Second Opinions and the Rise of REPAP: Part III

As the REPAP paper has just been published, now is a good time to finish up our earlier series on this innovative protocol. Briefly to summarize key points from the earlier works: 1) second opinion research is rare to non-existent in the sleep literature; 2) the use of retitrations to reverse CPAP failure is only slightly more evident in research publications; 3) the nature of CPAP failure is mostly defined by non-technological factors; whereas the notion of a technological solution for CPAP rejection is rarely considered in the literature or sleep medicine practice; 4) the key objective finding in second opinion patients is known as expiratory pressure intolerance (EPI), the problematic and uncomfortable sensation of breathing out against fixed (CPAP) pressure coming in; 5) sleep medicine professionals in general and reviewers of sleep research papers in particular show a lack of awareness or simply do not consider the potential for advanced PAP algorithms in newer pressurized air technology to solve the problem of EPI.

In working with the reviewers who critiqued our paper for Respiratory Care, we found much greater receptivity to discussions on all of the above points, and through the process of revise and resubmission we were able to make the recommended changes and finalize the manuscript into a publishable form. However, an early sticking had to be resolved, because we once again realized that reviewers here and before were concerned the REPAP protocol was promoted too rapidly for too many patients. To clarify this specific complaint, in our final submission we included the following information: 

We strictly follow AASM guidelines and conventional wisdom in our patient management approach. We developed an intensive and extensive follow-up program involving patient encounters in the clinic as well as through telephone and email communications. At no point will a patient end up in the REPAP protocol unless and until he or she has gone through all other standard steps in the management of PAP therapy patients. We vigorously attempt to solve all mask issues in the clinical environment using extensive mask fittings and a highly accessible, patient-friendly mask loaner program. All mask problem-solving is coupled either with assessment of leak data from objective downloads or the use of our own in-lab technology to directly monitor leak while the patient tries a new mask. Patients are repeatedly encouraged to contact our center for any issues, complaints, or concerns about using PAP therapy. Outcomes data are used to determine level of responsiveness to treatment, but we also place great emphasis on patients’ subjective reports about their progress and their perspective on whether the treatment response has achieved an optimal level to the best of their ability to interpret this standard. Before a patient would even be considered for the REPAP protocol, they would have experienced multiple encounters with the sleep physician, the Results Coordinator, the Follow-up Coordinator, the Clinical Manager, or one or more sleep technologists. These encounters, as above occur in person, by phone or by email. Some of the encounters are indirect through our collaboration with DME companies.

Then, to follow-up our introduction in the revised submission, we provided interesting background information to indicate why we were surprised about the eventual development of a retitration protocol to rescue CPAP failure patients: 

Specific to the topic of the REPAP protocol, we would never have imagined the need for a retitrations approach to patient care in 2003 when we opened our private, community-based sleep center. Instead, based on our earlier research in the field, which eventually led to the formulation of the Sleep Dynamic Therapy™ model (tested on trauma survivors from the Cerro Grand Fire in Los Alamos, NM, 2000 Journal of Clinical Psychiatry, 2002),(1) we had already begun using cognitive-imagery distraction techniques and sleep-related emotional processing strategies in our PAP patients to ease their adaptation to pressurized airflow. Despite the clear value of these techniques, which are used every single day at our sleep center, we were perplexed during the period of 2003 to 2005 to learn that the physiological barrier of expiratory pressure intolerance (EPI) was frequently triggered when attempting to titrate respiratory effort-related arousals (RERAs: discrete breathing events of UARS). Soon thereafter, we recognized this barrier was a crucial element in CPAP failure. And, as much as we were able to lessen subjective EPI in some patients with imagery distraction techniques and emotion-focused therapies, objective EPI often re-emerged once the patient was asleep.

