The answer: It depends. A sleep physician discusses many aspects of HST.
Shortly after the landmark decision by CMS to reimburse for CPAP based on home sleep testing (HST), an article published in Sleep Review discussed several implications of this decision for the future of sleep medicine (“Home Sleep Testing: Present Position and Predictions,” June 2008). Many of these predictions have come to pass, including difficulty with profitability and slowing of expansion of facility-based sleep testing, while some of the more optimistic projections have not. Almost 7 years later, many aspects of HST still need to be addressed including advantages and limitations, indications, technical considerations, provider qualifications, terminology, business models, analysis of results, challenges, and other issues. The intent of this article is to help answer the question: Should, would, or could providers begin or continue doing HST? The short answer: It depends.
HST Advantages & Limitations
Advantages: HST (aka portable or limited-channel unattended polysomnography) costs less than facility-based attended polysomnography (PSG), often does not require preauthorization, is more acceptable to patients, more accessible, easier to schedule, and performed in the comfort of the home or other location and may reduce the first night effect. Some patients with obstructive sleep apnea (OSA), if properly selected, may begin auto-adjusting CPAP (APAP) or oral appliance therapy (OAT) without the need for a CPAP titration. Some argue that the long-term costs of HST are the same or greater and outcomes from treatments based on HST are no different than traditional PSG and PAP titration. However, a recent study concluded that following diagnosis of OSA by HST, treatment with APAP results in equivalent PAP adherence and improvement in sleepiness compared to PSG titration and CPAP treatment (Berry RB, Sriram P. Auto-adjusting positive airway pressure treatment for sleep apnea diagnosed by HST. J Clin Sleep Med. 2014;10(12):1269-1275).
Limitations: When compared with PSG, HST limitations include the inability to stage sleep, diagnose coexisting sleep disorders, and ensure the chain of custody (COC). Most HST devices do not record EEG, so true total sleep time (TST) is not measured. Some units record motion or other indicators of estimated sleep time, but there is disagreement regarding the degree to which these estimates of sleep time are reliable. The apnea-hypopnea index for HST (“AHI”) is calculated using total recording time (TRT) or some indirect estimate of TST, so “AHI” derived from HST = apneas + hypopneas/TRT (or an estimate of TRT). If “AHI” determined by HST is ? 5, the true AHI is often underestimated, and the lower the sleep efficiency (TST/TRT x 100) the greater the underestimate. False negatives (ie, AHI < 5) could also occur. As with any single-night study, including PSG, the severity of OSA may also be underestimated due to lack of or limited supine and/or REM sleep and/or night-to-night variability of sleep habits, medication effects, alcohol, caffeine, smoking, and environmental factors.
Some consider delay of diagnosis and treatment based on HST to be a limitation, particularly if the equipment is delivered by mail. However, the diagnosis of OSA is rarely an emergency and the condition has likely existed for several years. HST can help reduce the real delay in the diagnosis and treatment of OSA—which is due to some combination of accessibility, cost, and education of healthcare providers and the public—for millions of undiagnosed individuals.
Criteria for HST
The specific indication for HST is to confirm diagnosis and severity in individuals at high risk for OSA. In some instances, HST may be helpful to confirm the efficacy of treatments such as OAT, weight loss, and certain surgical procedures.
Objective parameters for accessing the risk of OSA include BMI, neck size, age, craniofacial anatomy, and upper airway evaluation. Other risk factors including various symptoms and coexisting conditions can be documented using subjective questionnaires such as the Epworth, STOP-BANG, Berlin, and/or other subjective screening tools. However, patients classified as “low risk’’ based on a particular set of signs and symptoms may in fact be at very high risk for OSA based on a different set of parameters. There should also be criteria for exclusion of patients with certain comorbidities or suspected coexisting sleep disorders when home sleep testing may not be appropriate. Unfortunately, exclusion criteria for COPD, CHF, and other cardiac or medical problems have not been well defined with respect to severity. No questionnaire or set of physical findings is perfect, and the decision for testing should be based on the overall evaluation by the treating physician. Specific inclusion criteria for medical necessity vary among payors and may be important for reimbursement, an example being the April 1, 2014 UnitedHealthcare policy 2014T033P for “Polysomnography and Portable Monitoring for Evaluation of Sleep Related Breathing Disorders.”
