Issue Navigator

Volume 11 No. 08
Earn CME
Accepted Papers

Special Articles

Dealing with a Paradigm Shift

Allan I. Pack, MBChB, PhD
Center for Sleep and Circadian Neurobiology, University of Pennsylvania Perelman School of Medicine, Philadelphia, PA


Recent changes in policies by insurance companies with respect to mandating home sleep apnea testing rather than in-laboratory studies have a large impact on the financial viability of clinical sleep centers. Coping with this disruptive change requires forward planning. First, it is important to be well positioned with respect to facilities so that these can be quickly downsized to control costs. There is also a need to develop, in advance, an accredited home sleep study program so that centers can respond to the rapidly changing environment. Following the change there is a need to control costs by rapidly downsizing the technology workforce. Technologists can be retrained for other essential roles. Centralizing the precertification process with knowledgeable, well-trained staff and a robust auditing process is an essential component. The approach taken at the University of Pennsylvania to this change is described as is how one can ensure continued financial viability of a comprehensive sleep center program in a major academic medical center.


Pack AI. Dealing with a paradigm shift. J Clin Sleep Med 2015;11(8):925–929.

It is without doubt that the new discipline of sleep medicine in the United States is challenged. This has been documented in recent editorials in this journal.1,2 There has been a drop in applications for sleep medicine fellowship programs leading to concerns about whether there will be a sufficient workforce for sleep medicine in the future.1 There has been closure of many sleep centers due to financial reasons. The most noticeable of these were a Texas company with 58 sleep sites nationally declaring chapter 7 bankruptcy in 2011, and the 2012 chapter 7 bankruptcy of a prominent sleep company in Massachusetts that provided vertically integrated services in numerous owned centers and serviced a broad spectrum of academic and community hospitals.2 Sleep medicine was accepted by the Accreditation Council for Graduate Medical Education (ACGME) as an independent discipline in 2005. I was on the group the ACGME established to review the application for sleep medicine to become an independent specialty. It was not a difficult decision. There was a major body of scientific and clinical knowledge, a large number of conditions with different diagnostic approaches and therapies. The field relied, however, on polysomnography that was largely for the diagnosis of sleep apnea. This made the field vulnerable to technological advancements and policy changes that diminished utilization of this procedure. This has weakened the financial underpinning of the field. In this special article, the impact of the paradigm shift on the sleep medicine program at the University of Pennsylvania, how to plan for it and deal with it, and briefly about the future are all discussed. There is optimism about the future for this new field and a rationale is provided for this viewpoint.


I first provide a brief history of Penn Sleep to put changes we faced in context. In 1991, the Dean of the School of Medicine—Dr. William Kelley—established the first medical school-wide academic research sleep center in the USA—the Center for Sleep and Respiratory Neurobiology (now called the Center for Sleep and Circadian Neurobiology). The 1990s were a time of great development of our academic and research programs with key recruitments such as Amita Sehgal and Terri Weaver, and establishment of a sleep center at the Children's Hospital of Philadelphia (Founding Director, Dr. Raanan Arens). While our academic programs and research support grew rapidly, our clinical program did not. Our clinical faculty (Drs. Allan Pack, Richard Schwab, and Sigrid Veasey) were all part of the Pulmonary Division. Although the faculty in the Pulmonary, Allergy and Critical Care Division doubled in size from 1991 to 2001, there was no growth in sleep faculty. Our sleep lab had 4 beds open 5 nights/week. Hospital administration would not even support staffing for 7 nights. In academic medical centers it is difficult for sleep medicine to get on the radar screen. Health system administrators are, not unreasonably, focused on big-ticket items—transplant surgery, cardiac surgery, cancer, neurosurgery, orthopedics, etc. The waiting list for a sleep study grew to 6 months!

In 2000 I persuaded the Chair of Medicine to establish a network of sleep labs in the Greater Delaware Valley region. He wisely suggested that we should outsource this. A formal Request for Proposals was sent out by the University of Pennsylvania Health System and we selected a corporate partner. The selected corporate partner proposed a management service model where the labs and staff remained part of the University of Pennsylvania. They provided management oversight and resources to develop our program. This development was followed by establishment in 2002 of the first independent clinical Sleep Medicine Division in the United States. This gave us the ability to grow our clinical faculty, which we did. We placed faculty not only in the major academic medical center, but also as embedded specialists in locations in the suburbs within Penn primary care and internal medicine physicians' practices.

This proved to be enormously successfully. Over a seven-year period, the number of beds in our sleep lab network grew from 4 to 26, and our annual lab study volume grew from around 1,000 to 5,500 studies.


