OBJECTIVE. Hypoxia is a common reason for hospital admission in infants and children with acute bronchiolitis. No study has evaluated discharge from the emergency department (ED) on home oxygen. This study evaluated the feasibility and safety of ED discharge on home oxygen in the treatment of acute bronchiolitis.
METHODS. This was a prospective, randomized trial of infants and children with acute bronchiolitis and hypoxia (room-air saturations of ≤87%) aged 2 to 24 months presenting to an urban, academic, tertiary care children's hospital ED from December 1998 to April 2001. Subjects received inpatient admission or home oxygen after an 8-hour observation period in the ED. We measured the failure to meet discharge criteria during the observation period, return for hospital admission, and incidence of serious complications.
RESULTS. Ninety-two patients were enrolled. Fifty three (58%) were randomly assigned to home and 39 (42%) to inpatient admission. There were no differences between the groups in age, initial room-air saturation, and respiratory distress severity score. Of 53 patients, 37 (70%) randomly assigned to home oxygen completed the observation period and were discharged from the hospital. The remaining 16 patients were excluded from the study (6), resolved their oxygen requirement (5), or failed to meet the discharge criteria and were admitted (5). One discharged patient (2.7%) returned to the hospital and was admitted for a cyanotic spell at home after the 24-hour follow-up appointment. The patient had an uncomplicated hospital course with a length of stay of 45 hours. The remaining 36 patients (97%) were treated successfully as outpatients with home oxygen. Satisfaction with home oxygen was high from the caregiver and the primary care provider.
CONCLUSIONS. Discharge from the ED on home oxygen after a period of observation is an option for patients with acute bronchiolitis. Secondary to the low incidence of complications, the safety of this practice will require a larger study.
bronchiolitis is the most common cause of hospital admission in children <1 year of age. An estimated 80000 to 120000 children <1 year of age are admitted annually. Hospitalization rates have been increasing dramatically, with estimates of an increase of 239% in children <6 months from 1980 to 1996.1 One of the hypotheses for this dramatic increase is the almost universal use of pulse-oximetry screening and the implementation of an oxygen-saturation cut off that warrants hospital admission for supplemental oxygen. The need for supplemental oxygen therapy in a previously healthy patient is considered by many to require mandatory admission and is included as one of the therapies defining mandatory admission in studies evaluating a pediatric risk of admission score.2, 3 It has also been demonstrated that physicians use the oxygen-saturation information frequently to determine whether a patient requires hospital admission.4 A recent study addressed how often patients remain in the hospital only for supplemental oxygen after other clinical parameters (oral intake and work of breathing) have improved and found that 26% of patients have a prolonged stay of an average of 1.6 days.5
In the years preceding this study, community primary care providers (PCPs) were faced with a rapidly expanding population of children and a large burden of patients with bronchiolitis requiring oxygen. Hospital beds around the community were becoming very difficult to find, and the practice of arranging home oxygen from the office setting began to emerge. This was especially true in practices that had close relationships with their patients and felt comfortable that if any deterioration occurred, the caregivers would quickly notify them and present to a local emergency department (ED). To date, there are no publications evaluating the use of home oxygen therapy from the ED or the primary care clinic in patients with acute bronchiolitis requiring supplemental oxygen. We undertook a prospective, randomized pilot study of home oxygen from the ED compared with traditional inpatient hospitalization to assess for the feasibility and safety of this practice.
We conducted a prospective, randomized trial of traditional inpatient hospitalization versus home oxygen from the ED in a convenience sample of patients with acute bronchiolitis and hypoxia. In preparation for conducting the study, we sent out information letters to all PCPs who refer patients to our ED to let them know what the study would require of them if a patient in their practice was enrolled.
