Neonatal Use of Acute Care Services During the COVID-19 Pandemic

We conducted a retrospective cohort study of neonatal visits to acute care settings during the COVID-19 pandemic and a comparator period. This study was performed at a tertiary children’s hospital system that provides acute medical care through 2 EDs, 7 UCs, the main hospital’s inpatient units and neonatal, pediatric, and cardiac intensive care units (ICUs). Newborn deliveries do not occur at the facility.

We included infants born during the pandemic (March 13, 2020–October 31, 2020) and the same timeframe within the preceding year. A neonatal acute care visit was defined as any visit or admission to 1 of the acute care locations within the first 30 days of life. Newborns transferred between acute care settings during the same encounter with the health care system (ie, transferred from the ED to inpatient unit) had a visit count for each setting in which they received care. Newborns presenting for acute care multiple times within the first 30 days of life were also included. Visits were excluded if newborns were transferred from a referring nursery, inpatient unit, or ICU, because we could not confirm if they were discharged from the birth hospital.

Data collected from the electronic medical record for each unique visit included demographics, newborn characteristics, and visit characteristics. Demographic variables included age, sex, race, primary language, and insurance status. Newborn characteristics included gestational age and birth weight. Characteristics from UC and ED visits included the primary discharge diagnosis and disposition. Characteristics from inpatient and intensive care admissions included primary discharge diagnosis. The authors reviewed all primary discharge diagnoses in the dataset and collapsed them into 19 clinically synonymous categories by consensus (Supplemental Table 4). The remaining diagnoses were classified as ‘Other.’ The diagnoses were then arranged by acute care setting and similar settings were grouped together (UC and ED, inpatient units, and ICUs). For a more meaningful analysis, the 10 most common discharge diagnosis categories in the prepandemic cohort were maintained, and the remaining 9 categories were combined with the ‘Other’ uncategorized diagnoses.

Visit characteristics were analyzed at the visit level and reported for each acute care setting, whereas demographics and newborn characteristics were analyzed at the patient encounter level and reported across all acute care settings. Descriptive statistics were summarized and presented in tables as frequencies and percentages. Absolute change and percent changes were calculated and presented in tables. χ² tests determined statistically significant differences in proportions for categorical variables by prepandemic and COVID-19 pandemic cohorts. P values <.05 were considered statistically significant. All analyses were conducted by using SAS version 9.4 (SAS Institute Inc., Cary, NC, USA). Institutional review board approval was obtained for this study.

A total of 4439 neonatal acute care visits were included in the study, of which 2677 occurred in the prepandemic period and 1762 in the COVID-19 pandemic period, representing a 34.2% decline. Demographic and patient characteristics were similar between periods with the exception of the distribution of patient age (P < .001; Table 1), which included a larger proportion of neonates ≤7 days presenting during the pandemic (27.6% vs 33.6%). The decline in newborn visits in individual care settings during the pandemic ranged from 9.1% to 43.5% (Table 2). The distribution of neonatal visits among acute care settings did not change significantly between the prepandemic and COVID-19 pandemic periods (P = .26). Similarly, the proportion of newborns discharged from the hospital or transferred to a higher level of care from an UC or ED did not change significantly.

Consistent with the overall decline in visits, counts of diagnoses decreased in most categories during the COVID-19 pandemic. The largest reduction observed in the UC and ED setting was in visits for infectious conditions (49%), and the proportion of these visits declined significantly (P = .02; Table 3). The largest clinically significant reductions in the inpatient units were for infectious and respiratory diagnoses (48% and 52%, respectively), and the proportions of these hospitalizations significantly declined as well (P = .009 and 0.03, respectively). Despite a small reduction in hospitalizations for jaundice (3%), the proportion of jaundice hospitalizations significantly increased by 5.7% (P = .02). Conversely, cardiac diagnoses increased across settings during the pandemic, and the proportions of these visits also significantly increased. Diagnoses from ICUs were not reported as volumes in these settings were low (comprising <4% of all visits), making clinically or statistically meaningful comparisons difficult.