At the beginning of this expansion, discussions were held with our corporate partner about strategies to grow our program. The use of home studies was seen as an inevitable development in uncomplicated patients with high pre-test probability of OSA. At that time, however, none of our local payors reimbursed for home sleep apnea testing. Discussions were held with the payors but they did not want to do this since they were concerned that providing home studies could lead to an explosive growth in volume, not only of tests, but also CPAP utilization. Our facilities were developed, however, in a way that minimized the impact of a likely future switch from lab studies to home. Being left with a large inventory of unused sleep rooms was not a desirable outcome. This flexibility was achieved in two ways: (a) sleep labs were established in hotels using short-term leases that could be terminated relatively quickly and (b) large clinical exam rooms in new facilities were equipped with fold-down Murphy beds and adjacent toilets and showers. Thus, the room was an examination room by day and a sleep study bedroom at night. Thus, if the utilization of beds for in-lab studies decreased, there would still be a highly productive space in which to see clinic patients

When the paradigm shift by insurers from in-lab to home studies occurs, one is typically given 30–90 days' notice. It affects a large percentage of tests being performed after a pre-certification process is put in place. Thus, the first thing that any group planning for this paradigm shift needs to ask is the following: What can you do with your facilities if your volume of sleep studies suddenly drops to 40% of what you are currently doing? Programs dependent on laboratory-based studies utilizing expensive real estate with long-term commitments cannot adjust quickly. These programs have proven to be the most vulnerable to this paradigm shift.

The second preparatory step is to develop your own home sleep apnea testing program. This requires evaluating the different types of equipment, purchasing what you believe is optimal, and training your technological staff. It also requires developing the appropriate IT infrastructure to support this and obtaining contracts with individual payors to become a provider of home sleep apnea testing. Payors may require American Academy of Sleep Medicine (AASM) Out of Center accreditation as a prerequisite for providing service to subscribers. It is difficult to do all of this quickly. The problem is that when the paradigm shift comes, it occurs quickly. You need to be prepared for the rapid migration of diagnostic work from laboratory to the home.

The third preparatory step is to recognize that administrative costs will rise in proportion to the percentage of orders requiring precertification. Additional physician time must also be afforded to the process. The precertification process requires a thorough review of patient medical records and diagnostic study orders, and a complete understanding of the precertification requirements. These requirements vary by benefits management companies and are continuously evolving. Thus, designating persons with specific training and expertise to facilitate the pre-certification process is essential. This will expedite the process, although even with this kind of acumen, many pre-certifications take up to 3 weeks dependent on the circumstances. Effective coordination and management of the precertification process can help to mitigate physician time invested, but physicians should still expect to spend additional time on peer to peer reviews of orders and coordination with internal administrative personnel.

It is also important to communicate to all relevant leaders—health system, practice administrators, hospitals, clinical departments, etc.—that when the paradigm shift occurs there will be a sudden drop in revenue such that a program that is cash positive will overnight generate losses. This is discussed in more detail in the next section.

Two things that experience indicates should also be done are the following: (a) developing relevant order sets in the EMR, e.g., EPIC. The problem with EPIC is that you cannot program this yourself even if you have trained personnel. Programs need to get in a long line and wait their turn. Thus, doing this after the paradigm shift hits leads to inevitable delays; and (b) centralizing the pre-certification process. This is discussed in the next section in more detail.


When the paradigm shift occurs one is given little notice and little time to adjust. This change results in a sudden large drop in reimbursement. It is the large difference in reimbursement for home compared to lab studies that makes this paradigm shift so difficult to cope with.

Apart from differences in reimbursement, insurance carriers involve benefit management companies in the process to pre-certify studies. The rules are complex, constantly changing and vary from carrier to carrier.

As indicated, this occurs almost overnight. In our own practice, prior to the disruption, a number of knowledgeable internal medicine physicians, as well our bariatric surgery program, ordered sleep studies directly and the patients were seen after the study to discuss the results and organize therapy. Following the paradigm shift, this became too complicated for referring physicians to do and they switched almost overnight to ordering clinical evaluations, not studies. This resulted in a dramatic drop in orders for sleep studies but an increase in requests for consultations. Thus, the effect of this paradigm shift was compounded by a sudden shift in practice pattern. Moreover, prior to the paradigm shift, insurance companies insisted that we needed recent sleep studies to provide new CPAP machines to our long-term existing patients. Since our practice was very mature, a number of re-titration studies were done on our long-term patients. All of this went away with the sudden change in policy. The dramatic drop in lab sleep studies is documented in Figure 1.

