Association of NPO Status and Type of Nutritional Support on Weight and Length of Stay in Infants Hospitalized With Bronchiolitis
Objectives: We investigated the association of nil per os (NPO) status and subsequent nutritional support with patient weight and length of stay (LOS) during admission for bronchiolitis in patients <2 years old.
Methods: A retrospective chart review was performed of all patients <2 years old admitted to an academic pediatric hospital between November 2009 and June 2011 with a Current Procedural Terminology code of bronchiolitis. Data extracted from the medical record included respiratory rate, per os/NPO status, use of intravenous fluids, use of enteral tube feedings, weight, and LOS. Patients who did not have 2 weights recorded were excluded. The major outcome measures were weight change during admission and LOS.
Results: The study included 149 patients. The mean ± SD patient age was 3.7 ± 3.8 months, with a median age of 2 months. The median length of stay was 4 days (interquartile range: 3–6). Overall, 16% of patients were made NPO, 75% received intravenous fluids, and 9% received enteral tube feedings. The mean weight loss for all patients was 38 (289) g during the hospitalization, which was not statistically significant. No significant association was found between weight loss and LOS, per os/NPO status, or use of intravenous fluids. However, NPO status was associated with a significant increase in LOS.
Conclusions: The infants admitted for bronchiolitis did not demonstrate weight loss in this study; however, an association was seen between NPO status and prolonged LOS.
Respiratory distress due to acute viral bronchiolitis is one of the most common indications for hospital admission in infancy. There were an estimated 149 000 admissions for bronchiolitis in the United States in 2002, resulting in annual costs of more than $500 million.1 A significant amount of research has been conducted on the therapeutic management of bronchiolitis; however, few studies have looked at the impact of fluid and nutrition management on the weight of infants hospitalized with bronchiolitis or the effect on length of stay (LOS). Patients admitted with viral bronchiolitis often have limitations, such as nil per os (NPO) status, placed on their oral intake due to an increased respiratory rate (RR) and concerns about aspiration, which is supported by evidence on bronchoscopy.2–4
Due to limited oral feedings and suspected increased evaporative losses from fever and respiratory effort, many infants receive support with intravenous (IV) fluids or enteral tube feedings to support their hydration and nutrition during hospital admission. Of hospitalized infants in New Zealand, 21% received nasogastric (NG) fluids and 22% received IV fluids.5,6 In 1 study assessing treatment of pediatric patients with moderate dehydration in US emergency departments, IV hydration was relatively more common than NG hydration.7 In Australia and New Zealand, the optimal method of hydration is being addressed by the CRIB (Comparative Rehydration in Bronchiolitis) study, an ongoing, prospective, randomized, multicenter study that will compare IV and NG hydration for 750 patients.8
In addition, ensuring adequate nutrition in hospitalized infants may influence disease outcome. In the developing world, preexisting malnutrition is a strong predictor of mortality from acute lower respiratory tract infection in preschool-aged children.9 Acute nutritional support is also important for previously healthy and well-nourished children. Early enteral feedings have been independently associated with survival in critically ill children.10 Increased protein and energy requirements are needed to achieve protein anabolism in infants admitted to the PICU with viral bronchiolitis.11 Two recent studies demonstrated that poor caloric intake early in the hospital course by infants admitted for bronchiolitis may predict a longer total LOS.12,13
Few studies have explored correlations among NPO status, subsequent support with IV fluid hydration or enteral nutrition, and patient weight. The current study examined the nutrition and hydration support provided to patients acutely admitted for bronchiolitis and compared them with patients’ weight change during admission and LOS. We hypothesized that patients with increased restrictions on per os (PO) intake without supplemental enteral feedings would lose more weight. A secondary aim was to establish the frequency of use of enteral feedings during admission in our hospital for bronchiolitis. We did not specifically attempt to define optimal nutritional management for this population.
