OBJECTIVES: To describe the duration of illness in infants with first-time bronchiolitis who present to an emergency department (ED) and assess the burden of the illness on caregivers and families.
METHODS: This was a prospective cohort study of infants younger than 12 months who presented to a tertiary care children's hospital ED with a first episode of bronchiolitis. Subjects were excluded if they had a history of bronchodilator use or immunocompromise. Demographic and clinical data were collected in the ED. Outcomes data were collected by weekly telephone interviews for 4 weeks or until the subject was free of cough for 24 hours.
RESULTS: Ninety-five infants were enrolled from November 2007 to March 2008. Median duration of symptoms was 15 days; 25% of the infants remained symptomatic after 21 days. Subjects with a history of eczema trended toward a longer median duration of symptoms when compared with those who did not (18 days [interquartile range (IQR): 15.5–24] and 15 days [IQR: 11–19], respectively; P = .055). Duration of symptoms did not significantly vary with regards to respiratory syncytial virus status or secondhand smoke exposure. Subjects missed a median of 2.5 days (IQR: 0.5–5.5) of day care, and caregivers missed a median of 2 days (IQR: 1–4) of work. Of these infants, 37.1% (95% confidence interval: 24.3–44.1) had a subsequent unscheduled medical visit.
CONCLUSIONS: Infants seen in the ED for bronchiolitis have a prolonged disease course, with substantial burden to the family. Symptom duration may be influenced by a propensity toward atopy. Clinicians may use this information for counseling families.
WHAT'S KNOWN ON THIS SUBJECT:
Viral bronchiolitis is one of the most common illnesses of infancy. The majority of children evaluated in the ED for bronchiolitis are discharged from the hospital, where the burden of treatment falls on the caregivers.
WHAT THIS STUDY ADDS:
This report describes the duration of symptoms in children with bronchiolitis and provides information that clinicians can use to improve the accuracy of anticipatory guidance given to caregivers of children with bronchiolitis.
Viral respiratory tract infections are among the most common infections in children. Specifically in the infant population, the majority of these infections are caused by respiratory syncytial virus (RSV).1 In the United States, a preponderance of infants will remain outpatients for the duration of their illness; only an estimated 2% to 3% will require hospitalization.1,2 For infants evaluated in an emergency department (ED), caregivers rely on the anticipatory guidance provided before discharge as a guide for the expected course of illness in their child.
A search of the literature, however, has revealed that, although the disease course in hospitalized patients has been well described, less is known about the duration of symptoms of ambulatory patients.3,4 Pediatric and infectious disease texts vary in the stated disease course, from 3 to 7 days to 3 weeks.5,–,8 In the current literature there are only 2 previous studies that reported the length of symptoms in nonhospitalized infants with bronchiolitis.9,10 The generalizability of these studies is difficult, however, because the first study took place in an outpatient clinic in Cape Town, South Africa, and the second study was designed to evaluate viral etiology of bronchiolitis rather than illness duration.9,10
The variability of the natural history of bronchiolitis as stated in available resources can lead to inconsistent and inaccurate anticipatory guidance, which can lead to many unnecessary return visits to health care providers. The primary objective of this study was to describe the duration of illness in infants with first-time viral bronchiolitis who presented to an ED located in a major metropolitan area in the United States. Data were also collected on the impact of this illness on the daily life of both infants and caregivers; the number of missed days of work and/or missed days of day care by the caregiver and infant, respectively, and the number of unscheduled medical visits made during the illness were evaluated.
This was a prospective cohort study of infants younger than 12 months with first-time bronchiolitis (Fig 1). Potentially eligible infants were identified from a convenience sample of infants who presented to the Children's Hospital of Wisconsin ED from November 1, 2007, through March 31, 2008, with a first episode of acute bronchiolitis. The enrolling ED is within a large, tertiary care children's hospital and has an annual patient volume of ∼60 000 visits. The study was approved by the Children's Hospital of Wisconsin human research review board.
To be eligible for enrollment, infants had to be younger than 12 months and had to have a clinical diagnosis of bronchiolitis made by the treating ED physician. Infants were determined to have bronchiolitis if they had lower airway disease (tachypnea, crackles, wheezing, and/or retractions) and a history of previous upper airway infection (nasal congestion and/or rhinorrhea).11 Infants were excluded for any of the following reasons: any previous episode of wheezing; bronchiolitis or asthma (as diagnosed by a physician); previous bronchodilator use; treatment with corticosteroids in the 14 days before the current illness; immunosuppression; immunodeficiency; or a diagnosis of croup or pneumonia by the treating ED physician. Infants were also excluded if the primary caregiver did not speak English. Eligible infants were approached for enrollment by trained clinical research assistants who screened all patients when on duty (15 hours/day on weekdays and 10 hours/day on weekends and holidays). Patients who presented to the ED outside of this time frame were identified for comparison to the study cohort by a medical record search for infants younger than 12 months with an International Classification of Diseases, Ninth Revision diagnosis of bronchiolitis who were seen in the ED during the enrollment season but during nonenrolling hours. Aggregate data including age, gender, race, and disposition for the nonenrolled cohort was then abstracted from the records.