Number of laboratory sleep studies per month from July 2012 to June 2013 (top line, black) and from July 2013 to end of June 2014 (bottom line, red) and from July to December 2014 (bottom line, blue).

This red line illustrates the initial precipitous drop in laboratory sleep studies following the disruptive paradigm shift in September 2013. The number declined to about 50% of what we had been doing and has shown a slow recovery from the nadir in February 2014. The blue line indicates that laboratory studies are slowly increasing towards the pre-disruptive level (top black line).


Figure 1

Number of laboratory sleep studies per month from July 2012 to June 2013 (top line, black) and from July 2013 to end of June 2014 (bottom line, red) and from July to December 2014 (bottom line, blue). This red line illustrates the initial precipitous drop in laboratory sleep studies following the...

(more ...)

The varying pre-certification rules between different insurance companies led to confusion. It was realized that it was impossible for physicians to keep track of all of this and deal with the major administrative burden. A centralized pre-certification process was developed staffed by two of our previous technologists who were given specific training. They became expert in the different rules and processes. Physicians order a sleep study, fill out the request form and the pre-cert staff get all of the relevant information from the EMR. Since studies are ordered in the physician's name, there is a robust audit process to check all outcomes to ensure that the appropriate test is performed. This strategy is working well and is a strategy that is recommended. There is now a new pre-certification process. The local insurance company is now using their benefits management company to require pre-certification for CPAP mask refills. This component has been added to the new centralized process. Physicians may be on the phone appealing decisions not to give our patients a new CPAP mask. Thus, the strategy of the insurers in this area is obvious—increase the administrative burden and cost. This new process will again add unreimbursed administrative costs to the clinical practice of sleep medicine.

Given the dramatic drop in requests for in-lab studies, one quickly finds that you are over-staffed with technologists. Since the situation is so dynamic, it is difficult to do careful analyses to decide the size of the technologist pool that is required. One needs to use a best guess strategy. In our own case we had to terminate a number of technologists (n = 9), reassign others to other duties (n = 3), and not fill positions when others resigned (n = 3). This resulted in a 40% reduction in our technology workforce. This is difficult to do and there are inevitable delays in giving notice. This results in financial losses. Some positions in the Sleep Medicine Division were also terminated to reduce costs. Downsizing of staff to reduce costs needs to be done quickly. Having a drip drip approach to staff reduction leads to further reduction in morale. It is, therefore, important to do this quickly and decisively.

The dramatic drop in requests for lab studies also leads to unused facilities. Again, careful analyses are difficult to accomplish, and one needs to move quickly to reduce losses. Notice was given to two facilities. One was in the Sheraton Hotel in Philadelphia—8 bedrooms (had 180 days to terminate the lease) and two bedrooms in a Homewood Suites Hotel (90 days to terminate). Again, there are costs involved, since prior to closure there are rental costs without matching revenue. Right sizing the facilities is difficult since, as Figure 1 shows, the demand is a moving target. We may have overcut and we have now reactivated a 2-bedroom footprint in the Sheraton Hotel. This is very challenging.

All of this leads to a dramatic drop in revenue with a delay in expense reductions—hence substantial losses. In a complex health system, deficits attract the attention of financial administrative staff (health system, Clinical Practices of the University of Pennsylvania, Department of Medicine). Regular meetings with this group were organized to discuss the Sleep Medicine budget. Essentially, a financial control board was established. This limited self-determination to deal with the crisis. Fortunately, given all of the steps that have been taken, our financial situation has been turned around. We are on track to be cash positive at the end of this fiscal year (June 30, 2015).

To minimize the impact of this on your program, it would be helpful to do your own financial projections on the losses you might expect. You could use data in Figure 1 to help make those projections. (Other data can be provided if requested.) Ideally any program should maintain financial reserves to cover acute losses. The losses experienced by us were less than 10% of the positive margins we had generated in the last 10 years. At a minimum you need to communicate to all relevant administrative leadership your projections. Our situation was compounded by the fact that the Chair of the Department of Medicine and the Director of Business Development for the Health System, both of whom we had extensively briefed about this to warn them of the future, left in the middle of this. It is not helpful to surprise ones new bosses with deficits.