A sample-size calculation was performed before the start of the study. The success of the home oxygen protocol was defined as (1) not requiring hospitalization after ED discharge and (2) not having any serious complications. Success with hospitalized patients was defined as (1) not requiring transfer to the intensive care unit (ICU), (2) not being readmitted to the hospital after discharge, and (3) not having any serious complications. Serious complications were defined as respiratory arrest, apnea, cyanotic spell, and respiratory failure. The success rate for hospitalized patients was estimated at 99%. Using a minimum success rate of 90% for home oxygen, we estimated that ∼160 patients (80 in each group) would be needed with an α value of .05 and a β value of .20.
Patients were enrolled over 3 consecutive winter seasons from December 1998 to April 2001 at an urban, academic, tertiary care children's hospital at an elevation of 5280 ft (1609 m). All of the patients received deep nasal suctioning and 2 albuterol treatments (2.5 mg) via nebulizer over the first 60 minutes. All of the patients had their respiratory distress severity score (RDSS) (Table 1) documented at the start of the study, after suctioning, and after the nebulizer treatments. The RDSS has 4 categories: respiratory rate, wheezing, aeration, and retractions, and each category has a scoring system of 0, 1, or 2 points. The total score is obtained by adding the score from each category. All of the patients then received a chest radiograph, and, if read by the attending radiologist as consistent with an acute bacterial pneumonia, the patient was not approached for enrollment in the study.
Caregivers were approached during times at which a research assistant was available to assess whether they met the inclusion criteria or had any of the exclusion criteria (Table 2). Informed consent was obtained, and patients were randomly assigned to traditional inpatient hospitalization or to home oxygen therapy. Randomization was performed by using a block-randomization table created before the start of the study. Research assistants were blinded to allocation at the time of enrollment. The study packet with the assignment could not be opened until informed consent had been obtained. Patients randomly assigned to the inpatient arm of the study had an RDSS recorded every 2 hours for the first 8 hours of hospitalization. Other treatments in the hospital were at the discretion of the inpatient attending, and the families were contacted at 1 week by telephone and were asked to answer questions from a standardized questionnaire. A chart review was also performed to extract hospital complications and length of stay data.
Patients randomly assigned to the home oxygen arm of the study were placed in the ED observation unit for an 8-hour observation period. Standard observation consisted of continuous pulse oximetry and vital-sign checks every 2 hours. An RDDS was assigned every 2 hours. Four separate pulse-oximetry measurements were recorded while on oxygen (asleep and feeding were mandatory). At the end of the observation period, the following defined discharge criteria had to be met: saturations of ≥90% on ≤1 L/min nasal cannula oxygen while the patient was awake, asleep, and feeding; the patient must be able to maintain hydration; the patient must have no signs of deteriorating respiratory status; the attending physician and caregiver must be comfortable with discharge; and 24-hour follow-up must be arranged. Patients were taught the operation of the portable home oxygen unit and were given the unit before discharge. Patients were discharged on ≤1 L nasal cannula oxygen.
Standardized questionnaires were completed by the PCP at the 24- and 48-hour visit and were faxed back to the study coordinator. At 72 hours, a telephone follow-up questionnaire was administered to both the PCP and the caregiver. A 1-week telephone follow-up questionnaire was also administered to the caregiver. Fig 1 displays the treatment protocol.
Primary outcome measures included failure to meet discharge criteria during the observation period, return for hospital admission after successful discharge, and incidence of serious complications. Secondary outcome measures were caregiver satisfaction, caregiver preference, PCP satisfaction, and PCP preference.
Data were analyzed by using SPSS 12.0 (SPSS Inc, Chicago, IL). Differences between groups were analyzed by the Student's t test for continuous variables and the χ2 test for categorical variables. This study was approved by the Colorado Multiple Institutional Review Board and the Kaiser Permanente Institutional Review Board.
Ninety-two patients were enrolled over 3 consecutive bronchiolitis seasons. A modified consolidated standards of reporting trials diagram is presented in Fig 2. These patients had a mean age of 7.8 months, a mean room-air saturation of 85%, a mean RDSS of 4.23, and a mean oxygen requirement of 0.485 L/min.
