- 學位論文
海洛因成癮者自美沙冬維持治療流失之預測因子研究
The Predictors for Duration of Retention in Methadone Maintenance Treatment among Heroin-Dependent Individuals.
摘要
本研究之目的在調查南台灣接受美沙冬維持療法的海洛因成癮者之遵從醫囑留存於治療中的情形，並進一步探討能預測海洛因成癮者自美沙冬維持療法流失的因素。 研究採用次級資料進行分析，原調查資料乃收集南台灣三所醫療院所中接受美沙冬門診治療的個案資料及為期18個月的追蹤資料，共計有368名海洛因成癮者被納入本研究之對象。 所收集之資料以Kaplan Meier survival estimate來探討個案在不同的美沙冬治療追蹤時間點，留存於治療中的機率。以Cox比例風險迴歸及邏輯斯迴歸檢驗能預測個案後來自美沙冬維持療法流失的因素，納入模式中的變項包括：治療前的各項個案特性與家庭支持度、開始治療後三個月的美沙冬劑量。 研究結果顯示：本研究對象在美沙冬維持治療中一年的留存機率為41.6％，18個月為31.3%；排除非自願流失個案後，一年的留存機率為47.1％，18個月為39.5%。治療前每月花費在購買海洛因金額較高、自陳使用海洛因造成危害嚴重度越高、家庭支持程度越低者，在追蹤期間自美沙冬維持治療流失的危險性就越高。將進入治療三個月後進行第一次追蹤時所使用的美沙冬劑量納入考量後，結果顯示進入治療三個月後所使用的美沙冬劑量低於50mg、自陳使用海洛因造成危害嚴重度越高者，在追蹤期間自美沙冬維持治療流失的危險性就越高。 最後根據本研究結果進行討論，並提出具體建議，以提供臨床擬訂藥癮治療計劃及未來研究之參考。
並列摘要
Methadone maintenance treatment (MMT) is an effective harm reduction treatment method for heroin-dependent individuals. Duration of retention in MMT is a predictor of treatment outcomes. The aims of this study were to examine the duration of retention in MMT and to identify the predictors for duration of retention in MMT among heroin-dependent individuals in southern Taiwan. The current study used the secondary data for analysis. Primary data was gathered from three MMT units in southern Taiwan. The sample consisted of 368 heroin-dependent individuals who were investigated at baseline and followed during the treatment period. The duration of retention in MMT (in days) was the main outcome of the study. The predictor variables were categorized into client characteristics, family support and methadone dose after 3 months of admission. The study used Kaplan-Meier product-limit method to analyze retention rate in the MMT program, and used Cox proportional hazard regression and logistic regression to examine the predictors for drop-out from the MMT program. The results found that cumulative probability of retention at 360-days and 540-days were 41.6% and 31.3%, respectively. However, if the involuntary dropouts were excluded, cumulative probability of retention at 360-days and 540-days were 47.1% and 39.5%, respectively. Among pre-treatment variables, the predictors for drop-out from MMT program included: higher heroin cost, higher severity of harm caused by heroin use, and poorer family support. If the methadone dose was included in the regression model, methadone dose <50mg and higher severity of harm caused by heroin use were the significant predictor variables for drop-out. Implications for treatment were discussed.