Relationship Between Caloric Intake and Length of Hospital Stay for Infants With Bronchiolitis
Objective: Poor oral intake is a common presenting symptom among infants hospitalized with bronchiolitis. The prevalence, degree, and duration of diminished caloric intake in these infants have not been studied. Our goal was to determine the daily caloric intake among infants admitted with bronchiolitis and to evaluate the relationship between early hospital caloric intake and length of stay (LOS).
Methods: We conducted a retrospective chart review of infants aged <1 year admitted to Children’s Hospital of Wisconsin with bronchiolitis who were placed in the bronchiolitis treatment protocol during the 2004–2005 season. Patient-, disease-, respiratory-, and nutrition-specific data were abstracted.
Results: A total of 273 patients with bronchiolitis were admitted between November 1, 2004, and April 15, 2005; placed on the bronchiolitis protocol; and included in the study. Median caloric intake was diminished on day 1 (53 kcal/kg per day) and day 2 (64 kcal/kg per day). Caloric intake was slower to normalize in infants with progressively longer LOS, and a slower rate of increase from day 1 to day 2 was significantly correlated with longer LOS (r = –0.18; P = .002). Subgroup analysis revealed significant correlations between hospital day 2 caloric intake and LOS in formula-fed infants, breastfed infants, infants aged <183 days, and infants aged ≥183 days.
Conclusions: Caloric intake was diminished in the early course of hospitalization for infants who had bronchiolitis and slowest to normalize in infants with the longest LOS. Interventions aimed at decreasing LOS among infants admitted with bronchiolitis should consider the potential significance of nutrition for severely affected infants with this condition.
Bronchiolitis is the most common reason for admission to the hospital among infants aged <1 year.1 It causes substantial morbidity for patients as well as causing poorer health and increased stress for caretakers and families.2 The cost of bronchiolitis is substantial: approximately $3800 dollars per admission and $500 million per year in the United States.3 Unfortunately, most of the commonly used management modalities have not been effective in improving the clinical course for infants who have bronchiolitis.4
Nutrition is an area of potential importance in the recovery of the infant with bronchiolitis. Clinical signs and symptoms of bronchiolitis include tachypnea and increased respiratory effort,4 which can increase oxygen consumption and energy requirements.5 Despite having higher energy requirements, many infants admitted with bronchiolitis fall short of their normal requirements. One study reported 82% of infants admitted with bronchiolitis had poor feeding before admission and 26% of infants had inadequate feeding, extending the length of stay (LOS) beyond the resolution of their oxygen requirement.6 Our previous research has shown that the caloric intake of infants hospitalized with bronchiolitis was significantly lower on hospital day 2 in those who had a prolonged LOS >5 days compared with those with a shorter LOS.7 Poor nutrition may slow recovery from respiratory illness because of the adverse effects on respiratory muscle strength and host immune function.8 Conversely, good nutrition has positive effects for ill patients. Premature infants who had respiratory distress syndrome with early enteral feedings regained their birth weight faster, spent fewer days on a ventilator, and spent fewer days in the hospital compared with those with delayed enteral feedings.9 Patients with bronchiolitis in the ICU receiving adequate protein-energy had better-preserved anabolism and nitrogen balance.10,11
This research suggests the potential importance of nutrition as a therapeutic modality in the management of bronchiolitis, particularly in the most severely affected infants. The scope of the problem of poor nutrition in infants admitted with bronchiolitis, however, is not well understood. Although we found 1 study noting the prevalence of poor feeding before admission,6 detailed information concerning the magnitude of decreased caloric intake, rate of change, and its relation to length of recovery and hospital stay has not been described previously. Accurate information concerning the natural variation in caloric intake, including rate of change during the hospital course, is an important first step in the process of exploring the potential of nutritional interventions as a modality aimed at improving the recovery of infants hospitalized with bronchiolitis.
Our objectives were as follows: (1) characterize the daily caloric intake and rate of change in infants admitted with bronchiolitis with varying LOS; and (2) evaluate the relationship between early hospital caloric intake (days 1 and 2) and LOS. We hypothesized that infants with the most diminished early caloric intake would take the longest time to return to normal and have the longest LOS.
Study Population and Setting
We conducted a retrospective cohort study during a single bronchiolitis season to identify factors predictive of increased morbidity and a prolonged LOS, and these data have been published previously.7 The current study is a more detailed analysis of the nutrition-specific data whereas the previous article focused on factors predictive of prolonged LOS.