Despite these difficulties, the future of sleep medicine is bright and full of opportunities. This view is based on the following: (a) sleep disorders are extremely common and have important consequences on multiple domains of health; (b) there are documented effective treatments that have positive clinical outcomes, including improved quality of life; (c) prevalence of sleep disorders is increasing and there is growing recognition of their role in multiple disorders, e.g., diabetes, heart disease, neurodegeneration, cancer, based on extensive clinical research; (d) sleep disorders lend themselves to outcomes-based approaches to chronic care; (e) there is wonderful mobile technology to monitor compliance and effectiveness of therapy; (f) electronic medical records provide the opportunity for simple screening for sleep disorders; (g) as a result, sleep medicine lends itself to telemedicine approaches; and (h) most importantly, sleep disorders and inadequate sleep have not only consequences for health but also for safety, e.g., in the transportation sector, as well as for productivity. Thus, many employers, in different parts of our economy have a vested interest in the sleep health of their employees. Sleep fits well into the wellness movement in corporate America.

The first step for the future is stabilizing and growing the current model (see Table 1). This occurred at the University of Pennsylvania, and new patient visits are up 12% from before the paradigm shift. This continues to increase. This is driven by several factors: (a) altered practice patterns as described above; (b) closure of competing sleep centers due to bankruptcy; and (c) increasing demand for services based on clinical research and the growing recognition of the relevance of sleep apnea. In the last few months discussions have been held with colleagues in obstetrics (studies in high-risk pregnancies), cardiology (both heart failure and ablation for atrial fibrillation), orthopedics (identifying and treating sleep apnea in patients getting joint replacement surgery), and in neurology (neurodegenerative disorders). All of these areas represent opportunities not only for our program but more importantly to improve outcomes of patients with these disorders. Building specialized programs with tracking of relevant referrals is a helpful step. This is facilitated by use of the electronic medical record. These steps, together with centralizing the pre-cert process, have led to a slow increase in our in-lab study volume (see Figure 1).

Future developments in clinical sleep medicine.



table icon
Table 1

Future developments in clinical sleep medicine.

(more ...)

An important part of the strategy is to consider how to interact with primary care and internal medicine colleagues. They could, in theory, take on management of patients with uncomplicated sleep apnea. They do not, however, have the time to do so. They are already overwhelmed with keeping up with all of the new quality standards. Thus, if they suspect a sleep disorder, they want to simply hit a button in the EMR and Sleep Medicine takes care of the rest. To facilitate this, sleep medicine physicians and nurse practitioners have been embedded in the primary care and multidisciplinary practice sites of the University of Pennsylvania. They are at multiple locations within the Delaware Valley. Our DME companies provide CPAP “mask clinics” at each of these sites. The faculty in the Sleep Medicine Division build strong relationships with referring physicians. This is an ideal model to extend the reach of the major academic medical center. Nurse practitioners do follow-up care for sleep apnea patients and are trained to do CBTI for insomnia.

The Sleep Medicine program at the University of Pennsylvania utilizes selected DME companies for provision of CPAP. Developing an internal DME program within the health system was considered. Leadership of the health system was not enthusiastic, given the large regulatory burden. Moreover, following the results of the bidding process by Medicare, the financial situation changed. Thus, a decision was made not to pursue this option.

While these steps are helpful, a new integrated program focused on outcomes and chronic care is being built. This was discussed at the outstanding Future of Sleep Medicine Conference that was held by the American Academy of Sleep Medicine in November 2013. (A white paper on this is to be published shortly.) The group came to a number of important conclusions. In brief, they are: (a) sleep medicine is a care management, not a diagnostic, discipline; (b) sleep medicine is for all sleep disorders, not just sleep apnea; (c) sleep medicine should have defined outcomes of care for all disorders; and (d) sleep medicine lends itself to telemedicine approaches. Implementing the outcomes for different sleep disorders identified by the AASM task force is the next logical step.3 Integrated programs that line up with this vision exist, albeit not within fee-for-service insurance companies. Our colleague, Dr. Sam Kuna, at the Veterans Affairs Medical Center in Philadelphia, has developed such a program—REVAMP. It has the following attributes: (a) patients fill out a detailed questionnaire on a novel web-based platform; (b) these responses are reviewed by a provider (physician or nurse practitioner); (c) the provider interviews the patient using a telemedicine link; (d) based on responses to questionnaires, an initial note is created in the patient's electronic chart; (e) the telemedicine link is used to provide education about sleep apnea and conduct of home sleep apnea testing; (f) results from home studies are interpreted and data put on the customized website; (g) when CPAP is prescribed, compliance, mask leak, residual AHI are obtained from the CPAP machine remotely; (h) these data are provided both to the provider and patient. There is a patient portal on the website; and (i) the website also contains educational videos and frequently asked questions. This approach enhances CPAP compliance.4 In particular, the provision of adherence data to the patients enhances initial CPAP compliance.4 It allows diagnosis and delivery of care over a wide geographical region. In-lab studies are done based on specific indications. This occurs in about 10% to 15% of patients with suspected OSA.

