I just returned from Sao Paulo to give lectures at the 6th International Symposium on Snoring and Sleep Apnea at the Hospital Israelita Albert Einstein. Surprisingly, the organizing team keeps asking me to return and are not sick of hearing from me yet. Brazil is home to some of the leading sleep surgeons and sleep researchers in the world, individuals who are innovators in surgical techniques as well as alternative treatments for snoring and obstructive sleep apnea. Although my Brazilian Portuguese remains in its early stages, this has not limited the sharing of ideas with wonderful colleagues and enjoyment of outstanding Brazilian food (churrasco and feijoada are highly recommended). The most common topics on which I lecture are: (1) patient evaluation and the selection of procedures, (2) surgical outcomes, and (3) new treatment options for snoring and sleep apnea. For these topics, the question always arises as to how to evaluate surgical outcomes. This is much more complicated than it appears.
Snoring Is in the Ear of the Beholder–or Is It?
For snoring, most studies of procedures have focused on subjective reports from bed partners, who grade snoring on a 0-10 scale before and after treatment. It turns out that this is not an ideal evaluation, as studies that have also used objective sound measurements (usually with microphones and sound analysis) have shown that subjective reports show much greater “benefit” than the objective sound measurements do. While many patients achieve substantial improvement or resolution with snoring procedures, some studies suggest that you can use a single procedure for everyone that snores, with great success. In fact, that is not true. Because different patients have snoring for very different reasons, careful evaluation and thorough discussion of options with each patient is essential to determining the right treatment. Once a patient undergoes a procedure, I want to know whether and how much of a reduction in snoring occurs and, in what may be the most important outcome, whether that allows others to sleep peacefully without being disturbed by the snoring.
Sleep Apnea Treatment Outcomes: Now It Really Gets Interesting
For obstructive sleep apnea, the uncertainty surrounding outcome evaluation is not confined to surgery. Positive airway pressure therapy (with one type known as CPAP) works by delivering air pressure to balloon open the throat enough to prevent the collapse that occurs otherwise. Different patients need different pressures, so there is often what is called a titration study, where the appropriate pressure is determined, based on what normalizes breathing patterns (typically to an apnea-hypopnea index, or AHI, below 5 events/hour). Everyone assumes that this set pressure is perfect and that patients have normal breathing when using CPAP. It turns out that this is not true for most patients. In a study of CPAP efficacy of 70 patients from 3 highly-respected centers who were using CPAP consistently and were happy with it (basically, an ideal population), three-fourths actually had obstructive sleep apnea (AHI > 5), and 20% of the total had moderate to severe obstructive sleep apnea (AHI > 15). There are some other, smaller studies with similar findings. In addition to this potential issue of residual sleep apnea while on CPAP, it is extremely unclear how much CPAP use is sufficient. There is some evidence suggesting that at least 4 hours per night on 5 nights per week provides improvement. However, patients vary dramatically (some get benefit at that level and some do not), we do know that more use is better, and more studies are needed.
How about Sleep Apnea Surgery Outcomes?
These questions are also extremely important for surgery, and answers are also very limited. The traditional measure of surgery outcomes has focused on comparing breathing patterns (from sleep studies) before and after surgery. This is critical but too limiting; we treat patients, not numbers. Patients with sleep apnea seek treatment not only because of their sleep study results but also because they often have symptoms such as sleepiness, decreases in quality of life, or other problems such as difficulty with concentration or attention. Outcomes must be measured by clinical measures (including symptoms) as well as their breathing patterns from sleep studies. Sleep study results are important for many reason and the important thing is that they have been reported consistently in sleep surgery studies. The traditional threshold used to define surgical “response” or “success” has included two criteria: reducing the number of breathing events per hour (AHI) by at least 50% and having no more than a certain level of breathing events per hour (AHI < 20 or 15 in most studies). Some have criticized this as irresponsible, as it still leaves individuals with “dangerous” or “deadly” disease.
Not only are these words exaggerations, but they also ignore the reality that the comparison for surgery should not be breathing patterns while on CPAP in the idealized setting of that single titration study night. First, we now know that CPAP, even when worn, leaves many patients with some degree of obstructive sleep apnea (although usually much improved), so requiring surgery to resolve all sleep apnea actually holds it to a higher standard than CPAP. Second, patients undergoing surgery have almost always demonstrated that they do not respond or do not tolerate CPAP, making the real comparison to surgery no treatment at all. The goal of surgery is obviously the resolution of all sleep apnea permanently, but the reality is that it is reasonable to accept a substantial improvement in sleep apnea, which is the basis for the traditional definition. I personally favor a 50% reduction in the AHI to a level no worse than 15 events/hour, especially if the patient has a good clinical response; this would leave someone with no worse than mild obstructive sleep apnea, where the health risks of sleep apnea are much less concerning than for moderate to severe disease.
A Framework for Reporting Surgery Trial Results
Last year, a group of us developed a framework for the reporting of results for sleep apnea surgery trials. We discussed a number of topics that were occasionally missing from studies, including reporting preoperative patient characteristics and their association with results (to show which patients did better or worse), using appropriate statistical tests, and presenting changes in sleep study results as well as clinical outcomes. There is not enough evidence to determine clearly how we should define response to sleep apnea surgery; our hope was that future studies could provide more guidance.
I had taken a similar approach in a published study of combined palate surgery and tongue region surgery (genioglossus advancement, with or without hyoid suspension) surgery that was supported by the National Institutes of Health. In this study, certain patients (body mass index below 32 kg/m2) were more likely to respond to surgery. Importantly, those patients who had a greater improvement in their breathing patterns (based on their postoperative sleep studies) also had greater improvements in quality of life and a blood test called C-reactive protein that may reflect the risks of hypertension, stroke, and heart attack seen in moderate to severe obstructive sleep apnea. The purpose of the analyses in these studies and this framework is to provide more-detailed information to use in selecting procedures and counseling patients about the potential surgical outcomes. I use these results all the time in making decisions with my patients and use the information in my lectures to other surgeons.
What I Do in My Own Practice
In my own practice, I encourage my patients to be “difficult”, telling me if they are not feeling improved and not just the positives because they think that is what I want to hear, in addition to postoperative sleep studies when treating sleep apnea. Careful, critical evaluation of your own results is the key to any medical treatment, and the surgical treatment of snoring and sleep apnea is no different. Although my goal is to have as few patients as possible who do not respond to surgery, I speak and write often about how surgical outcomes are far from perfect. I learned early on not to push away the patients who do not respond to treatment but welcome them because they often need more treatment. They also provide the most insight about why a given procedures may not work well in certain patients, improving my own practice and allowing me to share what I learn with patients and colleagues.