Treatment Preference and Patient Satisfaction in Chronic Insomnia

Transcription

1 INSOMNIA Treatment Preference and Patient Satisfaction in Chronic Insomnia Norah Vincent PhD, and Carrie Lionberg, MA Dept. Of Clinical Health Psychology, University of Manitoba Study Objectives: The purposes of this study were to examine treatment preference and satisfaction with group treatment in individual with chronic insomnia. Design: Correlational. Setting: The study was conducted in an outpatient hospital setting. Patients or Participants: Participants were 43 adult volunteers from the community. Interventions: N/A Measurements and Results: Prior to treatment, participants were presented with descriptions of behavioral and pharmacological treatment for the problem of insomnia and asked to rate the acceptability, presumed effectiveness, and presumed side-effects of treatment. A sub-sample of these individuals (n = 37) participated in a 6-week cognitive behavioral treatment group for insomnia. Sleep diary and questionnaire data were collected prior to and following treatment. Results showed that cognitivebehavioral therapy was significantly preferred over pharmacological therapy at pre-treatment and that more favorable assessments of cognitivebehavioral therapy at pre-treatment were associated with better adherence but not improved outcome. Of treatment techniques, participants least liked sleep restriction and most liked sleep hygiene. Results indicated that more favorable ratings of the usefulness of sleep restriction were associated with improvements in sleep efficiency, sleep-related impairment, and quality of life. Conclusions: Implications of these findings are that patient preference is important to assess prior to treating insomnia and that more work may be needed to increase patients awareness of the benefits of sleep restriction. Key words: Insomnia; treatment preference; patient satisfaction; cognitive behavioral treatment; pharmacological treatment INTRODUCTION CHRONIC INSOMNIA IS A PREVALENT AND DISTRESS- ING PROBLEM, REPORTED TO AFFECT 9% 10% OF THE POPULATION. 1,2 Currently, there are two viable treatment options for chronic insomnia (i.e., pharmacological, psychological), and yet there is a paucity of research on treatment preference and patient satisfaction. The type of treatment which an individual receives is often dependent on the referring physician s awareness of treatment options and the availability of such services rather than on patient preference. The primary care physician usually provides pharmacological treatment; 3 however, the single study in this area showed that those with insomnia tend to prefer behavioral treatment over pharmacological treatment even when the treatments are described as equally effective. 4 In recent years, there has been a growing emphasis on the importance of assessing treatment preference and satisfaction with treatment among consumers of healthcare. 5,6,7,8 Reasons for assessing treatment preference include heightened concern about obtaining informed consent from individuals 9 and findings which indirectly suggest that treatment preference impacts on willingness to initiate and adhere to treatment in clinical trials for some mental health problems (e.g., panic disorder). 10 Whether this is also true for problems such as chronic insomnia remains unknown. Satisfaction with treatment is important to investigate because even preferred treatments are unlikely to be implemented if they produce negative side-effects or are too costly. Cox, Fergus, and Swinson 11 discuss that treatment strategies which are not viewed favorably may not be practiced or maintained over time. For chronic health problems such as insomnia, continued practice of Accepted for publication January 2001 Address correspondence to:dr. Norah Vincent, PZ-251 PsycHealth Centre, 771 Bannatyne Avenue, Winnipeg, Manitoba, R3E 3N4; Tel: (204) ; Fax: (204) ; NVincent@exchange.hsc.mb.ca SLEEP, Vol. 24, No. 4, treatment strategies is essential to preventing relapse. Therefore, patient satisfaction is an important area of study in insomnia. The aims of this study were to replicate and extend Morin et al. s 4 findings regarding the treatment preferences of those with chronic insomnia and to examine whether treatment preference impacts on adherence and/or outcome. A second purpose of the research was to describe patient satisfaction with treatment. METHODS Participants Participants were 43 adults with chronic insomnia. Inclusion criteria were total sleep time (TST) <6.5 hours per night, either sleep-onset latency (SOL)>45 minutes and/or time wake after sleep onset (WASO)>30 minutes, problems