560
A Telemedicine Program for Diagnosis and Management of Sleep-Disordered Breathing: The Fast-Track for Sleep Apnea Tele-Sleep Program Max Hirshkowitz, PhD, DABSM1,2
Amir Sharafkhaneh, MD, PhD, DABSM1
1 Department of Medicine, Baylor College of Medicine, Houston, Texas 2 Sleep Disorders and Research Center, Michael E. DeBakey Veterans
Address for correspondence Max Hirshkowitz, PhD, DABSM, 4380 Vine Hill Road, Sebastopol, CA 95472 (e-mail: [email protected]).
Affairs Medical Center, Pulmonary, Critical Care & Sleep Section, Houston, Texas
Abstract
Keywords
► sleep apnea ► home sleep testing ► positive airway pressure ► telehealth ► tele-medicine
The objective of this study was to facilitate access to sleep health care for veterans. We designed and implemented a Telehealth program for diagnosing and treating sleeprelated breathing disorders (SRBDs). Building on our ongoing out-of-laboratory “Fast Track for Sleep Apnea” program, procedures were modified to accommodate remote operations. This Tele-sleep program was set up at the medical center’s communitybased outpatient clinics. Home sleep testing and positive airway pressure device technological advances enabled realizing this application for Telehealth. In addition to obtaining appropriated teleconferencing equipment, the program involved implementing systematic processes for (1) six types of clinic visits, (2) training remote-site personnel, (3) making recommendations for inventory management, and (4) evaluating patient satisfaction. Over the past year, we have updated and refined our procedures to optimize program performance and efficiency. To achieve the next step, that is, increasing program scale beyond its current state (e.g., to region-wide), we will need to further develop and formalize quality control indicators to more efficiently monitor operations. The program has helped relieve clinical load at the central sleep program, improved local access to sleep care for veterans, and improved patient satisfaction with health care for SRBDs.
Sleep-related breathing disorders (SRBDs) include a spectrum of respiratory conditions occurring during and disrupting sleep. The condition manifests as cessations of breathing (for 10 or more seconds), shallow breaths provoking oxyhemoglobin desaturations, and/or partial airway obstructions leading to respiratory effort–related arousals.1 Patients may also suffer from hypoxemia, sometimes with significant time spent at or below 88% oxyhemoglobin saturation levels. The traditional diagnostic approach involves making recordings of brain, movement, cardiographic, and respiratory activity while the patient sleeps.2 These recordings typically
Issue Theme Clinical Consequences and Management of Sleep Disordered Breathing; Guest Editors, Ravi Aysola, MD, and Teofilo L. Lee-Chiong Jr., MD
involve the patient sleeping overnight in a hospital or freestanding clinic’s laboratory. Another overnight sleep study usually follows to adjust and optimize treatment with positive airway pressure. Technological advances now make it possible to diagnose and titrate some patients at home using portable devices.3,4 Thus, application of these techniques may improve access to care, especially in rural localities. Telehealth, in general, seeks to facilitate access to care, change the location where care is provided, and improve care management using health informatics and telecommunication technologies.5 The Veterans Affairs (VA) Telehealth
Copyright © 2014 by Thieme Medical Publishers, Inc., 333 Seventh Avenue, New York, NY 10001, USA. Tel: +1(212) 584-4662.
DOI http://dx.doi.org/ 10.1055/s-0034-1390069. ISSN 1069-3424.
Downloaded by: WEST VIRGINIA UNIVERSITY. Copyrighted material.
Semin Respir Crit Care Med 2014;35:560–70.
Service is a leader in deploying this approach in a wide variety of medical specialties. More specifically, Telehealth uses “videoconferencing, the internet, store-and-forward imaging, streaming media, and terrestrial and wireless communications.”6 Telemedicine focuses more narrowly on remote diagnosis and treatment (also called Telecare). A randomized controlled trial in the United Kingdom found Telemedicine for diabetes, chronic heart failure, and chronic obstructive pulmonary disease (COPD) reduced mortality by 45%, emergency admissions by 20%, and bed days by 14%.7 Studies also show economic advantage. For example, analysis of the Alaska Federal Health Care Access Network reported an $8.5 million dollars Medicaid saving for travel costs alone in 2012.8 SRBD represents the primary indication for ordering a sleep study.9 Laboratory polysomnographic testing requires specialized equipment and highly trained personnel; consequently, it is a limited resource. Demand for SRBD diagnosis steadily increased during the past two decades making it difficult to provide timely service. This situation prompted our developing a home sleep testing (HST) program that we later coupled with an automatic self-adjusting positive airway pressure (APAP) program. In a sense, one objective was to unburden the sleep laboratory with straightforward, uncomplicated SRBD cases that could be diagnosed and managed with clinical and home-based procedures.10 The other objective was to improve access to care by shifting point-of-service to the local outpatient clinics. The remainder of this article will describe the rational and nuts-and-bolts of the Fast-Track for Sleep Apnea Tele-sleep program implemented at the Michael E. DeBakey VA Medical Center to serve the community-based outpatient clinics (CBOCs) associated with our program.
Technologies Involved Home Sleep Testing Technology Portable sleep recording technology evolved over many years culminating in small, reliable, lightweight cardiopulmonary Holter-like devices. Typically, these cardiopulmonary recorders (CPRs) monitor and digitally store one to several nights of data for subsequent review using standard computer systems. Most qualify as Level-III devices according to American Academy of Sleep Medicine (AASM) criteria because they provide detailed, viewable “raw-data” of the individual’s (1) airflow, (2) respiratory effort, (3) snore sounds, and (4) oxyhemoglobin saturation. Surrogate measures of airflow usually derive from thermistor and/or nasal pressure changes; effort from diaphragmatic movement, snore sounds from a microphone, and oxyhemoglobin saturation from a pulse oximeter. Body position and heart rate (and/or cardiac rhythm) are also often included. For a comparison of devices, see the review by Hesselbacher et al.11 CPRs provide adequate information to score apnea and hypopnea episodes. The clinician can also determine whether events are obstructive or nonobstructive. Because CPRs are not equipped with electroencephalographic channels, actual sleep time and central nervous system (CNS) arousals cannot be scored. Consequently, apnea and hypopnea must be in-
Hirshkowitz, Sharafkhaneh
dexed by time in bed rather than total sleep time. Furthermore, neither respiratory effort–related arousals nor overall respiratory disturbance index can be tabulated. These limitations potentially reduce sensitivity by diluting the metrics applied against diagnostic criteria. Factors to consider when selecting a device include (1) diagnostic sensitivity, (2) the typical severity of SRBD in patients in your practice, (3) device durability, (4) simplicity of use, (5) initial cost, and (6) recurring costs for supplies and sensors. Our sleep medicine programs at the VA currently service mainly men, age 50 or older, who often have hypertension. Thus, we not only expect generally high pretest probability for sleep apnea but also a high volume needing testing. Consequently, we could trade off some diagnostic sensitivity for lower consumable supply costs. Devices need to be durable but we also expected our population to be somewhat adept at learning to operate the device (having experience being trained to use a variety of devices in their armed forces tours).