Fifty-three patients (58%) were randomly assigned to home, and 39 (42%) were randomly assigned to inpatient admission. There were no differences between the groups in age, initial room-air saturation, and initial RDSS. There was a statistically significant difference in the amount of oxygen required at 8 hours for the patients who had been admitted versus those who were discharged (Table 3). These relationships remained the same when patients completing the study were analyzed (Table 4). Based on the actual number of patients enrolled and a success rate of 99% in the hospitalized patients, we would detect an 80% success rate in the home oxygen patients as a statistically significant difference.
Of the 53 patients, 37 (70%) randomly assigned to home oxygen completed the observation period and were discharged from the hospital. Six patients (11.3%) failed to complete the observation period: 2 patients were diagnosed with pneumonia from their chest radiograph that was obtained after enrollment instead of before randomization, 3 withdrew secondary to caregiver anxiety, and 1 was excluded secondary to transfer from another facility, which did not allow for the ascertainment of initial data.
Ten patients completed the observation period but were unable to complete the remainder of the study: 5 resolved their oxygen requirement during the observation period and were discharged with no supplemental oxygen. These 5 patients did not return to the study institution in the following week for any additional care. The remaining 5 patients did not meet discharge criteria and were subsequently admitted to the hospital for further care: 2 patients had concerns from the nursing staff and attending physician for inadequate oral intake, 2 patients had concerns for increased work of breathing, and 1 patient was given a change in diagnosis to reactive airways disease and was given oral corticosteroids. Thirty-seven patients were successfully sent home with home oxygen.
One patient (2.7%; 95% confidence interval: 0.6–13.8%) of the 37 discharged from the ED returned to the hospital and was admitted for a cyanotic spell at home after the 24-hour follow-up appointment. This was a 4-month-old who was sent to the ED from the 24-hour follow-up visit with a history of a possible cyanotic spell at home. The patient had an uncomplicated hospital course with a length of stay of 45 hours. This is the only patient who had a serious complication in the study. The remaining 36 patients (97%; 95% confidence interval: 86–99%) were treated successfully as outpatients with home oxygen therapy.
Of the 37 patients, 33 (89%) had a 24- or 48-hour survey completed by their PCP; 33 (89%) had a completed 72-hour telephone follow-up; and 35 (95%) had a completed 1-week telephone follow-up. All 37 (100%) of the patients had at least 1 of the 3 follow-up surveys completed (Table 5).
At the 24/48-hour follow-up, 97% of the caregivers were satisfied with their child at home, and 94% of the PCPs were satisfied with the patient being at home. At the 72-hour telephone follow-up we asked about preference, and 79% of the caregivers stated that they preferred to be at home, and 15% of the caregivers would have preferred to be in the hospital. Sixty-four percent of the PCPs stated they preferred their patients to be at home, whereas 27% of them would have preferred the hospital. At the 72-hour survey, 100% of the caretakers felt they had received adequate instruction on the use of home oxygen. Ninety-seven percent (34 of 35) of the patients at the 7-day follow-up were satisfied with home oxygen for their child. Four patients (11%) were still on oxygen at a week follow-up (Table 5).
Thirty-nine patients (42%) were randomly assigned to traditional inpatient hospitalization. Six patients were excluded from the study: 2 patients were diagnosed with pneumonia from their chest radiograph that was obtained after enrollment instead of before randomization, 1 patient resolved the oxygen requirement before reaching the inpatient bed, 1 family became uncomfortable with being in a research study, and 2 patients had incomplete medical charts for review. The patient discharged who no longer required supplemental oxygen did not return in the following week for additional care to the study institution. The remaining 33 patients were available for analysis.
The hospitalized patients had a mean length of stay of 1.8 days, with a range of 0.58 to 6.3 days. Five of the patients (15.2%) had a hospital stay of >3 days (Table 6). Of these 5 patients, 2 were diagnosed with pneumonia on subsequent chest radiograph, and 1 was diagnosed with pneumonia and required intravenous fluids. One patient was a former 29-week infant who had no complications noted, and the remaining patient had no complications noted. Twenty six of the patients (79%) were discharged from the inpatient medical service on home oxygen. There were no patients who required transfer to the ICU, and there were no readmissions within 1 week after discharge. All 33 of the patient's caregivers were successfully contacted at the 1-week telephone follow-up; 36% of the caregivers would have preferred home care, 49% preferred hospital care, and 15% had no opinion (Table 7).