After institutional review board approval, the authors (E.E.H, N.C., and S.E.) performed a retrospective chart review of infants admitted to an academic pediatric hospital with bronchiolitis between November 2009 and June 2011. Inclusion criteria included age <2 years at admission and primary diagnosis indicated by Current Procedural Terminology codes 466.1 (acute bronchiolitis), 466.11 (acute bronchiolitis due to respiratory syncytial virus), or 466.19 (acute bronchiolitis due to other infectious organisms). Patients admitted to the NICU were excluded; because very few infectious patients at our institution are admitted to the NICU, they were not considered to be a representative patient sample. Those who spent all or part of their admission in the PICU were included. The medical records were reviewed for patient characteristics (age, gender, and comorbid medical conditions), indicators of respiratory status (RR and need for supplemental oxygen), nutritional status (PO/NPO), and nutritional support (use of IV fluids and use of enteral tube feedings). Our institutional practice is to restrict oral feedings for patients with RR >60 breaths per minute, although this order is ultimately at the discretion of the clinician. Patient weights recorded on admission and throughout the hospitalization were extracted. The weight closest to the day of discharge was used to determine weight change during the admission. Patients who did not have a second weight recorded were excluded from the analysis.
Given the skewed distribution of LOS, continuous LOS is presented as the median and interquartile range (IQR) in addition to mean and SD and is analyzed by using nonparametric tests. LOS was determined by comparing date of discharge with date of admission, with each of those dates counting as a full day in the LOS calculation. All time-based variables (percentage of time spent NPO, with an RR >60 breaths per minute, and receiving IV fluids) were analyzed as continuous variables according to Spearman’s rank correlation coefficient to look for associations with both weight change and LOS.
This analysis was also repeated by using categorical variables. Total LOS was recorded for each patient and was used to divide patients into tertiles of 1 to 3 days, 4 to 6 days, and >6 days, arbitrarily defined based on the national expected mean LOS for patients admitted for bronchiolitis of 2.96 days (UHC data, personal communication to S.E.E., October 2011). The mean change in weight during admission for each LOS tertile was tested by using an analysis of variance (data not shown). In addition, the age of the patient (≤1 month, 2 months, or ≥3 months), the number of days a patient was NPO (≥1 compared with none), the number of days they received IV fluids, and the number of days with an RR >60 breaths per minute (≥1 compared with none) were examined as discrete variables versus weight change (according to the Student's t test) and versus LOS (according to the Wilcoxon rank sum test).
Our initial review identified 272 potential subjects, but 119 charts documented only 1 recorded weight and were excluded. Four patients were also excluded due to NICU admission; the final analysis therefore included 149 patients. The mean patient age was 3.7 months, with a median age of 2 months; 88 (59%) were male. Preexisting medical conditions (27%) and concurrent acute medical issues (42%) were common. Demographic information is included in Table 1.
Data on patients’ disease course and nutritional status are shown in Table 2. LOS ranged from 2 to 21 days. All patients receiving enteral tube feedings had an LOS >6 days except for 2 patients who had preexisting gastrostomy tubes and were continued on their home feeding regimen.
Tables 3 and 4 address weight change during hospital admission. Overall, patients had no significant loss of weight, with a mean ± SD weight change of –38 ± 289 g (P = .11). Mean LOS was 5.3 days, with a median LOS of 4 days (IQR: 3–6). When considered according to LOS tertiles (data not shown) or age (youngest to oldest infants), no significant differences were seen in weight loss. Neither categorical (data not shown) nor continuous measures of duration of NPO status or days receiving IV fluids were significantly associated with weight change. Too few patients received enteral feedings to allow for meaningful comparison with patients receiving IV fluids.
To further explore factors contributing to LOS in this population, we used the same analytic strategy to look for potential associations between nutritional support, RR, and medical comorbidities when comparing LOS (Tables 5 and 6). Longer LOS was seen in patients with longer percentage of time spent NPO (P < .01), greater percentage of time with RR >60 breaths per minute (P < .01), medical comorbidities (P < .01), and acute illness (P < .01). To better understand these associations with LOS, the time receiving IV fluids was analyzed as a discrete variable and as a continuous variable. When examined as a discrete variable, time receiving IV fluids was associated with increased LOS but not when examined as a continuous variable.
We had hypothesized that children made NPO during hospitalization would lose weight if supported only with IV fluids rather than with supplemental enteral feedings. However, our study population did not have a statistically significant change in their weight during hospitalization. On further analysis, we saw no association between patient age, longer LOS, or NPO status with weight change. We found that IV fluid support was a consistent management strategy in our population, with 75% receiving IV fluids despite only 16% being NPO. There are several possible explanations for the lack of weight loss seen in this study. The high rate of IV fluid use may have increased the discharge weights of our population. Admission weights, which are often obtained in the emergency department before any fluid resuscitation, could be reflective of dehydration and therefore falsely low; dehydration associated with bronchiolitis may independently increase the likelihood of hospital admission14 and thus inclusion in our study. Finally, the infants’ oral intake may generally be sufficient to allow weight gain or at least weight maintenance; we did not specifically assess this factor in our study. Only a few patients received enteral tube feedings for support during their hospitalization, several of whom already had gastrostomy tubes in place.
