Article Review

Leder SB, Siner JM, Bizzarro MJ, McGinley BM, Lefton-Greif MA.

Oral alimentation in neonatal and adult populations requiring high-flow oxygen via nasal cannula. Dysphagia 2016, 31;154-159.

article available through Google Scholar

I am quite concerned by the conclusions of Leder et al regarding safety of oral feeding for preterm infants requiring CPAP.

Advocating for safety for these infants is a critical one for SLPs in the NICU and PICU. Current NICU technology has advanced to the point that more infants are surviving and yet many are requiring extended periods of CPAP and HFNC. Many extremely preterm infants in our NICU with CLD at post-term (41 weeks PMA +) remain dependent on CPAP or HFNC. Our team has had good collaborative conversations about the benefits of ST being involved to maintain a positive oral-sensory environment, promoting the oral-sensory-motor components that are the underpinning’s for future PO feeding, beginning early to foreshadow for parents the swallowing, breathing and postural skills needed, and helping families also support those components, versus attempting PO feeding when the infant clearly is struggling with respiratory stability. Clearly, medical co-morbidities predispose an infant in the NICU to PO feeding problems. Multiple papers have studied that. Those infants with the greatest respiratory co-morbidities, often those born < 28 weeks’ gestation and BW < 1000 grams, are most likely to require CPAP and/or HFNC at those post-menstrual ages when PO feeding is often attempted. Sick newborns may also present similar issues. Secondary to their co-morbidities.

If the infant has such respiratory needs that he requires CPAP, or a HFNC, one must ask if PO feeding is really a priority for that infant at that time. The ability to reconfigure the pharynx from a respiratory tract and back to an alimentary tract with precise timing and coordination surrounding each swallow is a concern. When we look objectively in radiology during an instrumental assessment of swallowing physiology, even infants with CLD stable on RA have altered or impaired swallowing physiology as a direct result of their CLD. The bolus mis-direction and resulting aspiration we often observe is typically silent. This is quite concerning given the conclusion from Leder et al’s study does not even mention this possibility. In the adult population in the most recent information I have seen (Garon et al, 2009 Journal of Neuroscience Nursing) reported that of 2000 adults studied with a variety of co-morbid conditions, including COPD, 54.5% of those who aspirated did so silently. Should the potential for silent aspiration not at least be mentioned? The data I have collected thus far for NICU infants suggests strongly to me that even the data from Arvedson et al in 1994 likely underestimated the tendency for infants to silently aspirate. In addition, her study population was not only less involved from a respiratory perspective back in 1994 than the population we see today, but it also was a population composed of not just infants. The need for an “urgent breath” often can predispose an infant with increased work of breathing to silently mis-direct the bolus into the airway during the swallow. The ability of the infant to close the glottis against the driving force of the respiratory support, while breathing with increased effort or with an increased respiratory rate, which effectively creates air hunger, and yet still maintain glottic closure throughout the duration of the swallow, would likely be precarious. Given the infant’s likelihood of baseline tachypnea and increased WOB, the dynamic adjustments of the airway surrounding the swallow are likely to be disrupted, and create uncoupling of swallowing and breathing. Without objective data on the impact of CPAP or HFNC on swallowing physiology we cannot conclude that feeding under these conditions is “safe”. Indeed, infants for whom we do not necessarily capture aspiration during a dynamic swallow study may indeed show alterations in swallowing physiology that may indeed predispose them to aspiration under “the right conditions” during PO feeding (changes in nipple flow, changes in position, changes in respiratory support for example) so it isn’t even just about aspiration but the potential impact of CPAP and HFNC on swallowing physiology. The fact that the infants “ate” and “were fed” and “transferred volume” does not equate to “safe feeding”. Leder et al’s conclusions don’t unfortunately take this into account.

We must of course consider as well the physiologic stress likely to occur when the infant experiences “feeding” under such circumstances. It is highly possible the stress of trying to breathe and coordinate a swallow may actually lay down neural pathways that move the infant away from wanting to eat, by wiring those sensory-motor pathways that lead to current and/or future maladaptive feeding behaviors. We know that studies looking at stress in preterms have shown an association with adverse changes in brain structure on MRIs.

My colleague, Dr. Suzanne Thoyre, a neonatal nurse feeding researcher, raised some excellent questions about the study design some of which I will try to share in my own words, as she is not a part of our list serve

What did Leder et al deem for inclusion criteria that would be “stable respiratory status” and “Cardio-respiratory stability”? Typically, infants requiring CPAP or HFNC have both increased WOB and increased RR. RR may appear artificially low when rates are captured by a monitor for example that does not sense shallow rapid breaths. Was WOB a consideration? This is a key component of infant clinical assessment of breathing in preparation for attempting as well as during PO feeding. Excessive WOB even in the presence of a less elevated RR can alter timing of swallowing and breathing and often lead to silent bolus mis-direction in the NICU population, much as it might in adults with respiratory work. It is common that infants requiring even low flow oxygen cannot feed without a co-occurring increase in their work of breathing. Two of my OT colleagues in NICUs out East who each observed MD-directed RNs feeding infants on CPAP and HFNC reported seeing significant increase in WOB, RR and signs of behavioral stress.

How was successful oral feeding measured? Did they collect physiologic and behavioral data that demonstrated no indicators of stress, change in oxygen saturation, change in HR, change in respiratory status? The tool they used to measure this needs to be provided. If the feeder simply filled out a checklist, it would be important to know how those observers were trained, how all of these parameters were described and what definitions were used for change in these parameters.
Are the infants in the study on CPAP and HFNC truly being PO fed at 32 weeks PMA? Or is the term “corrected GA” being used incorrectly? I ask because typically even healthy preterms without co-morbidities stable in RA are not being fed PO in most NICUs, with some rare exception. This needs clarification and perhaps makes the conclusions even more concerning.

Unfortunately, Leder et al’s paper will reinforce an incorrect not uncommon assumption. Many neonatologists incorrectly assume that there is a window within which our preterms must “experience” PO feeding or they will “miss that critical window and never learn”. So despite co-morbidities and often respiratory needs that are paramount, infants are being asked to feed. That well-intentioned paradigm is based on writings from Gesell back in the 60s that talked about a “critical window” for learning to eat. Those times were different in many ways as was the population being described. Early intervention now in NICUs to support readiness, neonatal care that is neuro-protective and promotes positive overcomes, and recognition of safety issues inherent in the complex task of PO feeding even when weaned from CPAP and HFNC clearly call for reconsideration of that paradigm, which, perhaps to a large part, underlies the thinking that leads to “pushing PO” and orders to PO on CPAP and HFNC. Many of our former preterms do learn to feed orally at later ages, once weaned, and from my experience do so with much less stress and much more safely.

The dialogue needs to continue and we need measures of oral feeding that go beyond intake, and methods of assessment that actually capture critical variables, including objective assessment of physiology.

I hope this is helpful.