with sleep on at least four of seven nights per week for at least the previous six months, and at least two daytime consequences which the individual attributed to sleep loss (e.g., fatigue). None of the participants met the study s exclusion criteria which were the presence of shift work, the report of symptoms suggestive of an alternative sleep disorder (e.g., sleep apnea), history of brain injury, bipolar disorder, schizophrenia, or serious medical condition (e.g., chronic pain), and concurrent or prior cognitive behavioral treatment for insomnia. Although potential participants who were female were not routinely asked about menopausal status, all participants had received a recent medical examination and any participant who reported awakening in the night with excessive perspiration and/or feelings of warmth was queried as to whether they had received specific investigation into hormonal status. Two participants indicated a suspicion that menopause might explain their insomnia, however, in both cases their primary care physicians had ruled out this explanation. Inclusion and exclusion criteria were verified by sleep diaries, a clinical interview, and questionnaire data. Of the sample, 70% 30 were females and 30% 13 were males,

2 Table 1 Preference ratings for behavioral and pharmacological treatment for chronic insomnia Pharmacological Treatment Behavioral Treatment Preference Scale Items M (SD) M (SD) t p 1. Acceptability for self 3.33 (1.51) 5.35 (.89) Acceptability for others 3.66 (1.13) 5.04 (.91) Willingness to adhere 3.89 (1.59) 5.55 (.72) a). Suitability for sleep-onset problem 4.48 (1.55) 4.76 ( 1.16) b). Suitability for sleepmaintenance problem 4.00 (1.40) 4.11 (1.14) Short-term effectiveness 4.43 (1.56) 3.96 (.97) Long-term effectiveness 2.35 (1.18) 4.46 (.81) Effectiveness for improving daytime functioning 3.33 (1.40) 4.85 (.92) Extent to which treatment would produce negative side-effects 2.70 (1.17) 5.41 (.98) Overall acceptability (8.73) (5.38) Note: Individual items from the Treatment Acceptability Scale range from 1 to 6, with higher ratings indicating a more acceptable or efficacious treatment rating. For item 8, ratings range from 1 (very strong side-effects) to 6 (no expected side-effects). Ratings of "Overall Acceptability" range from 8 to 48, with higher ratings reflecting more overall acceptability for that type of treatment. The internal consistency (Chronbach's alpha) of the two subscales was as follows: Pharmacological Treatment sub-scale (.87), Behavioral Treatment sub-scale (.80). 99% were Caucasian, and 58% were married. The mean age of the sample was years (SD=12.37) and 57% 25 had some post-secondary education. Of the sample, 37% (16) had another, secondary, co-morbid DSM-IV axis I diagnosis and the most common of these was generalized anxiety disorder (30%). The mean duration of insomnia was 16.1 years (SD=15.7) and average sleep efficiency was 56% (SD=14.83). The average total sleep time per night was 5.02 hours (SD=1.19), the average sleep onset latency was 1.02 hours (SD=1.03), the average number of awakenings per night was 2.25 (SD=1.52), and the average number of hours awake at night was.56 (SD=.39). Of those who were using medications for sleep (n=20), 10 were taking a benzodiazepine and 10 were taking Zopiclone. Procedure All of the measures and procedures used in the study were approved by a human ethics committee from the University of Manitoba. The Insomnia Treatment Acceptability Scale 4 is a 16- item questionnaire which assessed preferences for psychological and pharmacological treatment of insomnia. The scale provided brief but balanced paragraphs describing each of the two treatments for insomnia (i.e., pharmacological, psychological). See Morin et al. 4 for a description of the paragraphs. Participants were asked to rate each treatment on a variety of dimensions (e.g., presumed acceptability, presumed effectiveness, and presumed side effects) and two global scores were obtained from the scale (i.e., acceptability of medication treatment, acceptability of SLEEP, Vol. 24, No. 4, behavioral treatment). A standard sleep diary was used to evaluate pre-post treatment changes in total sleep time and sleep efficiency and the Sleep Impairment Index (SII) 12 was used to assess pre-post treatment changes in impairment due to sleep. SII scores ranged from 5 to 35, with higher scores reflecting more impairment due to sleep. The Group Therapy Questionnaire (GTQ), 11 originally developed for use with individuals receiving treatment for panic disorder, was modified for use in the study. It was modified by replacing anxiety treatment components