SUMMARY

The authors consider it clearly demonstrated that, in Sleep Apnea Hypopnea Syndrome (SAHS) of moderate or greater severity, continuous positive airway pressure (CPAP) benefits the major daytime symptoms of sleepiness, impaired cognitive performance, and generally poor well-being. However, they consider the case less well proven for mild SAHS with some daytime symptoms. Their study was designed to test CPAP efficacy in this group. The subjects were recruited from consecutive outpatients referred to the Sleep Clinic for investigation of SAHS. They were required to have two or more symptoms of SAHS and an AHI in the range of 5.0-14.9 apneas+hypopneas per hour of sleep. Twenty-eight subjects were invited to participate, of whom 18 accepted, two declining because of intolerance to CPAP during the pretreatment CPAP titration trial. Furthermore, two more subjects withdrew during the study itself, both while on CPAP, and were excluded from further analysis, one because of intolerance to the CPAP noise. So the final group analyzed consisted of 16 subjects, 12 men and 4 women, averaging in age 52 years (SD=+/-2) with an average Body Mass Index (see Glossary) 0f 29.8 (SD=+/-1.8) and an average AHI (see Glossary) of 11 (SD=+/-1), mean minimum oxygen saturation of 86% (SD=+/-1) and mean rate of desaturations of at least 4% averaging 4 (SD=+/-1) per hour. Their mean Epworth Sleepiness Scale Scores (see Glossary) moderately elevated at 14 (SD=+/-1). In addition, subjects were noted to have an average of 24 “microarousals” per hours of sleep, defined by increased EEG wave frequency lasting at least 1.5 seconds, along with any increase in muscular electrical activity measured by the electromyogram. Two treatment groups of eight subjects each then spent four weeks on CPAP therapy or four weeks on an oral placebo pill (Ranitidine, 300 mg two at bedtime), which patients were told might improve their upper respiratory function. On the final day of the first four weeks, subjects were evaluated for daytime function, then crossed over to the opposite treatment for another four weeks. Prior to treatment on the study, subjects had undergone CPAP titration to establish the therapeutic pressure. They were also contacted two weeks into the CPAP trial to check progress and manage problems. On the assessment day at the conclusion of each treatment period, numerous tests were administered, some familiar to most of us and some quite unfamiliar, presumably of local derivation. I will attempt to describe in the text the latter, rather than set up Glossary descriptions at this point. Sleepiness was assessed using the Multiple Sleep Latency Test (see Glossary), the Epworth Sleepiness Scale (see Glossary reference above), and an “energetic arousal“ scale drawn from a mood adjective checklist. Cognitive tests included subtests excerpted from the Wechsler Adult Intelligence Scale—Revised, specifically Digit Symbol Substitution and Block Design, plus the National Adult Reading Test, TrailMaking Task B, a 30-minute driving-based attention task called SteerClear, a Paced Auditory Serial Addition Test, a Rapid Visual Information Processing Test, eight choice reaction time, Borkowski’s test to assess verbal fluency and Benton’s Revised Visual Retention Test to assess memory. ”Minor” psychiatric symptoms were self-rated on the Hospital Anxiety and Depression Scale, the General Health Questionnaire, and the Nottingham Health Profile. Finally, patients were asked which treatment they preferred. CPAP compliance was monitored by time clocks hidden in the units measuring the total duration the machines were switched on and the presure was being delivered effectively to the mask. It turned out that CPAP units were run for an average of 3.2 hours (SD+=/-0.7) per night and used “effectively“ for 2.8 hours (SD=+/-0.6) per night. The patient group was split by median “effective“ CPAP use into better and poorer compliers, 8 in each group, with an average nightly CPAP use for the better compliers of 5.0 hours (SD=+/- 0.6), and for the poorer compliers 1.1 hour (SD=+/- 0.2). Comparing CPAP treatment with placebo across all subjects, there was no difference in either MSLT sleep latency or subjective sleepiness, though a total symptom score was lower with CPAP than placebo. The only one of the cognitive tests that showed a difference between CPAP and placebo was the TrainMaking B task of mental flexibility. Self-rated depresion was lower with CPAP than with placebo. Ten of the 16 patients preferred CPAP to placebo; these continued on long-term CPAP treatment. Breaking the subjects down into better and worse compliers showed the compliant patients to have improvement with CPAP in total symptom score, TrailMaking B performance, depression rating, and overall quality of life. Six of these eight better compliers preferred CPAP. Of various pretreatment measures tested as predictors of CPAP compliance, higher micoarousal indices (32.3 in better compliers versus 16.6 in poorer compliers) and higher AHI scores (12.5 in better compliers versus 9.6 in poorer compliers) both predicted better CPAP compliance, but not overall symptom score or sleepiness. The authors considered their results suggestive of benefits from CPAP in patients with mild SAHS, who should therefore be given a trial of CPAP. Most patients preferred CPAP to the placebo and wished to continue it. However, not all the tests showed advantages for CPAP over placebo, especially the majority of cognitive tests, and interestingly the Epworth Sleepiness Scale. They pointed out that both CPAP and placebo had beneficial effects on the latter. At the same time, the small size of the sample used to test cognitive effects limited the power of statistical tests to show significant change. The multiple outcome ratings raised the problem of spurious positive findings occurring as a result of chance (so-called Type I false positive errors). Note especially that neither of the two measures of daytime sleepiness (MSLT or ESS) showed a significant advantage for CPAP; prior to treatment, these measures approached the normal range, whereas after either treatment both scores increased. They questioned the usefulness of the MSLT in treatment studies of CPAP and sleep apnea in general, since in any case even in more severe cases CPAP generally fails to normalize this measure. They cautioned against extending their findings with mild SAHS to patients with Upper Airway Resistance Syndrome. They noted with interest that frequency of microarousals seemed more relevant to CPAP compliance than AHI, supporting the idea that criteria for arousals may need adjustment to make them more sensitive.

