改良「職能復健評估綜合評量表」應用於社區思覺失調症個案

Abstract

研究背景：思覺失調症個案常伴隨多種功能的缺損。職能復健評估綜合評量表 (Comprehensive Occupational Rehabilitation Evaluation Scale, CORES)為臺北市一些社區精神復健機構常用的例行評估工具之一，以評量個案在復健期間的整體功能表現。然而，CORES於不同機構間的評量項目與計分標準不同，且其心理計量特性大多未知，嚴重影響評估結果的解釋以及專業間的溝通。 研究目的：本研究目的為增加及修改CORES評量項目，使其更符合社區精神復健機構之思覺失調症個案，並驗證CORES修改後之心理計量特性，包含：內在一致性、施測者間信度與建構效度。 研究方法：本研究分為三階段：（一）新增及修改CORES評量項目：研究者透過專家會議討論臨床治療師使用CORES之經驗，以新增及修改CORES的評量項目，並確保其內容具備內容與表面效度。（二）實地測試：研究者邀請2位治療師實際使用上一階段修訂的改良版CORES (Modified CORES, M-CORES)於50名思覺失調症個案，並修正施測過程所發現之問題，以完成最終版的M-CORES。（三）驗證M-CORES之心理計量特性：研究者將M-CORES應用於3家社區精復機構之第一季 (2017/01-03)例行評估中，其中1家精復機構加入另1位治療師同時評量機構內的30位個案，以驗證施測者間信度。最後，研究者綜合第三階段的評估資料驗證M-CORES之內在一致性及建構效度。 結果：第一階段經過5次專家會議完成新增及修改CORES評量項目。第二階段實地測試後，進一步增加評量項目內容之說明、評量方式（觀察或訪談）及調整填寫格式。第三階段於3家社區精復機構招募思覺失調症個案190位，3位職能治療師實際完成共181份M-CORES評量紀錄；心理計量特性驗證結果顯示有良好的施測者間信度 (ICC=0.71-0.94)，多數主要面向（獨立生活功能、休閒功能、職業功能和社會功能）有可接受的內在一致性 (=0.76-0.87)；驗證型因素分析結果顯示卡方與自由度比值 (x /df=2.16)符合良好的適配度，其它指標（使用合併/刪除後的22個項目，漸進誤差均方根為0.09、比較適配度指標為0.87及標準化殘差均方根為0.10）則指出M-CORES之適配度不足。 結論：結果顯示M-CORES具有可接受的內在一致性及良好的施測者間信度，但建構效度不理想。研究者建議刪除或修改「身心健康狀況」及「社會支持系統」面向下之評量項目，以促進該量表於社區思覺失調症個案之評估效果。

Background: Clients with schizophrenia are usually accompanied by functional deficits. The Comprehensive Occupational Rehabilitation Evaluation Scale (CORES) is one of the routine assessments at some institutions of community psychiatric rehabilitation in Taipei, and is used to evaluate clients’ general function. However, the scoring and items of the CORES varies from institution to institution. Also, its psychometric properties are unknown. These factors affect the interpretation of results and communication between occupational therapists. Objective: First, we aim to increase and revise the items of the CORES in order to fit the characteristics of clients with schizophrenia in 3 Taipei City Hospital Songde Branch affiliated institutions of community psychiatric rehabilitation in Taipei. Next is to assess the internal consistency, inter-rater reliability, and construct validity of the modified CORES (M-CORES). Methods: There are three stages in the process of modifying the CORES. In the 1st stage, the items of the CORES are increased and revised by researchers. We initially identified the differences among clinical practitioners in administering the CORES, then increased and revised the items of the CORES, and furthermore to ensure its content validity. In the 2nd stage, the M-CORES was used in clinical practice. Two therapists were invited to evaluate 50 clients with schizophrenia independently at 2 institutions of community psychiatric rehabilitation. Simultaneously, we solved the problems found in the process of assessment in order to finish the final version of the M-CORES. In the 3rd stage the psychometric properties of the M-CORES are examined. We applied the M-CORES in routine assessments in the first phase (2017/01-2017/03) at 3 institutions of community psychiatric rehabilitation. In this period, one additional therapist was invited to help assess 30 clients using the M-CORES to evaluate the inter-rater reliability. Finally, the data from the 3rd stage was analyzed to examine the internal consistency, and construct validity of the M-CORES. Results: The items of the CORES were increased and revised through 5 expert consensus meetings in the 1st stage. In the 2nd stage, we added a new manual, and modified the format of the M-CORES. Then, we recruited 190 clients with schizophrenia at 3 institutions of community psychiatric rehabilitation in the 3rd stage. One hundred and eighty-one records of the M-CORES assessment were completed by 3 occupational therapists. According to analyzing the data from the 3rd stage, the M-CORES exhibited good inter-rater reliability (ICC=0.71-0.94), acceptable internal consistency (=0.76-0.87) in many of the major domains (IADL, leisure, work, and social function). Nevertheless, confirmatory factor analysis (CFA) showed that the M-CORES had insufficient goodness of fit (x /df=2.16, RMSEA=0.09, CFI=0.87 and SRMR=0.1 by analyzing 22 items). Conclusion: In summary, the M-CORES had good interrater reliability and acceptable internal consistency. Nevertheless, the M-CORES was insufficient in goodness of fit. Therefore, the items in the “physical and mental health” and “social support system” domains needed to be removed or revised. Then, it would become more useful in routine assessments at the institutions of community psychiatric rehabilitation in order to assess the general functional performance of clients with schizophrenia.