with sleep treatment components. The GTQ assessed participants likes and dislikes associated with group treatment, perceived usefulness of treatment, perceptions of therapist, and perceptions of improvement in lifestyle activities (e.g., leisure, family, job-related, social). Finally, the Counseling Satisfaction Questionnaire (CSQ) 13 measured global satisfaction with treatment. Participants responded to an advertisement placed in a local newspaper entitled Treatment for Chronic Insomnia. The advertisement indicated that researchers were recruiting individuals with insomnia to evaluate treatment but did not specify what type of treatment was to be made available (i.e., pharmacological, psychological). Approximately 169 individuals contacted the investigator about the study and participated in a brief telephone screening interview. Of these, 115 did not meet inclusion criteria for chronic insomnia, five indicated a lack of interest in group treatment, and six could not be reached by telephone. This left 43 individuals who met criteria for the study and were able to participate. At this time, participants were told about the type of treatment to be offered (i.e., behavioral). Prior to having an in-

3 person appointment, these individuals received two weeks of sleep diaries and a questionnaire package by mail (which included the Insomnia Treatment Acceptability Scale and the SII). During the intake appointment, sleep diary and questionnaire information was collected, a sleep interview was conducted, and a modified version of a structured clinical interview for Diagnostic and Statistical Manual of Mental Disorders (fourth edition)(dsm-iv) axis I disorders (SCID) 14 was administered. Of the initial 43 participants, 37 began a six-week behavioral treatment group for insomnia which was modeled after Morin s approach. 12 The other six were waiting to begin treatment at the time of this report. The number of participants per group varied from five to seven and the authors were the group leaders. Group leaders were blind to pre-treatment ratings on the Insomnia Treatment Acceptability Scale. Treatment components consisted of the following (in this order): sleep hygiene (i.e., education about the effects of diet, exercise, caffeine, alcohol, and environment on sleep), stimulus control (i.e., education about the effects of conditioning and assistance to strengthen the association between the bed/bedroom with sleep rather than with frustration/anxiety associated with sleeplessness), sleep restriction (i.e., curtailing time in bed to the actual sleep time), relaxation training, cognitive therapy, reducing and eliminating hypnotic medication, and relapse prevention. There were seven drop-outs from treatment, leaving a total of 30 individuals who completed the group. Reasons for dropping out included scheduling conflicts (n = 4), the development of a serious health problem (n=1), a significant interpersonal stressor (n=1), and in one case was unknown. Following treatment, two weeks of sleep diaries, and the SII, GTQ, and CSQ were administered. RESULTS The data was analyzed using t-tests, correlations, and multiple regression. When multiple t-tests or correlations were conducted, the Bonferonni correction procedure was used. Treatment Preference Results in Table 1 indicated that there were significant differences in the overall acceptability ratings assigned to psychological and pharmacological treatment. Psychological treatment, compared to pharmacological treatment, was judged to be more acceptable for self and others, more effective in the long-term, more likely to improve daytime functioning, and less likely to produce negative side effects. Participants reported the expectation that there was no difference between psychological and pharmacological treatment in short-term effectiveness. To evaluate whether treatment acceptability was associated with demographic variables, a standard multiple regression was conducted. Demographic predictor variables were age, gender, marital status, co-morbidity status (i.e., whether an individual met criteria for another axis I anxiety or mood disorder), and medication status (i.e., whether an individual was taking a medication for sleep). Several participants did not complete one of more questions pertaining to demographic variables and so were not included in this analysis. Results indicated that the full model did not explain significant amounts of variance in the criterion variables (i.e., acceptability of psychological treatment F(5, 33)=.54, p=.75, R=.27, R 2 =.08, AdjR 2 =.07; acceptability of pharmacological treatment F(5, 34)=2.13, p=.08, R=.49, R 