Analysis of self-reported data on missed days from work showed a trend in less missed days from work in the home oxygen group versus the hospitalized group (mean: 1.72 vs 2.69, respectively). This finding was not statistically significant.
This study describes the discharge on home oxygen from the ED after an 8-hour period of observation of infants with uncomplicated bronchiolitis. This practice is becoming more common in many sectors of the medical community and in PCP offices without benefit of measurement of return rates, subsequent ED visit for worsening disease, or hospital admission. The reasons for this change in clinical practice are not known but could reflect a perceived overcrowding of EDs, the belief that bronchiolitis is a mild disease in most infants, that the need for oxygen is often the only reason for hospital admission, ease of care of patients on home oxygen, and the fact that home oxygen has become easily accessible from home health care companies in recent years.
The use of home oxygen therapy and subsequent early nursery discharge has been evaluated in the management of premature infants with bronchopulmonary dysplasia and chronic lung disease and has been shown to be safe and effective, with an associated decrease in cost and no increase in morbidity or readmission rates.6, 7 Families, when surveyed, responded 94% of the time that they would again take an infant home on oxygen, and this practice is now routine.8, 9 This approach has not yet become the standard of care for older infants and children with acute respiratory tract illness. In 2001, Wilson et al10 evaluated care for patients <1 year of age admitted with bronchiolitis in 10 children's medical centers and found that care varied widely, that variations could not be explained by difference in disease severity, and that there was no evidence that a greater intensity of care affected morbidity or mortality. The use of home oxygen was not measured. In a letter to the editor of Pediatrics concerning this publication, Weiss and Annamalai11 queried the issue of standardized discharge criteria from the hospital for patients with bronchiolitis. They report on a survey of chief residents at 30 children's hospitals, and, of the 17 responses, only 2 hospitals reported that they routinely send patients home with supplemental oxygen for bronchiolitis. Other authors have indicated that the management of bronchiolitis varies widely in both the ED setting, as well as the inpatient setting, and is not based on clear evidence of efficacy of treatment options.12–14
The implementation of clinical care pathways have been shown to decrease ancillary testing and overall cost and length of stay; however, use of home oxygen after ED evaluation and observation have not been part of published guidelines to date.15–18 Hospitalization rates for bronchiolitis have increased 239% over the past 2 decades for children <6 months of age.1 During that time period, mortality rates remained relatively constant.19 This increase in admission frequency has been postulated to relate to perhaps an increased reliance on pulse-oximetry measurements.4 In a retrospective analysis of 62 patients admitted for bronchiolitis, Schroeder et al5 showed that hospitalization was prolonged by an average of 1.6 days in 16 patients (26%) because of oxygenation concerns based on pulse-oximetry readings after all of the other discharge criteria were met. These prolonged hospitalizations are not without potential additional costs and the risk of adverse events.20
To enroll patients with uncomplicated cases of bronchiolitis, we chose to obtain a chest radiograph on all of the patients to exclude possible bacterial pneumonia, foreign bodies, pulmonary edema, and other cardiac or respiratory tract abnormalities. Standard care at the time of our study included deep nasal suction and a trial of beta agonists. We realize that these aspects of bronchiolitis care, with the exception of nasal suctioning, may no longer be necessary. We used a respiratory distress score modeled from a previously published respiratory distress score to have an objective measure of respiratory distress to allow for comparison between the 2 groups at enrollment and discharge home.21 However, we did not incorporate this score as either inclusion or exclusion criteria or as discharge criteria. The study investigator team felt that incorporating an absolute cut-off score for discharge home would be difficult and that the clinical measures listed in the discharge criteria would be more clinically relevant.