We also examined factors affecting LOS, as studies suggest that poor nutrition may extend hospital stay.12,13 Our study population had a mean LOS of 5.3 days. The expected mean LOS for this diagnostic group is 2.96 days (UHC data, personal communication to S.E.E.). Our mean LOS may have been prolonged by the high percentage of patients with comorbid medical conditions (27%) and/or concurrent acute illness (42%). Both of these populations had an mean LOS > 6 days. When these patients are excluded from the analysis, the median LOS decreased to 4 days with a mean of 3.9 ± 2.1 days, which remains longer than the national mean. The reasons for this difference are unclear but may limit the generalizability of these results. We found that the rate of IV fluid use at our institution was higher than that reported by other researchers.5,6 Our analysis to look for an association between duration of IV fluids and LOS produced mixed results. When examined as a discrete variable, longer duration of IV fluids was associated with a longer LOS; however, when reformulated as a continuous variable (percentage of time receiving IV fluids), no association was seen. Our study did not specifically examine the daily need for IV fluids in this population. It is unclear whether the high rate of IV fluid use contributed to the longer LOS seen at our institution; a focused prospective study would help investigate this possibility further.
We found a positive association between LOS and the percentage of time patients were NPO. This observation is concordant with the analysis of Weisgerber et al,12 in which poor caloric intake on hospital day 2 was associated with a prolonged LOS. Furthermore, there was a strong negative association between LOS and the percentage of time with PO nutrition. We do not know from our study whether this finding is due to the generally milder illness in patients who can continue with oral nutrition or if the enhanced nutrition itself allows patients to recover more quickly. Physicians should be aware that patients with longer LOS are more likely to be made NPO and therefore may require closer attention to their nutritional status.
This report was a retrospective chart review, and thus our data collection was limited to whatever information was in the charts. Although designed to identify patients with the primary diagnosis of bronchiolitis, our population had a high percentage of children with comorbid acute and chronic medical illnesses, which may limit the generalizability of our results. Analysis of weight change required at least 1 weight measurement in addition to the admission weight; however, nearly one-half of our original patient population had only 1 weight recorded and thus could not be included in the analysis. Furthermore, we were unable to determine how standardized the process of weight collection is in our inpatient units. Too few of our patients received enteral tube feedings to allow for a meaningful analysis of enteral versus IV fluid nutrition and weight loss. Most of these patients also received IV fluids during their admission, further complicating this analysis. We could not retrospectively determine specific daily caloric and protein needs and intake, precluding a detailed nutritional analysis. Thus, we did not design the study to define optimal nutritional management for patients with bronchiolitis.
Patients admitted to our hospital with bronchiolitis had no statistically significant change in their weight over their hospital stay. Our patient population was unique in that they had a longer LOS than that reported in other studies and had multiple medical comorbidities, which may limit the generalizability of our results. However, NPO status in these patients was associated with a prolonged LOS. Patients primarily received IV fluids independent of their NPO status. Additional questions raised by this study include whether there is a role for enteral tube feedings in the nutritional management of patients with bronchiolitis, whose clinical status may put them at risk for aspiration of oral feedings, and whether enteral tube feeding can shorten the prolonged LOS in patients who are NPO. The validity of a potential association between duration of IV fluids and increased LOS is also unclear at this time. A more comprehensive and prospective study should be performed to address how NPO status affects weight and LOS in infants hospitalized with bronchiolitis.
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.
FUNDED: No external funding.
POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.
- interquartile range
- length of stay
- nil per os
- per os
- respiratory rate
- Pelletier AJ,
- Mansbach JM,
- Camargo CA Jr.
- Khoshoo V,
- Edell D
- Oakley E,
- Babl FE,
- Acworth J,
- et al
- Johnson WB,
- Aderele WI,
- Gbadero DA
- de Betue CT,
- van Waardenburg DA,
- Deutz NE,
- et al
- Weisgerber MC,
- Lye PS,
- Nugent M,
- et al
- Copyright © 2013 by the American Academy of Pediatrics