2 =.24, AdjR 2 =.13), however, the latter regression may have reached statistical significance with a larger sample size. The relationship between treatment acceptability and premature attrition was next examined. At pre-treatment, results show that drop-outs rated psychological treatment less favorably than completers however there were no differences between drop-outs and completers on ratings of pharmacological treatment (See Table 2). These findings should be interpreted cautiously, however, due to the small sample size of drop-outs. There was no association between number of sessions attended and overall treatment acceptability [i.e., acceptability of psychological treatment r(29)=.27, p=.16, acceptability of pharmacological treatment r(29)=-.15, p=.44]. There were no significant differences between drop-outs and completers in frequency of medication use t(10)=-.78, p=.45 or in pre-treatment satisfaction with sleep as measured using an item from the SII t(35)=-.57, p=.57. Of the seven drop-outs, five were female, three were taking a hypnotic medication, and three had a comorbid psychiatric diagnosis. Due to the small number of drop-outs, it was not possible to use statistical techniques to determine whether variables of gender, medication use, or psychiatric comorbidity were associated with dropping out. At post-treatment, using paired t-tests, there was a statistically significant difference in total sleep time t(25)=5.88, p=.0001 (Mchange=1 hour, SD=.90), sleep efficiency t(25)=8.23, p=.0001 (Mchange=22.1%, SD=13.7), sleep-related impairment (SII) t(27)=5.86, p=.0001 (Mchange=5.38, SD=4.85), and frequency of medication use t(27)=2.97, p=.006 (Mchange=18.46, SD=32.89) and these changes were very similar in magnitude to those reported by prominent clinicians in this area. 3,15 At posttreatment, 23.3% of the sample was continuing to use hypnotic medication and no participant began using a new hypnotic medication during the course of the study. Using the GTQ, ratings of improvement in quality of life ranged from 1 (not at all) to 8 (very much). With treatment, there were modest improvements in quality of life (e.g., leisure activities M = 3.93 SD=2.09, family activities M=3.89 SD=2.02, job-related activities M=3.84 SD=2.03, social activities M=3.59 SD=1.97). Results showed that there were no significant correlations between pre-treatment acceptability of psychological or pharmacological treatment and improvement in total sleep time (TST), sleep efficiency, sleep- Table 2 Pre-treatment acceptability and adherence to behavioral treatment Drop-Outs (n=7) Completers (n=30) Pre-Treatment Acceptability M (SD) M (SD) t p Acceptability of Behavioral Treatment (4.71) (4.27) Acceptability of Pharmacological Treatment (13.47) (8.03) SLEEP, Vol. 24, No. 4,

4 Table 3 Satisfaction with cognitive behavioral treatment for insomnia How Useful Personally Like Item M (SD) M (SD) Sleep Hygiene 6.45 (1.62) 6.62 (1.37) Stimulus Control 6.07 (1.44) 6.36 (1.55) Cognitive Therapy 6.03 (1.39) 6.48 (1.38) Relaxation Training 5.88 (1.37) 6.25 (1.71) Help with Tapering Medication 5.30 (2.52) 5.31 (2.38) Sleep Restriction 4.88 (1.94) 5.05 (1.73) Note. Using the GTQ, responses to "How Useful" and "Personally Like" range from 1 (not at all) to 8 (very much). Only those participants currently on hypnotic medication rated the medication component of treatment. related impairment (SII), or any of the lifestyle outcome measures (e.g., improvement in leisure, family, job, or social activities). Thus, pre-treatment acceptability did not appear to impact on outcome. More favorable pre-treatment ratings of the acceptability of psychological treatment were associated with more favorable post-treatment ratings of the usefulness of stimulus control r(28)=.60, p=.001. Results showed that pre-treatment acceptability of psychological treatment was associated with higher post-treatment ratings of therapist understanding r(29)=.57, p=.002 and therapist knowledge r(29)=.53, p=.005; Those who rated the psychological treatment more positively at pre-treatment rated the therapist more positively at post-treatment. Satisfaction with Psychological Treatment Results in Table 3 list participants perceptions of the treatment components. Of treatment components, sleep hygiene was rated as the most liked and most useful. Although the ratings for sleep restriction were still above the midpoint of the scale, sleep restriction was rated as the least liked and least useful of treatment components. There were no significant differences between how much participants liked a particular treatment component and how useful they found it (i.e., those