INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The sleep apnea/hypopnea syndrome (SAHS) occurs in from 1 to 6% of the middle-aged population (1, 2), causing daytime sleepiness, impaired concentration, and altered mood. Continuous positive airway pressure (CPAP) therapy is the current treatment of choice for SAHS, and has been shown in randomized controlled trials to alleviate each of these clinical features (3). However, CPAP therapy is relatively obtrusive, and perhaps not surprisingly, patients have been shown to use CPAP for much less of the night than had been hoped (6), with average durations of use ranging from 3.2 to 4.7 h/night. CPAP is widely used, with estimated prescription rates in Germany, France, the United States (9), and Scotland (our own data) exceeding 100 units/million population/yr.

Because it is both medically and economically important to establish whether CPAP usage can be increased, and whether this improves health outcomes, we tested the hypothesis that provision of additional education and support to SAHS patients beginning CPAP therapy would improve CPAP use and treatment outcomes. Because we have encountered negative attitudes toward CPAP in some SAHS patients who were cajoled into seeking medical advice by a partner, our secondary aim was to examine the hypothesis that CPAP use is lower among patients whose referral was initiated by a partner.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

We conducted a prospective, randomized, controlled, parallel group trial of standard support versus intensive support for new SAHS patients beginning CPAP therapy.

Patients

We studied 80 consecutive patients with SAHS defined by > 15 apneas plus hypopneas/h of sleep as revealed by routine polysomnography done with our standard techniques (10), plus either daytime sleepiness or two other major symptoms of the syndrome (11). None of the patients had previously received CPAP therapy; none had coexisting chronic obstructive pulmonary disease (COPD), asthma, or neurologic problems; and all lived within 50 miles of Edinburgh. The resulting patient population (78 males and two females; mean age 51 ± 11 yr [mean ± SD]; body mass index [BMI] 33 ± 7 kg/m²; apnea-hypopnea index [AHI] 58 ± 33) were then randomized into two groups, one to receive standard CPAP support and the other intensive support. Randomization of each patient was done with predetermined balanced blocks generated by tossing a coin. Patients were blinded to the group to which they were allocated.

Prior to initiation of CPAP therapy, patient referrals were categorized into patient or partner, according to the response to the question, "Who prompted you to seek help, yourself or your partner?"

Support Protocols

Standard support. After undergoing baseline polysomnography, each patient was given a full explanation of the need for and benefits of CPAP therapy by a sleep physician and a sleep nurse. This was reinforced by a 20-min video education program. Each patient was then sized for an appropriate CPAP mask from among a range of masks of more than 20 sizes provided by three different manufacturers. The patients were then acclimated to wearing the CPAP mask for at least 20 min during the daytime, before being booked for a full-night CPAP titration study in which CPAP was manually adjusted to abolish apneas/hypopneas and snoring, and to minimize arousals. During that night and on the following morning, any problems with CPAP were addressed by the CPAP nurse. The patients went home with their CPAP units, and were asked to report any side effects immediately by using a 24-h telephone line to the sleep nurses. The nurses contacted the patients by telephone on Days 2 and 21 to try to identify and resolve any problems with CPAP. The patients were then seen in the hospital by the CPAP nurse and by a sleep physician at 1, 3, and 6 mo after the full night titration study, and any problems with CPAP that the patients had encountered were treated appropriately.

Intensive support. The intensive support protocol consisted of all of the features named earlier for the standard support protocol, plus the following:

Initial CPAP education was given in the patient's home, with involvement of the patient's partner, with the belief that convincing partners of the need for CPAP and its continued use was important.

The patients had an additional 2 nights (3 nights in total) of CPAP titration in the sleep center in an effort to resolve early problems encountered with CPAP, since there was some evidence (12) that patients who spent 3 nights in the hospital at the inititation of CPAP did well.