Children’s Hospital of Wisconsin (CHW) is a 294-bed tertiary care academic center. The charts of all infants discharged from CHW who met all of the following criteria were reviewed: (1) age <365 days; (2) admitted between November 1, 2004, and April 15, 2005; (3) discharge diagnosis of bronchiolitis according to the International Classification of Diseases, Ninth Revision discharge codes 466.11 (acute bronchiolitis due to respiratory syncytial virus) or 466.19 (acute bronchiolitis due to other infectious organisms); and (4) placement on the CHW bronchiolitis treatment protocol. Only patients placed on this treatment protocol were studied because these infants have a consistent model of care modeled after American Academy of Pediatrics’ guidelines.4 Inclusion in the protocol requires a physician making a clinical diagnosis of bronchiolitis and ordering the protocol. Protocol monitoring and therapeutic decisions are guided by clinical respiratory scores, which influence decisions on whether to initiate or continue therapies such as bronchodilators or positive expiratory pressure treatments. Typically, 70% of infants admitted to CHW with bronchiolitis are placed on this protocol. The other 30% are composed of patients whose diagnoses are less certain, whose conditions are medically complex and involve other comorbidities, who have other preexisting respiratory problems in which the clinical scores would be confounded and not be able to be used for appropriate management, or who have admitting physicians who choose not to order the protocol.
Infants with events or surgeries that occurred during the admission not related to bronchiolitis and affecting LOS were excluded. An example of this would include gastrostomy tube placement during admission. Infants admitted to the ICU were included if placed in the bronchiolitis treatment protocol.
Five trained abstractors (3 were study authors) abstracted information from patient records (including written charts and electronic data sources). Variables abstracted from patient charts included the following: demographic data; clinical respiratory status data; comorbidity data; and nutritional data including weight, height, Waterlow classification (percentage of ideal body weight) and calculated daily caloric intake. After data entry, a statistician informed the primary author of missing data and statistical outliers. These findings were double-checked for accuracy. Selected charts were reviewed for accuracy by the primary reviewer in the training process of new abstractors to ensure that a standardized procedure was used.
Caloric intake was defined as the number of kilocalories consumed or estimated on a given hospital day and expressed in kilocalories per kilogram per day (kcal/kg per day). A hospital day was defined according to nursing flow sheets from 6 AM to 5:59 AM the following day. Total kilocalories for any infant were determined by summing the measurable calories from formula or expressed breast milk with an estimate of calories obtained from sessions of active breastfeeding (Table 1). If the infant was in the hospital <24 hours (typically on the day of admission or discharge), the caloric intake was extrapolated to a 24-hour day and expressed as kcal/kg per day by dividing the measured intake by the number of hours they were in the hospital multiplied by 24.
It was assumed that the majority of caloric intake was from formula or breast milk. Intravenous fluids were not included. Calories from solid food intake were not included because flow sheet data include qualitative but not quantitative descriptions of solid food intake, and these data are insufficient to produce a caloric estimate.
The caloric intake obtained via active breastfeeding was estimated by using an equation created by an institutional expert panel after careful literature review. The estimate equation involved a low, medium, and high estimate of calories likely obtained per feeding (Table 1). The high estimate was equivalent to the volume of calories per feeding to meet age-appropriate norms.12,13 The low estimate assumed that the infant had no intake (zero calories). The medium estimate was mid-way between the low and high estimates. The 3 estimates allowed for a crude sensitivity analysis for the possible variable contributions from breastfeeding, which were added to measurable caloric intake. For all baseline analyses, the medium estimate of the breastfeeding contribution to caloric intake was used. For the primary outcome, sensitivity analyses were run using the low, medium, and high estimates. The primary outcome was correlation between caloric intake on hospital day 2 and LOS, defined as the number of hours from the time a subject arrived in the hospital unit to the time of last nursing documentation at time of discharge. Hospital day 2 was chosen rather than hospital day 1 because it was the first complete day of caloric intake information. For many patients admitted at night, hospital day 1 only comprised a few hours. In addition, many outcomes from our previous work demonstrated stronger associations with day 2 variables; thus, day 2 caloric intake was chosen for our primary outcome.7
When accounting for baseline malnutrition status, the Waterlow classification was chosen as an expression of chronic malnutrition status by using the percentage of ideal body weight: class 1, >90% (normal); class 2, 81% to 90% (mild malnutrition); class 3, 71% to 80% (moderate malnutrition); and class 4, <70% (severe malnutrition).14
Analysis and Sample
Standard descriptive statistics were used for describing daily caloric intake on hospital days 1 through 5 for infants of varying LOS. Because age and caloric intake were likely to be nonnormally distributed, median and interquartile range (IQRs) were used.
Correlation between day 2 caloric intake and LOS was calculated. To identify a modest correlation by using the Pearson correlation coefficient between caloric intake and LOS (correlation coefficient of ≥0.2), a sample of 287 subjects was needed. Power was chosen to achieve 90%, and a lower level of significance of P < .01 was used to account for multiple comparisons. Stratified subgroup analyses were performed for these outcomes to identify potential confounding and effect modification from breastfeeding status, age, and preexisting chronic malnutrition.
