支持清潔針具計畫量表之設計及信效度評估

Support for a Needle Syringe Program in Taiwan:A Questionnaire's Design and Evaluation of Validity and Reliability

2004年國內愛滋疫情在靜脈注射藥癮者（intravenous drug users,IDUs）暴增，於2006年7月全國實施清潔針具計畫，目前國內愛滋疫情已減緩。IDUs、衛生醫療人員及檢警獄人員對清潔針具計畫的支持態度均會影響其推動的成效，因沒有完整探討這方面的量表，所以設計支持清潔針具計畫態度量表，評估其信效度，以利後續研究。量表共6題三個因素（地點反對、疑慮及認同），分數愈高表示支持態度愈正向。10位專家的平均內容效度指標（content validity index, CVI）為77.8%。邀請IDUs（2008年11-12月）、衛生醫療人員及檢警獄人員（2007年10-12月）填寫問卷，共回收有效問卷1345份，回收率為85.6%。IDUs、衛生醫療人員及檢警獄人員在三個構面的平均分數有達統計顯著差異，表具已知團體差異的建構效度。驗證性因素分析中模式適配度指標，如GFI（適配度指數）、AGFI（調整後適配度指數）、NFI（規準適配指數）均>0.9，RMSEA（漸進殘差均方和平方根）=0.036，顯示模式合適。各因素組合信度>0.5，在接受的範圍內。本研究量表具良好的內容效度、建構效度及內在一致性，歡迎大家採用這份問卷，成果也能互相比較。

After 2004, Taiwan experienced a sharp rise in AIDS/HIV cases among intravenous drug users (IDUs). Beginning in July 2006, a Needle Syringe Program (NSP) was implemented nationwide, and since then, the number of IDUs with HIV/AIDS infection has slowed. Understanding how the various stakeholders support this NSP is important in determining the success of the program. These stakeholders include the IDUs themselves, healthcare workers, and correctional workers including police, prosecutors and jailers. Since no information has ever been gathered to quantify the levels of support for the NSP, we developed a questionnaire for these three groups for future study. The questionnaire is based on the literature and consists of 6 items using three subscales (objections to the locations, worries about the program, and agreement with the program). The high scores on the questionnaire represent positive support for the NSP. The mean content validity index (CVI) from 10 experts was 77.8%. Recruitment for the study included IDUs (November to December, 2008), healthcare workers, and correctional workers (October to December, 2007). A total of 1345 valid and anonymous questionnaires was returned with a response rate of 85.6%. The scores for the three subscales and for the total reached statistical significance difference among the three study groups, indicating that our questionnaire has good known-groups of construct validity. Confirmatory factor analysis verified the appropriateness of the three factors for the questionnaire with goodness-of-fit index (GFI), adjusted goodnessof-fit index (AGFI), and normal fix index (NFI) > 0.9, root mean square error of approximation (RMSEA) = 0.036, and composite reliability among each subscale > 0.5. Our questionnaire has an acceptable psychometric property, including good content validity, internal consistency, and construct validity. Other researchers are invited to use the questionnaire so that findings can be compared on the same basis.