SARFO: The syndrome of acute respiratory failure in obesity

A 56-year-old male with history of morbid obesity (150kg), CHF, OSA, OHS, and COPD presents to the ED with altered mental status. His initial blood gas shows signs of hypoxia and acute-on-chronic hypercapnia. Chest CT rules out PE, but shows widespread atelectasis, small pleural effusions, and vascular congestion. He is intubated and admitted to the ICU.

Over the next 5 days, he receives diuretics for CHF, antibiotics for possible pneumonia, and steroids and bronchodilators for a possible COPD exacerbation. His respiratory function improves; he is extubated and transferred to a step-down bed. However, within 24 hours he is again found obtunded with recurrence of hypercarbia. He is placed on BiPap with good response.

Over the next 2 weeks, he proves unable to wean from BiPap for more than several hours at a time; he takes breaks on high-flow nasal cannula but inevitably becomes either tachypneic or hypercarbic without positive pressure. Further diuresis results in prerenal azotemia. There are no further signs of pneumonia. Eventually, he is trached and transferred to a long-term acute care center.

Have you seen these patients? I have, and I find them very challenging. In fact, it’s a sufficiently unique phenotype that I’ve started to think of them as a syndrome—let’s call it the Syndrome of Acute Respiratory Failure Syndrome in Obesity (or SARFO).

The textbooks don’t have much to say about SARFO. What are the defining features?

1. A mixed respiratory failure

These patients typically present with both hypoxia and hypercarbia, although the hypercarbia often predominates. This is typically an acute-on-chronic finding, with signs of chronic compensation (i.e. an elevated serum bicarbonate).

2. A severely atelectatic chest

Imaging typically reveals lungs that could be summarized as “terrible.” They are underinflated, congested, and perhaps bracketed by small effusions. Most striking, however, is widespread atelectasis. Is it due to compression, consolidation, hydrostatic edema, mucus plugging? Hard to say; in reality, all of the above are probably present.

3. Numerous misdiagnoses

As a rule, SARFO patients receive the full boilerplate of respiratory labels.

They are invariably diagnosed with a CHF exacerbation, although in reality their systolic function is often preserved, or at least not remarkably poor; diastolic failure is probably more common. They are then diuresed until they turn into a raisin.

COPD exacerbation will also be invoked, and the patient will receive steroids, bronchodilators, and azithromycin; however, they are often not wheezing on auscultation, and in some cases, there have never had a PFT-confirmed diagnosis of COPD at all; it is these “exacerbations” alone that led to the label.

Obstructive sleep apnea (OSA) and obesity hypoventilation syndrome (OHS) are usually mentioned, and the patient kept on CPAP or BiPAP at night. A good thing, no doubt. But it doesn’t fix the larger problem.

4. A slow, refractory course

More than anything, the hallmark of the SARFO patient may simply be their failure to improve. They often spend days if not weeks trying and failing to wean from positive pressure ventilation. These failures are frustrating, tethering patients continuously to a mask that limits their ability to eat and speak, as well as predisposing to skin breakdown—but each time they try and fail also accords some risk. They may precipitously deteriorate to the point of intubation, or even cardiac arrest.

What’s their deal?

It’s difficult to fully understand these patients, who are not easily categorized. The etiology of their respiratory failure seems to be multifactorial.

The closest diagnostic fit is obesity hypoventilation syndrome. But the general OHS literature has relatively little to say (1, 2, 3) about management during these acute exacerbations. Also, classic definitions of OHS are predominantly hypercarbic in nature, with any hypoxemia due to hypoventilation alone, while SARFO patients often exhibit “real” hypoxia from shunt physiology as well.

Obstructive sleep apnea is, of course, often present, but usually manageable; the daytime symptoms are more troubling.

Atelectasis is clearly a key component, caused by a combination of external lung compression and immobility.

Finally, a miscellany of comorbidities can and do contribute. Some degree of heart failure probably does exist in many cases, often diastolic in nature. More importantly, pulmonary hypertension (WHO Group 3) is common. I suspect that this latter is responsible for the poor reserve seen in some SARFO patients, as it may predispose them to cardiovascular collapse and PEA arrest when a small insult (such as overaggressive weaning or sedative meds) nudges them over the edge. True COPD can be present, of course. And a degree of renal failure is very common, often complicating volume management.