treatment components rated as useful also tended to be liked ). Participants liked several treatment components significantly more than others. For example, sleep hygiene t(28)=4.33, p=.0001, relaxation training t(27)=3.10, p=.005, and stimulus control t(28)=4.02, p =.0001 were each liked significantly more than sleep restriction. Also, respondents found sleep hygiene t(27)=4.08, p=.0001 and stimulus control t(27)=2.94, p=.007 to be significantly more useful than sleep restriction. Despite participants dislike of sleep restriction relative to the other treatment components, sleep restriction was associated with improved sleep efficiency, sleeprelated impairment (SII), and quality of life. There was a trend towards females (M=6.39, SD=1.30) to report cognitive therapy to be more useful than males (M=4.6, SD=.89) t(26)=2.92, p=.007. Otherwise, there were no gender differences in ratings of usefulness or liking of different treatment components. There were no significant differences in how useful or how well-liked treatment components were as a function of psychiatric comorbidity or whether an individual was using hypnotic medication at the beginning of treatment. SLEEP, Vol. 24, No. 4, Using the GTQ, the majority of participants (64.3%)(n=18) reported that the six sessions was about right, 35.7% (n=10) indicated that the six sessions was not enough, and no one reported that six sessions was too much. Two participants did not respond to this question. When asked which treatment components participants would have liked to spend more time on, results showed that 59% wanted to spend more time on cognitive therapy, 38% reported wanting to spend more time on relaxation training, 35% indicated wanting to spend more time on sleep hygiene, 31% wanted to spend more time on sleep restriction, 17% wanted more time spent on medication usage, and 14% requested spending more time on stimulus control. When queried about treatment components which participants would have liked to spend less time on, results showed that 24% would have liked to spend less time on medication use, 14% requested spending less time on stimulus control, 10% would have liked to spend less time on sleep restriction, and 7% would have liked to spend less time on both sleep hygiene and relaxation training. After group completion, using the GTQ with a scale ranging from 1 (not at all) to 10 (very much), participants reported that they had more understanding of their own problem with sleep (M=8.21 SD=2.13) and that they felt reassured that they were not alone in having a sleep problem (M=8.83 SD=1.56). Psychiatric comorbidity, medication use, and gender did not significantly impact on ratings of self-understanding and feelings of being alone with a sleep problem. When asked Did the primary therapist in your group understand your own particular sleep problem? and Did the primary therapist in your group have adequate knowledge about the nature and treatment of sleep problems?, results showed that participants viewed the therapist as quite knowledgeable (M=9.19 SD=1.06) and understanding (M=8.55 SD=1.68). There were no significant differences in ratings of therapist characteristics as a function of pre-treatment hypnotic medication use, psychiatric co-morbidity, or gender. There was a significant relationship between ratings of therapist understanding and improvement in family activity r(26)=.49, p=.008; Those who rated the therapist as more understanding tended to report more improvements in family activities with treatment. Pre-treatment ratings of the acceptability of psychological r(27=.29, p=.14 and pharmacological treatment r(27=.20, p=.30 were not significantly correlated with overall satisfaction with treatment (CSQ). Table 4 shows that CSQ scores were significantly associated with the perceived usefulness of sleep hygiene, sleep restriction, and stimulus control. DISCUSSION This study was designed to assess patients treatment preferences and perceptions of behavioral treatment for insomnia. As such, it was not a treatment outcome study. One of the main findings of the study was that psychological treatment is preferred over pharmacologic treatment for the problem of chronic insomnia. This study replicated and extended the findings of Morin et al. 4 Morin s sample was composed of older (M=66.5 years), well-educated (M=14.1 years of education) adults from the community, 20% of whom had a comorbid axis I diagnosis. The sample used in this study was younger, had more psychopathology, and had less efficient sleep. This suggests that Morin s findings can be extended to a broader sample of individuals with chronic insomnia.

