In-hospital Neonatal Falls: An Unintended Consequence of Efforts to Improve Breastfeeding

This project was reviewed and exempted by the University of Maryland School of Medicine Institutional Review Board. Cases were identified through the hospital’s adverse event reporting system from January 2011 to February 2018. Medical records of identified cases were reviewed to determine injuries sustained and the presence of known risk factors for falls, including delivery type, time of fall event, time from birth to fall event, the administration of pain medication within 4 hours before fall, or other factors that might have contributed to the fall event. Any sustained injuries were recorded. Metrics related to breastfeeding were prospectively collected by chart audit or patient survey while ongoing efforts to improve breastfeeding occurred (September 2015–August 2017). Data included breastfeeding (infants ever breastfed), rooming-in (infants who spent <1 hour out of the mother’s room), skin-to-skin contact (infants who were placed skin-to-skin immediately and uninterrupted until after the first breastfeed or at least 1 hour if formula fed), breastfeeding education (mothers who were taught hand expression of breastmilk and cue-based feeding) and the percent of mothers who received assistance with breastfeeding among those who ever nursed. Monthly data were plotted on charts, annotating the relationship between metrics and the fall events. Each monthly data point represented a sample of 30 chart audits or patient surveys.

J.T. was born to a 24-year-old gravida 3, para 1011 mother who was unsuccessful in breastfeeding her first child. The infant was born at 1:45 am by repeat elective cesarean delivery at term. Breastfeeding was initiated at 4:18 pm on hospital day 1. On hospital day 2, the mother briefly discontinued breastfeeding because of fatigue and discomfort but resumed exclusive breastfeeding with lactation support. On hospital day 4, the mother notified her nurse that she had woken up to her infant crying on the floor, unswaddled. She described breastfeeding the infant and falling asleep. Neonatal examination was normal. Computed tomography (CT) of the head revealed a nondisplaced left parietal bone fracture extending to the anterior fontanelle. The infant was asymptomatic and discharged from the hospital that evening. At 7 weeks old, the infant presented to the emergency department with clinical seizures, confirmed by EEG. Magnetic resonance, CT imaging, and laboratory investigations revealed no signs of infectious, metabolic, or toxic etiology. The patient was discharged on anticonvulsant therapy for posttraumatic seizures associated with previous skull fracture. At 13 months old, the patient was seizure-free, developing normally and discontinued from anticonvulsant therapy.

A.W. was born to a 14-year-old gravida 1, para 0 mother. Maternal history is notable for being in foster care and depression associated with bullying. The infant was born at 11:10 pm by vaginal delivery at term. Breastfeeding was initiated on hospital day 2 at 3:25 am and lactation support was provided. Breastfeeding was discontinued at the mother’s request. On hospital day 4, the mother reported she had fallen asleep while burping the infant. She was startled by the infant crying on her lap and witnessed the infant fall to the floor. Neonatal examination was notable for a 5 cm crescent-shaped region of erythema on the right temple with 2 small fresh abrasions. Head ultrasound revealed a right frontal soft tissue hematoma. A CT scan confirmed the hematoma overlying the right parietal bone, but no underlying fracture or intracranial hemorrhage was demonstrated. The infant remained asymptomatic and was discharged from the hospital that evening.

P.R. was born to a 26-year-old gravida 3, para 1011 mother. Hospital presentation was notable for ongoing depression with suicidal ideation, triggered by domestic abuse and enslavement. The infant was born at 10:02 am by cesarean delivery at term. At 7:30 pm, breastfeeding was initiated. On hospital day 2, the mother reported feeling drowsy from her pain medication. Her spouse and a toddler not yet weaned from breastfeeding were present. The infant was brought to the nursery at the mother’s request and lactation support continued. On hospital day 3, the mother reported that she was breastfeeding the infant and fell asleep. The infant fell onto the floor, landing on a pillow. Neonatal examination was unremarkable. A CT scan of the head revealed no acute pathology. The infant was asymptomatic and was discharged on hospital day 5.

