Is home testing as good as in-lab PSG? A review of recent home testing literature attempts to answer the question.
By Dennis Hwang, MD
Use of home sleep testing (HST) is increasing, and many predict that it may eventually become the primary testing modality over in-laboratory polysomnography (PSG). Thus, a critical question must be asked: “Does the evidence support the use of home sleep testing?” Early studies focused on validating portable monitoring (PM) technology in regard to accuracy of diagnosing obstructive sleep apnea (OSA) and identifying severity. Auto-titrating CPAP (APAP) technology has likewise been validated as an effective titration method and also as a primary treatment modality.1,2 Recent research in home sleep testing has evolved into trying to answer questions relevant to delivery of care. These questions include: 1) How should HST be applied in clinical practice, and are the clinical outcomes similar compared to PSG? 2) Is HST truly more cost efficient than in-lab PSG? Recent articles by Kuna et al, Rosen et al, Masa et al, and Pietzsch et al attempt to answer these questions and further establish under what conditions use of HST is appropriate.
Noninferiority of Functional Outcome in Ambulatory Management of Obstructive Sleep Apnea
The objective of this study3 was to compare functional outcomes and CPAP adherence in patients with suspected OSA randomized into a home pathway versus a PSG pathway. For the study, 143 patients underwent home sleep testing; 108 (75%) had at least moderate OSA (AHI ?15) and underwent home APAP titration for 4 to 5 nights, while 35 (25%) had an AHI less than 15 and underwent subsequent PSG with most resulting in a diagnosis of OSA and started on CPAP (either qualifying for split-night or after a home APAP titration). These patients were included in the home pathway group as an intention-to-treat analysis. In all, 113 (79%) in the home pathway were started on CPAP. The PSG pathway originally started with 148 patients, and 110 (74%) were initiated on CPAP after split-night or dedicated CPAP titration PSG.
While both groups demonstrated a significant improvement at 3 months follow-up in the Functional Outcomes of Sleep Questionnaire (FOSQ), Epworth Sleepiness Scale (ESS), Psychomotor Vigilance Test (PVT), and questionnaires evaluating physical and mental health, there was no difference in degree of effect between the two pathways. CPAP adherence may have actually favored the home sleep testing group with 3.5 hours compared to 2.9 hours of average daily use (P=0.08). All patients were treated with fixed CPAP with an average pressure of 11.1 in the home pathway versus 9.4 in the in-lab pathway (P<0.001).
The authors concluded that the home testing pathway is not clinically inferior to the traditional PSG pathway. It should be noted that the patients were enrolled from two Veterans Health Administration (VA) facilities and were almost exclusively middle-aged obese men, thus this may limit the ability to generalize the conclusions to other populations.
A Multisite Randomized Trial of Portable Sleep Studies and Positive Airway Pressure Autotitration Versus Laboratory-Based Polysomnography for the Diagnosis and Treatment of Obstructive Sleep Apnea: The HomePAP Study
The HomePAP Study4 is another recent landmark study comparing CPAP adherence and functional outcomes between home sleep testing and PSG pathways. In the study, 373 patients were randomized into the two pathways from seven sleep centers and prescreened for moderate to severe OSA using a clinical algorithm based on symptoms, medical history, and physical parameters.
Like the VA study, those with moderate to severe OSA on a portable monitor underwent a home APAP titration (5-7 days), while those with an AHI less than 15 received in-lab PSG with subsequent home APAP titration if repeat study revealed at least moderate OSA. Those with unsuccessful home APAP titration underwent in-lab CPAP titration but were analyzed as part of the home sleep testing group. Unlike the VA study, which included all patients on CPAP regardless of OSA severity, this study included only moderate to severe OSA patients.
There was no difference between the two groups comparing titration success and CPAP acceptance. About 92% of patients in both pathways with moderate to severe OSA successfully completed a titration study, and 94% in both groups accepted CPAP therapy. Three-month follow-up showed patients in the home pathway had better “average nightly usage” (281 versus 219 minutes; P=0.01), better “nights used at least 4 hours” (63% versus 49%; P=0.02), and better results for “percent of patients meeting Medicare compliance” (50% versus 39%; P=0.22).