This study was approved by the CHW institutional review board, and the requirement for informed consent was waived.
A total of 273 (79% of all bronchiolitis admissions) infants met inclusion criteria and their data reviewed. One patient was excluded because of gastrostomy tube placement during the admission. Two patients had missing data for caloric intake on days 1 and 2. Twenty-three patients did not have length obtained, and therefore their Waterlow classification could not be determined.
Baseline and selected characteristics of infants admitted with bronchiolitis are described in Table 2. The majority of subjects (77%) were term infants, were <6 months old (63%), and were formula fed (16% being breastfed some or all of the time).
Median caloric intake for infants of varying LOS for the first 5 days of the hospitalization using the medium estimate (estimate of caloric intake for breastfed infants) is described in Table 3 and depicted in Fig 1. The 270 patients with complete caloric intake data were included in this analysis. We noted a trend between change in caloric intake from day 1 to day 2 in which it seemed that shorter LOS stay groups were faster to improve their caloric intake. However, there was substantial variation in the range of caloric intake as noted by a wide IQR in caloric intake among the LOS groups. There was a modest inverse correlation (r = –0.18; P = .002) between change in caloric intake (day 2 – day 1 kcal/kg per day) and LOS (hours) (lower caloric intake was associated with longer LOS).
There was a modest inverse correlation between the caloric intake of formula-fed–only infants and LOS (r = –0.28; P < .001) (Table 4). These formula-fed infants had their entire caloric intake measured without the use of estimate equations for contributions from breastfeeding. When the caloric intake of all infants was evaluated (including the use of breastfeeding caloric estimates for those who had breastfed), there was a modest significant inverse correlation between caloric intake on day 2 and LOS in all infants by using the low, medium, and high estimates for the contribution of breastfeeding.
Stratified subgroup analysis of day 2 caloric intake and correlation with LOS was performed in the following strata: infants aged <183 days versus those aged ≥183 days, any breastfeeding versus none, and no/mild chronic malnutrition versus moderate/severe chronic malnutrition. All subgroups had significant mild correlations between day 2 caloric intake and LOS (Table 4). The only subgroup comparison with a significant difference in day 2 caloric intake was in infants ≥183 days old who had significantly lower median caloric intake than infants <183 days old.
To the best of our knowledge, this study is the first to report detailed information about the daily caloric intake of a large cohort of infants admitted to the hospital with bronchiolitis. We report the daily caloric intake of the first 5 days of admission for 6 different LOS groups (Fig 1). This study confirmed our hypothesis that low caloric intake is common for many infants who have bronchiolitis and for some is persistent through many days of the hospitalization. It also confirmed there is a correlation between LOS and both lower early hospital caloric intake and slower rate of improvement in caloric intake.
Previous research has shown that many infants with bronchiolitis have poor feeding before admission.6 An important part of our work was to evaluate the spectrum of poor caloric intake extending into the hospitalization as well. Descriptive statistics revealed that the median caloric intake of infants on day 1 of the admission was ∼49% of their caloric needs. It was fairly similar in groups of infants of varying LOS, ranging from 39% to 60% of their needs assuming the normal state, although we would expect with increased respiratory rate and effort, that their actual energy expenditure would be higher. By day 2 of the admission, median intake was similar (53% of needs); however, there was increased separation of the median caloric intake among infants of varying LOS. The rate of change in caloric intake from day 1 to day 2 modestly correlated with LOS. The shortest LOS groups had rapid increases in caloric intake toward 70 to 80 kcal/kg per day by the day of discharge, whereas progressively longer LOS groups reached this level of caloric intake more slowly. In the longest LOS group, caloric intake remained diminished even 5 days into the admission. In addition, infants had diminished oral intake at home for an average of 1 day before admission. In our previous study, we found that infants with LOS >5 days had diminished oral intake for 3 days before admission.7 Although this study cannot determine the effect of up to 8 days with a growing calorie deficit between consumption and expenditure, we speculate a deficit of this magnitude might slow recovery. Our findings correlating low caloric intake with longer LOS aligns with findings in our previous research, which identified very low caloric intake (<23.5 kcal/kg per day) on day 2 of the hospitalization as 1 of 5 factors predictive of a prolonged LOS in a recursive partitioning model.