Automatic (Self-Adjusting) Positive Airway Pressure Technology Continuous positive airway pressure as a treatment for obstructive sleep apnea was pioneered by Sullivan et al in 1981. 12 They were able to eliminate sleep-associated oropharynx collapse using positive pressure when pressure was raised high enough to pneumatically splint the airway. Their discovery revolutionized sleep apnea treatment; however, determining the appropriate pressure involved patients sleeping in the laboratory to accomplish efficacious titration. As microprocessor technology progressed, it became feasible to digitally monitor a PAP machine’s pressure and the pressure change oscillation associated with inhalation and exhalation. These oscillations provided a surrogate measure for airflow allowing a device to simulate laboratory titration by increasing pressure when the flow signal flattened or ceased. Such self-adjusting PAP devices would also be expected to accommodate to changing pressure needs (e.g., raising pressures when the sleeper is supine compared with laterally recumbent). More advanced models incorporate oximetry to continuously record oxyhemoglobin saturation level.
Home Sleep Testing Informatics for Level-III Cardiopulmonary Recorders Key parameters derive from HST devices relate to diagnosis. Compared with laboratory polysomnography (PSG), the HST provides a small subset of information describing sleep. In some way, it is like driving at night—even though your visual field is extremely impoverished compared with the daytime, enough information exits to manage the task. Nonetheless, these recorders still provide more than adequate of information with six parameters providing pivotal metrics (see ►Table 1). Key parameters from the patient history and interview include comorbid conditions and sleepiness. Sleepiness can derive from the clinician’s impression or based on a standardized sleepiness scale. Seminars in Respiratory and Critical Care Medicine
Vol. 35
No. 5/2014
561
Downloaded by: WEST VIRGINIA UNIVERSITY. Copyrighted material.
Tele-Sleep for Sleep Apnea
Tele-Sleep for Sleep Apnea
Hirshkowitz, Sharafkhaneh
Table 1 HST level-III device informatics needed for the Fast-Track for Sleep Apnea Tele-sleep program Parameter
Description
Rational for use
TRT
The total number of minutes that the HST device was used and data were actually recording.
To assure diagnostic integrity, a minimum of 2 hours is required. However, we typically feel more comfortable when 4, or more, hours of data are available.
AHIa
The number of sleep apnea and sleep hypopnea episodes per hour of TRT. Where apnea is defined as a 10-second (or longer) 90% drop (or greater) in the nasal/oral airflow channel’s peak-to-trough amplitude persisting for at least 90% of the event’s duration and hypopnea is defined as a 10 seconds, 30% drop in nasal pressure signal amplitude (compared with baseline) persisting for 90% of the event’s duration and associated with a 4% drop in O2.
The AHI represents the pivotal diagnostic criteria for SRBD. When AHI is 5 or above, the patient can be diagnosed with a sleep-related breathing disorder. A noncentral AHI of 10, or more, provides justification for APAP titration. Treatment recommendations for AHI between 5 (inclusive) and 10 (exclusive) involve consideration of other.
Proportion of central SRBD events to total number of events
Each apnea should be classified as either central (no respiratory effort present during flow cessation) or noncentral (respiratory effort present during at least part of the flow cessation time).
The proportion of central to all events provides the basis for differentially diagnosing central, complex, or obstructive sleep apnea. Nonobstructive types of SRBD constitute rule-outs for APAP therapy.
TOxT
The number of minutes during the study that the oximeter was functioning.
This parameter is used for validation. There is no agreed upon minimum; however, oximetry is needed to score hypopnea. Consequently, the lower TOxT, the higher likelihood that AHI (and overall SRBD severity) is underestimated.
Time below 88% (time 88% Sat)
The number of minutes that SaO2 was at or below 88% oxyhemoglobin saturation level.
This parameter is used to determine if the patient also has hypoxemia.
SaO2 Nadir
The lowest valid oxyhemoglobin saturation level observed during the recording.
This parameter helps define severity of hypoxemia (when present).
Abbreviations: AHI, apnea þ hypopnea index; APAP, automatic self-adjusting positive airway pressure; HST, home sleep testing; SRBD, sleep-related breathing disorders; TOxT, total oximetry time; TRT, total recording time. a The definitions here are derived from the AASM Scoring Manual.
PAP Informatics One key development making Tele-sleep possible involves advances in informatics available from APAP devices. Critical data examined in our program address (1) therapeutic adherence, (2) sleep-disordered breathing event rate (index), (3) mask leak rates, and (4) APAP with concurrent oximetry. These data allow clinicians to properly follow-up and care for patients in accordance with currently accepted standards of practice.
available. These data serve several purposes, including (1) allowing clinicians to gauge overall therapeutic adherence, (2) allowing clinicians to determine if enough data are available to make appropriate machine adjustments, (3) providing statistics required for Centers for Medicare and Medicaid Services (CMS) “compliance” reports, and (4) representing regulatory used for licensing in the transportation industry.
Sleep-Disordered Breathing Indices Therapeutic Adherence A primitive way to determine usage involves recording the machine’s blower’s total operating time. However, an audit trail of actual usage, sometimes called “time-on-mask,” represents the required statistic. Time-on-mask is determined by the presence of an oscillating back pressure sensed by the machine. This back pressure, when fluctuating at a frequency consistent with breathing and/or contiguous with sleepdisordered breathing events, confirms usage. Time-on-mask should be recorded and stored for a minimum of 90 days; be displayed graphically in 24-hour blocks; and be time and date stamped. A daily 24-hour usage amount; a total and mean use per day over some user-specified time frame; and the percentage of days with 4 hours of use (or more) must be Seminars in Respiratory and Critical Care Medicine
Vol. 35
No. 5/2014
Sleep-disordered breathing events detectable with APAP machines include flow cessations and flow limitation. If the device is equipped with a plug-in oximeter, data concerning oxyhemoglobin desaturation events, the lowest saturation level, and the amount of time spent at or below 88%a saturation are also available. It is critically important to not equate events detected by APAP machines with those scored according to standard technique (AASM, 2007). Confusion arises because APAP machines label flow cessations “apneas” and flow limitations as “hypopnea.” Luckily, APAP scored flow cessations correlate a
Some systems report 85% or 90%. We prefer 88% because it defines respiratory insufficiency.
Downloaded by: WEST VIRGINIA UNIVERSITY. Copyrighted material.
