PAP vs Surgery: Which is Better?

Recently, I had the pleasure of listening to a podcast recorded by my friend and colleague, Dr. Steven Park, who is also the author of an interesting sleep disorders’ book, Sleep Interrupted. I have known Dr. Park for several years, and he is a very well respected Ear, Nose & Throat Surgeon who has developed a tremendous interest in the management of sleep disorders. Unlike most sleep physicians, he is acutely aware of the UARS component of sleep-disordered breathing, and as a result, he is very concerned that the UARS aspect in many patients goes untreated whether they use PAP, OAT or undergo surgical interventions.

In Dr. Park’s talk, he lays out a very interesting argument for the superiority of surgery over CPAP. His theory boils down to simple math. Far too many patients do not use their CPAP devices, and among those who do use the device it is not uncommon for them to remove the mask halfway through the night, thus total use is only 3 to 5 hours of sleep per night. As he astutely points out, the most severe sleep breathing events occur in the finally third of the night when many OSA/UARS patients have already removed the device, which adds to the overall poor response to many CPAP users.

Mathematically, he then compares two hypothetical patients, both starting out with say 30 breathing events per hour. For the sake of this discussion we’ll just say AHI or RDI = 30. Post-surgery, the intervention decreased the number of events to 15/hour or a 50% reduction in our surgery patient. Now, if we imagine this patient sleeping 7 hours per night, then their AHI/RDI dropped from 210 events down to 105 events per night, which for one week would equate to 735 events per week instead of 1470.

I trust you see where this tally is headed. If a CPAP patient only wears the device for 3.5 hours per night (a not uncommon experience) and perhaps once per week misses an entire night without any use, then his total number of breathing events for one night or one week likely equals or exceeds the number of events in the surgery-treated patient. To calculate the numbers during CPAP, we start with 3.5 hours per night with less than 5 breathing events per hour; let’s make that tally 10 events total in these 3.5 hours. Then, in the second half of the night without the mask, the events would equal 3.5 x 30 or 105, but since breathing is worse in the final third of the night, the number is probably higher, so let’s say 120 events. In sum, this person is averaging 130 (10 + 120) breathing events per night and therefore already exceeds the 105 events per night of the surgery patient as well as the weekly count because with just the first six days of the week, the total is at least 780 (130 x 6) events compared to the seven day count of 735 described above. Throw in the additional 210 events for the skipped night of PAP (a very common experience among PAP users), and now we are up to 990 breathing events per week, a substantially worse value of 35% more events in the CPAP patient compared to the surgery patient.

Do these calculations support the use of surgery in more OSA/UARS patients? How completely do these values paint a picture of the average OSA/UARS patient who is failing PAP? And, from a cost-effectiveness analysis is the surgery option ultimately better than the PAP option given the high rate of PAP failure? 

To analyze this information we must first consider how to properly evaluate CPAP failure. At our center, at any given time, 90% of patients who fill a prescription for PAP therapy are using the device. Notice the term “using” instead of the term “compliance,” the latter a technical metric adopted by insurance carriers to determine whether or not coverage will be provided. Among these 90% using the device, we find 60% are compliant but another 25% appear to be on their way to meeting compliance but require additional time to attain this goal. Thus, 10% are not using at all, and among users there are about 15% where the use is low enough to question whether they would ever become compliant. 

All told then roughly 75% of patients are likely to use the device to gain sufficient benefits. Moreover, at our center it is very common to see breathing event indexes on data downloads to reach 0 for the AHI. And, as an additional advantage, the majority of our patients are using PAP therapy every night for 5 to 6 hour per night. If you have read my other posts on advanced auto-adjusting technologies, you know how the use of ASV and ABPAP are prime reasons we can lower the AHI so well and how patients adapt more readily to pressurized airflow. Thus, our data on our specific cohort of patients would be much more competitive and routinely superior to the surgery intervention patients. Furthermore, because patients are sleeping so well with ASV or ABPAP, they may no longer need to sleep 7 to 8 hours per night; they may discover 6 hours per night is more than satisfactory, therefore they are not running up more untreated breathing events in the morning hours without the mask, because they only sleep every hour of sleep with the mask.

Then again, we could not argue against Dr. Park’s point about the other roughly one-quarter of our patients who are clearly failing PAP by not using at all or using too little. Some may eventually return to PAP; we see this scenario quite a bit over time, especially among patients who received modest benefits from surgery and now are ready to give PAP a second chance. But, a sizeable number of patients are happy to move on to oral appliance therapy (OAT), and in these cases the numbers are also very competitive with surgery, because a large proportion of these patients may wear the device every night, all night long; and, the OAT often produces reductions in the AHI/RDI between 50 to 80% from baseline.