Interestingly, both pathways had similar CPAP adherence at 1 month, but while adherence stayed about the same when rechecked at 3 months in the PSG pathway, adherence actually improved in the home pathway by about 15%. The reason for this is unclear; while the home pathway had double the number of patient phone calls for trouble-shooting during the follow-up period (82 versus 45) and one can speculate this additional education could enhance adherence, the vast majority of patients in either arm did not make any phone calls. As in the VA study, there was significant improvement in functional outcomes (including ESS and FOSQ) without difference between the two groups. Cost analysis was performed taking into account cost of PSG, home testing, repeat studies, and follow-up visits or phone calls. Total cost was 25% less in the home testing pathway ($139,000 versus $186,000).
The authors concluded that a home-based strategy for diagnosis and treatment of OSA is not inferior to the traditional in-lab pathway in regard to CPAP adherence, titration pressures, time to treatment, and functional outcomes. It should be noted that the demographic data of the study patients were more diverse and may be more representative of a typical sleep center referral population.
Therapeutic Decision-making for Sleep Apnea and Hypopnea Syndrome Using Home Respiratory Polygraphy—A Large Multicentric Study
The objective of this study5 from the Spanish Sleep Network was to investigate whether clinical decision-making would differ if using the results of portable monitoring versus PSG.
For the study, 366 patients with suspected OSA underwent both portable monitoring and PSG within 3 days of each other—about half with portable monitoring first and half with PSG first. Research clinicians were given a patient’s baseline clinical history with their portable monitor results and asked to recommend “CPAP,” “no CPAP,” or “impossible decision” (although impossible decision was never selected). At a different time in a random sequence, the same patient history and PSG results were given and asked to repeat therapy recommendation. This exercise was repeated 3 months later. Under the assumption that PSG is the gold standard, the agreement in therapy decision using the PSG results between the two time points was used to create a reference. The degree of agreement was slightly lower (about 90%) when AHI was lower and higher (over 95%) when AHI was higher, but overall agreement was good. On the other hand, the agreement in therapy recommendation based on portable monitoring compared to PSG was only 76%. However, the degree of agreement improved with increasing sleep apnea severity; when the AHI ?30, agreement was excellent (91%) and nearly matched the reference.
The results of this study largely mimic those of a similar study a group of researchers performed at New York University.6 The study methods were similar and also found excellent agreement in diagnosis and therapy decisions between portable monitoring and PSG when the respiratory index was high and poor when the index was low. However, the study additionally found that agreement among research clinicians was low even when only gold standard PSG results were used when respiratory index was low. The study concluded that clinical decision-making was more confident when significant OSA is present. When less OSA is present, greater variability in clinical judgment is present, not due to the type of test performed but due to greater reliance on clinical history, which is more subject to variable interpretation. The study’s conclusions may be applicable to the Spanish study, although the Spanish study had better agreement when only PSG results were used. This may be due to their limited treatment options (“CPAP” or “no CPAP”) from which the research clinicians could choose while the NYU study had multiple diagnostic and therapy options.
An Integrated Health-Economic Analysis of Diagnostic and Therapeutic Strategies in the Treatment of Moderate-to-Severe Obstructive Sleep Apnea
This study7 was a computer-based model comparing cost and effectiveness of 1) full-night PSG, 2) split-night PSG, and 3) portable monitoring diagnostic pathways in managing patients specifically with moderate to severe OSA. In the portable monitoring pathway, the diagnostic study was assumed to be definitive in ruling in or ruling out moderate to severe OSA without need for PSG even for patients with AHI less than 15. Assumptions were made about the prevalence of OSA (50%), cardiovascular disease risk and motor vehicle risk attributed to OSA, cost of sleep studies and follow-up management, true positive and negative rates for moderate to severe OSA with each diagnostic pathway, technical failure rates for home sleep testing (22%), initial acceptance rate of CPAP therapy, extended CPAP compliance (68%), and quality of life measures.
This model determined that over a lifetime horizon, CPAP versus no treatment had an incremental cost-effectiveness ratio (ICER) of $15,915 per quality-adjusted life year (QALY), which is the cost of adding a year of life weighted by quality of life factors. Interventions that have an ICER less than $50,000 are considered to be cost-effective, thus CPAP is very cost-effective based on this standard if the underlying assumptions are accurate. Furthermore, the study determined that full-night PSG is more cost-effective than split-night and home testing due to “superior diagnostic accuracy” and limiting false-positive and false-negative studies that result in unnecessary therapies and untreated cardiovascular risk.