562
highly with apnea scored on PSGs and HSTs. The same is not true for hypopnea. APAP machines regard decreased flow as indicative of hypopnea with no regard to oxyhemoglobin saturation or CNS arousal. While this may accurately accord with physiologic conceptualizations of hypopnea, they do not properly represent the sleep pathophysiology designated as hypopnea. CMS considers sleep-related hypopnea pathophysiological because they are associated with a 4% (or greater) oxyhemoglobin desaturation. AASM also includes flow limitations that provoke CNS arousals as pathophysiological because they fragment sleep. Simple flow reductions, without hypoxemia or sleep disturbance, are normal (you have them when you eat and speak). These simple flow reductions are what APAP machines report. Therefore, clinicians must be vigilantly aware that terms are being used differently, to avoid making therapeutic errors. This unfortunate misuse of the terms sleep-related apnea and hypopnea has created serious ambiguity. ►Table 1 summarizes parameters needed for our Telesleep program. The summary data listed and described on the table should be available for each night and means should be calculated over any operator-selected range of nights.
Mask Leak Rates Every mask intrinsically leaks through exhalation ports. Leak rate depends on the type of mask and the pressure setting. For example, at 10 cmH2O pressure, Respironics ComfortGel Blue and ResMed SoftGel port 26 and 31 lpm, respectively. By contrast, these two popular masks port 37 and 45 lpm at 20 cmH2O pressure. Overall, to our knowledge, no popular adult
Hirshkowitz, Sharafkhaneh
masks or nasal pillows port more than 55 lpm at pressures up to 20 cmH2O pressure. Therefore, we typically regard leaks greater than 60 lpm as significant. Total time with significant leak becomes important when interpreting pressure and sleep-disorder breathing event statistics. When leak rate increases, many APAP machines will interpret the change as a flow limitation and raise the pressure. The Tele-sleep clinician therefore needs a graphic representation illustrating leak rates on the same time axis as pressure setting, flow cessations, nonobstructive flow cessations, and flow limitations (see ►Fig. 1). This type of graphic allows quick assessment of a night’s data integrity and interpretability. For example, if maximum pressure occurred during leak, it can be disregarded. We also commonly see high leak episodes marked by sleep-disordered breathing events increase (in response to which the machine raises pressures). This pattern usually inflates automatically calculated indices (that do not consider leak rates). Clinical judgment must then be applied when selecting prescriptive pressure settings.
APAP with Concurrent Oximetry (APAP-Oxy) APAP machine capability to concurrently record oximetry represented a major advance for developing our Tele-sleep program. A large proportion of patients were excluded from our initial APAP program because of practice standards concerns. Indeed, the AASM APAP practice guideline (which was largely authored by VA members of the AASM including myself) excluded patients with lung disease and other conditions associated with oxyhemoglobin desaturation. The rationale was clear for this rule-out. A patient might feel
Fig. 1 Leak, sleep-disordered breathing events, and pressure across the night.
Seminars in Respiratory and Critical Care Medicine
Vol. 35
No. 5/2014
563
Downloaded by: WEST VIRGINIA UNIVERSITY. Copyrighted material.
Tele-Sleep for Sleep Apnea
Tele-Sleep for Sleep Apnea
Hirshkowitz, Sharafkhaneh
much better, have greatly reduced apnea and hypopnea, but still have severe hypoxemia during rapid eye movement (REM) sleep and/or when sleeping supine, and the clinician would never know. Once concurrent oximetry became available, this problem was rendered moot. Having the oximetry graphically displayed along the same axis as pressures, sleepdisordered breathing events, pressure, and leak allow clinicians to rapidly determine if persistent hypoxemia exists. A quick glance at each night’s histogram adequately satisfies such concerns (see ►Fig. 2).
Program Development Developing Fast-Track for Sleep Apnea Protocols— General Approach Traditional diagnosis and treatment for SRBD is a costly and time-consuming endeavor. Over time, we phased-in HST and APAP programs to streamline the process in our sleep center so that we could provide care more efficiently. The groundwork for this development effort involved our clinical sleep specialists setting up clinics to follow-up patients undergoing HST diagnosis and APAP interventions. In this manner, we could keep close watch on these new procedures and troubleshoot any problems directly and immediately. We thereby gained the necessary insight into what worked and what to monitor for quality assurance. We compared notes and designed the Fast-Track program for Sleep Apnea. Protocols were then developed and clinical extenders trained. The resulting program was monitored closely for the 1st year and minor adjustments made, as needed. The procedures were gradually adapted for small group sessions (1–6 patients) and ultimately for larger groups (up to 10–20 pa-
tients). The following paragraphs describe the overarching concepts and some of the specifics incorporated into the FastTrack for Sleep Apnea program.
When Is HST Appropriate? Currently, HST is only useful to confirm SRBD diagnosis.3 If there is clinical suspicion for Narcolepsy, Nocturnal Seizures, Parasomnias, or non-SRBD forms of hypersomnolence, HST should not be used. CPRs qualify as Level-III HST devices, according to the AASM classification system. As such, they can detect sleepdisordered breathing events and therefore provide diagnostic information; however, they are less sensitive than laboratory polysomnography. Therefore, only patients with a high pretest probability for SRBD should be referred for testing. HST cannot rule-out SRBD!13 HST is a tool of confirmation; therefore, negative tests must be verified using more sensitive technique (i.e., laboratory-PSG). Thus, referring patients with a low pretest probability for SRBD is inefficient. HST’s lower diagnostic sensitivity compared with PSG stems for several factors. The first is simply arithmetic. Indices are calculated using total recording time (rather than total sleep time) as a denominator. Because total sleep time is always smaller than total recording time (unless sleep efficiency is 100%), the calculated index from Level-III HST recorders are invariably lower. For example, a patient with 30 SRBD events during an 8-hour HST recording will have an index of 3.75 events per hour (which does not reach diagnostic criteria). However, if he or she slept only 5 hours during the recording, the actual SRBD rate would be 6 per hour (which would meet diagnostic criteria). Another factor relates to information sparseness. Level-III HST devices only detect
Fig. 2 Oxyhemoglobin saturation, leak, sleep-disordered breathing events, and pressure for a patient with (right panel) and without (left panel) episodic hypoxemia. Seminars in Respiratory and Critical Care Medicine
Vol. 35
No. 5/2014
Downloaded by: WEST VIRGINIA UNIVERSITY. Copyrighted material.
564
Tele-Sleep for Sleep Apnea
When Is APAP-Oxy Appropriate? In 2002, the AASM standards of practice guidelines recognized APAP as a treatment option to initially determine pressures for fixed CPAP or APAP.4 However, APAP guidelines excluded patients with congestive heart failure (CHF), lung disease (e.g., COPD), and other conditions associated with oxyhemoglobin desaturation. Patients with a history of palate surgery were also excluded. These exclusions arise from the fact that machines at that time (1) had no capability for monitoring oxyhemoglobin desaturations, (2) had not facility to distinguish obstructive from central apnea (and therefore would increase pressure unnecessarily), and (3) some machines used vibration in their algorithm when adjusting pressures (and palatal surgery would thus compromise titration functionality). During the past decade, most of these problems have been obviated. Several current major APAP manufacturers now incorporate optional oximetry. At least two current APAP systems use pressure or ultrasonic pulses to
565
test airway patency to avoid unnecessarily raising pressures. Finally, modern APAP machines seldom (if ever) use vibration algorithms to determine pressure change. As a result of this advancing technology, our list of “ruleouts” has evolved. Patients with the following comorbid conditions are usually referred for laboratory titration and generally excluded from APAP-Oxy titration: • • • • • •
COPD, daytime hypoxemia, or hypercapnia CHF or other heart disease Morbid obesity (BMI 40) Prior uvulopalatopharyngoplasty (UPPP) Opioid analgesics use Neuromuscular, neurological, or neurodegenerative disease; history of stroke; or seizure disorder • Mental or physical conditions compromising machine and interface use.