The sleep nurses made home visits to the patient and the patient's partner at 7, 14, and 28 d and at 4 mo after CPAP initiation, in order to minimize problems with treatment. The early period was targeted because of evidence (6) that CPAP usage in the first month is a major determinant of CPAP use over the longer term. The 4-mo visit was for long-term reinforcement.

CPAP units. Elite CPAP units (ResMed, Sydney, Australia) were used in the study. This allowed measurement of the time during which CPAP units were applying the appropriate pressure on a night-by-night basis.

Outcome measurements. At each clinic visit (1, 3, and 6 mo), the patients' CPAP use data were downloaded to a personal computer using ResMed SCAN software.

Immediately before starting CPAP and after 6 mo of CPAP, all patients underwent a full day of daytime function testing. This included rating of the severity of sleep apnea symptoms with an in-house questionnaire, and of sleepiness, using the Epworth (13) and Stanford (14) sleepiness scales and the objective Maintenance of Wakefulness Test (15). Mood was measured with the Hospital Anxiety and Depression Scale (16), and quality of life with the Short Form 36 (17) and Nottingham Health Profile Part 2 (18) scales. Cognitive function tests included the paced auditory serial addition test, trailmaking test, simple unprepared reaction time, and steerclear, block design, and digit-symbol substitution tests (3). The study had the approval of the Lothian Area Ethical Advisory Committee, and all patients gave written informed consent.

Statistical analysis

Data were analyzed on an intention-to-treat basis. Statistical analysis of CPAP use data collected at 1, 3, and 6 mo was done through three-way analysis of variance (ANOVA), with educational intensity and self or partner referral as intersubject factors and time as an intrasubject factor. For 6-mo outcomes, the unpaired t test was used for normally distributed variables and Wilcoxon's rank sum test for nonnormally distributed data, with results for the standard and intensive support groups compared at each time point.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Eighty patients entered the trial and underwent baseline daytime function assessments. There was no difference between the standard and intensive support groups at baseline in apnea + hypopnea frequency or in any of the outcome measures (Table 1). Thirty-nine of the patients were classified as self-referrals and 41 as partner referrals. The self-referred patients had a higher mean Epworth score than did the partner-referred patients (15 ± 1 versus 12 ± 1 [mean ± SEM]; p < 0.01). Eighteen of the self-referred patients were randomized to the intensive support and 21 to the control group.

Seven of the 80 patients (four receiving standard and three intensive support) were unavailable for daytime function retesting at 6 mo; the four patients in the standard support group had stopped using CPAP, one patient in the intensive support group died of lung carcinoma diagnosed after randomization, another stopped using CPAP, and one defaulted from daytime testing at 6 months. All 80 patients had their CPAP usage over the 6-mo trial period recorded and analyzed.

CPAP Usage

Patients who received intensive CPAP support had higher (p = 0.003) CPAP usage than those in the standard support group (Table 2; Figure 1). CPAP usage was higher (p = 0.002) in the patients who self-referred than in those whose referrals were initiated by a partner (6-mo mean: self-referred: 5.3 ± 0.4; partner-referred: 4.0 ± 0.4 [mean ± SEM]; Figure 2). Overall, duration of CPAP therapy had no effect on CPAP use, but there was an interaction (p = 0.04) between duration of CPAP use and referral group, indicating a decline in CPAP use with time in the partner-referred group.

View larger version (38K): [in this window] [in a new window]   Figure 1.   Mean CPAP usage at 1 mo, 3 mo (incorporating 1- to 3-mo data), and 6 mo (incorporating 3- to 6-mo data) for patients receiving intensive (solid symbols) and standard (hatched symbols) support. Means ± SEM are shown. The difference between groups was significant (p = 0.003).

View larger version (39K): [in this window] [in a new window]   Figure 2.   Mean CPAP usage at 1 mo, 3 mo (incorporating 1- to 3-mo data), and 6 mo (incorporating 3- to 6-mo data) for patients who initiated their own referral (solid symbols) and patients whose referral was initiated by their partners (hatched symbols). The difference between groups was significant (p = 0.002).