These studies suggest that home sleep testing is as good as the traditional PSG pathway, and CPAP adherence may actually be better with the home pathway. However, a few questions remain:
1) The studies primarily focused on moderate to severe OSA, and the relevancy of the findings to mild OSA is uncertain. In both of the comparative effectiveness studies, in-lab PSG was performed on all patients with an AHI less than 15 to limit unrecognized moderate to severe OSA, but is this really necessary? Considering the night-to-night variability inherent with OSA, maybe the argument should not be portable monitoring versus PSG, but rather whether multinight portable monitoring should be recommended to increase the sensitivity of identifying moderate OSA.
2) What is the true cost-effectiveness of home sleep testing? The HomePAP study showed a lower cost with home testing, while the Pietzsch model suggests that PSG may be more cost-effective, despite higher up-front costs, by limiting extended health risks and associated health care utilization assuming superior diagnostic accuracy.
3) Finally, home sleep testing versus PSG needs to be further evaluated taking into account the delivery of care model. Effectiveness of portable monitoring used within primary care clinic models needs to be considered,8,9 especially in remote areas without access to tertiary care. In an excellent commentary by Dr Ayas, where he challenged the accuracy of several of Pietzsch’s assumptions, his final statement seems to be an appropriate conclusion, “The fundamental issue is not whether ambulatory studies are better or worse than PSG, but rather under what conditions use of home studies [is] appropriate.”10
Dennis Hwang, MD, is medical director of the Sleep Medicine Department at the Kaiser Permanente Fontana Medical Center and co-chair of sleep medicine for the Southern California Permanente Medical Group. He is also faculty at the NYU School of Medicine, Department of Pulmonary, Critical Care, and Sleep Medicine.
1. McArdle N, Singh B, Murphy M, et al. Continuous positive airway pressure titration for obstructive sleep apnoea: automatic versus manual titration. Thorax. 2010;65:606-611.
2. Meurice JC, Cornette A, Philip-Joet F, et al; ANTADIR “PIC” Working Group. Evaluation of autoCPAP devices in home treatment of sleep apnea/hypopnea syndrome. Sleep Med. 2007;8(7-8):695-703.
3. Kuna ST, Gurubhagavatula I, Maislin G, et al. Noninferiority of functional outcome in ambulatory management of obstructive sleep apnea. Am J Respir Crit Care Med. 2011;183:1238-1244.
4. Rosen CL, Auckley D, Benca R, et al. A multisite randomized trial of portable sleep studies and positive airway pressure autotitration versus laboratory-based polysomnography for the diagnosis and treatment of obstructive sleep apnea: the HomePAP Study. Sleep. 2012;35(6):757-767
5. Masa JF, Corral J, Pereira R, et al. Therapeutic decision-making for sleep apnea and hypopnea syndrome using home respiratory polygraphy—a large multicentric study. Am J Respir Crit Care Med. 2011;184:964-971.
6. Masdeu MJ, Ayappa I, Hwang D, Mooney AM, Rapoport DM. Impact of clinical assessment on use of data from unattended limited monitoring as opposed to full-in lab PSG in sleep disordered breathing. J Clin Sleep Med. 2010;6(1):51-58.
7. Pietzsch JB, Garner A, Cipriano LE, Linehan JH. An integrated health-economic analysis of diagnostic and therapeutic strategies in the treatment of moderate-to-severe obstructive sleep apnea. Sleep. 2011;34(6):695-709.
8. Chai-Coetzer CL, Antic NA, Rowland LS, et al. A simplified model of screening questionnaire and home monitoring for obstructive sleep apnoea in primary care. Thorax. 2011;66(3):213-9.
9. Antic NA, Buchan C, Esterman A, et al. A randomized controlled trial of nurse-led care for symptomatic moderate-severe obstructive sleep apnea. Am J Respir Crit Care Med. 2009;179(6):501-8.
10. Ayas NT, Pack A, Marra C. The demise of portable monitoring to diagnose OSA? Not so fast! Sleep. 2011;34(6):691-2.