Explanatory Diagnoses Following Hospitalization for a Brief Resolved Unexplained Event

This is a multicenter retrospective cohort study of infants <1 year of age with a BRUE who presented to 1 of 15 EDs participating in the BRUE Research and Quality Improvement Network between October 1, 2015, and September 30, 2018. The time frame for the study was chosen to coincide with the transition from the International Classification of Diseases, Ninth Revision, to the International Classification of Diseases, Tenth Revision, (ICD-10), which occurred in October 2015. The institutional review board of each participating hospital approved the study.

Hospital administrative data and medical record review were used to identify potential subjects. In 11 of the hospitals, the Pediatric Health Information System (Children's Hospital Association, Lenexa, KS) was used to identify infants for inclusion in the cohort. The 4 non–Pediatric Health Information System hospitals used an identical query of their institutional electronic medical record to identify subjects. Deidentified data were entered into Research Electronic Data Capture database.⁷ 

Potential subjects were hospitalized infants who met the AAP CPG definition of a BRUE.¹  A BRUE was defined as an episode of cyanosis or pallor, absent decreased or irregular breathing, marked change in tone, and/or altered responsiveness occurring in an infant >1 year of age that was brief, resolved, and not explained by an identifiable medical condition after an appropriate history and physical examination.

Potential subjects were identified by using the ICD-10 diagnosis codes. The data sources and search strategy are described in detail in a previous publication by the BRUE Research and Quality Improvement Network.⁸  The search strategy involved a weighted convenience sampling of subjects from the following 4 groups: patients with apparent life-threatening event or BRUE (R68.13) discharge code (group 1); patients with discharge codes for common BRUE symptoms, such as “altered consciousness” (group 2); patients with discharge codes for serious conditions, such as child abuse (group 3); and patients with less serious discharge codes, such as GER (group 4). The sample was weighted on the basis of perceived probability of a BRUE to maximize sensitivity and specificity. Groups 1 and 2 were oversampled, given their specificity for a BRUE. Groups 3 and 4 were included to reduce underclassification of rare causes and overclassification from the BRUE ICD-10 code use.

Manual chart review was used to confirm the AAP and study BRUE criteria, identify admission and discharge diagnoses, outpatient revisits, specialty visits, readmissions, and observed events in the ED or during hospitalization and determine results from any diagnostic tests and specialty consultations. Individual chart reviewers were trained on the data abstraction tool until an interrater reliability of 0.8 was achieved.

Investigators from each site manually reviewed medical records to determine the admission, discharge, and follow-up diagnoses and ascribe the level of certainty (definite, probable, possible, or unclear) associated with each diagnosis. A “definite” diagnosis was defined as a single diagnosis listed in the medical decision-making portion of the ED or, if applicable, inpatient documentation. A “probable” diagnosis was defined as being mentioned as the most likely cause of the event. A “possible” diagnosis was defined as 1 of many other causes listed in the differential diagnoses. A committee of investigators from each site independently reviewed each case with a “probable” or “definite” discharge diagnosis to determine if the diagnosis could plausibly explain the event. Site investigators provided additional chart details, when necessary, to adjudicate disagreements. Each infant with a “probable” or “definite” diagnosis that explained the event was classified as having an explanatory diagnosis; those without were classified as having a BRUE. Infants with an explanatory diagnosis at admission were excluded.

A committee of investigators identified whether each explanatory diagnosis established during or after hospitalization qualified as a serious underlying condition. A serious diagnosis was defined as a condition in which a delay in diagnosis or treatment could potentially increase morbidity or mortality (eg, child abuse).⁸  To identify explanatory diagnoses made after hospital discharge, all ED visits, clinic visits, and hospitalizations were manually reviewed until the child reached 1 year of age.