In sum, we have gone down the path of focusing on the need for a physiological intervention to solve this particular side-effect that may occur in many (but certainly not all) patients who require PAP therapy. We have described an extraordinarily high rate of this problem set in psychiatric patients, mostly due to their anxiety in trying to use pressurized air. Overall, our psychiatric patients receive a great deal of standard care as well as specialized care before they are scheduled for retitrations in the REPAP protocol.

Several additional points from the review process were quite illuminating and will be discussed in the remainder of this post.

The issue of CPAP failure as we described in the earlier post can be noted in different ways. Thus, when the reviewer asked about patients who were still using CPAP but responding poorly, the question arose as to why we did not include them in the protocol. We responded: 

We began our chart review and analysis with nearly 1500 2nd opinion patients seen in a 7-year period (2006 – 2013) and eventually found 900+ who met various criteria for CPAP failure irrespective of current use. Then, we further explained the same information posted earlier and why we wanted to include only a “pure” sample of current non-users and then we followed with four rationales for approaching the problem in this manner. First, there is no generally accepted definition of CPAP failure in the literature or from the insurance carriers who greatly influence this area of care. Second, as we looked over current CPAP users suffering from low levels of use, we perceived distinctive features between these users with low hours and a poor response and those not using CPAP at all. Third, we intend to examine low hours/poor response patients in another manuscript on the REPAP protocol, because these patients require different approaches and pathways compared to CPAP rejectors. Last, by focusing on patients who have unequivocally rejected PAP therapy, we believe we have removed any doubt about categorizing them as CPAP failure. 

Another very clinically relevant point was offered by one of the reviewers to suggest that the intensive model of care we were pursuing with the REPAP protocol “was” or “should be” standard of care at the majority of sleep centers. The reviewer then asked the pivotal question of what is different about the REPAP protocol other than using precise pressure changes during the retitration? To which, we replied: 

If we are correctly interpreting the reviewer’s comment/question, it seems the REPAP protocol is something that would or should already be in use at all accredited sleep centers. Our struggle with this specific interpretation is that most of the 2nd opinion patients we treated had originally been diagnosed and treated at accredited sleep centers (>90% in New Mexico). As the best example, the vast majority of patients were never offered a trial of bilevel devices not to mention ABPAP or ASV after reporting difficulty or failure with CPAP. Nearly all these patients reported no discussions occurred on the problems of subjective pressure intolerance at any point in their care, and greater than 90% reported that the concept of residual breathing events such as RERAs were never broached. Many were never offered new mask choices from the sleep centers or their DMEs. While we are reporting information alleged by the patients, we hear these comments repeatedly among 2nd opinion patients, expressed succinctly by the phrase: ‘I was not informed of these other options.’

Regarding the rounding of the airflow curve, we published a paper describing exactly this protocol in our clinical practice whereby we titrate in small increments, and we cite Condos et al 1994 paper (2) that to our knowledge first described the clinical relevance of rounding the airflow curve. Their work showed rounding the airflow curve led to further reductions in daytime sleepiness in their research sample. To reiterate, the delicate nature of raising and lowering pressures also is required to prevent expiratory pressure intolerance, which appeared as a central factor in preventing patients from using CPAP devices and appears to have contributed to their CPAP failure.

Another salient comment asked whether or not patients would constantly need to change their pressures modes after each titration, necessitating far too much transitioning from one device to another, to which we replied: 

There is the possibility that a new device would be needed after every titration, but we conduct our titrations similar to Kuzniar et al’s protocol (2007)(3) to avoid this issue from arising in the majority of cases. Using the multi-modal protocol, we progress through a sequence that might include CPAP, APAP, CPAP w/EPR, APAP w/EPR, BPAP, ABPAP on the very first titration as indicated in any given patient. Thus, the most common scenario is that a patient fails the other modes by persistent RERAs or EPI or both and initiates treatment with ABPAP. In the smaller proportion of cases that ultimately need ASV, they usually are switched to this mode after having used ABPAP and then returning for a retitration due to poor outcomes and low use.