Baseline Data Collection
After informed consent was obtained from the parent or guardian, 3 categories of baseline data were collected: demographic, historical, and clinical. Demographic data including age, gender, race, and ethnicity were collected from the ED medical record and directly from caregivers before patient discharge. Historical data obtained included the presence of risk factors for a severe bronchiolitis course (prematurity, cardiac disease, and pulmonary disease), past history of atopy, and any family history of asthma or atopy (defined as having a first-degree relative with asthma, eczema, or food allergies). Clinical data collected in the ED included the number of days of symptoms before presentation to the ED and the severity of current symptoms. To determine onset of illness, caregivers were asked, “How many days ago did this illness start?” Severity of symptoms was determined by the treating physician using the wheezing and retraction scales from the Respiratory Distress Assessment Instrument (RDAI).12 Infants also had a nasal swab collected during the ED stay for RSV testing. Caregivers were given the option to refuse RSV testing but still remain in the study.
Follow-up Data Collection
Caregivers were contacted for evaluation of cough, unplanned medical treatment, and missed days of work and/or day care via weekly telephone interviews for 4 weeks after discharge from the ED or until the illness had resolved. ED “discharge” is used to indicate the end of the ED stay and applies both to infants discharged from the hospital and those admitted for inpatient management. Before discharge from the ED, the parent or guardian was given a symptom diary for daily recording of symptoms and outcomes. Each page of the diary consisted of the exact questions that were asked of the caregiver during the follow-up telephone interviews. Caregivers were asked to refer to this diary during the interview to help minimize recall bias. Infants were considered lost to follow-up if 3 consecutive attempts to contact the parent or guardian at appropriate times were unsuccessful over a period of at least 3 days for 2 consecutive weeks.
The primary outcome was time from onset of symptoms to resolution of illness. The illness was considered resolved when the subject was free of cough for a 24-hour period. Cough was chosen as the index of illness duration for several reasons. First, it has been used in previous literature to estimate bronchiolitis symptoms.9,13 Second, it is a symptom that is often worrisome to parents. Third, it was chosen on the basis of the ease with which it can be monitored at home. Secondary outcomes were determination of difference in duration of illness between RSV-positive and RSV-negative bronchiolitis, as well as estimation of the burden of illness. Measures of this burden included days of missed day care by the infant, number of days of alternate or emergency day care arranged for by the caregiver, days of missed work by the caregiver, and frequency of unplanned return medical visits that resulted from the illness.
Standard descriptive statistics were used for demographic, historical, and clinical data and for the primary and secondary outcomes. χ2 testing was used to compare demographic and disposition data between the enrolled and missed eligible cohorts. Additional subgroup analysis for the primary outcome was performed by using Wilcoxon rank-sum tests, because duration of illness was found to have a nonnormal distribution. Survival analysis was also performed to evaluate time to symptom resolution both in the entire cohort and in subgroups. By expressing duration of illness as a median rather than a mean, and by performing survival analysis, statistical analysis was performed on the entire cohort without the exclusion of censored data.
Minimum sample size was based on the primary study aim: the ability to estimate illness duration in the overall sample with a 95% confidence interval (CI) of <1 day. On the basis of data available from Swingler et al,9 we assumed an average duration of 12 days with an SD of 3. A sample size of at least 50 would yield a 95% CI of <1 day, and 60 would be sufficient with an SD of 4. The study was also powered to find a clinically important difference in duration of illness between RSV-positive and RSV-negative infants, assuming a 2:1 ratio of RSV-positive to RSV-negative infants. To detect a clinically relevant difference in duration of illness of 2 days between RSV groups with 90% power and an α value of .05, we estimated a total sample size of 108.
During the enrollment period, 112 infants were enrolled onto the study (Fig 2). After enrollment, 9 of these infants were diagnosed by the treating physician as having pneumonia and, therefore, were subsequently excluded from the study, which left 103 enrolled children. Eight infants (8%) were lost to follow-up during the study period, which left 95 infants who completed the study and were included in data analysis.
