Transition to Weight-Based High-Flow Nasal Cannula Use Outside of the ICU for Bronchiolitis

This cohort study assesses changes in intensive care unit (ICU) admission rate, hospital length of stay, and use of noninvasive and invasive ventilation after the adoption of weight-based high-flow nasal cannula protocol in hospitalized patients aged 0 to 24 months.


Introduction
High-flow nasal cannula (HFNC) is a common respiratory support modality in children hospitalized with bronchiolitis [1][2][3][4][5] ; more than 50% of children were exposed to HFNC in a recent multicenter study. 6Use of HFNC was initially restricted to the intensive care unit (ICU), but over time it has expanded to pediatric wards to reduce ICU use for patients with bronchiolitis. 2,5Early protocols for ward-based HFNC applied flow rates according to age (eg, 8 L of flow for children younger than 12 months). 1An emerging alternative to age-based HFNC protocols are weight-based HFNC protocols (eg, 2 L of flow per kilogram of body weight).
The present study focused on weight-based non-ICU HFNC use for 2 principal reasons.First, multicenter studies have examined the transition to age-based non-ICU HFNC use, but similar studies for weight-based protocols are limited. 7,8Second, a recent survey of children's hospitals that participate in the Pediatric Health Information Systems (PHIS) found that age-based HFNC protocols are being replaced by weight-based protocols; in a 2021 survey, 75% of hospitals with non-ICU HFNC protocols used weight-based flow rates 5 compared with 27% of hospitals in a 2017 survey. 2 The objective of this study was to measure the association between hospital transition to weight-based non-ICU HFNC use and subsequent ICU admission.

Methods
We conducted a multicenter retrospective cohort study involving 18 children's hospitals that contribute patient-level data to the PHIS database, including demographic characteristics, procedure codes, and discharge diagnosis codes.The University of Utah Institutional Review Board deemed this study exempt from review and the informed consent requirement because it was not human participant research.We followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.

Patient Characteristics
Patients aged 0 to 24 months who were hospitalized between January 1, 2010, and December 31, 2021, were included in the cohort if they had any International Classification of Diseases, Ninth Revision (ICD-9) or International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10) bronchiolitis code in any diagnostic position. 9,10Patients were excluded if they received care in the neonatal ICU or had a length of stay (LOS) greater than 30 days.

Exposure and Outcomes
The study exposure was hospital-level transition from ICU-only to weight-based use of HFNC in non-ICU wards.Because HFNC use is not reliably documented in the PHIS database, we obtained data from our 2021 survey 5 (which had an 82% response rate) to identify hospitals that restricted HFNC use to the ICU (ICU-only group) and hospitals that transitioned from using HFNC only in the ICU to adopting weight-based HFNC use in non-ICU wards (weight-based protocol group).Ten hospitals were included in the ICU-only group, and 8 hospitals were in the weight-based protocol group.Together, these 18 hospitals were the source of the patient-level data we obtained from the PHIS database.The PHIS database includes a hospital indicator, allowing linkage to the 2021 survey results and patient-level data.
The primary outcome was the proportion of patients with bronchiolitis admitted to the ICU.
Secondary outcomes included mean total hospital LOS, the proportion of patients who received noninvasive positive pressure ventilation (NIPPV), and the proportion of patients who received invasive mechanical ventilation (IMV).Patients were classified as receiving NIPPV if they had an ICD-9 1][12] A crosswalk available from the Centers for

JAMA Network Open | Pediatrics
Transition to Weight-Based High-Flow Nasal Cannula Use Outside of the ICU for Bronchiolitis Medicare and Medicaid Services was used to convert previously published definitions that contained ICD-9 codes to ICD-10 codes.

Changes in Outcomes
We used a controlled interrupted time series approach to measure changes in outcomes associated with the transition from ICU-only to weight-based HFNC use, comparing 3 years before with 3 years after transition.The transition point for the weight-based protocol group was determined by survey responses. 5Because the ICU-only group did not have a true transition point, we established the transition point by matching the ICU-only hospitals with the weight-based protocol hospitals according to size (number of beds) and geographic location.Each ICU-only hospital was given the transition point of its matched weight-based protocol hospital.Next, we measured the immediate change at the transition point and the change in slope after the transition point. 13 identified the immediate change and the change over time associated with the transition by comparing changes in the slopes observed in the weight-based protocol group with changes observed in the ICU-only group. 14This approach reduces unmeasured confounding by controlling for other interventions and events concurrent with the study exposure. 14

Statistical Analysis
Statistical analyses were performed with 2 models: a mixed-effects linear regression model and a time series model.First, we ran the mixed-effects linear regression model using an interaction between time relative to transition and weight-based protocol adoption status, accounting for patient clustering within hospitals.We adjusted for patient age, sex, race, ethnicity, insurance type, and whether hospitalization occurred during the COVID-19 pandemic.We defined hospitalization during the COVID-19 pandemic as starting in April 2020 and ending in December 2021.Race and ethnicity (categorized as American Indian, Asian, Black, Hispanic and non-Hispanic, Native Hawaiian, White, and other [not specified]) were identified in hospitals and obtained from the PHIS database.
Race and ethnicity data were collected and analyzed in the study because of prior research showing race-based disparities in receipt of bronchiolitis guideline-adherent care. 15cond, we used adjusted means for each outcome from the mixed-effects model as inputs for a controlled ordinary least squares time series model.We adjusted for autocorrelation with Newey-West SEs. 16Two-sided P < .05indicated statistical significance.Analyses were performed from July 2023 to January 2024 using Stata, version 16 (StataCorp LLC).

Results
A total of 86 046 patients with bronchiolitis received care at 10 hospitals in the ICU-only group (n = 47 336) and 8 hospitals in the weight-based protocol group (n = 38 710).Age and sex were similar between the ICU-only group (  1).Hospitals in the ICU-only group vs weight-based protocol group had higher proportions of Black (26.2% vs 19.8%) and non-Hispanic (81.6% vs 63.8%) patients and patients with governmental insurance (68.1% vs 65.9%).Hospital characteristics including freestanding status, size, mean daily census, and geographic region were similar between the 2 groups (Table 2).

JAMA Network Open | Pediatrics
Transition to Weight-Based High-Flow Nasal Cannula Use Outside of the ICU for Bronchiolitis

Figure .
Figure.Comparison of Outcomes Between Intensive Care Unit (ICU)-Only and Weight-Based Protocol High-Flow Nasal Cannula (HFNC) Use The dashed line indicates the transition point from using HFNC at ICU only to adopting a weight-based HFNC protocol.Error bars represent 95% CIs.

Table 1 .
Patient Demographic Characteristics by Group

Table 2 .
Hospital Characteristics by Group

Table 3 .
Outcomes by Group a Statistically significant results.