A similar program has been developed by Dr. Dennis Hwang in the Kaiser system. He has developed a team approach to care management with physicians, nurse practitioners and retrained sleep technologists who work during the day dealing with clinical issues with patients. These individuals—“sleep medicine coordinators”—likely represent a new important part of the sleep medicine workforce. Given that the number of sleep medicine physicians in training is likely inadequate to meet clinical needs,1 creative team approaches are needed involving sleep medicine physicians, nurse practitioners, and sleep medicine coordinators.

Thus, there are models for integrated outcomes-based care using new IT platforms. The challenge is not what needs to be done, but rather how to get this established in the current fee-for-service reimbursement system used by private insurers. Currently they seem to be focused on one thing—reducing costs of diagnosis. But if individuals do not get treated and their therapy monitored so that it is effective what is the point? It seems that the proposal that insurers can simply save money by use of home rather than lab studies completely misses the point. It is conceivable that return on expenditures with this model is actually less, if a much smaller percentage of patients get effective treatment. Money is being wasted on diagnostic tests with no follow through. It is not diagnostic strategy that matters but rather focusing on treatment. Conceivably new forms of reimbursement will emerge, such as bundled payment models or pay for performance. Costing the bundle will be challenging.

Given that treatment of sleep disorders also reduces future health care costs,5 disability claims,6 absenteeism,7 and crashes,7,8 it is highly likely that these new models will be first adopted in the self-insured employer market. A developing literature shows the benefit to them of doing this. This is already occurring in transportation companies such as Snyder and JB Hunt, who are screening their commercial driver employees for OSA and treating them with continued monitoring of CPAP compliance. It seems likely that this strategy will spread not only to other transportation companies in all modes of transportation but also to other sectors of our economy.


In this special article, the reasons for the paradigm shift has been explained with increased use of home sleep apnea testing with mandated pre-certification for lab studies. How to prepare for it and manage it has also been described. It is a disruption that leads initially to a dramatic drop in revenue combined with an increase in administrative costs. There are key steps to be taken to “right the ship.” The future for sleep medicine, however, is likely to be bright if new models of care are established. A focus on educating insurers on the benefits of this is essential. Sleep medicine can, with the changes described above, continue to be a vibrant field of medicine with a sound financial base.


This was not an industry supported study. Dr. Pack has indicated no financial conflicts of interest. Dr. Pack is currently President of the Sleep Research Society. The views expressed in this article are his own and do not in any way reflect the views of the Sleep Research Society.



Quan SF, author. Graduate medical education in sleep medicine: did the canary just die? J Clin Sleep Med. 2013;9:101. [PubMed Central][PubMed]


Quan SF, Epstein LJ, authors. A warning shot across the bow: the changing face of sleep medicine. J Clin Sleep Med. 2013;9:301–2. [PubMed Central][PubMed]


Morgenthaler TI, Aronsky AJ, Carden KA, Chervin RD, Thomas ST, Watson NT, authors. Measurement of quality to improve care in sleep medicine. J Clin Sleep Med. 2015;11:279–91. [PubMed]


Kuna ST, Shuttleworth D, Chi L, et al., authors. Web-based access to positive airway pressure usage with or without an initial financial incentive improves treatment use in patients with obstructive sleep apnea. Sleep. 2015;38:1229–36.


Potts KJ, Butterfield DT, Sims P, Henderson M, Shames CB, authors. Cost savings associated with an education campaign on the diagnosis and management of sleep-disordered breathing: a retrospective, claims-based US study. Popul Health Manag. 2013;16:7–13. [PubMed]


Hoffman B, Wingenbach DD, Kagey AN, Schaneman JL, Kasper D, authors. The long-term health plan and disability cost benefit of obstructive sleep apnea treatment in a commercial motor vehicle driver population. J Occup Environ Med. 2010;52:473–7. [PubMed]


Sassani A, Findley LJ, Kryger M, Goldlust E, George C, Davidson TM, authors. Reducing motor-vehicle collisions, costs, and fatalities by treating obstructive sleep apnea syndrome. Sleep. 2004;27:453–8. [PubMed]


Tregear S, Reston J, Schoelles K, Phillips B, authors. Obstructive sleep apnea and risk of motor vehicle crash: systematic review and meta-analysis. J Clin Sleep Med. 2009;5:573–81. [PubMed Central][PubMed]