So, is surgery just as good if not better than CPAP? Although I respect Dr. Park’s positon on the subject, I believe there are two primary factors and one other potential factor that more strongly favor PAP therapy over surgery. The first and most pressing is the cost and adverse side-effects that arise from surgery. The second is the recurring theme of UARS and how surgery often does not address this problem. And, the third is the speculative theory based on a bit of evidence that suggests when you use a CPAP device in the first half of the night, it may produce carry-forward effects in which your breathing is better than it would have been had you not used CPAP at all.

ENT surgery may not be that expensive if we are calculating costs for a nasal repair, but it becomes increasingly expensive as more complexity is added into the intervention. One could argue that the cost may be largely covered by insurance, but nevertheless, the cost far exceeds that of a CPAP device by a factor of five- to ten-fold if not more. However, more important than the money is that surgery always involves a healing and recovery period, where people suffer from pain or other side effects. They miss work or they are otherwise incapacitated for a period of time. In the worst case scenario, some patients die following ear, nose, and throat surgery due to bleeding or breathing complications. In my career as a sleep physician, I know of at least 3 cases of adults with very thick necks, very large tongues and very large tonsils who could never tolerate CPAP and then died in the hospital within 24 hours of surgery. As confided to me by one ENT surgeon, the problem appears to be that the individual is in so much agony from the pain that opiates are administered in these patients, which depresses respiration and ultimately leads to depressed oxygenation, cardiac arrhythmias, and sudden death.

Another side effect infrequently mentioned by surgeons is the well-documented fact that oral airway surgery, particularly interventions involving the soft palate or uvula in the back of throat, routinely makes it more difficult for patients to use PAP therapy. These same patients usually received no benefits from surgery, a common result of the UPPP procedure, so they are immediately sent back to the sleep center to try CPAP again. Then, depending upon the sleep center practices, these patients can make a circuit going through multiple sleep centers seeking relief and yet never being introduced to a more sophisticated PAP device, such as ABPAP or ASV, which in our experience are extremely efficacious in helping this particular cohort of patients. In fact, when someone completes an intake at our center and marks having undergone the UPPP procedure, we immediately inform them CPAP is a waste of their time, effort and money, because it rarely works in someone post-UPPP. We recommend the advanced technology auto-adjusting dual pressure devices (ABPAP, ASV) and discuss the rationale behind their greater success rates.

The second major area of concern is the UARS phenomenon. These breathing events, subtle though they appear to some untrained sleep professionals, must be aggressively treated in order to attempt to achieve an optimal response. Yet, in most surgical intervention studies, UARS is rarely addressed. You will notice in Dr. Park’s podcast, to his credit, how he points out this problem among surgeons who report a decrease in the AHI post-surgery yet fail to appreciate the persistently elevated RDI, that is, the residual UARS breathing events beyond the reductions of the apneas and hypopneas. In these cases, unless the surgeon categorically declares the surgical procedure will address UARS, then the results will not prove as beneficial as anticipated. This point brings up a sticky situation, because Dr. Park states or implies in his podcast that more aggressive surgical interventions may further improve UARS, thus he is proposing a theory in parallel to our use of advanced technology. Taken together, both of us are offering ways to more aggressively treat UARS residual events.

Not being a surgeon and not knowing the exact details of how Dr. Park is measuring UARS in his post-surgical patients, we will need to leave this point for future debate or discussion, but based purely on my respect for my colleague, I have to believe he must be seeing some very obvious objective and subjective results that have persuaded him to approach UARS surgically in ways other surgeries may not appreciate. 

Regarding the third point on CPAP effects post-mask removal, there has been at least one study published in the scientific literature that suggests using the device for a few hours yields synergistic effects due to a carry-forward effect on overall breathing. I was unable to find the citation for this post, but it is an interesting enough theory that we should expect more research in this area as we see more comparative research between CPAP vs surgery patients.

A final area of more obvious compromise between the competing therapies are the surgeries that do not seek to cure sleep apnea, but rather are performed to facilitate a patient’s eventual use of PAP therapies. These types are often highly successful and include surgeries in the nasal passages to correct a deviated septum or shrink the turbinate tissues; these interventions are probably the most common and useful ones to undergo. Tonsil removal surgery in adults is not common but may prove beneficial, although it is often very painful unless it involves newer surgical techniques. Last, sinus surgeries may prove extremely beneficial if they reduce chronic congestion problems, which almost always interfere with efforts to use PAP therapy.

Summing up, the debate is not over on this very important area in sleep medicine and will likely continue for many years. The good news for patients is this type of competition among different physicians often leads to newer and better interventions, either in PAP therapies or surgeries or in some whole new area of treatment invented by someone motivated by the notable limitations observed in current therapeutic paradigms.

Barry Krakow MD


Dr Krakow’s 27 years of sleep research have focused on the complex relationship between physiological and psychological sleep disorders. Dr Krakow currently operates private sleep medical center, Maimonides Sleep Arts & Sciences, Ltd., and serves as Classic SleepCare’s paid Medical Director.

Leave a comment

Comments have to be approved before showing up.