To establish which diagnostic tests and specialty consultations potentially contributed to an established diagnosis, we first grouped individual diagnoses into the following diagnostic categories: neurology, airway and/or gastrointestinal (GI), respiratory infections, apneas, nonspecific color or breathing changes, and nonspecific movement or changes in consciousness. Diagnoses that did not fit into one of these categories were classified as other. Relevant testing and specialty consultations for each diagnostic category were then determined through the group before reviewing patient charts (Table 1). The reason for grouping individual diagnoses into diagnostic categories was to be systematic when assigning relevant testing and specialty consultations to related diagnoses for which variations in nomenclature likely exists between individual clinicians. For example, choking, gagging, laryngospasm, and oral pharyngeal dysphagia are all diagnoses that may be used interchangeably by clinicians to explain a BRUE and therefore we classified them as airway or GI when considering relevant testing and specialty consultations. Because the diagnoses in the other category were unrelated to each other, relevant diagnostic testing and consultation for this group were determined individually for each diagnosis.

A diagnostic test was considered relevant to the diagnostic category if the test result could be used by a medical provider to rule in or rule out the diagnosis (Table 1). For example, an EEG was considered relevant for establishing diagnoses in the neurology category. A specialty consultation was considered relevant if the specialty service has expertise related to the diagnostic category or included diagnoses. For example, otolaryngology and pulmonary consults were considered relevant to the apnea diagnostic category. In determining relevance, we included testing and consultations that were supportive or contributory to the diagnosis but oftentimes not necessary to make the diagnosis. For example, an upper GI series is not necessary to make the diagnosis of GER but, if performed, can potentially contribute to the diagnosis. An infant was classified as having an observed event in the ED or hospital on the basis of documentation of the event in the electronic medical record.

For all infants admitted to the hospital, demographic characteristics were summarized by medians and interquartile ranges or frequencies and percentages. Given the lack of normality for continuous data elements, we used nonparametric tests and measures of central tendency. The proportion of infants hospitalized with a BRUE who received an explanatory diagnosis during hospitalization and after discharge was determined. Explanatory diagnoses were stratified by severity (serious versus less serious). For infants who received an explanatory diagnosis during hospitalization, the rate of relevant diagnostic testing, specialty consultation and observed events was determined. Relevant diagnostic testing, specialty consultations, and observed events were analyzed separately based on diagnosis severity.

Of the 2036 infants presenting to a participating ED for a BRUE, 1286 (63.2%) were hospitalized. A total of 980 (76.2%) hospitalized infants had no explanatory diagnosis at the time of admission, constituting our study cohort (Fig 1). Demographic characteristics for these infants are presented in Table 2. The median age was 36 days (interquartile range: 15–81), and 475 (48.5%) were male. There were 659 (67.2%) infants <60 days old, and 900 (91.8%) met the AAP higher-risk BRUE criteria.

An explanatory diagnosis was identified during hospitalization for 363 (37.0%) infants. Most explanatory diagnoses identified during hospitalization were classified as not serious (348 [95.9% of explanatory diagnoses; 35.5% of infants hospitalized with a BRUE]). GER and feeding difficulties (overfeeding, choking, gaging, and laryngospasm) composed over two-thirds of the diagnoses established during hospitalization. There were 15 patients who received a serious diagnosis (4.1% of explanatory diagnoses; 1.5% of infants hospitalized with a BRUE). The most frequent serious diagnoses made during hospitalization were seizures and infantile spasms, severe viral lower respiratory tract infections that required intensive care, and respiratory pressure or oxygen support and airway abnormalities, occurring in 4 (0.4%), 3 (0.3%), and 2 (0.2%) infants, respectively (Table 3).

Diagnostic testing, specialty consultations, and observed events contributed to an explanatory diagnosis in 69 (7.0%), 98 (10.0%), and 69 (7.0%) of patients hospitalized with a BRUE, respectively (Table 4). Diagnostic testing, specialty consultations, and observed events did not contribute to an explanatory diagnosis in 805 (82.1%) patients.