In our study of approximately 4500 neonatal visits, we observed a significant reduction in acute care service use across all settings at our pediatric hospital system during the early COVID-19 pandemic, consistent with our hypothesis. The reduction is consistent with previous reports of substantial declines in pediatric visits to EDs and hospitalizations during the pandemic. DeLaroche et al reported a 46% decline in ED visits to 27 US Children’s hospitals during the pandemic.¹¹  Similarly, Markham et al³  reported a 35% reduction in admissions across 45 pediatric hospitals during the pandemic. Although our results align with previous studies, our study is specific to the neonatal population and details the decline in health care use across multiple acute care settings within 1 hospital system, a novel approach to studying pediatric health care use during the pandemic.

Although we did not measure newborn outcomes directly, we found no significant change in the disposition of neonates presenting to UC or ED or the proportion of visits to an ICU to suggest neonatal illness severity was drastically impacted on a population level. The decline in neonatal infectious and respiratory diagnoses seen in our study is similar to other reported trends. Markham et al³  identified decreases in pediatric hospitalizations for asthma, bronchiolitis, pneumonia, and upper respiratory tract infections. Kadambari et al¹²  reported a 54% reduction in infection-related admissions extending beyond respiratory conditions, including bacterial osteoarticular processes and cellulitis. Linking the decrease in transmissible disease with social distancing recommendations and stay-at-home orders has been frequently cited.^3,4,11,12  Our study reinforces these findings but is specific to the neonatal population.

Jaundice is one of the most common reasons that newborns require medical attention and readmission.^13–15  During the pandemic there was a substantial reduction in visits to UCs and EDs within our hospital system for jaundice. Providers may have hesitated to refer newborns to acute care who were near, but below, phototherapy treatment level out of concern for exposure to SARS-CoV-2 in health care settings or used home phototherapy more often. There was also a small decrease in hospitalizations for jaundice; however, after accounting for local and nationally reported declines in birth rates from 2019 to 2020,^16,17  the rate of hospitalizations for jaundice did not change. Although the proportion of hospitalizations for jaundice increased during the pandemic, this is likely a result of more notable declines in other diagnosis categories. These results may alleviate concerns that early postnatal discharge,⁹  decreased ambulatory lactation services,^18,19  and the decline in primary care visits during the pandemic²⁰  adversely affect rates of jaundice-related hospitalizations.

Despite overall low volumes of cardiac diagnoses, the increase in cardiac visits to UCs, EDs, and inpatient units highlights the need for future investigation in this area to determine the clinical significance and implications. It was reassuring that the number of visits to the cardiac ICU did not increase, but these findings should be further explored by multiinstitutional data to better understand the nature of neonatal cardiac conditions presenting during the pandemic given our small visit numbers.

This study has several limitations. First, our data came from a single tertiary care pediatric center which may limit generalizability. Second, regarding primary discharge diagnoses, it is possible diagnoses were not prioritized by the provider, resulting in secondary or tertiary diagnoses aligning more accurately with the primary concern. It was reassuring that the top 5 discharge diagnosis categories far outnumbered any of the others, suggesting misclassification bias to be small. The data also does not capture information pertaining to primary care visits, limiting interpretation of its impact on the use of acute care services. It is possible that shifts to less traditional types of services (ie telehealth) contributed to the decline in neonatal visits to acute care settings. Lastly, this retrospective study lends itself to other biases inherent to such study designs, ie, selection and information bias, thereby limiting interpretation of findings to correlation with the pandemic and not causation.

In conclusion, the impact of the COVID-19 pandemic on neonatal use of acute health care services mirrored declines seen in the pediatric population at large. This study further adds to the literature by linking trends across a spectrum of acute care settings and revealing substantial reductions in neonatal hospitalizations related to infectious and respiratory diagnoses. These results reveal opportunities to mitigate potentially preventable newborn visits to acute care, lending support to the development of telehealth services, the availability of home phototherapy, and infection control measures during surges of the pandemic. Importantly, further research is needed to determine the downstream effects these visit declines have on patient outcomes. Other areas for future analysis include postnatal nursery length of stay and the use of ambulatory services to understand their role in affecting neonatal use of acute care services during the pandemic.