In addition to examining groups of varying LOS, we examined subgroups based on type of feeding, age, and the presence of malnutrition. Caloric intake in these subgroups has not been previously reported and is important in gaining a complete understanding of caloric intake among infants hospitalized with bronchiolitis. Infants who were strictly formula-fed had a significant modest correlation between caloric intake and LOS. This was an important group because it encompassed the majority of infants in our study, and these infants had caloric intake that was entirely measured without the need for breastfeeding estimates. Infants who were breastfed had similar correlations between caloric intake and LOS. We found a significantly lower day 2 caloric intake among infants aged ≥183 days (6 months) compared with those aged <183 days. One possibility for this finding is an underestimation of calories because of unquantified calories from solid foods in older infants. These 2 age groups, however, had similar correlations between caloric intake and LOS. Patients with moderate to severe malnutrition had caloric intake more strongly correlated with LOS. However, with only 10 of these patients in our study, this correlation was underpowered and did not reach statistical significance. This trend does suggest that early caloric intake may be even more important in malnourished patients who have bronchiolitis.
This study has several limitations. First, it was a retrospective study of a single bronchiolitis season at a single institution. It is possible results would differ during a different season or at another institution. Second, the authors also served as data abstractors and could have been biased because they were not blinded. Third, intake from breastfeeding was estimated by using a nonvalidated measure because no validated measure exists. In addition, because the minority of infants were breastfed, it is possible we could have missed a differential effect of breastfeeding on caloric intake or relationship with LOS. Fourth, solid food intake was not quantified and is therefore absent from caloric estimates. For infants <183 days old, this would unlikely affect the results. For older infants, it could have led to an underestimation of caloric intake. Fifth, this study did not evaluate the contribution of intravenous fluids as a provider of calories or modifier of oral intake. However, the proportion of infants receiving intravenous fluids was similar in both short and prolonged stay groups, and intravenous fluids provide very few calories. Sixth, this study evaluated nutrition by measuring only 1 aspect (caloric intake). There are other specific nutrients such as fatty acids, vitamins, and various nutrients that may have an impact on recovery from bronchiolitis that could not be evaluated in our study.15–19 Seventh, because adequate caloric intake is part of the discharge criteria at our institution as well as many others, it is logical that low caloric intake will contribute to a prolonged stay merely based on the fact that it is 1 of the discharge criteria. Eighth, our sample size was slightly less than thought to be needed in our power calculations (272 vs 288); therefore, it is possible some outcomes fell short of significance because they were underpowered. Ninth, because not all infants with bronchiolitis were placed in the bronchiolitis protocol, a systematic bias could have been introduced, and it is important to note these results may not generalize to highly medically complex infants with bronchiolitis who were underrepresented in the bronchiolitis protocol.
Finally, it is important to recognize this study cannot differentiate whether lower caloric intake itself had an independent role in prolonging recovery in bronchiolitis or whether it merely served as a proxy for severity of illness. We suspect that severity of illness is the predominant factor, which causes diminished caloric intake in infants admitted with bronchiolitis. For many infants, this is a short-lived problem unlikely to contribute to morbidity. However, for the most severely affected infants with many days of diminished caloric intake, we suspect poor nutritional intake may become an additional source of morbidity and slow recovery.
This study lays the groundwork needed for further exploration of the potential benefit of nutritional interventions for infants hospitalized with bronchiolitis. There is a group of infants with bronchiolitis and prolonged LOS who have poor caloric intake readily identifiable early in their hospitalization. With few effective interventions known to reduce morbidity in bronchiolitis, the impact of nutrition ought to be considered.
Caloric intake is often diminished in infants admitted with bronchiolitis. It seems to normalize more slowly in infants with progressively longer LOS. As early as hospital day 2, it was significantly lower in infants with a prolonged LOS and inversely and modestly correlated with LOS. These findings suggest that efforts to improve nutrition early in the course of bronchiolitis are worthy of additional study, particularly in the most severely affected patients in whom diminished caloric intake remains prolonged and theoretically could prolong recovery.
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.
FUNDING: No external funding.
- Children’s Hospital of Wisconsin
- interquartile range
- length of stay
- Leidy NK,
- Margolis MK,
- Marcin JP,
- et al
- Pelletier AJ,
- Mansbach JM,
- Camargo CA Jr.
- 4.↵American Academy of Pediatrics Subcommittee on Diagnosis and Management of Bronchiolitis. Diagnosis and management of bronchiolitis. Pediatrics. 2006;118(4):1774–1793.
- Premer DM,
- Georgieff MK
- Unger S,
- Cunningham S
- Pingleton SK
- de Betue CT,
- van Waardenburg DA,
- Deutz NE,
- et al
- 12.↵Subcommittee on the Tenth Edition of the RDAs Food and Nutrition Board Commission on Life Sciences. Recommended Dietary Allowances. 10th ed. Washington, DC: National Research Council National Academy Press; 1989.
- Kelts DG,
- Jones EG
- Waterlow JC
- Robertson J,
- Shilkofski N
- Samour PQ,
- Helm KK,
- Lang CE
- Copyright © 2013 by the American Academy of Pediatrics