Analysis of HST Results
First it’s important to screen individuals at high OSA risk and exclude those with certain comorbid conditions. Then consistent protocols should be used for screening, analyzing, and acting on the results of the HST so further diagnostic testing is recommended when indicated. Protocols should include recommendations for prescribing and monitoring treatment. A suggested protocol/algorithm is published in the November 2014 issue of Sleep Review (“Sleep Apnea and Trucking: Where Are We Now?”). Regardless of the protocol used, here are a few points to consider:
Terminology used in HST reports is inconsistent, making comparison of results difficult. Definitions for and calculations of AHI and respiratory disturbance index (RDI) vary among HST device manufacturers, payors, published articles, and PAP machine data downloads. Certain payors define RDI based on actual sleep time (PSG or Type II HST) or recording time (Type III and IV HST). Sometimes AHI and RDI are used interchangeably and often RDI and/or AHI may be based on recording time. Some facilities use the term RDI to describe a calculation that includes other sleep disturbances, such as flow limitation based on waveform shape or snoring. Also various payors, professional societies, and PAP data downloads use different definitions of hypopneas and different sensors and/or combinations of sensors.
Here is the current “standard” terminology: The term “true sleep” as determined by EEG is used here. AHI = apneas + hypopneas per hour of true sleep; RERA = respiratory effort related arousal (requires EEG); RDI = apneas + hypopneas + RERAs per hour of true sleep. The term respiratory event index (REI) = apneas + hypopneas per hour of true sleep or recording time (actually three possible denominators if indirect estimates of sleep time are included). A detailed analysis of many of these issues has been published in the Journal of Clinical Sleep Medicine in 2011 (Collop NA et al, Vol 7, No 5).
It is important that the HST is scored manually and that the raw data is reviewed by a qualified sleep physician. An average AHI < 5 or ? 5 (HST or PSG) may also underestimate the severity of OSA when periods of more frequent respiratory events and O2 desaturations occur. An example would be a 30 minute portion of the recording perhaps during supine and/or REM sleep, during which 20 respiratory events are scored, equivalent to an AHI or RDI of 40 for that segment, and which may be associated with O2 desaturations of 80% or less. Such episodes can be identified in the raw data, although some professional associations and payors have not made formal recommendations in this regard.
Only individuals with high OSA risk would (should) be tested via HST. So if AHI < 5 for any reason including true and false negatives, absence of respiratory events during the HST, or there has been an attempt to alter the HST outcome, PSG should be recommended. This is consistent with the position of the American Academy of Sleep Medicine (AASM) and, in effect, ensures the COC, important when employment is at stake. The COC, in my view, is a procedural rather than technical issue since there is no incentive to influence the outcome of the HST as it would result in the recommendation for a PSG (and likely a titration study), which would incur more cost and time. Some will spend time and money on HST and have to be retested but, if properly selected for high OSA risk, this number should be small.
Situations that may require facility-based testing if AHI ? 5 on the HST include: CPAP titration is recommended due to the severity of OSA and/or desaturations, central or complex sleep apnea is suspected, OSA is detected and an additional sleep disorder suspected, the HST is suggestive of OSA but the testing device does not detect respiratory effort, or if symptoms persist after a reasonable trial of APAP or OAT.
Payors and Regulations
CMS publishes minimal requirements (national coverage determinations or NCD) for coverage of medical services nationally. Specific guidelines for clarification published by local contractors for geographic areas (local coverage determinations or LCD) such as provider qualifications, CPT and ICD-9 codes, and requirements for medical necessity, are often more restrictive than the NCD.
