Gap-Kalamazoo溝通技巧評估表之專家版於生理職能治療人員之心理計量特性驗證

Validation of the Gap-Kalamazoo Communication Skills Assessment Form-Clinical/Faculty Version in Occupational Therapy Practitioners for Adults

蔡劭彤(Shao-Tong Tsai)

指導教授 ： 薛漪平

國立臺灣大學/醫學院/職能治療研究所/碩士(2022年)

https://doi.org/10.6342/NTU202200082

摘要

背景及目的：職能治療人員的醫病溝通技巧甚為重要，其為影響醫病關係與治療成效的關鍵之一。良好的溝通技巧評估工具，可確認治療人員溝通技巧之表現與結果之穩定度與正確性。現今評估醫療人員之醫病溝通技巧評估工具中，Gap-Kalamazoo溝通技巧評估表 (Gap-Kalamazoo Communication Skills Assessment Form, GKCSAF) 最具潛力，共有3種版本（專家版、病人版、自評版），且其評估內容完整性較高，共有9個溝通技巧，包括：建立醫病關係、開啟議題討論、收集資料、了解病人/家屬的觀點、確保病人/家屬理解訊息、達成共識、總結並結束、展現同理、傳遞正確資訊。各溝通技巧分別有2至5個參考項目，以作為各溝通技巧的評分參考，且各溝通技巧均以5點量尺計分，可分別呈現9個溝通技巧的分數與總分，因此可確認該人員於各別與整體溝通技巧之表現。由於GKCSAF的心理計量特性如信度與效度驗證較不足，為提升此工具的臨床應用性，本研究主在驗證GKCSAF中文版之專家版於生理疾患職能治療人員之評估者間信度、評估者內信度、隨機測量誤差、編碼一致性、表面效度與內容效度。
方法：評估的資料來自2個研究計畫，以錄音方式分別蒐集治療師與病人及實習生與病人間之臨床治療訪談內容，並轉譯為逐字稿，以利評估者的詳細評估。研究流程有3部分：(1) 溝通資料之收集；(2) 信度驗證，先做資料合適度篩選，再進行評估者訓練，最後為正式評估；(3) 效度驗證。本研究將使用立意取樣，共招募5位評估者（直接評分3位、編碼後評分2位）獨立評分。「直接評分」即按照原工具之評估方式評分。「編碼後評分」為使用ATLAS.ti 9，將符合特定溝通技巧之敘述編碼，再依編碼內容評分。前測完成後，研究者於該2種評分方式各選擇1位評估者完成後測，前測與後測中間間隔8週。評估者間與評估者內信度之資料分析，於整體分數以組內相關係數 (intraclass correlation coefficient, ICC) 檢驗，個別分數以一致性百分比 (Percentage agreement) 與weighted kappa檢驗。隨機測量誤差以測量標準誤 (standard error of measurement, SEM)，編碼一致性以holsti index檢驗。表面效度與內容效度均以同意百分比檢驗。
結果：本研究的「直接評分」組與「編碼後評分」於基本信度與效度的結果顯示：(1) 評估者間信度的總分的ICC值分別為0.57與0.52，均為不佳的一致性，而各大題加權kappa值分別為 -0.03 ~ 0.34與 -0.03 ~ 0.38，均為差至普通的一致性。(2) 評估者內信度的總分的ICC值分別為0.86與0.82，均為好的一致性，而各大題加權kappa值分別為 -0.02 ~ 0.49與 -0.02 ~ 0.64，「直接評分」組為差至中度的一致性，而「編碼後評分」組為差至高度的一致性。(3) 隨機測量誤差的SEM% 值於評估者間分別為8.1% 與7.0%，為可接受之標準。(4) 編碼一致性的整體結果為24.0%，而各大題的結果為10.5% ~ 63.8%；評估者內於整體結果為61.1%，而各大題的結果為56.8% ~ 80.2%。(5) 表面效度之同意百分比為80% ~ 100%，皆通過檢核，具備表面效度 (6) 內容效度之同意百分比為20% ~ 100%，其中有4個大題未達標準，其餘皆通過檢核，代表不完全具備內容效度。
結論：本研究目的為驗證GKCSAF專家版之中文版，於生理疾病職能治療人員的心理計量特性。於信度部分，此工具於各大題大部分結果顯示信度不佳，僅少部分達普通以上之一致性；總分為好的一致性；而效度部分於表面效度顯示為可接受，而內容效度則為部分可接受。此外，由於大部分的大題於「直接評分」組與「編碼後評分」組相似，因此此工具可使用「直接評分」的方式評估即可。研究者建議未來需將此工具之內容進行改良，將各項目與評分標準定義清楚，並調整可能較不適用於職能治療之項目，且未來亦需完整驗證其心理計量特性，以確認其於職能治療的適用性。

