比較個案自填方式不同對於生活品質問卷再測信度之影響

A comparison of test-retest reliability of self-reported SF-36, WHOQOL, and EQ-5D questionnaires based on different administration approaches

林育任(Yu-Jen Lin)

指導教授 ： 王榮德

共同指導教授 ： 陳健弘

國立臺灣大學/公共衛生學院/職業醫學與工業衛生研究所/碩士(2009年)

https://doi.org/10.6342/NTU.2009.03238

摘要

目的：
本研究目的為使用SF-36, WHOQOL-BREF, 和 EQ-5D三種國際間認可且常用之生活品質問卷，探討慢性肝病病人是否會因為個案收案的方式不同而導致再測可信度測試的結果有所差異。

方法：
本研究於台灣北部某大型教學醫院，於門診時收集慢性肝病病人自填式的生活品質問卷，其中採取兩種不同的收案方式進行，第一種方式：先於醫院門診填寫前測問卷，於兩週後讓個案在家中填寫後測問卷，再以郵寄方式寄回。第二種收案方式：則是前測與後測問卷都是讓個案帶回家填寫後以郵寄方式，寄回間隔也是兩週。本研究經過台大醫院研究倫理委員會審查通過後進行。第一種方式同意參與的個案人數為69人，完成前測與後測兩次問卷填寫之有效問卷共52份 (回覆率=75.4%)。第二種方式同意參與的個案為180人，完成前測與後測兩次問卷填寫之有效問卷共127份 (回覆率=70.6%)。問卷分數經計算後，以paired-t test和 intra-class correlation coefficient (ICC)比較三種問卷再測信度，及使用Bootstrap re-sampling的方式檢定兩種收案方式的再測信度是否具統計上的差異。使用independent t test比較兩種收案方式之前後分數差異，再以多迴歸分析控制相關潛在性干擾因子進行分析探討。

結果：
第一種收案方式問卷再測信度ICC值最低為0.37最高達0.81；第二種收案方式則為0.61-0.87，經分析比較後發現這三份問卷在第二種收案方式中各面向的ICC分數均高於第一種收案方式。在Independent t test兩種不同收案方式的比較上，除了EQ-5D疼痛狀態有達統計上顯著差異以外 (分別為0.1 ± 0.4 vs. 0.0 ± 0.3, p=0.04)，其他問卷中狀態變項皆無發現有所差異，並進一步使用多迴歸分析，在控制年齡、性別、教育程度後，並無統計上的顯著差異（p=0.06）。比較前測的三種問卷後，發現EQ-5D有較大的天花板效應(75.0%-100.0%)而且平均utility值很高(0.95±0.08)，證明這個問卷有較少的敏感性去量測疾病程度較輕的狀況，WHOQOL-BREF產生的天花板效應較小（1.9%）而且沒有地板效應。

結論：
本研究的結果發現以郵寄方式收案及使用WHOQOL-BREF和SF-36問卷較適合慢性肝病病人的生活品質調查，建議未來針對此族群進行生活品質研究者，可將本研究結果當作參考。

關鍵字

訪視者效應 ； 生活品質 ； 填寫方式

並列摘要

OBJECTIVES: This aim of this study is to examine whether, and to what extent the test-retest reliability of Short-Form-36 (SF-36), World Health Organization Quality of Life (WHOQOL-BREF), and EuroQol-5D (EQ-5D) questionnaires are affected by the administration approaches in patients with chronic liver disease.

METHODS: Patients with chronic liver disease were recruited from the outpatient department of a medical center in Taiwan. Their self-reported questionnaires were collected by two approaches. The first approach is that patients received an interview and filled the questionnaire in hospital. Retest questionnaires were returned by mail two weeks later. The other approach is that patients filled both test and retest questionnaires at home, under researcher’s instruction, during the two-week period. Of 69 patients recruited for the first approach; 52 persons completed both questionnaires (75.4%), while the response rate of the second approach is 70.6% (127 of 180). The response rate in the second approach was better (two mail surveys) than that in the first approach (interview at the out patient clinics followed by mail).
After scoring questionnaires, a paired-t test and intra-class correlation coefficient (ICC) were conducted to compare test-retest reliability for three questionnaires. The difference of mean score between two approaches was examined by independent t test. Analyses of mean score differences of different domains were performed by multiple linear regressions with different administration methods, age, gender, and education as covariates.

RESULTS: The test-retest reliabilities ranged from 0.37 to 0.81 in the first approach, while they ranged 0.61-0.87 in the 2nd approach. Most ICCs were higher in the 2nd approach than the 1st approach. There was also no significant difference in the mean difference of test-retest results in two approaches, except in the dimension of ‘pain/discomfort’ mean difference (0.1±0.4 and 0.0 ± 0.3, p=0.04) by EQ-5D but the result were not found by multiple linear regression, after controlling variables of age, sex, and education (p=0.06). Among the three questionnaires, the large ceiling effect (75.0%-100.0%) and the high mean score for utility (0.95±0.08) of the EQ-5D suggest it could be less responsive to detect any difference when the severity of illness is low. In contrast, the scores of WHOQOL-BREF have wider ranges, less ceiling effect （1.9%） and no floor effect.

CONCLUSIONS: The high reliabilities of the present study supported that mail survey method is acceptable for measuring quality of life, and the WHOQOL-BREF and the SF-36 questionnaires of instrument could be used for assessing HRQoL in patients with asymptotic carrier.

並列關鍵字

interviewer's effect ； Quality of life ； mode of administration

參考文獻

22. Tseng, H. M., Lu, J. F. R.,& Tsai, Y. J. Assessment of Health-related Quality of Life in Taiwan (Ⅱ): Norming and Validation of SF-36 Taiwan Version. Taiwan Journal Public Health 2003;22:512-8.

13. Lu, J. F. R, Tseng, H. M., & Tsai, Y. J. Assessment of Health-related Quality of Life in Taiwan (I): Development and Psychometric Testing of SF-36 Taiwan Version. Taiwan Journal Public Health 2003;22:501-11.

5. Hsu, C., Wang, J. D., Hwang, J. S., Tien, H. F., Chang, S. M., Cheng, A. L., Chen, Y. C. & Tang, J. L. Survival-weighted health profile for long-term survivors of acute myelogenous leukemia. Qual Life Res 2003;12:503-17.

24. Fuh, J. L., Wang, S. J., Lu, S. R., Juang, K. D., & Lee, S. J. Psychometric evaluation of a Chinese (Taiwanese) version of the SF-36 health survey amongst middle-aged women from a rural community. Qual Life Res 2000;9:675-83.

1. Chen, D. S., Kuo, G. C., Sung, J. L., Lai, M. Y., Sheu, J. C., Chen, P. J., Yang, P. M., Hsu, H. M., Chang, M. H., & Chen, C.J., et al. Hepatitis C virus infection in an area hyperendemic for hepatitis B and chronic liver disease: the Taiwan experience. J Infect Dis 1990;162:817-22.