5 Table 4 Correlations between improvement in lifestyle activity, satisfaction with treatment, and perceived usefulness of treatment Components Usefulness of Treatment Component Item Cognitive Therapy Sleep Hygiene Medication use Relaxation Sleep Stimulus Training Restriction Control Improvement in Lifestyle Activity Leisure-related a.49 Family-related a.39 Job-related a.38 Social-related a.29 Symptom Improvement TST Sleep Efficiency a.17 SII a.24 Counseling Satisfaction (CSQ) a a.63a Note: Values reflect Pearson Product-Moment Correlations. Those with a superscript are significantly correlated after Bonferonni adjustment (p<.001). TST = total sleep time, SII=Sleep Impairment Index. Research has shown that many individuals indicate a stronger preference for psychological than pharmacological treatment for mental health problems such as panic disorder 16 and hypochondriasis. 8 Based on the findings of the current study, it appears that this preference also applies to the problem of insomnia. In the current investigation, individuals tended to have more distinct preferences (often negative) about pharmacological treatment compared to psychological treatment as observed by the relatively larger standard deviations in the acceptability ratings assigned to medication treatment. Horne 17 discusses that individuals tend to have a negative view of medicines, perceiving them to be unnatural and potentially harmful. In the current study, psychological treatment was viewed as more acceptable than pharmacological treatment primarily because of the presumption that psychological treatment would result in improved daytime functioning, have better long-term effectiveness, and include fewer negative side effects. Based on clinical impression, some of the reticence about medication treatment for insomnia in the current study was that such medication would lead to daytime drowsiness and tolerance/dependence. Although this has been noted with some hypnotic medications (e.g., benzodiazepines), newer hypnotic medications such as Zopiclone are reported to have fewer negative side effects and reduced tolerance properties. 18,19 Interestingly, Jonas and colleagues 20 reviewed 38 clinical and epidemiologic studies of hypnotic medication and concluded that some newer hypnotics (e.g., Zopiclone) were no more effective and no less likely to produce side effects compared to some older hypnotics (e.g., Triazolam). The second main finding of the study was that treatment preference was significantly associated with premature attrition although clearly this finding needs to be replicated with a larger sample size. Other variables which have been associated with premature attrition from a combined psychological and pharmacological treatment for insomnia include increased psychopathology, an earlier REM onset, increased concern about sleep problems, greater early morning fatigue, and gender (with males more likely to drop-out of treatment). 21 It appears that the acceptability of treatment may be a variable which impacts on willingness to complete behavioral treatment for insomnia. This suggests that opinions about treatment options should be assessed prior to making referrals for services. This is particularly salient because all participants in the current study were volunteers who received no compensation for participating and pre-treatment ratings of the acceptability of psychological treatment were quite high. Individuals with insomnia who are referred by family physicians and respirologists may be much more mixed in terms of the receptivity of behavioral interventions and so there may be an even greater need to assess patient preference in this group. We have recently begun to assess the treatment preferences of insomnia patients who have been referred by hospital-based respirologists. Our preliminary data indicate that respirology-referred patients (n=11) rate behavioral treatment (M=43, SD=5.9) as favorably as do self-referred individuals from the community (M=43.63, SD=5.3). Also, respirology-referred patients rate medication treatment (M=31.18, SD=0.1) as favorably as do those referred from the community (M=32.44, SD=8.7). Outside of the area of insomnia, research has examined patients adherence to medical treatment often focusing on patients personal and demographic features, social support, barriers to treatment, values, expectancies, and recently physician characteristics. 22,23,24,25 A recent longitudinal study showed that physician job satisfaction, degree to which the physician answered the patient s questions, the number of patients seen by a physician per week, and the scheduling of follow-up appointments each predicted adherence to treatment for various health problems. 22 This is interesting because it suggests that practitioner variables may have a significant impact on patient s adherence behavior. There is much more work to be done examining adherence to treatment for chronic insomnia patients. The third main finding to emerge from this study was that sleep restriction is the treatment component least preferred by those with insomnia (although approximately one-third indicated that they would like to spend more time on this strategy). Chambers 26 notes that sleep restriction is often at odds with SLEEP, Vol. 24, No. 4,

