hospital pediatrics services. For nursing education, our nursing champion used a multipronged approach that included creating educational folders for each unit, providing in-services to units on the proper setup for neonatal incubators and phototherapy equipment, and recording a presentation on “Care for the jaundiced infant” that was given during a nursing education day but could also be accessed by nurses via the hospital intranet.
The physician guideline was approved and implemented for use on the teams in January 2011. Copies of our guideline were posted in resident work areas, and attending physicians were encouraged to review this with their team. Initial data in early 2011 showed minimal improvement in our lactation consultations and variable improvement in AAP compliance. Thus, we focused on educating the night resident team, as they are the physicians who often have initial contact with infants admitted for jaundice. Although no formal nighttime resident curriculum was in existence during our initial interventions, we connected with the night teams by sending e-mails of our physician guideline and educating them regarding our efforts to improve the care of jaundiced infants. Residents were encouraged to provide feedback on the new guidelines as well.
Initial feedback from physicians and nurses suggested that existing phototherapy orders in our electronic medical record (Epic [Epic Systems, Verona, WI]) were incongruent with the physician guideline and confusing for residents and nurses. Thus, efforts were made to revise the existing phototherapy order while awaiting approval and implement of an admission order set for hyperbilirubinemia. Through our hospital quality office, we identified a champion in our information technology department to assist with any revisions in the electronic medical record. A revised phototherapy order was available in July 2011 while an admission order set was put into production in our electronic medical record in August 2011.
Methods of Evaluation
Using the Model for Improvement as a framework, we identified process and outcome measures to study the impact of our interventions.12 Baseline data were collected through manual chart reviews from January through December 2010. Our measures focused on compliance with AAP guidelines and breastfeeding support. Length of stay served as a balancing measure. To determine compliance with AAP guidelines, we developed an “AAP compliance score” that we used to score each patient admission for completion of a standard evaluation, avoidance of unnecessary intravenous (IV) fluids and peripheral IV line placement, avoidance of rebound bilirubin checks while in the hospital, and the bilirubin level at discharge (Table 1). Each domain was scored between 0 and 2 points, with a maximum AAP compliance score of 8 per patient. Because a subset of patients were admitted through the ED, where it is standard practice to place a peripheral IV line, we created a modified AAP compliance score for those patients that omitted the domain regarding peripheral IV lines and IV fluids, giving a maximum score of 6 for any patient admitted through the ED. As a marker for breastfeeding support, we monitored lactation consultations for breastfeeding mothers as the percentage of mothers who were breastfeeding on admission who received a lactation consultation. In addition, we attempted to monitor the percentage of formula supplementation in breastfed infants as well as the percentage of mothers who were still breastfeeding at discharge. Length of stay was monitored in hours and was counted from the time the patient arrived on the unit to the time he or she left the hospital.
Baseline data were collected from January through December 2010, with October through December serving as a transition period. Postintervention data were collected monthly from January 2011 through December 2011. Data were plotted on Excel-based templates used to create statistical process control charts as described by Benneyan et al.13 Lactation consultations were plotted by using a statistical process control chart (p-chart) and displayed as percent lactation consultations for eligible patients. Compliance with AAP guidelines was also plotted by using a p-chart that displayed monthly compliance scores as a percentage of the total points possible. Results of our progress were shared quarterly with the faculty in hospital pediatrics.
During our baseline period in 2010, a total of 99 infants were admitted to the pediatric ward for neonatal jaundice. Of these, 35 (35.4%) were admitted through the ED, and 87 (87.9%) were breastfed. In 2011, 86 infants were admitted to the pediatric ward for neonatal jaundice. Of these, 30 (34.9%) were admitted through the ED, and 72 (83.7%) were breastfed.
Monthly compliance scores with AAP guidelines are presented in Fig 1. Our baseline AAP compliance was 60.5%. We experienced a transition period that coincided with the working group formation and persisted through the first 2 months of our interventions. A second baseline was established from March to December 2012, revealing a mean AAP composite score of 90.4% and meeting our goal of >80% compliance with the AAP guidelines. Although there is some monthly variation in compliance scores related to different physician teams, the magnitude of variation decreased during the intervention period compared with the baseline period. Of note, we also reviewed the percentage of patients who received a “perfect” AAP compliance score. During 2010, only 15 patients (15.1%) received the total possible points, whereas in 2011, a total of 41 patients (47.7%) received the total possible points (P = .0001, Fisher’s exact test.)
The percentage of lactation consultations offered for eligible patients are displayed in Fig 2. Consultations increased from 48% during the baseline period to 63% during the transition and early intervention periods. During the last 5 months of our intervention period, a significant increase in lactation consultations was noted, with an average lactation consultation percentage of 90% (P = .004, Fisher’s exact test.) Although the sustainability of this increase has not been determined, this result exceeded our goal of >80% of breastfeeding mothers being offered a lactation consultation while in the hospital. In comparing rates of formula supplementation across the baseline and intervention periods, it was apparent that administration of formula versus expressed breast milk was not accurately documented in our electronic medical record. We reviewed physician documentation and found evidence of a plan to supplement with formula in breastfed infants in 56.3% of cases in 2010 (with another 27.6% of cases undetermined based on documentation.) This finding compares with only 47.2% of infants with a plan to supplement formula during the 2011 time period and only 1 patient in whom it was undetermined (P = .1868, Fisher’s exact test for documented supplementation). Maintenance of breastfeeding on discharge was similar across the 2 time periods, with 96.5% of mothers breastfeeding on discharge during the baseline period and 94.5% of mothers breastfeeding at discharge during the intervention period.