HST produces several possible outcomes. For patients who do not have any APAP rule-outs, ►Table 2 enumerates possible HST outcomes and primary treatment recommendations. A positive HST with an apnea hypopnea index (AHI) of 10 or more events per hour of recording time justifies attempting out-of-laboratory titration with APAP-Oxy. If AHI ranges upward from 5 but falls below 10 events per hour, we recommended APAP-Oxy when significant oxyhemoglobin desaturation occurs.b Otherwise, in cases with mild SRBD without significant desaturation, oral appliance and nasal expiratory valves are considered.
Adapting Fast-Track for Sleep Apnea to TeleSleep We began with Fast-Track for Sleep Apnea (see ►Fig. 3). ►Fig. 3 shows a schematic of the overall clinical algorithm used in Fast-Track for Sleep Apnea. Our current Tele-sleep program embodies an adaptation and extension of Fast-Track for Sleep Apnea. From a system design approach this required creating a modular structure that was both scalable and manageable within staffing levels provided. Translating and testing protocols for scalability and manageability was required before initial implementation of each of the following processes: (1) clinic visits, (2) personnel training, (3) inventory management, and (4) evaluation of patient satisfaction. Our implementation also requires a high-quality video– audio link between the central site and the outlying clinics. A conference-style room that could accommodate multiple patients was equipped with such a system at each CBOC site. Three systems that could link into these clinics were set up at the central site. At the central site, a split screen displayed all linked in CBOCs. Current, large-screen videophone technology available at our VAMC is more than adequate for our Tele-sleep application. b
We define significant desaturation as an SaO2 nadir at or below 85% or SaO2 at or below 88 for 5 minutes, or more. Other clinicians may chose different criteria based on their population, comfort level, and health concerns. Seminars in Respiratory and Critical Care Medicine
Vol. 35
No. 5/2014
Downloaded by: WEST VIRGINIA UNIVERSITY. Copyrighted material.
apnea and hypopnea. Respiratory effort related arousal (RERA) cannot be scored because electroencephalography is not recorded (making CNS arousals undetectable). Consequently, overall respiratory disturbance index cannot be calculated. Many instruments are available for determining high clinical suspicion for SRBD, including the STOP-BANG test,14 the Berlin Questionnaire,15 and the multivariable apnea prediction scale.16 Other tests also exist. To triage patients for HST referral we use the STOP-BANG because it is simple, embeds easily into our clinical flow, and easy to calculate. The STOP-BANG is an index calculated from eight items for which you assign one point for each proposition affirmed. The eight items are (1) do you snore loudly (loud enough to be heard through closed doors), (2) are you tired, fatigued, or sleeping during the daytime, (3) has anyone observed you stop breathing in your sleep, (4) do you have or are you being treated for high blood pressure, (5) is your body mass index (BMI) greater than 35 kg/m, (6) are you older than 50 years, (7) is your neck circumference greater than 40 cm, and (8) are you male. The original article indicates that a total score above 3 indicates a high risk for SRBD; however, we began by using a cutting score of 5 for HST referral with good effect. We subsequently lowered it to 3 with little loss of diagnostic efficacy. In addition, we obtain information about BMI, sex, and hypertension from the electronic medical record; the remaining items are determined by questioning the patient. The greatest challenge for HST revolves around acquiring usable data. Patients self-attach sensors and sleep in an uncontrolled environment. Air flow sensors failure renders a test uninterpretable but thankfully that seldom occurs. Our biggest problem, however, involves oximetry. When the oximeter fails to record, hypopnea cannot be scored. Therefore, unless the patient has mostly apnea episodes or has very severe SRBD, a false-negative test results. Oximetry traces can begin well and then disappear, record intermittently, or be absent throughout. The clinician reviewing an HST must consider the recordings oximetry status and extrapolate results, when needed.
Hirshkowitz, Sharafkhaneh
Tele-Sleep for Sleep Apnea
Hirshkowitz, Sharafkhaneh
Table 2 APAP machine informatics needed for the Fast-Track for Sleep Apnea Tele-sleep program Parameter
Description
Number of flow cessations
Flow cessations are the absence of airflow for 10 seconds, or more. This parameter closely approximates what is more commonly called an episode of sleep apnea.
Number of flow cessations judged to be nonobstructive
Some APAP systems have algorithms to help determine if a flow cessation is obstructive or nonobstructive. While the reliability may vary, this information may be useful clinically.
Number of flow limitations or flow-curve flattenings
Flow reduction is defined differently by different manufactures. Moreover, most systems incorrectly refer to flow reductions as sleep-related hypopnea.a In general, these events constitute decreased signal amplitude or flow-curve flattenings derived from pressures in the device’s pneumotachometer.
FLC index
Flow cessations þ flow limitations per hour of time on mask. This parameter is usually (and erroneously) labeled apnea þ hypopnea index.
Mean or median pressure
Mean and median pressures serve as estimates of the central tendency of a patient’s pressure requirement. In principle, the machine will dwell at a particular pressure until flow cessations or limitations occur. Thus, the average pressure provides a global representation of a night’s actuarial outcome. Since the APAP machine raises and lowers pressures in response to flow-curve changes, nightly the mean or median can also inform a clinician about night-tonight variability in pressure needs. Some systems provide the mean (parametric) while others report median (nonparametric); either appears sufficient in this application.
90th percentile pressureb
The 90th percentile pressure indicates at what pressure the machine dwelled at or below for 90% of the time. Thus, it represents one way to estimate a single pressure that should relieve the vast majority of the patient’s flow cessations and limitations (as successfully as possible using the machine’s titration algorithm).
Maximum pressure
The maximum pressure is the highest pressure the machine self-adjusted to during the course of a night.
Abbreviations: APAP, automatic self-adjusting positive airway pressure; FLC, flow limitation and cessation. a See text for details relating to this issue. b Some APAP machines provide the 95th percentile pressure.
Clinic Visits The core process guiding overall system design revolved around defining procedures conducted during each clinical visit. All other processes were derived from there. Consequently, our starting point was to adapt and formalize protocols and processes to handle (1) HST setup visits, (2) HST follow-up visits, (3) APAP-Oxy initial setup visits, (4) APAPOxy first follow-up visits, (5) routine follow-up visits, and (6) re-titration follow-up visits.