Among 617 infants discharged from a hospitalization with a BRUE without an explanatory diagnosis, 58 (9.4%) received a diagnosis before 1 year of age. A total of 13 (2.1%) of these infants received a serious diagnosis, accounting for 22.4% of all explanatory diagnoses made after discharge. The most frequent serious diagnoses made after discharge were seizures and airway abnormalities, occurring in 5 (0.8%) and 3 (0.5%) infants, respectively. There were 45 (7.3%) infants who received a less serious diagnosis, accounting for 77.6% of explanatory diagnoses made after hospital discharge. The most common less serious diagnosis was GER, occurring in 20 (3.2%) infants.

Findings from this large retrospective cohort study provide insight into the rate of explanatory diagnosis determined during hospitalization after a BRUE. Although approximately one-third of infants who are hospitalized with a BRUE received an explanatory diagnosis during hospitalization, most diagnoses were self-limited and not serious. A total of 15 (1.5%) infants hospitalized with a BRUE were diagnosed with a serious underlying condition during hospitalization.

The rate of identification of serious conditions presenting as a BRUE was lower in our study than in other investigations. In 1 single-center study, the authors reported 4 (4.1%) infants with a BRUE had a serious etiology, including bacteremia, seizure, enteroviral meningitis, and influenza.²  In another study, the authors found that 21 (6.4%) infants with a BRUE received a serious diagnosis.³  Serious diagnoses included sleep apnea, sepsis, hydrocephalus, metabolic disorder, seizure, subglottic stenosis, and sudden infant death syndrome. In a third study, the authors observed that 22 (44.9%) infants received an explanatory diagnosis, including 4 infants with seizures and 1 with a prolonged QT interval.⁴ 

Among infants in whom an explanatory diagnosis was established during hospitalization, >95% had benign self-limited conditions. GER and choking or gagging were the most common, composing two-thirds of all diagnoses established during hospitalization. Diagnostic testing and consultation infrequently contributed to the ultimate diagnosis. Our findings are consistent with that of previous studies. Brand et al,⁹  for example, found that only 5.9% of diagnostic tests helped to identify a diagnosis in infants hospitalized with an apparent life-threatening event. In a more recent single-center study, 0.6% of the laboratory tests, 0% of imaging, and 1.5% of ancillary studies contributed to the diagnosis of a serious underlying condition after a BRUE.² 

The findings from our study may be useful for pediatricians and ED clinicians when counseling families about the benefit of hospitalization to determine an explanatory diagnosis. Parents are often concerned that the event is life-threatening, and clinicians may offer hospitalization and monitoring to reassure the family, particularly when the explanation for an event is uncertain.^10,11  With our study, we confirm that serious explanatory diagnoses are rarely established during hospitalization. Conveying the low likelihood of diagnosing a serious underlying condition for a BRUE during hospitalization and the potential costs associated with hospitalization (eg, nosocomial infections, false alarms on monitoring, false-positive tests, and lost wages by caregivers) can help facilitate shared decision-making with parents about the need for hospitalization.

The fear of missing a serious underlying condition that could result in mortality may be a major factor for a clinician’s decision to hospitalize an infant presenting with a BRUE. Additionally, some families are fearful of leaving the hospital after an ED visit for a BRUE, believing that their child may have a life-threatening event or diagnosis in the future. However, in a meta-analysis, researchers found no increased risk of death after a BRUE, compared with the baseline risk of death in the first year of life.¹²  In another study in which researchers managed infants hospitalized with a BRUE for 5 years, excellent long-term prognosis was observed.¹³  Our findings add to the evidence that most patients with a BRUE do not receive a serious explanatory diagnosis, despite high rates of diagnostic testing and consultation to identify a cause for the event.

This study has several limitations common to retrospective studies that rely on medical record documentation. Factors may have existed that were not documented in determining the need for hospitalization, diagnoses, and diagnostic certainty. Subjectivity during individual chart reviewer interpretation of the data may have also introduced bias. Reviewer training and adjudication was used to address these potential sources of bias. Additionally, patients were only managed after hospital discharge up until 1 year of age. Follow-up encounters not occurring at the study institutions were also not captured. As a result, we likely underestimated the explanatory diagnoses among children discharged from hospitalization. Lastly, our study period started 7 months before the publication of the BRUE guidelines, which may have affected the decision to hospitalize.