關鍵字

溝通技巧 ； 信度 ； 效度 ； 職能治療

並列摘要

Background and purposes: Therapist-patient communication skills are important for occupational therapists. It is one of the keys to affect the therapist-patient relationship and the effectiveness of treatment. A good communication skills’ assessment tool can be used to confirm the performance of the therapists’ communication skills. Among current assessment tools, Gap-Kalamazoo Communication Skills Assessment Form (GKCSAF) has the most potential for evaluating medical staffs’ communication skills. There are 3 versions (clinician/faculty version, patient/family version, and self-assessment version). Compared to other tools, the assessment content of GKCSAF is relatively completed. GKCSAF includes 9 communication skills: builds a relationship, opens the discussion, gathers information, understands the patient’s and family’s perspective, shares information, reaches agreement, provides closure, demonstrates empathy, and communicates accurate information. Each communication skill has 2 to 5 items, used as references for scoring each communication skill, and each communication skill is scored on a 5-point scale, generating 9 independent scores and a total score. These scores can be used to ascertain each and an overall communication skill’s performance of a person. Due to the insufficient verification of GKCSAF’s reliability, validity and the expectation to improve the quality of the evaluation results, this study used 2 evaluation methods: " scoring after coding " and "direct scoring" to verify the inter-rater reliability, intra-rater reliability, random measurement error of the Chinese GKCSAF clinician/faculty version among the occupational therapy practitioners for adults. Besides, the coding agreement, face validity and content validity are verified as well.
Methods: The assessable materials came from 2 research projects. The contents are clinical interviews between therapist and patient or intern and patient. These materials were presented in transcripts and audios. This study comprised of 3 phases: (1) collection of assessable materials; (2) validation of reliability; (3) validation of validity. Purposive sampling would be used to recruit 5 raters (3 for “direct scoring” and 2 for scoring after coding”). Each rater scores independently. “Direct scoring” means scoring in the same way as the original tool's method. “Scoring after coding” is to use ATLAS.ti 9 to code the content that met the specific communication skills, and then to score according to the coded content. After the pre-test is completed, the researcher would select 1 rater from each group to complete the post-test (8 weeks apart). Data analysis of inter-rater and intra-rater reliability of the total scores would be investigated by intraclass correlation coefficient (ICC). As for each communication skill’s score, the researcher would use percentage agreement and weighted kappa to investigate the inter-rater and intra-rater reliability. In addition, random measurement error would be examined by the standard error of measurement (SEM). As for coding agreement, it would be examined by holsti index. Moreover, face validity and content validity, percentage agreement would be used to investigate them.
Results: The results of “Direct scoring” and “Scoring after coding” indicate that (1) both groups have poor inter-rater reliability of total score (ICC: 0.57 0.52), and poor-to-average reliability among 9 communication skills (weighted kappa: -0.03 ~ 0.34 -0.03 ~ 0.38). (2) Both groups have good intra-rater reliability of total score (ICC: 0.86 0.82), and poor-to-medium reliability among 9 communication skills (weighted kappa: -0.02 ~ 0.49 -0.02 ~ 0.64). (3) The random measurement error of the inter-rater reliability and intra-rater reliability’s SEM% are 8.1% 7.0% and 3.5% 4.7%, indicating acceptable random measurement error. (4) The overall coding agreements among and within raters are 24.0% and 61.1%. As for the 9 communication skills, the results are 10.5% ~ 63.8% and 56.8% ~ 80.2%. (5) The face validity showed that 80% ~ 100% of occupational therapists agreed with the translation, indicating that it has face validity. (6) The content validity showed that 20% ~ 100% of raters agreed with the content, indicating that some parts of this tool have content validity while some don’t.
Conclusion: This research aims to validate the Chinese GKCSAF clinician/faculty version among the occupational therapy practitioners for adults. As for the reliability, the results indicate that most communication skills have poor reliability, and only few are shown to be average. However, the results of the total scores indicate good reliability. As for the validity, the face validity is shown to be acceptable while not all parts of the tool have acceptable content validity. Besides, most communication skills between “Direct scoring” and “Scoring after coding” have similar results. Therefore, raters may then use “Direct scoring”. The researcher suggests that this tool may be modified for relevantly clear definitions of the content and the rating scale. Moreover, researchers may modify some items that may not be suitable for occupational therapy’s scenario, and have further verification of the psychometric properties to confirm its applicability in occupational therapy.

並列關鍵字

communication skills ； reliability ； validity ； occupational therapy

參考文獻

黃怡靜、林恭宏、吳姿誼、呂文賢、陳怡妙、謝清麟（2013）。中風個案自陳職能治療服務需求與滿意度之探討。臺灣職能治療研究與實務雜誌，9（2），99-112。

Arnall, F. A., Koumantakis, G. A., Oldham, J. A., Cooper, R. G. J. C. R. (2002). Between-days reliability of electromyographic measures of paraspinal muscle fatigue at 40, 50 and 60% levels of maximal voluntary contractile force. 16(7), 761-771.

Atkinson, G., Nevill, A. M. J. S. m. (1998). Statistical methods for assessing measurement error (reliability) in variables relevant to sports medicine. 26(4), 217-238.

Bai, S., Wu, B., Yao, Z., Zhu, X., Jiang, Y., Chang, Q., . . . Tong, T. J. B. m. e. (2019). Effectiveness of a modified doctor–patient communication training Programme designed for surgical residents in China: a prospective, large-volume study at a single Centre. 19(1), 338.

Basehore, P. M., Pomerantz, S. C., Gentile, M. J. M. t. (2014). Reliability and benefits of medical student peers in rating complex clinical skills. 36(5), 409-414.