The most attention to this muddled phenotype has been from Paul Marik, who actually beat me to the neologistic punch, coining the term “malignant OHS” to describe these refractory cases that make it to the ICU. (This review is his most comprehensive, even though it was actually retracted due to similarity to another piece he wrote; I’ll still shamelessly link to it.) He went on to analyze two cohorts of “MOHS” patients with marked similarities to our example (1, 2). There, we see again the very large proportion of his population that was diagnosed with COPD — which they did not have — while a surprisingly small number carried diagnoses of OHS and/or OSA, which they most likely did have. Pulmonary hypertension and myocardial hypertrophy, CKD, hepatic failure (NASH), and vitamin D deficiencies were all common, and diabetes was almost universal; he recommends screening for all of this, as well as aggressive use of APRV, weight loss (with a high protein diet), even sildenafil.

In one cohort, Marik sank esophageal manometers into 18 of the intubated patients, demonstrating a mean intrapleural pressure of 17 cmH2O. While similar figures are not unheard of in mechanically ventilated patients, it underscores the importance of an adequate PEEP or EPAP; obviously an end-expiratory intra-alveolar pressure of 5 is unlikely to prevent atelectasis in such a person.

So what do we do?

A splendid question. The one thing that’s clear is that there aren’t any easy answers in these tough patients. Here is what tends to work for me:

  1. BiPAP: Non-invasive ventilation is the first-line therapy, and most patients will respond briskly. A fairly high EPAP may be necessary. In principle, the delta (driving pressure) need not be very high if you can recruit them effectively, but in practice I find that needs vary widely, and pressures should be titrated to achieve adequate tidal volumes and blood gasses. A low-normal SpO2 goal of 88–92% or thereabouts is probably appropriate.
  2. Diuresis: SARFO is not a CHF exacerbation, but they probably benefit from being as dry as possible. Given their body habitus and chronically wet-appearing chest imaging, it can be very hard to know when you overshoot, which you probably will. An increasing BUN may be your best clue.
  3. Ruling out other problems: Make sure there isn’t a PE or a true pneumonia hiding in the lungs. While usually not present, the diagnostic confounders can make them easy to miss, so err on the side of caution.
  4. Recruit aggressively: In intubated patients, the number one goal is recruitment, which will require high airway pressures. During inspiration, you may need to disable peak pressure limits to achieve acceptable tidal volumes, accepting a “dangerous” Pplat with the belief that the patient’s poor compliance is largely caused by extrinsic compression, and therefore that the transpulmonary gradient is actually normal. (Of course, if you have esophageal manometry available, you could prove this belief and titrate more precisely.) On the expiratory side, high PEEP will be needed. Best of all is probably the early use of APRV, which can be a real boon.
  5. Try to wean: Cautious attempts at weaning from NIPPV should be undertaken once they’re medically optimized. Try for a few additional hours off support each day, starting at mealtimes, and going back on the mask overnight. High-flow nasal cannula is probably not nearly as effective, but can be tried as a bridge. Don’t push them too hard; remember that when they fail, they sometimes fail fast, and may be difficult to reintubate. Sedation should be avoided at all costs, as it can easily torch their respiratory drive.
  6. Accept the trach: The inevitable endpoint of many SARFO patients is a tracheostomy, which allows safe ongoing ventilation, reduces work-of-breathing, and instantly eliminates OSA. Many of them were functional and independent prior to admission (ambulatory, living at home, working), so this is often unappealing to both patients and providers. But if a patient has been stuck in a high-acuity setting for weeks with little sign of improvement, or has suffered repeated episodes of deterioration with resultant morbidity (numerous bouncebacks to the ICU and reintubation, for instance), it’s probably time; the string connecting them to the ventilator is never going to stretch very far unless they can break out of their current loop.

Common interventions of slimmer benefit include:

  1. Mobilization: If you can pull it off, ambulation can dramatically improve respiratory mechanics, decrease atelectasis, mobilize secretions, attenuate DVT risk, and more. But it’s extremely difficult to do. Give it your best go with a large team, and consider specialized equipment — such as “tilt-table” beds that allow a full upright, load-bearing position with zero patient effort.
  2. Pulmonary toilet: You can thump on their chest, encourage incentive spirometry and flutter valves, and even bronch these patients. In my experience, unless there is a true pneumonia in the mix, it rarely amounts to much. Bronchoscopy can also complicate the picture with decruitment and sedation, and in general is best avoided.
  3. COPD treatments: Unless the patient truly has COPD, nebulizers and steroids don’t seem to accomplish much. The latter in particular can also worsen fluid retention, glucose control, and myopathy, all of which are an extra burden you don’t need.