There was a concern about whether patients had the opportunity to try less sophisticated modes of PAP prior to advancing to ABPAP or ASV, despite the fact that all these patients were currently failing modes, most typically CPAP. Nonetheless, this point is a valid one, and we responded as follows:

We followed AASM guidelines for titrations in all patients. Thus, all patients always attempted CPAP at some point in their care at our center. The two most common places where CPAP was re-attempted occurred during the desensitization period of the retitration or in the early period of the retitration itself. Once objective findings of residual breathing events or expiratory pressure intolerance or both could not be resolved, then a switch to a dual pressure device would be initiated. APAP was used rarely because it suffers the same problem as fixed CPAP, since APAP is fixed CPAP at any given point in time and space. To resolve expiratory pressure intolerance, we discovered dual pressures provide much less stress on the patient while preventing the emergence of RERA breathing events. 

During the time of our work on the revisions, a paper was published raising questions about the value of full face masks (FFM), and this issue may have prompted the reviewer to ask why we showed the tendency for even greater use of FFM. In responding to this comment, we took the opportunity to also note our surprisingly greater use of FFM in psychiatric patients who would have presumably suffered a higher likelihood for claustrophobic responses. Our response was as follows:

Regarding FFM, we find the recent evidence weighing against the full face mask to be confounded, because nearly all this research involves patients using fixed CPAP. In other words, regardless of mask type, when pressures are raised with a CPAP device, there is likely to be a triggering or exacerbation of expiratory pressure intolerance, which in practice should lead to lower levels of adherence. In contrast, this mask issue did not carry over in our sample when they used the combination of FFM with dual pressure auto-adjusting devices, perhaps because these modes allow for raising inspiratory pressures to eliminate RERAs while maintaining lower expiratory pressures to prevent or minimize expiratory pressure intolerance.

In a related issue, it might be anticipated that our predominant sample of psychiatric patients would suffer greater claustrophobia or claustrophobic-like responses to FFM. Ironically, we find the exact opposite. Many of these patients, who need considerable coaxing to try FFM, reported FFM relieves their anxieties, because they know and feel they can breathe through their mouths if a panicky feeling emerges. Since acute mouth opening and mouth breathing are classic responses in claustrophobic patients trying to gain more air, these patients realized FFM is actually a more natural fit, especially compared to the use of a chinstrap, which in many cases adds to the restrictive nature of feeling foreign objects on one’s face. The majority of psychiatric patients do not start with FFMs for the obvious reason it looks like a guaranteed trigger for claustrophobia, but as patients learn more about and adapt to PAP therapy, we have been surprised how many eventually chose FFM. Remarkably, we saw the combo of FFM and chinstrap in 39 patients.

A great question was raised on how we were able to precisely sort out the problem of expiratory pressure intolerance, to which we replied:

In our 2013 paper,(4) we published a graph showing how we observe objective EPI in the sleep lab, which may arise with any mode of PAP but especially with fixed pressure devices such as CPAP or APAP; even BPAP is technically fixed at given points in time and space for inspiratory and expiratory settings. The irregularities we see on the expiratory limb of the airflow curve have been the most consistent objective indicator of EPI in our clinical practice. But, as part of our protocol, the technologist is required to insure other factors such as nasal congestion, mask leak, mask flutter, mouth breathing, mouth sputtering, and dropped jaw displacement are not causing an artefact in the tracing. Switching to ABPAP or ASV and then switching back to standard PAP modes has also been a reliable indicator of objective EPI. Subjective EPI requires a much more simplistic approach, because we rely on the patient to communicate specific discomfort when breathing out against pressurized air coming in. During the desensitization period, patients are instructed on imagery distraction techniques to avoid focusing on breathing as well as receiving a “control of breathing” educational session to prepare them to not try to synchronize (usually over-control) their breathing with the PAP device. Together, these two instructions, delivered to all patients, lead to some diminishment of subjective EPI in many patients, yet once asleep many individuals still demonstrate objective EPI.”

