學生版自助式失眠認知行為治療結合個人化自動反饋之療效研究

Abstract

根據研究調查統計，睡眠問題影響著近三分之一人口的日常生活與工作，對社會經濟與醫療都造成負擔。目前主流的失眠治療分為藥物治療和心理治療(如，認知行為治療，CBT-I)兩種形式，兩者短期療效相當，但藥物治療伴隨著副作用及易成癮的特性，且CBT-I的長期效果更佳，目前CBT-I已被建議用作原發性失眠的首選治療方案之一。面對面形式的CBT-I存在著供應不足、時間空間不便利等侷限，於是自助形式的CBT-I應運而生。自助CBT-I的療效相較於治療師的面對面CBT-I仍略有差距，而研究發現，當自助CBT-I搭配以治療師的線上反饋時，治療效果有所提升。但過去線上反饋的形式多數是治療師通過電話提供，每個個案每週約15分鐘，即仍需要消耗較多的治療資源。為此，本研究將建立一個規則化的自動反饋框架，將參與者睡眠日誌中的參數套用至規則中，每週提供個人化的自動反饋，從而提高自助式CBT-I的療效而不會耗費更多的治療資源。\n本研究於大專院校校園內招募共92位自陳具有睡眠問題的學生作為參與者，並採用隨機分配的方式，將參與者隨機分為自助治療搭配反饋組(反饋組，n=31)，自助治療無反饋組(自助組，n=31)以及暫不開始治療的等候組(等候組，n=30)，所有參與者需要在治療前、後完成一週的線上睡眠日誌及睡眠相關問卷(失眠嚴重度量表、簡式睡眠失功能信念與態度量表、匹茲堡睡眠品質量表、睡眠衛生習慣量表、簡式憂鬱焦慮壓力量表)。接受治療的參與者治療的八週期間，參與者每週均能通過電子郵箱接收線上自助治療教材，同時每天填寫睡眠日誌。反饋組每週收到自助治療教材的同時還會額外收到對其上週睡眠狀況的反饋，反饋系統以睡眠三系統模型(恆定系統、晝夜節律系統、激發系統)為基礎編寫而成，將參與過去一週的睡眠參數代入反饋系統後，即自動生成反饋。\n由混合設計變異數分析以及成對樣本t檢定的結果顯示，經過八週的治療，反饋組與自助組在失眠嚴重程度、睡眠品質、睡眠信念、睡眠衛生習慣以及睡眠日誌上的入睡耗時、睡眠效率相較於前測都有顯著的改善效果。PSQI當中的日間功能分量表以及DBAS的與失眠後果相關信念分量表，反饋組的改善效果比自助組更爲顯著。而且反饋組在治療結束後對治療的滿意度更高。在流失率分析上，本研究反饋組流失率39%，略低於自助組的52%，但此一差異並沒達到統計上的顯著性。本研究驗證了線上CBT-I對睡眠問題的治療效果。而且額外的自動化反饋有利於給患者提供更具體的改善建議，並彌補了自助化教材的不足，從而進一步為睡眠帶來改善。

Insomnia is affecting approximately one third of the population, causing significant psychological, health and economic consequences. Currently, pharmacological and psychological (e.g. Cognitive Behavioral Treatment for Insomnia, CBT-I) treatments have been proved to be effective in short term. As CBT-I is a treatment with no or minimal adverse effect, and its therapeutic effect can be sustained, it becomes the treatment of choice that is recommended by various clinical guidelines. However, traditional face to face CBT-I is unavailable to most of the people due to limited well-trained professionals and the space or time limitation. Different forms of self-help treatment have been developed. Self-help CBT-I is not as effective as the traditional one, its therapeutic effect was shown to be improved by combining therapist feedback. Yet, most of the online feedback is provided through telephone. Even though it takes approximately 15 minutes weekly for each patient, it is adding up to be quite a lot of therapeutic resources. In this study, we developed a set of rules to provide personalized feedback for participants automatically according to their sleep logs. In this way, we attempt to improve service efficiency of self-helped CBT-I by reducing the consumption of extra therapeutic resources.\nNinety-two college students with self-reported sleep disturbances associated with distress or daytime impairment were recruited and randomly assigned to one of three groups: an online self-help CBT-I with feedback group (Feedback, n=31), an online self-help CBT-I group (Self-help, n=31), and a waitlist group (Waitlist, n=30). All three groups need to fill out an online sleep log for a week, several sleep-related questionnaires (including Insomnia Severity Index [ISI], Dysfunctional Beliefs and Attitudes about Sleep questionnaire [DBAS], Pittsburgh Sleep Quality Index [PSQI], Sleep Hygiene Practice Scale [SHPS], and Depression Anxiety, and Stress Scales, [DASS-21]) at baseline and posttreatment. During the 8-week treatment period, participants in the two treatment groups received 8 weekly CBT-I online materials via email and were required to fill out daily online sleep log. For the participants in the feedback group, they received a personalized feedback in addition every week. Feedback rules are developed based on the daily life and sleep patterns that may affect the three major neural systems that regulate sleep and wake, which are the homeostatic system, the circadian system, and the arousal system. Following the feedback rules, personalized feedback can be generalized automatically based on the sleep log data for each participant.\nRepeated-measures analysis of variance and paired sample t test showed that, after eight weeks of treatment, both the feedback group and the self-help group had significant improvement in insomnia severity as measured by the ISI, sleep quality as measured by the PSQI, sleep onset latency and sleep efficiency on sleep log, as well as sleep-related beliefs and behaviors as measured by the DBAS and the SHPS, respectively. The improvement on the Daytime Dysfunction subscale of the PSQI, and the Consequences of Insomnia subscale of the DBAS was more significant in the feedback group than in the self-help group. The drop-out rate in our study was about 39% in feedback group, slightly lower than 52% in self-help group. In conclusion, the study again validated the therapeutic effect of online CBT-I on sleep problems. And the additional automated feedback helped provide patients with more specific suggestions further improving insomnia-related beliefs and daytime dysfunction.