Symptoms

The patients who received intensive CPAP support had significantly lower symptom scores at 6 mo (7 ± 1 [mean ± SEM] on a 45-point severity scale) than did the patients receiving standard support (10 ± 1; p = 0.03).

Sleepiness

There was no difference in sleepiness at 6 mo in the intensive and standard support groups in terms of Stanford or Epworth Sleepiness scores or the Maintenance of Wakefulness Test (p > 0.1).

Mood

At 6 mo, the intensive support group had significantly better mood, as shown by lower depression scores on the Hospital Anxiety and Depression Scale. There was no difference between the two groups in the anxiety component of the Hospital Anxiety and Depression Scale.

Quality of life

There was no difference between the two groups in quality of life at 6 mo as assessed with either the Short Form 36 or the Nottingham Health Profile Part 2.

Cognitive function

The reaction time of the intensive support group was significantly shorter at the 95th centile than was that of the standard support group, with a trend (p = 0.09) toward a shorter mean reaction time (0.29 ± 0.01 s versus 0.30 ± 0.01 s [mean ± SEM]). No other cognitive function tests showed significant differences between the two groups.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This randomized controlled trial shows that intensive support improves CPAP usage and alleviates symptoms, benefits mood, and improves a measure of cognitive ability in patients with the SAHS. This is the first randomized controlled trial to show that objective CPAP usage and outcomes can be improved by intervention.

The mean CPAP usage documented in our standard support patients was similar to that found in previous studies of new patients beginning CPAP in which data were analyzed on the basis of intention to treat (6). Such studies will always incorporate some patients who initially decline or later abandon CPAP therapy, and the CPAP usage documented in such studies will therefore be lower than that documented in cross-sectional or retrospective studies of patients attending CPAP clinics and using CPAP (12). It must be stressed, however, that randomized controlled trials have shown that daytime function improves significantly with even the relatively low levels of CPAP usage found in our standard support patients (3). Furthermore, in patients with other chronic diseases, use of either tablets (19) or inhalers (20) is similar, with disappointingly low levels of compliance. Thus, it is significant that we were able to demonstrate 40% improvements in patient use of CPAP, that were sustained over the 6 mo follow-up period.

The greater improvements in patients' symptoms and mood with intensive support are important, since these are the features that drive patients to seek therapy. The improvement in reaction time in tests of cognitive function, reflected by the shorter 95th centile time in patients given intensive support, is a reversal of one of the characteristic consequences of sleep loss (21). The greater improvement in reaction time found with intensive support in our study could be relevant to reducing road traffic accidents. Because of low statistical power for these components, the study did not examine the effect of CPAP support programs on hypertension and road accidents. Because both of these factors relate to the frequency of apneas and hypopneas (22, 23), it is possible that improving CPAP usage by 40% will reduce their frequency, but this needs to be tested.

It is important to recognize that the intensive support group in our study achieved greater CPAP use and better outcomes than did a group given a usual support system that was itself more intensive than that routinely used in many sleep centers. Practices vary widely between centers and between countries, and it is therefore impossible to generalize. However, it is probable that the CPAP usage and outcomes achieved by the intensive supported group in the present study were markedly better than those obtained with the standard support used in many centers.