From the experts

These patients are so tricky that there’s no right answer — and when we don’t have a right answer, we should at least have lots of them. I asked a few sources in the FOAM community to share their own approaches.

Greg Martin, MD, thinks one of the fundamental problems is that:

… we see these patients with OSA/OHS in the exact scenario you describe, but we don’t understand what causes an “exacerbation of OSA/OHS” like we do for CHF and COPD.

I found that to be profound. Why isn’t “exacerbation of OHS/OSA” a concept just as familiar to us as an exacerbation of COPD? It might not get us 100% to understanding these patients, but it would at least avoid the multitude of misdiagnoses that accumulate while providers look for a label that fits.

He also reminds us that, despite some of these patients appearing reasonably functional prior to admission, that’s an illusion. SARFO didn’t come out of nowhere.

… all of these patients have had OSA/OHS for years at the point they have the episode you describe with refractory respiratory failure.  It’s important to recognize this case as one in a series of earlier, perhaps more minor, episodes where the patient continues to struggle with weight and deconditioning and multi-morbidity, and ultimately ends up with refractory respiratory failure. This is important because (1) we need to more aggressively manage these patients earlier in their course, and (2) this parallels every other disease we manage. It’s really not so different from CHF, COPD, cancer, CKD, etc where the disease progresses and ultimately the patient develops refractory symptoms and needs chronic inotropic therapy, chronic oxygen, chronic ventilation, salvage chemo, dialysis, etc.

Josh Farkas, MD underscores the value of APRV, and reminds us that…

… these folks can get protein-calorie malnourished, which is really frustrating and ironic (despite being massively overweight, they can rapidly catabolize protein). So providing some reasonable amount of nutrition is important. Exactly how much to use is super controversial.

This is particularly common in the “limbo” period that many of these patients spend on BiPap, which can last days with few opportunities for nutrition. Consider meal-time breaks, or squeezing in a small-bore NG tube.

Finally, Jeremy Amayo, PA-C reminds us that while they’re necessarily overweight, it’s not always the most super-obese patients that exhibit these features. He also finds that HFNC can be fruitful in the right patients, using low FiO2s and high flows; he also thinks diuresis can sometimes be facilitated by judicious use of acetazolamide. And he particularly emphasizes that some these folks are hiding true right heart failure from cor pulmonale:

… [they] sometimes they can be preload dependent even if they are volume overloaded. We’ve had some success with inotropic support either with silly-dose dopamine or legit milrinone/dobutamine while we try to get fluid off…. We have found that some of them get worsening renal failure with diuresis despite still appearing grossly overloaded. We think they may depend on that preload to perfuse the kidneys…

He recalls patients with this sort of occult cardio-renal phenomenon who remained stubbornly intractable until placed on inotrope drips, after which they briskly diuresed, with a corresponding respiratory improvement.

That’s all we’ve got

That about does it for our paltry insights. Please feel free to share your own methods and approaches for managing these challenging patients.

Special thanks to our external peer reviewers:

Greg Martin, MD, MSc
Professor at Emory University
Twitter: @gsmartinmd

Josh Farkas, MD
Website: Pulmcrit
Twitter: @Pulmcrit

Jeremy Amayo, PA-C
Emory University School of Medicine
Website: Pulmcast
Twitter: @MedEd_PA

2 Replies to “SARFO: The syndrome of acute respiratory failure in obesity”

  1. Great post!

    A word about Marik’s “retracted” paper – Marik was accused of “self-plagarism” due to similarity between this paper and a previous one that he wrote. He describes this experience in a rather amusing article here: https://onlinelibrary.wiley.com/doi/full/10.1111/eci.12473

    Personally I find this to be a bogus claim, as the practice of reusing pieces of text is *extremely* common among authors who write multiple review articles on the same topic. Regardless of whether you believe in “self-plagarism,” there are no questions regarding the validity of the article so I wouldn’t hesitate to cite it.

    Overall I think Marik deserves credit for shining light on a common problem which is generally ignored in the literature. As do you, for drawing more attention to this major problem (albeit one which isn’t flashy).

  2. Thanks Josh, interesting addition. I personally didn’t find the two pieces to be inappropriately similar; there is obviously shared content, but it’s not like he changed the title and sent it to another journal. There’s plenty of novel material in each; one is obviously more of a retrospective series and the other a narrative review on the topic.

Leave a Reply

Your email address will not be published. Required fields are marked *