Length of stay did not significantly decrease across the baseline and intervention time periods. Mean length of stay was 30.9 hours during the baseline period and 28.3 hours during the intervention period (P = .2561, unpaired t test.)
This study highlights the effect of interprofessional collaboration on improving the care of jaundiced neonates admitted to the hospital, specifically regarding improved compliance with AAP guidelines to >90% and improved lactation consultations for breastfeeding mothers to >90%. The crux of our intervention was adoption of inpatient care guidelines by attending and resident physicians. This guideline was used to draft an admission order set and to facilitate education of nurses and residents on the care and management of jaundiced infants.
Our improvements in compliance with AAP guidelines correspond with development and implementation of our institution-specific guideline. Physician practice seemed to change even before our guideline was adopted, suggesting that forming the working group and generating discussions about ideal management of jaundiced infants are effective in changing provider behavior. The greatest positive effects on lactation consultations were associated with the creation of an admission order set that lists lactation consultation for breastfeeding mothers as a potential order.
Sustaining our gains in the care of jaundiced infants will require continued education of new faculty, nurses, and resident physicians. We continue to use our physician guideline to orient and educate residents and students at the beginning of their service on the general pediatrics ward. Subsequent steps may include monitoring usage of our admission order set to ensure residents are aware of this patient care tool. We recently transitioned to a newly built hospital with a dedicated floor for hospital pediatrics and a renewed hope for geographic localization of our patients. With this change, education efforts can be more targeted to the nurses and staff who care for this population. In addition, previous efforts have not focused on care provided in the ED or care provided by community physicians in the hospital setting. Opportunities exist to engage providers in the ED to reduce potentially unnecessary procedures such as IV placement and to expedite initiation of phototherapy while in the ED. Efforts to engage the faculty and staff in the ED are underway, and we are working with the hospital administration to secure additional phototherapy equipment to ensure availability of supplies for patients in the ED.
This study adds to the armamentarium for improving care of jaundiced infants. Before 2012, studies published with regard to quality improvement and jaundice focused on timely detection of jaundiced newborns in the nursery or immediately postdischarge.6–9 Wolff et al10 recently published a quality improvement report based in the ED and focused on the use of a nursing clinical pathway to reduce the time to phototherapy initiation and decrease ED length of stay. Our study is the first to focus on patients needing intensive phototherapy after discharge from the nursery. Through our study, we have demonstrated that compliance with guidelines can be improved while encouraging breastfeeding and providing additional support for breastfeeding mothers.
Despite improvements in AAP compliance scores and lactation consultations, this quality improvement project has several limitations. Length of stay did not significantly decrease from baseline. One reason for this finding may be that interventions to increase efficiency (ie, early identification of high-risk patients by using a standard evaluation or prompt initiation of intensive phototherapy) may have been counteracted by mothers waiting to see a lactation consultant. In addition, the mean length of stay for jaundiced infants is already very short (ie, 1.28 days during our baseline period). Decreasing the length of stay to a level that is clinically significant may be difficult to accomplish. Another limitation was the use of lactation consultations as a measure of breastfeeding support. Limitations in our electronic medical record did not allow us to accurately determine the rates of formula supplementation in breastfed infants.
This study reinforces existing evidence that merely publishing national guidelines may not be sufficient to change practice patterns.5 Providers must be engaged on a local level to understand the rationale for a guideline and must reinforce the use of guidelines when interacting with trainees. With the success of our program in hyperbilirubinemia, we plan to apply our successes of creating a physician algorithm and order set to other common conditions on the pediatric ward, including apparent life-threatening events and failure to thrive. This practice could be used by other institutions who are struggling to streamline processes and provide consistent care for common pediatric diagnoses. Additional studies should identify and focus on both process and outcome measures for improving care.
We thank Art Wheeler for his assistance with data analysis and Katelyn Krivchenia for her initial work creating the physician guideline. We thank the hospital pediatrics faculty who reviewed and implemented the guideline, as well as the care coordinators, Keely Reis and Rebecca Callentine, for their ongoing support of the project. Finally, we thank the inpatient nursing staff and pediatric residents for their engagement in improvement efforts.
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.
FUNDING: No external funding.
- American Academy of Pediatrics
- emergency department
- Kaplan M,
- Muraca M,
- Hammerman C,
- et al
- 4.↵AAP Subcommittee on Neonatal Hyperbilirubinemia. Neonatal jaundice and kernicterus. Pediatrics. 2001;108(3):763–765.
- Palmer RH,
- Clanton M,
- Ezhuthachan S,
- et al
- Mercier CE,
- Barry SE,
- Paul K,
- et al
- Chou SC,
- Palmer RH,
- Ezhuthachan S,
- et al
- Wolff M,
- Schinasi DA,
- Lavelle J,
- Boorstein N,
- Zorc JJ
- Langley G,
- Nolan K,
- Nolan T,
- Norman C,
- Provost L
- Benneyan JC,
- Lloyd RC,
- Plsek PE
- Copyright © 2013 by the American Academy of Pediatrics