Our methodology also included a mandatory 8-hour observation period to allow for varying stages of disease at the time of enrollment. This observation period proved to be of significant benefit and affected care in 10 (21%) of 47 patients. Five patients were found to no longer require supplemental oxygen and were discharged from the hospital. These patients were an unexpected finding. We removed the patients from the study at that time, because it did not seem reasonable to require these families to adhere to the stringent follow-up requirements. They did receive standard bronchiolitis discharge instructions to return immediately for any concern of increased work of breathing, cyanosis, poor feeding, or any other concerns. Because these patients were observed for an extended period of time and were found to be hypoxic at the start of their evaluation, we performed chart review on these patients, and none of them returned to the study institution in the following week for additional care. The remaining 5 patients were subsequently admitted to the hospital. None of these patients were transferred to the ICU. Only 1 patient was readmitted for a questionable cyanotic episode at home and experienced no additional episodes during the subsequent 45-hour inpatient admission.
Our sample size is too small to evaluate the ideal period of observation, although the 8-hour time period we chose allowed some patients to improve and identified some patients that failed. Our results do support the use of an observation period rather than immediate discharge after evaluation and diagnosis.
Our study showed high rates of caregiver and PCP satisfaction with both inpatient admission and discharge after observation in the ED on home oxygen therapy. Only 15% of the caregivers of the patients discharged from the ED on oxygen therapy at the time of the 72-hour follow-up questionnaire stated that they would have preferred hospital admission, whereas 36% of caregivers of the hospitalized patients would have preferred to have been treated at home. Forty-nine percent of the caregivers of the hospitalized patients preferred hospital admission. The preference for home treatment was higher among caregivers than among PCPs. PCPs of the patients discharged from the ED would have preferred hospital admission 27% of the time. The reason for this is unclear. We feel that the comfort level and preference for discharge on oxygen by parents who experienced home management were because of specific concerns about the complexity of home oxygen use before discharge, the use of highly trained respiratory therapists to explain home care, and prearranged follow-up with their PCP. Caregivers of admitted patients did not receive this education; thus, they may prefer inpatient admission because of misgivings and misconceptions about the challenges of the use of home oxygen, although 79% of the hospitalized patients were eventually discharged from the hospital on home oxygen. It is also interesting to note that 45% of the parents who were randomly assigned to home oxygen felt that the observation period of 8 hours was too long, and only 1 patient (3%) felt it was too short. Although we did not attempt to ascertain why, this feeling could be because of adequate explanation and education about home oxygen, the fact that the majority of the patients did well, and could be reflective of a high acceptance by families of home therapies. We developed conservative inclusion criteria including the availability of transportation to return at 24 and 48 hours after discharge, a contact telephone number, living no more than 30 minutes from an emergency medical facility, and a smoke-free environment to maximize the safety of discharge home.
Bronchiolitis is the most common cause of lower respiratory disease that requires hospital admission and contributes significantly to health care costs.22 Home management may decrease these costs, as has been demonstrated with infants discharged on home oxygen from the newborn nursery. Although it did not reach significance, we also demonstrated a trend toward less missed days of work for families who were randomly assigned to home therapy. Family economic impact, ability to comply with prescribed care plans, and overall acceptance of home oxygen therapy for acute respiratory disease in infants and children has not been fully evaluated and warrants further study.
Our study has several limitations. This is a convenience sample of patients who presented with bronchiolitis, and we have no information on patients who were not enrolled; in addition, there is the potential of selection bias. Patients who required higher levels of oxygen, presented with more respiratory distress, or did not at first impression seem to have a good home environment may have been excluded. Our patient population is also a referred population, so it may not be reflective of patients who present to other sites of care such as a primary care clinic or office. Caregivers and ED physicians were not blinded to which treatment arm the patient was assigned, which may have affected treatment decisions. Our strict inclusion and exclusion criteria, as well as the performance of the study at an altitude of 5280 ft (1609 m), limit the external validity of the results.
