doi: 10.1161/01.str.0000168860.45858.10
(Stroke. 2005;36:1293.)
© 2005 American Heart Association, Inc.
Research Reports |
Editorial Comment: How Much Is a Good Night’s Sleep Worth?
Health Services Research and Development, Roudebush VAMC, and, Department of Neurology, Indiana University School of Medicine, Indianapolis, Ind
Neurology and Community & Preventive Medicine, University of Rochester School of Medicine and Dentistry, Rochester, NY
As Brown et al point out in this issue of Stroke, sleep disordered breathing is associated both with increased stroke risk and increased poststroke morbidity and mortality.1 Obstructive sleep apnea (OSA) is common after stroke, occurring in as many as 60 to 90% of patients, although the natural history of OSA in stroke patients, especially when OSA is first identified during the acute hospitalization, is not well studied. Because effective treatment, namely continuous positive airway pressure (CPAP), is effective at reducing sleep apnea and has been shown in small studies to be feasible in stroke patients, it is reasonable to consider planning trials to evaluate whether CPAP can improve stroke outcomes and reduce subsequent vascular events.
The authors used decision–analytic modeling to identify the magnitude of benefit that identification and treatment of OSA would need to demonstrate to be considered cost-effective. They estimated the magnitude of benefit of screening and treatment of OSA by using quality-adjusted life years (QALYs), a metric that combines one’s preference for a health state with the time that one lives in that health state. They estimated the cost-effectiveness by calculating an incremental cost-effectiveness ratio or the extra costs to screen and treat OSA compared to not screen and treat to gain QALYs. Although the level at which a treatment is considered cost-effective varies, in the US interventions that cost less than $100 000 to $200 000 per QALY are typically considered cost-effective.2
As far as decision–analytic models go, this model is simple and thus is missing some elements that would improve its face-validity.3 The three-month time horizon is not adequately justified, nor is the exclusion of mortality. Important model inputs include the prevalence estimates of OSA, the estimates of acceptance of CPAP, and the estimates of utility of stroke with and without OSA. Although the authors allowed these estimates to vary in sensitivity analyses, the range of estimates in some cases is not adequately justified. As the Cochrane collaboration points out, acceptance of CPAP is almost certainly lower than published studies suggest, so it would be helpful to allow CPAP acceptance to vary below the stated 50% acceptance rate.4 Importantly, 1 study in a stroke cohort found that only 4 of 34 patients with OSA demonstrated objective compliance with home CPAP over a 3-month period.5 Key variables that were not directly tested in this model include an estimate of the disutility of testing, and the effect of any reduction in sensitivity or specificity of polysomnography. This is especially important in stroke patients where testing may be done during hospitalization when sleep patterns are already disturbed by the associated environmental and possibly pharmacological milieu. Other treatments for OSA have also been shown to be effective and, in some studies, preferred by patients over CPAP, so including models of other treatment strategies would also be informative.6 Finally, a 50% increase in utility (from 0.4 to 0.6) associated with treatment may be a somewhat discouraging target. For example, utility differences for persons with Parkinson disease initially treated with levodopa showed a relative utility increase of only 5 to 10%.7 Further, although studies typically demonstrate reduction in symptoms with treatment of OSA, this symptom improvement does not always translate to a corresponding improvement in utilities.8,9
When evaluating any model, it is instructive to recall, "All models are wrong, but some are useful."10 There is no doubt that this model then, like all models, is wrong, but is it useful? We think it so. Not so much in its ability to predict truth or to inform current clinical decision-making, but rather in its approach to technology assessment. This study is an important illustration of the direction we need to move in considering cost-effectiveness modeling in clinical trial planning and ultimately clinical trial conduct. Pilot clinical trial data from studies in OSA could serve to inform not only the precision of the estimates of acceptance and efficacy of CPAP, but also the expected benefit in utility scores, and could allow for evaluation of various methods of assessing patient utilities which may also affect the ability to detect meaningful group differences. Ideally, clinical trial planning groups would include multidisciplinary teams of health economists and quality of life experts to address this critical issue, to ensure that clinical trials will not only address the question of "does this treatment work," but will also help policy makers and the public with the question of "is the treatment worth the expense?"
 |
References |
|---|
 Top References |
|---|
1. Brown DL, Chervin RD, Hickenbottom SL, Langa KM, Morgenstern LB. Screening for obstructive sleep apnea in stroke patients: a cost-effectiveness analysis. Stroke. 2005; 36: 1291–1294.
2. Ubel PA, Hirth RA, Chernew ME, and Fendrick AM. What is the price of life and why doesn’t it increase at the rate of inflation? Arch Intern Med. 2003: 163; 1637–1641.
3. Kassirer JP, Angell M. The journal’s policy on cost-effectiveness analyses. N Engl J Med. 1994; 331: 669–670.
4. Haniffa M, Lasserson TJ, Smith I. Interventions to improve compliance with continuous positive airway pressure for obstructive sleep apnoea. Cochrane Database Syst Rev. 2004; 4: CD003531.[Medline] [Order article via Infotrieve]
5. Hui DS, Choy DK, Wong LK, Ko RW, Li TS, Woo J, Kay R. Prevalence of sleep-disordered breathing and continuous positive airway pressure compliance: results in Chinese patients with first-ever ischemic stroke. Chest. 2002; 122: 852–860.[CrossRef][Medline] [Order article via Infotrieve]
6. White J, Cates C, Wright J. Continuous positive airways pressure for obstructive sleep apnoea. Cochrane Database Syst Rev 2001;4:CD001106.
7. Noyes K., Dick A., Holloway RG; and Parkinson Study Group. Pramipexole vs levodopa as initial treatment for Parkinson’s disease: a randomized clinical-economic trial. Med Dec Making. 2004; 24: 472–485.
8. Sandberg O, Franklin KA, Bucht G, Eriksson S, Gustafson Y. Nasal continuous positive airway pressure in stroke patients with sleep apnoea: a randomized treatment study. Eur Respir J. 2001; 18: 619–622.
9. Jenkinson C, Stradling J, Petersen S. Comparison of three measures of quality of life outcome in the evaluation of continuous positive airways pressure therapy for sleep apnoea. J Sleep Res. 1997; 6: 199–204[CrossRef][Medline] [Order article via Infotrieve]
10. Box GEP, Hunter WG, and Hunter JS. Statistics for Experimenters: an Introduction to Design, Data Analysis, and Model Building. New York: John Wiley & Sons; 1978.
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||