6 patients beliefs and attitudes about sleep (i.e., removing oneself from bed in order to eventually obtain more sleep). Although viewed negatively, perceptions of the usefulness of sleep restriction had the strongest relationship with outcome measures. This is consistent with meta-analytic findings which have shown sleep restriction to be one of the most powerful behavioral techniques available for the problem of insomnia. 27 It is plausible that patients do not recognize the degree to which sleep restriction is responsible for improvement in their sleep and daytime functioning. Such an attribution, however, may impact on the continued use of certain techniques after treatment has ended. Although sleep restriction is a powerful intervention, if it is difficult to implement or unappealing to those with insomnia, investigators/clinicians may want to consider how to make the technique more appealing or investigators may need to work on developing a stronger patient-therapist relationship so that individuals are more likely to attempt and to use sleep restriction. In contrast, sleep hygiene was viewed quite positively by participants and found to be very useful perhaps because it is relatively easy to implement. This finding is in contrast to one other study which found that sleep hygiene was viewed more negatively than either stimulus control or meditation. 28 In the current investigation, although reported to be useful, sleep hygiene was not associated with self-reported improvements in any outcome measure suggesting that individuals give sleep hygiene more credit for personal change than is warranted. These findings should not be taken to mean that investigators should discontinue using sleep restriction in favor of sleep hygiene, as research has shown that sleep hygiene in isolation is not a very effective strategy for chronic insomnia. 28 Instead, these results merely highlight that patients perceptions of sleep restriction may need clinical attention. Perceived therapist understanding, but not expertise, was associated with outcome in the treatment of insomnia. Perhaps surprisingly, this therapist factor appeared to be important in a relatively directive treatment. To speculate, individuals may have been more willing to disclose problems they were having with adhering to sleep strategies in the presence of someone they viewed as understanding and thus more progress could be made in completing the treatment techniques. Therefore, the interpersonal style of group leaders may be more important than the expertise of such leaders (although clearly the latter is important as well). Implications of these findings are that, after a thorough medical examination reveals that an individual has psychophysiologic insomnia, individuals should be fully informed about treatment options for insomnia. The field would benefit from the development of a high quality video which described the advantages and disadvantages of psychological and pharmacological treatment to help individuals make a more informed choice. The clinician needs to be aware that sleep restriction may be viewed unfavorably by some and that education regarding the importance of such a strategy may improve adherence. Those receiving treatment for chronic insomnia will likely benefit from validation of the difficulties and frustrations associated with working on techniques aimed at improving sleep. Limitations of this study include a reliance on subjective rather than objective assessments of insomnia, a small sample size, and a lack of attention to counterbalancing treatment components. Sleep diary data, although correlated with results from SLEEP, Vol. 24, No. 4, polysomnography, has been shown to provide an overestimation of sleep-latency and an underestimation of total sleep time. 29 Although it is recognized that relying on self-reported symptoms to diagnose insomnia is not error-free, sleep diaries are a costeffective alternative to polysomnographic assessment. 12 Also, sleep diary data capture the subjective impression of sleep which is very important in an individual s desire to seek treatment. Nevertheless, the corroboration of subjective data with objective findings would lend further support to study conclusions. Another limitation of the study was that treatment components were delivered in a specified order and so ratings of the usefulness of such techniques may have been influenced by the order of presentation. Another limitation of the study was that preferences for treatment were elicited after participants learned that the treatment to be offered was behavioral and not pharmacological. Thus, it is possible that participants favorably evaluated behavioral treatment due to demand characteristics of the situation. It should be noted that all respondents were reassured that the decision to withdraw or even begin a new hypnotic medication during the study would be up to them and that the investigator would support whatever decision that they made. Regardless, this experimental format may have impacted on participant ratings. Future research may wish to verify sleep diary data with polysomnography, to use larger samples to explore treatment preference and satisfaction, to present treatment components in a counterbalanced order, and to extend this research to pharmacological treatment. In the future, research should determine whether there is an association between the perceived usefulness of treatment components at post-treatment, continued use of treatment techniques at follow-up, and relapse. ACKNOWLEDGMENTS The authors would like to acknowledge Ms. Shannon Pudlubny, B.A. who provided assistance with telephone screening. REFERENCES 1. Simon GE, VonKorff SD. Prevalence, burden, and treatment of insomnia in primary care. Am J Psychiatry 1997;154(10): Ancoli-Israel S, Roth T. Characteristics of insomnia in the United States: Results of the 1991 National Sleep Foundation Survey. Sleep 1999;22(2): Morin CM, Hauri PJ, Espie CA, Spielman AJ, Buysse DJ, Bootzin RR. Nonpharmacologic treatment of chronic insomnia. Sleep 1999;22(8): Morin CM, Gaulier B, Barry T, Kowatch RA. Patients acceptance of psychological and pharmacological therapies for insomnia. Sleep 1992;15(4): Banken DM, Wilson GL. Treatment acceptability of alternative therapies for depression: A comparative analysis. Psychotherapy 1992;29(4): Davidson JR, Brundage MD, Feldman-Stewart D. Lung cancer treatment decisions: Patients desires for participation and information. Psycho-Oncology 1999;8(6): Irwin E, Arnold A, Whelan TJ, Reyno L, Cranton P. Offering a choice between two adjuvant chemotherapy regimes: a pilot study to develop a decision aid for women with breast cancer. Patient Education and Counseling 1999;37(3): Walker J, Vincent N, Furer P, Cox B, Kjernisted K. Treatment preference in hypochondriasis. J Behav Ther Exp Psychiatry 1999;30:251-8.