In a related technology question, a reviewer noted that “better sleep” the morning after and patient reports of “easier adaptability” were crucial factors in the statistical analysis regarding what predicted the patient’s likelihood of transitioning from non-user to user. This point came up several times, and we felt obliged to explain it from the point of view of what makes for better sleep and easier adaptability:

These questions are crucial to understanding our manuscript and teasing apart the clinically relevant practice points. That statistical analysis demonstrated consistent findings of better sleep ratings in the morning and greater capacity to use PAP raises the larger question of what actually caused the better sleep and what caused the greater sense of adaptability. We believe it is of clinical interest that our patients consistently reported sleeping better with the advanced PAP devices (ABPAP, ASV) and also found these modes easier to use than standard CPAP. Moreover, these same patients reported greater than 70% of adverse factors attributed to CPAP failure were due to related technological barriers (e.g. pressure intolerance, unable to adapt to PAP, mask issues, and PAP side-effects), thus further aligning with the association between use of advanced technology and better sleep or easier adaptability. That Users also show a significantly lower AHI and a trend for a lower RDI also aligns with the association between use and advanced technology.

One of the final concerns for one reviewer was how much the role of education might have played in these patients ultimately achieving user status, especially given all the complexities in attempting to help PAP patients to overcome barriers and gain compliance. Our response was as follows:

We could not agree more with the reviewer on the complexities involved in CPAP rejection and non-adherence, albeit recent data suggests education may not play as large a role as once thought, and certainly follow-up efforts depend a great deal on who is providing the follow-up and what issues are being addressed. As far as data downloads, it remains concerning that so many sleep centers are apparently ignoring the AASM mandated obligation to address the problem of residual RERAs. Unfortunately, RERAs are only indirectly measured if at all on many device data downloads. Thus, RERA-induced sleep fragmentation is a likely cause of poor subjective outcomes that may go unnoticed despite PAP use, and in our experience downloads that show residual “flattening” are very helpful in our decision-making to consider a retitration in the sleep lab.

Further, a key distinction in our theory about rejection and non-adherence is that “experience” appears to be significantly more important than “education” in resolving a great many adaptation issues with PAP therapy. As our first major step in developing an experiential model, we formally organized the PAP-NAP procedure to give reluctant and otherwise complex patients the opportunity to “test-drive” PAP therapy. As we fine-tuned this experiential model, we realized how crucial retitrations were for some (but not all) of these complex patients.

As you can see, reviewers often ask very probing and challenging questions to insure that a research paper is not only clear but arguably more importantly that all dimensions of the research including all the strengths and weaknesses are discussed at length to insure the manuscript does not attempt to claim more than what was achieved in the study. The process of revise and resubmission therefore is a very intense and rewarding experience.


  1. Krakow BJ, Melendrez DC, Johnston LG, Clark JO, Santana EM, Warner TD, Hollifield MA, Schrader R, Sisley BN, Lee SA. Sleep Dynamic Therapy for Cerro Grande Fire evacuees with posttraumatic stress symptoms: a preliminary report. J Clin Psychiatry. 2002 Aug;63(8):673-84.
  2. Condos R, Norman RG, Krishnasamy I, Peduzzi N, Goldring RM,Rapoport DM. Flow limitation as a noninvasive assessment of residual upper-airway resistance during continuous positive airway pressure therapy of obstructive sleep apnea. Am J Respir Crit Care Med 1994;150(2):475-480.
  3. Kuzniar TJ, Golbin JM, Morgenthaler TI. Moving beyond empiric continuous positive airway pressure (CPAP) trials for central sleep apnea: a multi-modality titration study. Sleep Breath. 2007 Dec;11(4):259-66.
  4. Krakow B, Ulibarri VA, Romero EA, Thomas RJ, McIver ND. Adaptive servo-ventilation therapy in a case series of patients with comorbid insomnia and sleep apnea. J Sleep Disord Treat Care 2013. doi: 10.4172/2325-9639.1000107.

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.