Providers of HST should be aware of the details for specific issues that are outside the scope of this article. These include: billing, coding, reimbursement, medical necessity requirements, federal and state regulatory or legal issues, training, certification or licensure requirements for personnel, HIPAA, payor requirements for physician and facility enrollment, accreditation, interstate issues, who can bill technical and professional components of the HST, advance beneficiary notice requirements, place of service, frequency of testing, Stark and AKS issues, and marketing. Remember that reimbursement for HST, co-pays, and deductibles may vary among Medicare and other third-party payors with location. Also, these specific LCD and other payor requirements are periodically updated and may require preauthorization and specific documentation. There is no guarantee that the HST, its interpretation, or any recommended equipment or treatment based on the outcome of the HST will be reimbursed in full or in part, so it is important (as with PSG) to have proper documentation of medical necessity.
The Medicare coverage database (MCD) contains all NCDs and LCDs, local articles, and proposed NCD decisions. For HST, NCD Pub 100-3 Section 240.4.1 TN 103 and the regional LCDs are easily searchable on the cms.gov Medicare coverage database.
Generally it’s safe to follow Medicare guidelines for other payors, but it’s strongly suggested that providers review the LCD of their regional Medicare contractor and the policies of other payors and consult with a healthcare attorney knowledgeable in sleep.
Specifics—such as data security transmission, storage, and retrieval; auto or hand scoring of raw data; comparison of specific HST devices; disposable or reusable components; and the technical and licensing policies of individual manufacturers’ scoring and reporting software—depend on business model. A recent article has addressed many other technical aspects of OSA devices for out-of-center testing and proposed a new classification system (ie, SCOPER) for measurement parameters (see Collop et al article in “Analysis of HST Results,” cited above).
Minimum channels for level III HST should include airflow, oxygen saturation, pulse rate, and respiratory effort. Optional channels may include snoring, body or head position, motion detection, pulse plethysmography, peripheral arterial tonometry, pulse transit time, extra lead(s) for some combination of EEG, EOG, EMG, or EKG, an indicator of signal quality, the ability to perform HST during PAP, and possibly a light sensor. The minimum channels are adequate to accomplish the purpose for which HST is intended—the diagnosis of OSA. The optional parameters may be helpful, particularly for research, but are often not economically feasible, and their practical clinical use is limited since most home sleep tests (and PSGs) are single-night studies with no control over a patient’s body position and sleep patterns during subsequent nights.
Business Aspects of HST
Should, would, or could providers begin or continue doing HST? The short answer: It depends. Let’s have a look at some real costs and reimbursement. This section addresses fixed and variable costs that should be considered in the decision to begin or continue doing HST.
Sleep Review regularly publishes a useful side-by-side comparison for most of the available HST devices including cost of the unit and consumables, physical specifications, power, memory storage, available channels, and other features—a good start but light years from the true costs, which depend on specific business models and other variables.
Variable or per test costs Device utilization, licensing policies of manufacturers’ scoring and reporting software, disposables, scoring and interpretation fees, lost or broken hardware, delayed return of units, insurance authorization, obtaining prescriptions and clinical documentation, order processing, billing, coding, device cleaning and storage, failed tests, scheduling, transmitting reports to physicians and payors, bad debt, credit card fees, and the testing facility’s policies for delivery and return of the device—including postage (if applicable).
It’s important to consider the cost of purchase or rental of the device based on utilization. Per-device physician providers and sleep labs realistically can do three tests a week. Providers that mail devices can in reality do two to three tests per month, and postage increases the cost. These models change if providers utilize multiple devices.
Fixed costs* Record and data processing, maintenance, backup, security, and storage; sales and marketing; training; other IT; 24/7 help/support line; initial and recurring accreditation fees; and HIPAA compliance—all in addition to the usual costs of running a business, eg, payroll, HR, rent, utilities, licenses and permits, phone, insurance, legal, accounting, and interest on debt. (*Some may be variable depending on business model.)