COMMENTS

I found this study disappointing in many respects, despite the authors’ optimistic assessment of its resuls as supporting usefulness of CPAP in mild sleep apneics. The many measures used and the few that showed significant differences left real differences in doubt. I blame this on substantial methodological problems, particularly the progressive attrition of an initially small subject sample to one where sensitivity to changes must be severely compromised. Furthermore, the widespread poor compliance underminded potential therapeutic effects. The use of a single phone call after two weeks is an example of really inadequate support for compliance and problem-solving in the crucial first weeks of CPAP treatment. Also, the use of an active placebo, ranitidine, which may actually have benefited sleep apneics with reflux esophagitis disturbing their sleep, makes me doubt that they were really comparing CPAP with a truly inactive treatment.Of course, despite the use of the term “placebo” which implies double-blind methodology, there could be nothing blind about this comparison. One feature I found of special interest was the negligible “effective” use of CPAP in the noncompliers, and the difference of almost an hour per night between the time the machine was turned on and the time it was used “effectively.” Presumably this represented times that patients left the machine on but took the mask off to go to the bathroom, get something to drink or eat, or just take a break from the discomfort of the mask. I also noticed with interest the suggested inadequacy of the MSLT in measuring sleepiness changes in sleep apneics under treatment. I personally believe that a much more comprehensive assessment of daytime sleepiness infolving, for example, activity monitoring, report of significant others, and even direct observation of the subject over a variety of activities, should be developed to assess this ever-more-important variable as the ability sleep apneics to work and drive is increasingly called into question. Fundamentally, I believe the authors could have made a much more convincing demonstration of CPAP efficacy in this group by (a) at least doubling the sample size; (b) using a plausibly blindable and inactive placebo such as setting the CPAP pressure very low for the placebo treatment; (c) dispensing with the “shotgun” approach to cognitive assessment and focussing on the few tests that have shown some consistent impairment across studies of sleep apneics; (d) developing a measure of “chief complaints” determined by each patient separately and perhaps with the participation of a signifianc tother, to be followed across the treatments; (e) developing a more comprehensive and meaningful set of measures to describe daytime alertness, its impairment, and functional consequences from various perspectives. Furthermore, more effort should be exerted to extract better compliance with CPAP whether it is set at therapeutic or placebo levels, perhaps including weekly clinic or home visits.





This page hosted by Get your own Free Home Page