7 9. Sider RC. The ethics of therapeutic modality choice. Am J Psychiatry 1984; 141(3): Hofmann SG, Barlow DH, Papp LA, Detweiler MF, Ray SE, Shear K, Woods et al. Pretreatment attrition in a comparative treatment outcome study on panic disorder. Am J Psychiatry 1998;155(1): Cox BJ, Fergus KD, Swinson RP. Patient satisfaction with behavioral treatments for panic disorder with agoraphobia. Journal of Anxiety Disorders 1994;8(3): Morin C. Relief from Insomnia: Getting the sleep of your dreams. New York, NY: Doubleday Publishing Larsen DL. Enhancing client utilization of community mental health outpatient services. Dissertation Abstracts International 1997; 39(4):4041B. 14. Spitzer RL, Williams JBW, Gibbon M, First MB. User s guide for the structured clinical interview for DSM-III-R. Washington, DC: American Psychiatric Press, Morin CM, Wooten V. Psychological and pharmacological approaches to treating insomnia: critical issues in assessing their separate and combined effects. Clinical Psychology Review 1996;16(6): Walker J, Eldridge GD, Hazen AL, Norton GR, Stein MB. Selfreported preference for type of treatment for panic disorder. Poster presented at the Anxiety Disorders Association of America Annual Conference, March Horne R. Editorial: Patients beliefs about treatment: the hidden determinant of treatment outcome? J Psychosomatic Res 1999;47(6): Fleming JA, McClure DJ, Mayes C, Phillips R, Bourgouin J. A comparison of the efficacy, safety and withdrawal effects of zopiclone and triazolam in the treatment of insomnia. International Clinical Psychopharmacology 1990;5(Suppl 2): Hajak G, Clarenbach P, Fischer W, Haase W, Ruther E. Zopiclone improves sleep quality and daytime well-being in insomniac patients: Comparison with triazolam, flunitrazepam and placebo. International Clinical Psychopharmacology 1994;9: Jonas JM, Coleman BS, Sheridan AQ, Kalinske RW. Comparative clinical profiles of triazolam versus other shorter-acting hypnotics. J Clin Psychiatry 1992;53(12, Suppl): Dashevsky B, Kramer M. Patients who discontinue combined behavioral and medicinal treatment of insomnia. Sleep Research 1997;26: DiMatteo MR, Sherbourne CD, Hays RD, Ordway L, Kravitz RL, McGlynn EA, Kaplan S, Rogers WH. Physicians characteristics influence patients adherence to medical treatment: results from the Medical Outcomes Study. Health Psychol 1993;12(2): Meichenbaum D, Turk DC. Facilitating treatment adherence: a practitioner s guidebook. New York: Plenum Press, DiMatteo MR, DiNicola DD. Achieving patient compliance: the psychology of the medical practitioner s role. New York: Pergamon Press, Janz NK, Becker MH. The health belief model: a decade later. Health Educ Q 1984;11: Chambers MJ. Therapeutic issues in the behavioral treatment of insomnia. Professional Psychology: Research and Practice 1992;23(2): Morin CM, Culbert JP, Schwartz SM. Nonpharmacological interventions for insomnia: a meta-analysis of treatment efficacy. Am J Psychiatry 1994;151(8): Schoicket SL, Bertelson AD, Lacks P. Is sleep hygiene a sufficient treatment for sleep-maintenance insomnia? Behavior Therapy 1988;19: Trinder J. Subjective insomnia without objective findings: a pseudo diagnostic classification? Psychol Bull 1988;103: SLEEP, Vol. 24, No. 4,