HST Setup Visit Our Tele-sleep HST setup visit involves a sleep specialist and HST specialist at the central site and an HST specialist and 5 to
Lab PSG
Y STOP BANG
A Telemedicine Program for Diagnosis and Management of Sleep-Disordered Breathing: The Fast-Track for Sleep Apnea Tele-Sleep Program Max Hirshkowitz, PhD, DABSM1,2
Amir Sharafkhaneh, MD, PhD, DABSM1
1 Department of Medicine, Baylor College of Medicine, Houston, Texas 2 Sleep Disorders and Research Center, Michael E. DeBakey Veterans
Address for correspondence Max Hirshkowitz, PhD, DABSM, 4380 Vine Hill Road, Sebastopol, CA 95472 (e-mail: [email protected]).
Affairs Medical Center, Pulmonary, Critical Care & Sleep Section, Houston, Texas
Abstract
Keywords
► sleep apnea ► home sleep testing ► positive airway pressure ► telehealth ► tele-medicine
The objective of this study was to facilitate access to sleep health care for veterans. We designed and implemented a Telehealth program for diagnosing and treating sleeprelated breathing disorders (SRBDs). Building on our ongoing out-of-laboratory “Fast Track for Sleep Apnea” program, procedures were modified to accommodate remote operations. This Tele-sleep program was set up at the medical center’s communitybased outpatient clinics. Home sleep testing and positive airway pressure device technological advances enabled realizing this application for Telehealth. In addition to obtaining appropriated teleconferencing equipment, the program involved implementing systematic processes for (1) six types of clinic visits, (2) training remote-site personnel, (3) making recommendations for inventory management, and (4) evaluating patient satisfaction. Over the past year, we have updated and refined our procedures to optimize program performance and efficiency. To achieve the next step, that is, increasing program scale beyond its current state (e.g., to region-wide), we will need to further develop and formalize quality control indicators to more efficiently monitor operations. The program has helped relieve clinical load at the central sleep program, improved local access to sleep care for veterans, and improved patient satisfaction with health care for SRBDs.
Sleep-related breathing disorders (SRBDs) include a spectrum of respiratory conditions occurring during and disrupting sleep. The condition manifests as cessations of breathing (for 10 or more seconds), shallow breaths provoking oxyhemoglobin desaturations, and/or partial airway obstructions leading to respiratory effort–related arousals.1 Patients may also suffer from hypoxemia, sometimes with significant time spent at or below 88% oxyhemoglobin saturation levels. The traditional diagnostic approach involves making recordings of brain, movement, cardiographic, and respiratory activity while the patient sleeps.2 These recordings typically
Issue Theme Clinical Consequences and Management of Sleep Disordered Breathing; Guest Editors, Ravi Aysola, MD, and Teofilo L. Lee-Chiong Jr., MD
involve the patient sleeping overnight in a hospital or freestanding clinic’s laboratory. Another overnight sleep study usually follows to adjust and optimize treatment with positive airway pressure. Technological advances now make it possible to diagnose and titrate some patients at home using portable devices.3,4 Thus, application of these techniques may improve access to care, especially in rural localities. Telehealth, in general, seeks to facilitate access to care, change the location where care is provided, and improve care management using health informatics and telecommunication technologies.5 The Veterans Affairs (VA) Telehealth
Copyright © 2014 by Thieme Medical Publishers, Inc., 333 Seventh Avenue, New York, NY 10001, USA. Tel: +1(212) 584-4662.
DOI http://dx.doi.org/ 10.1055/s-0034-1390069. ISSN 1069-3424.
Downloaded by: WEST VIRGINIA UNIVERSITY. Copyrighted material.
Semin Respir Crit Care Med 2014;35:560–70.
Service is a leader in deploying this approach in a wide variety of medical specialties. More specifically, Telehealth uses “videoconferencing, the internet, store-and-forward imaging, streaming media, and terrestrial and wireless communications.”6 Telemedicine focuses more narrowly on remote diagnosis and treatment (also called Telecare). A randomized controlled trial in the United Kingdom found Telemedicine for diabetes, chronic heart failure, and chronic obstructive pulmonary disease (COPD) reduced mortality by 45%, emergency admissions by 20%, and bed days by 14%.7 Studies also show economic advantage. For example, analysis of the Alaska Federal Health Care Access Network reported an $8.5 million dollars Medicaid saving for travel costs alone in 2012.8 SRBD represents the primary indication for ordering a sleep study.9 Laboratory polysomnographic testing requires specialized equipment and highly trained personnel; consequently, it is a limited resource. Demand for SRBD diagnosis steadily increased during the past two decades making it difficult to provide timely service. This situation prompted our developing a home sleep testing (HST) program that we later coupled with an automatic self-adjusting positive airway pressure (APAP) program. In a sense, one objective was to unburden the sleep laboratory with straightforward, uncomplicated SRBD cases that could be diagnosed and managed with clinical and home-based procedures.10 The other objective was to improve access to care by shifting point-of-service to the local outpatient clinics. The remainder of this article will describe the rational and nuts-and-bolts of the Fast-Track for Sleep Apnea Tele-sleep program implemented at the Michael E. DeBakey VA Medical Center to serve the community-based outpatient clinics (CBOCs) associated with our program.
Technologies Involved Home Sleep Testing Technology Portable sleep recording technology evolved over many years culminating in small, reliable, lightweight cardiopulmonary Holter-like devices. Typically, these cardiopulmonary recorders (CPRs) monitor and digitally store one to several nights of data for subsequent review using standard computer systems. Most qualify as Level-III devices according to American Academy of Sleep Medicine (AASM) criteria because they provide detailed, viewable “raw-data” of the individual’s (1) airflow, (2) respiratory effort, (3) snore sounds, and (4) oxyhemoglobin saturation. Surrogate measures of airflow usually derive from thermistor and/or nasal pressure changes; effort from diaphragmatic movement, snore sounds from a microphone, and oxyhemoglobin saturation from a pulse oximeter. Body position and heart rate (and/or cardiac rhythm) are also often included. For a comparison of devices, see the review by Hesselbacher et al.11 CPRs provide adequate information to score apnea and hypopnea episodes. The clinician can also determine whether events are obstructive or nonobstructive. Because CPRs are not equipped with electroencephalographic channels, actual sleep time and central nervous system (CNS) arousals cannot be scored. Consequently, apnea and hypopnea must be in-
Hirshkowitz, Sharafkhaneh
dexed by time in bed rather than total sleep time. Furthermore, neither respiratory effort–related arousals nor overall respiratory disturbance index can be tabulated. These limitations potentially reduce sensitivity by diluting the metrics applied against diagnostic criteria. Factors to consider when selecting a device include (1) diagnostic sensitivity, (2) the typical severity of SRBD in patients in your practice, (3) device durability, (4) simplicity of use, (5) initial cost, and (6) recurring costs for supplies and sensors. Our sleep medicine programs at the VA currently service mainly men, age 50 or older, who often have hypertension. Thus, we not only expect generally high pretest probability for sleep apnea but also a high volume needing testing. Consequently, we could trade off some diagnostic sensitivity for lower consumable supply costs. Devices need to be durable but we also expected our population to be somewhat adept at learning to operate the device (having experience being trained to use a variety of devices in their armed forces tours).