Subject characteristics are shown in Table 1. Just over one-third of the infants had passive exposure to tobacco smoke at home. Twelve infants (13%) had been given the diagnosis of eczema by a physician, and a family history of atopy was reported for the majority of them (72%). Of those tested, approximately three-quarters of the cohort were RSV-positive (74% [n = 85]). Overall, infants had mild-to-moderate disease, with a median RDAI score of 5 and approximately two-thirds of infants discharged from the ED. Rate of admission was not affected by history of atopy, history of prematurity, RSV status, or exposure to smoke in the home. Younger infants were admitted more frequently (60.9% of infants younger than 3 months admitted versus 26.3% of infants 3 to 6 months of age and 20.6% of infants 6 to 12 months of age; P = .004).
There was no difference found between the enrolled cohort and the missed eligible cohort with regards to age, gender, race, or disposition from the ED.
Median time to resolution of symptoms in the cohort was 15 days, and 25% of the infants continued to be symptomatic at day 20 of illness (Fig 3). At the end of the follow-up period, 11% of the infants continued to be symptomatic. Subgroup analysis (shown in Table 2) revealed a trend toward a prolonged duration of illness in those children with a history of eczema compared with those without a history of eczema (median: 18 days [interquartile range (IQR): 15.5–24] vs 15 days [IQR: 11–19], respectively; P = .0547). A similar trend was seen in those with family history of atopy versus those with no atopic family history (median: 16 days [IQR: 11.5–24] vs 14 days [IQR: 11–15], respectively; P = .0627). No difference in illness duration was found with regards to RSV status, history of prematurity, exposure to tobacco smoke, or disposition from the ED.
Table 3 shows the burden placed on families by the illness in the form of missed day care, missed work, and return medical visits. In our study population, 34% of the infants attended day care. Caregivers reported that the infants missed a median of 2.5 days (IQR: 0.5–5.5) of day care, and 75% of the children who attended day care missed at least 1 day because of the illness. Thirty-nine percent of the caregivers with children in day care reported having to make emergency alternate day care arrangements for at least 1 day. Of the 54% of caregivers who worked or attended school outside of the home, 78% reported missing at least 1 day of work or school because of their child's illness. The median number of days of work or school missed by caregivers was 2 days (IQR: 1–4).
After disposition from the ED, 59 (61%) of the infants were reevaluated by a physician. Of these infants, 36 (37%) had at least 1 subsequent medical visit that was not recommended in the discharge instructions. The majority (78%) of these visits were to the primary physician's office; only 22% presented to the ED for reevaluation.
Despite the ubiquity of viral bronchiolitis in the infant population, there remains no therapy that has proven to be consistently effective in ameliorating symptoms or shortening the duration of illness in children with bronchiolitis.14,–,17 Therefore, in the ED setting, after evaluation and stabilization, often the most important role of health care providers is to provide accurate anticipatory guidance to the caregivers before discharge.
To provide this guidance, however, providers must have accurate information in regards to the natural history of the disease in question. In this study, we found a median duration of illness of 15 days in infants seen in the ED, and almost 25% of the infants continued to be symptomatic 3 weeks after the onset of illness. To our knowledge, this is the first prospective cohort study designed specifically to evaluate the duration of illness in infants with bronchiolitis who present to an ED.
In the recent literature there are 2 previous studies that reported the length of symptoms in nonhospitalized infants with bronchiolitis.9,10 The first study, by Swingler et al,9 revealed a median duration of illness of 12 days and that ∼40% of the patients continued to be symptomatic by day 14 of illness. Although the authors did not evaluate the impact on caregiver work attendance or infant day care attendance, it did report that 34% of the infants were seen for unscheduled medical visits within 28 days of enrollment.9 The findings of the Swingler et al9 study may be difficult to generalize to the patient population seen in an urban ED in the United States, however. The study was conducted in an outpatient clinic in Cape Town, South Africa, which raises the question of demographic generalizability of the subject sample. Also, moderately to severely ill children were excluded from the study, which suggests that the study population examined was less ill than the typical population of infants with bronchiolitis seen in an ED. The second study, by Mansbach et al,10 reported a duration of illness of 3 to 8 days; however, this study was designed to evaluate viral etiology of bronchiolitis rather than duration of illness. It is unclear which symptoms the authors used to determine duration of illness or how often caregivers were contacted to assess progression of the illness.
