High Flow Nasal Cannulae Therapy in Infants with Bronchiolitis
Christine McKiernan, MD, Lee Chadrick Chua, MD, Paul F. Visintainer, PhD, and Holley Allen, MD
This study is the first published to assess the use of HFNC therapy to support infants with bronchiolitis admitted to a PICU. The goals of this study were to determine whether the availability of HFNC was associated with a decreased need for intubation and to describe the changes in respiratory measurements like respiratory rate (RR) and length of stay (LOS).
The study was designed as a retrospective chart review with a historical control. HFNC therapy became available at the study institution in September 2006. Data was collected just prior to the availability of HFNC therapy (HFNC-NA) and again after the system became available (HFNC-A). Chart review was performed for infants <24 months of age admitted to the PICU with a diagnosis of bronchiolitis, respiratory distress or respiratory failure. Infants were excluded if the primary diagnosis was not bronchiolitis (i.e. reactive airway disease or pneumonia), if they had a preexisting tracheostomy or were intubated in the prehospital setting. The groups were analyzed by the time period of the admissions and not necessarily by the treatment type. The majority of the infants in this study were admitted from the ED or were transferred from the general pediatric units.
Intubation and discharge from the PICU was at the discretion of the PICU physicians, but patients discharged to the floor were only allowed to have blow-by or standard nasal cannulae oxygen. The decision to use HFNC was at the discretion of the PICU attending and no specific protocol was utilized. Other external factors which add significant validity to the study included the following: PICU admission guidelines, PICU attending staff, and general bronchiolitis standards of care (i.e. nebulized medications, suctioning) did not change throughout the study. In addition, there was less than 10% turnover of respiratory therapy and nursing staff. This study utilized the Fisher Paykel heated HFNC system with low-resistance infant nasal cannulae with a maximum flow rate of 7L/min or the pediatric cannulae with a maximum flow rate of 8L/min. Of note, there are other commercially available systems such as Vapotherm HFNC and even “homemade” systems. The size of the nasal cannulae was chosen to fit the nares without occlusion, and the flow rates were usually started at the maximum rate for the respective cannulae.
115 patients met the enrollment criteria (57 HFNC-NA and 58 HFNC-A). The groups were similar regarding median age, weight, gestational age, admission oxygen saturation, Pediatric Index of Mortality 2 score, sex, history of chronic lung disease and respiratory syncytial virus (RSV) status. More children in the HFNC-NA versus HFNC-A were defined as premature (gestational age < 37 weeks) (40% vs 19%), but the median and mean gestational ages were not significantly different. The median admission respiratory rates (RR) were statistically higher in the HFNC-A group (61 [range 24-120]) versus the HFNC-NA group (47 [range 21-85]) (p=<0.001). 33/57 (57.9%) of the HFNC-NA received nasal oxygen as the primary respiratory intervention. Other respiratory treatments for HFNC-NA included room air (15.8%), blow-by oxygen (15.8%), simple face mask (1.8%), non-rebreather mask (3.5%) and nasal CPAP (5.3%). 51/58 (87.9%) of the HFNC-A received HFNC. Other treatments for HFNC-A included room air (1.7%), blow-by oxygen (1.7%) and nasal cannulae (8.6%).
Infants treated with HFNC had a significantly decreased RR at 1 hour after therapy was initiated (18+/-16 breaths/min versus 6+/-14 breaths/min; p=<0.001). This effect persisted after logistic regression controlling for age, weight, gestational age and whether the infant required intubation. 13/57 (23%) required intubation in the HFNC-NA group versus 5/58 (9%) in the HFNC-A group (p=0.043). This yielded a 14% absolute reduction in the need for intubation (NNT=7; 95% CI 4-110). HFNC reduced the need for intubation by 68% (unadjusted logistic regression) and after controlling for age, weight, and RSV status, it remained at 68%. Adding gestational age to the regression analysis led to a risk reduction of 65%. PICU LOS decreased from a median of 6 days to 4 days in the HFNC-A group (p=0.0058).
The authors felt that HFNC was well tolerated by the infants and appeared to be safe. Some of the infants were even able to tolerate oral feeds during HFNC treatment. There were no complications such as nasal or facial trauma and there were no deaths in either group. There were 2 pneumothaces cases – one in each group – and both of these patients had been intubated.
As noted, this was a retrospective chart review and there are inherent problems that can be associated with this type of study design. Was it quality data? How was it abstracted and by whom? Were the abstractors trained? There are no clear consensus-based standards for how these studies should all be designed, however understanding some of the limitations of a retrospective chart review is paramount. (4) This was a convenience sample obtained during a set period. A few things that were missing from the methods were that the authors did not calculate a predetermined sample size and they did not note how data was abstracted from the charts. This would have been nice, but since most of the data collected was objective and pretty routine data for patients going to a PICU, this is unlikely to have affected the study outcome.
This paper is a great starting point to improving our care for bronchiolitic patients. Whether it is truly better than other non-invasive forms of ventilation like CPAP remains to be determined. According to a recent Cochrane review, “there is a lack of well designed, controlled experiments of non-invasive modes of respiratory support for children.” (5) Theoretically, HFNC works by providing CPAP-like effects which leads to decreased work of breathing and fewer intubations. Of note, a recent abstract addressed the use of HFNC using data from the Australian and New Zealand Paediatric Intensive Care Registry (ANZPIC). (6) The investigators systematically reviewed patient data and the rate of use of non-invasive ventilation and invasive ventilation with the ANZPIC data. This was interesting since HFNC had been used routinely and prophylactically in all bronchiolitics at their institution since 2003. They showed a lesser need for invasive airway management in their population. Hopefully, the future will bring us a prospective randomized trial comparing HFNC with CPAP or a trial with more objective measures. We utilize this at our institution in the ED for patients going to the PICU. Think of this option and call your respiratory therapist to set this up if your patient is “sick,” needs an ICU, but does not yet need the ET tube. Maybe you will save them an intubation and provide them a shorter length of stay in your busy PICU.
2. Martinón-Torres F, Rodríguez-Núñez A, Martinón-Sánchez JM: Nasal continuous positive airway pressure with heliox versus air oxygen in infants with acute bronchiolitis: a crossover study. Pediatrics. 2008;121(5):e1190-5.
3. McKiernan C, Chua L, Visintainer P et al: High flow nasal cannulae therapy in infants with bronchiolitis. J Peds. 2009 Dec 24. [Epub ahead of print]
4. Worster A, Haines T: Advanced Statistics: Understanding Medical Record Review (MRR) Studies. Acad Emerg Med. 2004;11(2):187-192.
5. Shah P, Ohlsson A, Shah J: Continuous negative extrathoracic pressure or continuous positive airway pressure for acute hypoxemic respiratory failure in children. Cochrane Database of Systematic Reviews 2008; Issue 1.