Automatic (Self-Adjusting) Positive Airway Pressure Technology Continuous positive airway pressure as a treatment for obstructive sleep apnea was pioneered by Sullivan et al in 1981. 12 They were able to eliminate sleep-associated oropharynx collapse using positive pressure when pressure was raised high enough to pneumatically splint the airway. Their discovery revolutionized sleep apnea treatment; however, determining the appropriate pressure involved patients sleeping in the laboratory to accomplish efficacious titration. As microprocessor technology progressed, it became feasible to digitally monitor a PAP machine’s pressure and the pressure change oscillation associated with inhalation and exhalation. These oscillations provided a surrogate measure for airflow allowing a device to simulate laboratory titration by increasing pressure when the flow signal flattened or ceased. Such self-adjusting PAP devices would also be expected to accommodate to changing pressure needs (e.g., raising pressures when the sleeper is supine compared with laterally recumbent). More advanced models incorporate oximetry to continuously record oxyhemoglobin saturation level.
Home Sleep Testing Informatics for Level-III Cardiopulmonary Recorders Key parameters derive from HST devices relate to diagnosis. Compared with laboratory polysomnography (PSG), the HST provides a small subset of information describing sleep. In some way, it is like driving at night—even though your visual field is extremely impoverished compared with the daytime, enough information exits to manage the task. Nonetheless, these recorders still provide more than adequate of information with six parameters providing pivotal metrics (see ►Table 1). Key parameters from the patient history and interview include comorbid conditions and sleepiness. Sleepiness can derive from the clinician’s impression or based on a standardized sleepiness scale. Seminars in Respiratory and Critical Care Medicine
Vol. 35
No. 5/2014
561
Downloaded by: WEST VIRGINIA UNIVERSITY. Copyrighted material.
Tele-Sleep for Sleep Apnea
Tele-Sleep for Sleep Apnea
Hirshkowitz, Sharafkhaneh
Table 1 HST level-III device informatics needed for the Fast-Track for Sleep Apnea Tele-sleep program Parameter
Description
Rational for use
TRT
The total number of minutes that the HST device was used and data were actually recording.
To assure diagnostic integrity, a minimum of 2 hours is required. However, we typically feel more comfortable when 4, or more, hours of data are available.
AHIa
The number of sleep apnea and sleep hypopnea episodes per hour of TRT. Where apnea is defined as a 10-second (or longer) 90% drop (or greater) in the nasal/oral airflow channel’s peak-to-trough amplitude persisting for at least 90% of the event’s duration and hypopnea is defined as a 10 seconds, 30% drop in nasal pressure signal amplitude (compared with baseline) persisting for 90% of the event’s duration and associated with a 4% drop in O2.
The AHI represents the pivotal diagnostic criteria for SRBD. When AHI is 5 or above, the patient can be diagnosed with a sleep-related breathing disorder. A noncentral AHI of 10, or more, provides justification for APAP titration. Treatment recommendations for AHI between 5 (inclusive) and 10 (exclusive) involve consideration of other.
Proportion of central SRBD events to total number of events
Each apnea should be classified as either central (no respiratory effort present during flow cessation) or noncentral (respiratory effort present during at least part of the flow cessation time).
The proportion of central to all events provides the basis for differentially diagnosing central, complex, or obstructive sleep apnea. Nonobstructive types of SRBD constitute rule-outs for APAP therapy.
TOxT
The number of minutes during the study that the oximeter was functioning.
This parameter is used for validation. There is no agreed upon minimum; however, oximetry is needed to score hypopnea. Consequently, the lower TOxT, the higher likelihood that AHI (and overall SRBD severity) is underestimated.
Time below 88% (time 88% Sat)
The number of minutes that SaO2 was at or below 88% oxyhemoglobin saturation level.
This parameter is used to determine if the patient also has hypoxemia.
SaO2 Nadir
The lowest valid oxyhemoglobin saturation level observed during the recording.
This parameter helps define severity of hypoxemia (when present).
Abbreviations: AHI, apnea þ hypopnea index; APAP, automatic self-adjusting positive airway pressure; HST, home sleep testing; SRBD, sleep-related breathing disorders; TOxT, total oximetry time; TRT, total recording time. a The definitions here are derived from the AASM Scoring Manual.
PAP Informatics One key development making Tele-sleep possible involves advances in informatics available from APAP devices. Critical data examined in our program address (1) therapeutic adherence, (2) sleep-disordered breathing event rate (index), (3) mask leak rates, and (4) APAP with concurrent oximetry. These data allow clinicians to properly follow-up and care for patients in accordance with currently accepted standards of practice.
available. These data serve several purposes, including (1) allowing clinicians to gauge overall therapeutic adherence, (2) allowing clinicians to determine if enough data are available to make appropriate machine adjustments, (3) providing statistics required for Centers for Medicare and Medicaid Services (CMS) “compliance” reports, and (4) representing regulatory used for licensing in the transportation industry.
Sleep-Disordered Breathing Indices Therapeutic Adherence A primitive way to determine usage involves recording the machine’s blower’s total operating time. However, an audit trail of actual usage, sometimes called “time-on-mask,” represents the required statistic. Time-on-mask is determined by the presence of an oscillating back pressure sensed by the machine. This back pressure, when fluctuating at a frequency consistent with breathing and/or contiguous with sleepdisordered breathing events, confirms usage. Time-on-mask should be recorded and stored for a minimum of 90 days; be displayed graphically in 24-hour blocks; and be time and date stamped. A daily 24-hour usage amount; a total and mean use per day over some user-specified time frame; and the percentage of days with 4 hours of use (or more) must be Seminars in Respiratory and Critical Care Medicine
Vol. 35
No. 5/2014
Sleep-disordered breathing events detectable with APAP machines include flow cessations and flow limitation. If the device is equipped with a plug-in oximeter, data concerning oxyhemoglobin desaturation events, the lowest saturation level, and the amount of time spent at or below 88%a saturation are also available. It is critically important to not equate events detected by APAP machines with those scored according to standard technique (AASM, 2007). Confusion arises because APAP machines label flow cessations “apneas” and flow limitations as “hypopnea.” Luckily, APAP scored flow cessations correlate a
Some systems report 85% or 90%. We prefer 88% because it defines respiratory insufficiency.
Downloaded by: WEST VIRGINIA UNIVERSITY. Copyrighted material.