Results of previous studies have demonstrated that early infection with bronchiolitis is associated with prolonged and recurrent wheezing in some children.18,19 There is also epidemiologic evidence to suggest a link between bronchiolitis in infancy and development of asthma later in life.20,–,22 On the basis of these previous studies, one could hypothesize that those children with a genetic predisposition toward asthma or atopic disease would have a more prolonged disease course. In our cohort, we observed a trend toward prolonged illness in infants with a diagnosis of eczema when compared with those without an atopic history. Similarly, those infants with first-degree relatives diagnosed with atopic disease tended to have longer illness. Although these trends were not statistically significant, our study was not powered to detect a difference between these groups. Therefore, although these infants may truly be at risk for a more protracted illness, the combination of small sample size in each subgroup and lack of power to detect a difference between these subgroups increased the likelihood of type II error in these analyses.
As suggested by the duration of illness reported for this cohort, bronchiolitis can cause a significant disruption in routine home life of infants with the disease. Results of previous studies have shown a delayed return to normal family routine and a detrimental effect on quality of life for infants and families after hospitalization for bronchiolitis.3,23 Our data are consistent with these findings; approximately three-fourths of working caregivers in our cohort reported missed days of work or day care because of the illness. When one also considers that almost two-thirds of the infants were taken for subsequent medical evaluation at least 1 time during the course of the illness, and that more than half of these visits were unplanned, the significance of the financial and temporal burden placed on families becomes obvious.
This study has several limitations. In general, when infants are selected by convenience sampling, the possibility exists that the cohort analyzed is not representative of the general population. Specifically, in the case of respiratory illnesses in children, failing to enroll patients during the overnight hours could result in a study sample that is less symptomatic than the general population. Although we do not have RDAI scores for those infants who presented to the ED during the overnight hours, we found no difference between this cohort and the enrolled infants with regards to the proportion of infants who required admission. Using admission as a proxy for symptom severity, the study cohort enrolled during the daytime and evening hours seemed to have had a similar severity of illness as those missed during the overnight hours. Our cohort was enrolled at a single site during a single bronchiolitis season. As such, the possibility exists that the population from which our sample was drawn is not generalizable to other populations. Given the yearly variability seen in viral strains, it is possible that the season during which the cohort was selected was an unusually mild or severe season. However, by examining admission rates of infants with bronchiolitis seen in our ED from 2002 to 2006 (32% ± 2%), we see that the admission rate of the current cohort (33%) is similar to that seen in previous bronchiolitis seasons. Finally, our assessment of the end points in the study relied on the accuracy of caregiver recall with regards to infant symptoms and burden of illness. To address this issue, before discharge from the ED the parent or guardian was given a symptom diary for daily recording of symptoms and outcomes. Each page of the diary consisted of the exact questions that were asked of the caregiver during the follow-up telephone interviews. Caregivers were asked to refer to this diary during the interview to help minimize recall bias.
We have shown that infants evaluated in the ED with acute bronchiolitis have a prolonged disease course, with substantial burden placed on the family during the illness. Clinicians may use this information for counseling families in an attempt to reduce unscheduled return visits. Our data also suggest that symptom duration may be influenced by an infant's propensity toward atopy. Future study is required to determine the true role of atopic tendency in the duration of illness in bronchiolitis.
- Accepted April 27, 2010.
- Address correspondence to Frank D. Petruzella, MD, MS, Medical College of Wisconsin, Children's Corporate Center, Suite 550, 999 N 92nd St, Milwaukee, WI 53226. E-mail:
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.
- RSV =
- respiratory syncytial virus •
- ED =
- emergency department •
- RDAI =
- Respiratory Distress Assessment Instrument •
- CI =
- confidence interval •
- IQR =
- interquartile range
- Kliegman RM,
- Behrman RE,
- Jenson HB
- Watts KD,
- Goodman DM
- Mandell GL,
- Bennet JE,
- Dolin R
- Hall CB,
- McBride JT
- Rudolph CD,
- Rudolph AM,
- Hostetter MK
- Brady MT
- Mcmillan JA,
- Feigin RD,
- DeAngelis C
- Schweigh PJ
- Bennett BL,
- Garofalo RP,
- Cron SG,
- et al
- Corneli HM,
- Zorc JJ,
- Majahan P,
- Bronchiolitis Study Group of the Pediatric Emergency Care Applied Research Network (PECARN),
- et al.
- Blom D,
- Ermers M,
- Bont L,
- van Aalderen WM,
- van Woensel JB
- Gadomski AM,
- Bhasale AL
- Hartling L,
- Wiebe N,
- Russell K,
- Patel H,
- Klassen TP
- Patel H,
- Platt R,
- Lozano JM,
- Wang EE,
- Wang EEL
- Copyright © 2010 by the American Academy of Pediatrics