Table 1 summarizes the fall events. Figures 1–5 illustrate the timing of each neonatal fall event, represented by arrows, and the relationship to improving processes and outcome measures for breastfeeding as they relate to Steps 4, 5, 7, and 8 of the 10 Steps.

In-hospital neonatal falls are increasingly recognized as a safety risk during postpartum hospitalizations, with as many as 160 to 1600 newborn falls occurring annually.^3,4,7 Given the number of neonates at risk and the limited availability of data on prevention and management, additional reporting of these falls is deserved. We describe an increase in falls at a single birthing center, commensurate with systematic changes in policy and care delivery intended to facilitate breastfeeding. Although breastfeeding itself likely did not play a direct role in the falls, an emphasis on rooming-in and on-demand feeding may have contributed to maternal fatigue; all of these falls were associated with mothers falling asleep during newborn feeding. The additional emphasis on lactation education, assistance with breastfeeding, and early skin-to-skin support may have further reduced opportunities for maternal rest during the relatively short postpartum hospitalization.

Falls that occur when caregivers fall asleep is a consistent theme in reported series. In a review of patient safety data over a 9-year period, Wallace reported that 150 of 272 neonatal falls occurred after a caregiver fell asleep while holding the newborn.⁷ Several smaller series report that 36% to 66% of in-hospital neonatal falls involved caregivers falling asleep.^1,3,4,12 Not surprisingly, these falls transpired most commonly during night or early morning hours. Similar to our series, Galuska¹³ described 5 cases at a single center in which all neonates fell from their mothers’ arms in the early morning hours, continuing to implicate fatigue as a factor.

Our case series is the first in which an increase in neonatal falls coincided with improved metrics related to the 10 Steps. Of all the Step targets, rooming-in efforts resulted in our biggest practice change because of the number of clinical workflows and providers affected. Successful rooming-in requires extensive collaboration among providers across different specialties to minimize disruptions to maternal sleep. To further support a restful environment, coordination with services such as dietary and environmental services is necessary but may be difficult to achieve. Notably, 2 of our patients (cases 2 and 3) suffered from stressful social situations that may have resulted in more frequent interruptions to sleep by support staff. In case 3, social work visits were required to support the mother while avoiding the presence of a potentially abusive partner. In such situations, the timing of evaluations and treatment may be suboptimal. Thoughtful consideration should also be given to visitation hours during the postpartum period, balancing family centeredness with a restful experience. In case 3, the presence of a toddler in the mother’s room likely facilitated family centeredness by encouraging the ongoing breastfeeding of the older child but possibly contributed to maternal fatigue. Given the aforementioned challenges, implementation of rooming-in practices may pose an increased fall risk to newborns. Therefore, we recommend that neonatal falls be considered when working toward improved breastfeeding practices, particularly when rooming-in is a part of the approach. In addition, neonatal fall risk should be emphasized for every patient regardless of feeding type (breast or bottle) because multiple risks of maternal fatigue exist in the postpartum period.

In response to our neonatal falls, we employed a multipronged approach involving patient education, staff education, and changes to our care delivery. To educate parents on the risk of falls, we placed a safe sleep card in each bassinet, modified an existing safety contract, provided more family assistance during nursing, and provided companion cards that suggested how a father or other support person can assist mothers, especially at night. Staff were provided education on falls with a focus on using clinical judgement to weigh the benefits of rooming-in for each patient on the basis of perceived fatigue and the presence of a support person. Finally, the environment of care delivery was altered by instituting quiet hours to allow for napping, discouraging visitors after 9 pm, and reinforcing the need for frequent nurse rounding on night shift. Similar interventions have been described to assist hospitals with reducing neonatal falls while newborns room in during the postpartum period.^6,–8,13,14 Research on the efficacy of these interventions is needed to inform hospitals that promote rooming-in.

The majority of in-hospital neonatal falls are associated with caregiver fatigue. Common efforts to promote breastfeeding, such as rooming-in practices, may contribute to fatigue and increase the risk of falls during hospitalization. Surveillance of neonatal falls should be incorporated into projects aimed at improving breastfeeding outcomes.

CT

computed tomography