562
highly with apnea scored on PSGs and HSTs. The same is not true for hypopnea. APAP machines regard decreased flow as indicative of hypopnea with no regard to oxyhemoglobin saturation or CNS arousal. While this may accurately accord with physiologic conceptualizations of hypopnea, they do not properly represent the sleep pathophysiology designated as hypopnea. CMS considers sleep-related hypopnea pathophysiological because they are associated with a 4% (or greater) oxyhemoglobin desaturation. AASM also includes flow limitations that provoke CNS arousals as pathophysiological because they fragment sleep. Simple flow reductions, without hypoxemia or sleep disturbance, are normal (you have them when you eat and speak). These simple flow reductions are what APAP machines report. Therefore, clinicians must be vigilantly aware that terms are being used differently, to avoid making therapeutic errors. This unfortunate misuse of the terms sleep-related apnea and hypopnea has created serious ambiguity. ►Table 1 summarizes parameters needed for our Telesleep program. The summary data listed and described on the table should be available for each night and means should be calculated over any operator-selected range of nights.
Mask Leak Rates Every mask intrinsically leaks through exhalation ports. Leak rate depends on the type of mask and the pressure setting. For example, at 10 cmH2O pressure, Respironics ComfortGel Blue and ResMed SoftGel port 26 and 31 lpm, respectively. By contrast, these two popular masks port 37 and 45 lpm at 20 cmH2O pressure. Overall, to our knowledge, no popular adult
Hirshkowitz, Sharafkhaneh
masks or nasal pillows port more than 55 lpm at pressures up to 20 cmH2O pressure. Therefore, we typically regard leaks greater than 60 lpm as significant. Total time with significant leak becomes important when interpreting pressure and sleep-disorder breathing event statistics. When leak rate increases, many APAP machines will interpret the change as a flow limitation and raise the pressure. The Tele-sleep clinician therefore needs a graphic representation illustrating leak rates on the same time axis as pressure setting, flow cessations, nonobstructive flow cessations, and flow limitations (see ►Fig. 1). This type of graphic allows quick assessment of a night’s data integrity and interpretability. For example, if maximum pressure occurred during leak, it can be disregarded. We also commonly see high leak episodes marked by sleep-disordered breathing events increase (in response to which the machine raises pressures). This pattern usually inflates automatically calculated indices (that do not consider leak rates). Clinical judgment must then be applied when selecting prescriptive pressure settings.
APAP with Concurrent Oximetry (APAP-Oxy) APAP machine capability to concurrently record oximetry represented a major advance for developing our Tele-sleep program. A large proportion of patients were excluded from our initial APAP program because of practice standards concerns. Indeed, the AASM APAP practice guideline (which was largely authored by VA members of the AASM including myself) excluded patients with lung disease and other conditions associated with oxyhemoglobin desaturation. The rationale was clear for this rule-out. A patient might feel
Fig. 1 Leak, sleep-disordered breathing events, and pressure across the night.
Seminars in Respiratory and Critical Care Medicine
Vol. 35
No. 5/2014
563
Downloaded by: WEST VIRGINIA UNIVERSITY. Copyrighted material.
Tele-Sleep for Sleep Apnea
Tele-Sleep for Sleep Apnea
Hirshkowitz, Sharafkhaneh
much better, have greatly reduced apnea and hypopnea, but still have severe hypoxemia during rapid eye movement (REM) sleep and/or when sleeping supine, and the clinician would never know. Once concurrent oximetry became available, this problem was rendered moot. Having the oximetry graphically displayed along the same axis as pressures, sleepdisordered breathing events, pressure, and leak allow clinicians to rapidly determine if persistent hypoxemia exists. A quick glance at each night’s histogram adequately satisfies such concerns (see ►Fig. 2).
Program Development Developing Fast-Track for Sleep Apnea Protocols— General Approach Traditional diagnosis and treatment for SRBD is a costly and time-consuming endeavor. Over time, we phased-in HST and APAP programs to streamline the process in our sleep center so that we could provide care more efficiently. The groundwork for this development effort involved our clinical sleep specialists setting up clinics to follow-up patients undergoing HST diagnosis and APAP interventions. In this manner, we could keep close watch on these new procedures and troubleshoot any problems directly and immediately. We thereby gained the necessary insight into what worked and what to monitor for quality assurance. We compared notes and designed the Fast-Track program for Sleep Apnea. Protocols were then developed and clinical extenders trained. The resulting program was monitored closely for the 1st year and minor adjustments made, as needed. The procedures were gradually adapted for small group sessions (1–6 patients) and ultimately for larger groups (up to 10–20 pa-
tients). The following paragraphs describe the overarching concepts and some of the specifics incorporated into the FastTrack for Sleep Apnea program.
When Is HST Appropriate? Currently, HST is only useful to confirm SRBD diagnosis.3 If there is clinical suspicion for Narcolepsy, Nocturnal Seizures, Parasomnias, or non-SRBD forms of hypersomnolence, HST should not be used. CPRs qualify as Level-III HST devices, according to the AASM classification system. As such, they can detect sleepdisordered breathing events and therefore provide diagnostic information; however, they are less sensitive than laboratory polysomnography. Therefore, only patients with a high pretest probability for SRBD should be referred for testing. HST cannot rule-out SRBD!13 HST is a tool of confirmation; therefore, negative tests must be verified using more sensitive technique (i.e., laboratory-PSG). Thus, referring patients with a low pretest probability for SRBD is inefficient. HST’s lower diagnostic sensitivity compared with PSG stems for several factors. The first is simply arithmetic. Indices are calculated using total recording time (rather than total sleep time) as a denominator. Because total sleep time is always smaller than total recording time (unless sleep efficiency is 100%), the calculated index from Level-III HST recorders are invariably lower. For example, a patient with 30 SRBD events during an 8-hour HST recording will have an index of 3.75 events per hour (which does not reach diagnostic criteria). However, if he or she slept only 5 hours during the recording, the actual SRBD rate would be 6 per hour (which would meet diagnostic criteria). Another factor relates to information sparseness. Level-III HST devices only detect
Fig. 2 Oxyhemoglobin saturation, leak, sleep-disordered breathing events, and pressure for a patient with (right panel) and without (left panel) episodic hypoxemia. Seminars in Respiratory and Critical Care Medicine
Vol. 35
No. 5/2014
Downloaded by: WEST VIRGINIA UNIVERSITY. Copyrighted material.
564
Tele-Sleep for Sleep Apnea
When Is APAP-Oxy Appropriate? In 2002, the AASM standards of practice guidelines recognized APAP as a treatment option to initially determine pressures for fixed CPAP or APAP.4 However, APAP guidelines excluded patients with congestive heart failure (CHF), lung disease (e.g., COPD), and other conditions associated with oxyhemoglobin desaturation. Patients with a history of palate surgery were also excluded. These exclusions arise from the fact that machines at that time (1) had no capability for monitoring oxyhemoglobin desaturations, (2) had not facility to distinguish obstructive from central apnea (and therefore would increase pressure unnecessarily), and (3) some machines used vibration in their algorithm when adjusting pressures (and palatal surgery would thus compromise titration functionality). During the past decade, most of these problems have been obviated. Several current major APAP manufacturers now incorporate optional oximetry. At least two current APAP systems use pressure or ultrasonic pulses to
565
test airway patency to avoid unnecessarily raising pressures. Finally, modern APAP machines seldom (if ever) use vibration algorithms to determine pressure change. As a result of this advancing technology, our list of “ruleouts” has evolved. Patients with the following comorbid conditions are usually referred for laboratory titration and generally excluded from APAP-Oxy titration: • • • • • •
COPD, daytime hypoxemia, or hypercapnia CHF or other heart disease Morbid obesity (BMI 40) Prior uvulopalatopharyngoplasty (UPPP) Opioid analgesics use Neuromuscular, neurological, or neurodegenerative disease; history of stroke; or seizure disorder • Mental or physical conditions compromising machine and interface use.