The limitations of this study included limitations in the power to observe differences in outcomes between the two study groups. For example, with the data generated in this study, the power to detect a mean difference in the Maintenance of Wakefulness Test of 5 min was 0.70 at the 0.05 alpha level. Most of the outcome variables found to produce a greater change with intensive support were subjective, and this could reflect the greater involvement of patients given such support with sleep center staff members. Nevertheless, the greater improvements in symptom and depression scores occurred in two of the main modalities found in randomized, placebo-controlled trials to improve with CPAP therapy (3- 5), and involved two of the main reasons why patients seek treatment for SAHS. Another potential criticism is in the 20 endpoints tested in the study. However, the significant differences found in the primary endpoint of CPAP use, and in the three secondary endpoints that showed differences, were all in the postulated direction of improvement with intensive support, with both the number and direction of these differences suggesting that the findings did not occur by chance alone. Additionally, although the physicians involved in the routine clinical care of the patients were blinded to each patient's treatment group, the research nurses supporting the patients were not. In an attempt to minimize inadvertent bias caused to this unblinding, every effort was made to ensure that patient care was protocol driven. However, it is impossible to exclude this source of bias, although it seems unlikely to be an explanation for the observed differences, especially because the CPAP usage in the standard support group was similar to that in our clinic population when all members of the latter are analyzed (3). A further point is that this trial lasted only 6 mo, whereas CPAP is usually envisaged as life long therapy. Thus, longer-term trials are required to determine whether the benefit of intensive support is sustained, and these should be larger in order to increase the power to detect changes in sleepiness and cognitive function.

Three previous interventional studies have attempted to improve CPAP usage. Fletcher and Luckett (24) measured objective CPAP usage in a crossover study in which 20 patients received, in randomized order, regular telephone support or no additional support. Their study showed no difference in CPAP usage with the two approaches, but the fact that half the patients receiving no support had already had frequent supportive telephone calls while participating in the first (support) limb of the study would tend to obscure any beneficial effects. Chervin and colleagues (25) attempted to determine the effects of phone call support or educational literature on CPAP usage. However, CPAP usage in their study was reported by the patients rather than being measured, a minority of patients were naive to CPAP, the supported groups had apnea + hypopnea frequencies twice those in the control group, and differences reported in the study were not statistically significant. Likar and associates (26) conducted a retrospective analysis of a nonrandomized study and suggested that CPAP usage might be improved with education. None of these three studies examined the outcomes of CPAP therapy.

The present study also shows for the first time that a major determinant of CPAP use is whether the patient or the patient's partner initiated the referral for therapy. When the patient was the initiator, CPAP usage was significantly greater at all time points than when the partner initiated the referral process. Because patients who initiated their own referral had similar baseline symptom and sleepiness scores and similar apnea + hypopnea frequencies to patients whose referral was initiated by the patient's partner, differences in disease severity cannot explain the differences we found in CPAP usage. Furthermore, CPAP usage did not decline with time in the self-referred group, but did decline in the partner-referred group. Recognition of these features will allow more cost- effective targetting of resources in terms of CPAP machines and support staff.

Several differences existed between the protocols used for intensive support and for standard support in the study, and it is impossible to speculate usefully about which elements resulted in the increased CPAP usage in the intensive support group. However, it was our impression that the two additional nights spent in the sleep center contributed relatively little to this, since few problems were actively solved during this period. However, that is a subjective impression and would need to be tested by a prospective study to see whether this additional time in the sleep center, which was the most costly element of the intensive support protocol, was in fact beneficial. Nevertheless, relatively high CPAP use rates have been reported in some European studies in which several nights of hospitalization at CPAP initiation was the norm (12). It seems likely that the improved CPAP usage seen in our study was due to increased patient motivation, increased partner involvement, and the proactive seeking of solutions to problems.

It was not the intention of this study to include a cost-benefit analysis, particularly because such analyses are highly dependent on local costs, which vary considerably internationally. Furthermore, the costs involved will depend strongly on the distance the support nurses have to travel for home visits. However, to allow comparisons to be made, the additional costs for the intensive support group were approximately 1,500 United States dollars per patient, with half this cost arising from the two additional nights in the sleep center and a quarter from support nurses' salaries. We estimate that the provision of an intensive support system such as that described here requires the employment of approximately one additional support nurse for each 100 patients begun on CPAP per year.

In sum, the study reported here shows that CPAP use and outcomes can be improved by an active educational program.