Most infants hospitalized with a BRUE did not receive an explanation during the hospitalization for their episode, and a majority of those that did were benign or self-limited conditions. However, 15 (1.5%) were diagnosed with a serious underlying condition. More research is needed to identify which infants with a BRUE are the most likely to benefit from hospitalization by identifying an etiology for the event.

BRUE Research and Quality Improvement Network

We thank all members of the BRUE Research and Quality Improvement Network: Amy M. DeLaroche, MBBS, Division of Pediatric Emergency Medicine, Department of Pediatrics, Children’s Hospital of Michigan (Detroit, MI); Victoria Wilkins, MD, MPH, Division of Pediatric Hospital Medicine, University of Utah, Primary Children’s Hospital (Salt Lake City, UT); Emily Kane, MD, Children’s Hospital of Philadelphia, Perelman School of Medicine, University of Pennsylvania (Philadelphia, PA); Shobhit Jain, MD, Division of Emergency Medicine, Department of Pediatrics, Children’s Mercy Hospital (Kansas City, KS); Yiannis Kastogridakis, MD, Division of Emergency Medicine, Department of Pediatrics, Ann & Robert H. Lurie Children’s Hospital of Chicago, Feinberg School of Medicine, Northwestern University (Chicago, IL); Joyee G. Vachani, MD, Section of Hospital Medicine, Department of Pediatrics, Baylor College of Medicine/Texas Children’s Hospital (Houston, TX); Edward Kim, MD, Division of Pediatric Hospital Medicine, Department of Pediatrics, Riley Hospital for Children at Indiana University Health (Indianapolis, IN); Jessica Nicholson, MD, Carilion Children’s Hospital (Roanoke, VA); Kathleen Murphy, DO, Children’s Hospital of Philadelphia, Perelman School of Medicine, University of Pennsylvania (Philadelphia, PA); Miguel L. Knochel, MD, Division of Pediatric Hospital Medicine, University of Utah and Primary Children’s Hospital (Salt Lake City, UT); Roxanna Lefort, MD, MPH, Riley Hospital for Children, Indiana University Health (Indianapolis, IN); Ron L. Kaplan, MD, Division of Emergency Medicine, Seattle, Department of Pediatrics, Children's and the University of Washington School of Medicine (Seattle, WA); Adam Cohen, MD, Section of Hospital Medicine (Baylor College of Medicine and Texas Children’s Hospital, Houston, TX); Bruce E. Herman, MD, Division of Pediatric Emergency Medicine, School of Medicine, University of Utah (Salt Lake City, UT); Melanie K. Prusakowski, MD, Department of Emergency Medicine, Carilion Clinic (Roanoke, VA); Joshua L. Bonkowsky, MD, PhD, Division of Pediatric Neurology, School of Medicine, University of Utah (Salt Lake City, UT); Caleb Porter, MD, School of Medicine, University of Utah (Salt Lake City, UT); Kathryn Westphal, MD, Division of Hospital-Based Medicine, Ann & Robert H. Lurie Children’s Hospital of Chicago; Melissa Clemente, MD, Nicklaus Children’s Hospital (Miami, FL); Hannah C. Neubauer, MD, Section of Hospital Medicine, Department of Pediatrics, Baylor College of Medicine and Texas Children’s Hospital (Houston, TX); Lisa Uherick, Department of Emergency Medicine, Carilion Clinic (Roanoke, VA); Jennifer Y. Colgan, MD, Department of Pediatrics, Division of Emergency Medicine, Ann & Robert H. Lurie Children’s Hospital of Chicago, Feinberg School of Medicine, Northwestern University (Evanston, Il); Teena Hadvani, MD, Section of Hospital Medicine, Department of Pediatrics, Baylor College of Medicine and Texas Children’s Hospital, Houston, TX

AAP

American Academy of Pediatrics

BRUE

brief resolved unexplained event

CPG

clinical practice guideline

ED

emergency department

GER

gastroesophageal reflux

GI

gastrointestinal

ICD-10

International Classification of Diseases, Tenth Revision