The study was stopped before the enrollment of the desired number of patients in our sample-size calculation. This occurred for 2 main reasons. ED census was beginning to outstretch inpatient bed availability; therefore, many patients with bronchiolitis were being discharged on oxygen from the ED outside of the study. The hospital was also in the process of developing a short-stay unit distinct from the ED observation unit to attempt to meet the needs of this patient population. Therefore, the decision was made to stop the study protocol, and additional attempts to restart the study were unsuccessful.
Our sample size is small, and we cannot make statements about safety of discharge on home oxygen, optimal observation time period, or criteria that might predict which patients might fail and require hospital admission. A next step will be to systematically follow all of the patients discharged from the ED on home oxygen for adverse events and need for subsequent hospitalization.
This is the first study to demonstrate discharge home from the ED with supplemental oxygen in patients with bronchiolitis. We found that an 8-hour observation period identifies those patients who may resolve their oxygen requirement and those who may worsen and require hospital admission, and, because of a low incidence of complications, the 8-hour observation period is an option for management. High acceptance rates by caregivers and PCPs supports this approach. Additional study is necessary to determine safety and economic impact.
This study was funded by the Children's Hospital Research Institute and Kaiser Permanente. Funding was used to support research assistant time in enrolling patients, collecting data, performing telephone follow-up, and entering data.
Lalit Bajaj had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
- Accepted July 22, 2005.
- Address correspondence to Lalit Bajaj, MD, MPH, Department of Pediatrics/Section of Emergency Medicine, University of Colorado Health Sciences Center/Children's Hospital, 1056 E 19th Ave B251, Denver, CO 80218. E-mail:
The authors have indicated they have no financial relationships relevant to this article to disclose.
- ↵Shay DK, Holman RC, Newman RD, Liu LL, Stout JW, Anderson LJ. Bronchiolitis-associated hospitalizations among US children, 1980–1996. JAMA.1999;282:1140–1446
- ↵Chamberlain JM, Patel KM, Pollack MM. The pediatric risk of hospital admission score: a second-generation severity-of-illness score for pediatric emergency patients. Pediatrics.2005;115:388–395
- ↵Mallory MD, Shay DK, Garrett J, Bordley WC. Bronchiolitis management preferences and the influence of pulse oximetry and respiratory rate on the decision to admit. Pediatrics.2003;111(1). Available at: www.pediatrics.org/cgi/content/full/111/1/e45
- ↵Wilson DF, Horn SD, Hendley O, Smout R, Gassaway J. Effect of practice variation on resource utilization in infants for viral lower respiratory illness. Pediatrics.2001;108:851–855
- ↵Weiss J, Annamalai VR. Discharge criteria for bronchiolitis patients [letter]. Pediatrics.2003;111:445
- ↵Johnson DW, Adair C, Brant R, Holmwood J, Mitchell I. Differences in admission rates of children with bronchiolitis by pediatric and general emergency departments. Pediatrics.2002;110(4). Available at: www.pediatrics.org/cgi/content/full/110/4/e49
- ↵Christakis DA, Cowan CA, Garrison MM, Molteni R, Marcuse E, Zerr DM. Variation in inpatient testing and management of bronchiolitis. Pediatrics.2005;115:878–884
- ↵Perlstein PH, Kotagal UR, Bolling C, et al. Evaluation of an evidence-based guideline for bronchiolitis. Pediatrics.1999;104:1334–1341
- Perlstein PH, Kotagal UR, Schoettker PJ, et al. Sustaining the implementation of an evidence-based guideline for bronchiolitis. Arch Pediatr Adolesc Med.2002;154:1001–1107
- ↵Shay DK, Holman RC, Roosevelt GE, Clarke MJ, Anderson LJ. Bronchiolitis-associated mortality and estimates of respiratory syncytial virus-associated deaths among US children, 1979–1997. J Infect Dis.2001;183:16–22
- Copyright © 2006 by the American Academy of Pediatrics