HST produces several possible outcomes. For patients who do not have any APAP rule-outs, ►Table 2 enumerates possible HST outcomes and primary treatment recommendations. A positive HST with an apnea hypopnea index (AHI) of 10 or more events per hour of recording time justifies attempting out-of-laboratory titration with APAP-Oxy. If AHI ranges upward from 5 but falls below 10 events per hour, we recommended APAP-Oxy when significant oxyhemoglobin desaturation occurs.b Otherwise, in cases with mild SRBD without significant desaturation, oral appliance and nasal expiratory valves are considered.
Adapting Fast-Track for Sleep Apnea to TeleSleep We began with Fast-Track for Sleep Apnea (see ►Fig. 3). ►Fig. 3 shows a schematic of the overall clinical algorithm used in Fast-Track for Sleep Apnea. Our current Tele-sleep program embodies an adaptation and extension of Fast-Track for Sleep Apnea. From a system design approach this required creating a modular structure that was both scalable and manageable within staffing levels provided. Translating and testing protocols for scalability and manageability was required before initial implementation of each of the following processes: (1) clinic visits, (2) personnel training, (3) inventory management, and (4) evaluation of patient satisfaction. Our implementation also requires a high-quality video– audio link between the central site and the outlying clinics. A conference-style room that could accommodate multiple patients was equipped with such a system at each CBOC site. Three systems that could link into these clinics were set up at the central site. At the central site, a split screen displayed all linked in CBOCs. Current, large-screen videophone technology available at our VAMC is more than adequate for our Tele-sleep application. b
We define significant desaturation as an SaO2 nadir at or below 85% or SaO2 at or below 88 for 5 minutes, or more. Other clinicians may chose different criteria based on their population, comfort level, and health concerns. Seminars in Respiratory and Critical Care Medicine
Vol. 35
No. 5/2014
Downloaded by: WEST VIRGINIA UNIVERSITY. Copyrighted material.
apnea and hypopnea. Respiratory effort related arousal (RERA) cannot be scored because electroencephalography is not recorded (making CNS arousals undetectable). Consequently, overall respiratory disturbance index cannot be calculated. Many instruments are available for determining high clinical suspicion for SRBD, including the STOP-BANG test,14 the Berlin Questionnaire,15 and the multivariable apnea prediction scale.16 Other tests also exist. To triage patients for HST referral we use the STOP-BANG because it is simple, embeds easily into our clinical flow, and easy to calculate. The STOP-BANG is an index calculated from eight items for which you assign one point for each proposition affirmed. The eight items are (1) do you snore loudly (loud enough to be heard through closed doors), (2) are you tired, fatigued, or sleeping during the daytime, (3) has anyone observed you stop breathing in your sleep, (4) do you have or are you being treated for high blood pressure, (5) is your body mass index (BMI) greater than 35 kg/m, (6) are you older than 50 years, (7) is your neck circumference greater than 40 cm, and (8) are you male. The original article indicates that a total score above 3 indicates a high risk for SRBD; however, we began by using a cutting score of 5 for HST referral with good effect. We subsequently lowered it to 3 with little loss of diagnostic efficacy. In addition, we obtain information about BMI, sex, and hypertension from the electronic medical record; the remaining items are determined by questioning the patient. The greatest challenge for HST revolves around acquiring usable data. Patients self-attach sensors and sleep in an uncontrolled environment. Air flow sensors failure renders a test uninterpretable but thankfully that seldom occurs. Our biggest problem, however, involves oximetry. When the oximeter fails to record, hypopnea cannot be scored. Therefore, unless the patient has mostly apnea episodes or has very severe SRBD, a false-negative test results. Oximetry traces can begin well and then disappear, record intermittently, or be absent throughout. The clinician reviewing an HST must consider the recordings oximetry status and extrapolate results, when needed.
Hirshkowitz, Sharafkhaneh
Tele-Sleep for Sleep Apnea
Hirshkowitz, Sharafkhaneh
Table 2 APAP machine informatics needed for the Fast-Track for Sleep Apnea Tele-sleep program Parameter
Description
Number of flow cessations
Flow cessations are the absence of airflow for 10 seconds, or more. This parameter closely approximates what is more commonly called an episode of sleep apnea.
Number of flow cessations judged to be nonobstructive
Some APAP systems have algorithms to help determine if a flow cessation is obstructive or nonobstructive. While the reliability may vary, this information may be useful clinically.
Number of flow limitations or flow-curve flattenings
Flow reduction is defined differently by different manufactures. Moreover, most systems incorrectly refer to flow reductions as sleep-related hypopnea.a In general, these events constitute decreased signal amplitude or flow-curve flattenings derived from pressures in the device’s pneumotachometer.
FLC index
Flow cessations þ flow limitations per hour of time on mask. This parameter is usually (and erroneously) labeled apnea þ hypopnea index.
Mean or median pressure
Mean and median pressures serve as estimates of the central tendency of a patient’s pressure requirement. In principle, the machine will dwell at a particular pressure until flow cessations or limitations occur. Thus, the average pressure provides a global representation of a night’s actuarial outcome. Since the APAP machine raises and lowers pressures in response to flow-curve changes, nightly the mean or median can also inform a clinician about night-tonight variability in pressure needs. Some systems provide the mean (parametric) while others report median (nonparametric); either appears sufficient in this application.
90th percentile pressureb
The 90th percentile pressure indicates at what pressure the machine dwelled at or below for 90% of the time. Thus, it represents one way to estimate a single pressure that should relieve the vast majority of the patient’s flow cessations and limitations (as successfully as possible using the machine’s titration algorithm).
Maximum pressure
The maximum pressure is the highest pressure the machine self-adjusted to during the course of a night.
Abbreviations: APAP, automatic self-adjusting positive airway pressure; FLC, flow limitation and cessation. a See text for details relating to this issue. b Some APAP machines provide the 95th percentile pressure.
Clinic Visits The core process guiding overall system design revolved around defining procedures conducted during each clinical visit. All other processes were derived from there. Consequently, our starting point was to adapt and formalize protocols and processes to handle (1) HST setup visits, (2) HST follow-up visits, (3) APAP-Oxy initial setup visits, (4) APAPOxy first follow-up visits, (5) routine follow-up visits, and (6) re-titration follow-up visits.
HST Setup Visit Our Tele-sleep HST setup visit involves a sleep specialist and HST specialist at the central site and an HST specialist and 5 to
Lab PSG
Y STOP BANG