The bottom line Charges for HST range from $200 to $300, although some are higher. Reimbursements are highly dependent on the payor, specific billing code, cash or insurance, and exclusive provider contracts, and range from about $120 to $300, and average collected amount from $175 to $250. So profitability is a challenge.
While not a major revenue source, sleep physicians and sleep labs performing all aspects of HST on their own patients is likely doable. Non-sleep specialists…not so much, particularly if regulatory issues and qualifications are considered. From a commercial point of view, the economics of HST are more workable with large volume, multiple devices, effective IT systems, possibly outsourcing, and additional revenue from CPAP equipment and supplies.
Dentistry and HST
OAT can provide an alternative for patients unable or unwilling to use CPAP and is increasingly becoming a revenue source for dental practices. Many educational programs are available. However, implementation of a successful program is challenging—including the learning curve, proper selection of patients, marketing for patients, cost of treatment, expense to patients, reimbursement, adequate follow-up, obtaining sleep testing, and creating relationships with physicians. Regardless of specific state regulations, I recommend that dentists do not write prescriptions for sleep testing as they will be legally required to deal with the results, whether or not OAT is indicated. Also, some payors will not reimburse for therapy if some or all aspects of the sleep test are not performed in accordance with their requirements. The position of the American Academy of Dental Sleep Medicine (AADSM) on this issue is that all dentists should follow state licensing laws and is not aware of any state that permits a dentist to order or interpret a diagnostic sleep test. The AADSM December 7, 2012 “Policy Statement on the Diagnosis and Treatment of Obstructive Sleep Apnea” indicates that dental licensing laws and practice acts do not include provisions for the diagnosis of medical diseases and disorders or the treatment of (non-dental) diseases by dentists without a prescription from a physician.
Challenges and Conclusions
Media coverage has highlighted the consequences of undiagnosed and untreated OSA. High insurance deductibles and the awareness of the public, employers, regulators, attorneys, and insurance companies about costs, accessibility, and liability associated with OSA are driving the sleep medicine industry to perform more HST.
Most non-sleep physicians recognize the importance of diagnosis and treatment of OSA, but many do not screen or order testing for sleep apnea. Reasons may include busy practices, little formal training in sleep medicine, and/or lack of credentials to perform or interpret sleep tests. Some physicians do not order sleep tests when indicated as they may not be comfortable or motivated to prescribe, manage, monitor, and document adherence to and effectiveness of CPAP. Low reimbursement, onerous record keeping, and prohibitions to providing CPAP are deterrents as well. Some refer to pulmonologists or other sleep specialists, who may be too busy with their practices and hospital work.
In the US there are not sufficient resources for sleep specialists alone to serve the OSA population. There are only about 7,500 board certified sleep specialists, according to an AASM estimate, but there are many millions of people with undiagnosed sleep apnea. Non-sleep physicians should be able to utilize HST in overall patient management if they are educated to identify OSA risk and order HST to be interpreted by a sleep specialist, whose recommendations they could implement. Stand-alone testing programs and DME providers should develop associations or “sleep communities,” and network with sleep physicians, dentists, and more advanced facilities to assure continuity of care and follow-up and also keep the treating physician in the loop.
Sleep testing is only one component of overall care plans for patients with OSA, which should include education, screening, testing, treatment, monitoring, outcome evaluation, and follow-up. In spite of lingering questions, HST is here to stay, and, if used appropriately for its intended purpose—reducing costs and improving access while maintaining quality—a formal, consistent protocol for screening, analyzing, acting on the results of the HST, and monitoring treatment is necessary. SR
Edward D. Michaelson, MD FACP FCCP FAASM, is based in Ft Lauderdale, Fla, and a member of Sleep Review’s Editorial Advisory Board.