目的：自2000年健保施行總額預算後，少有探討醫療人員對健保總額預算影響醫療工作感受，也沒有這方面的量表。因此，本研究設計一份「健保總額預算影響醫療工作感受量表(WPGB)」，並進行信效度的評估，以利後續研究。 方法：WPGB題目的訂定是參考相關文獻、訪談及觀察，包括9個問題（健保施行總額預算前後的工作量、工作時間、人力調整、文書處理量、薪水、醫藥物品的運用、工作流程、醫療品質、與病人關係）。之後採專家效度、建構效度之假說驗證、內在一致性、再測信度評估WPGB的信效度。 結果：13位專家對WPGB的內容效度指標(CVI)平均值爲83%，在55-100%間。邀請北部某醫院的醫療人員填寫問卷，有效問卷348份，回收率爲48.3%。以工作滿意度（中文版職業壓力指標第二版(OSI-2)）來評估建構效度之假說驗證，結果相關係數r爲-0.35，達顯著意義。24位醫療人員兩週內填寫WPGB兩次，再測信度ICC=0.78。 結論：我們設計的WPGB具良好的內容效度、建構效度的假說驗證及兩週的再測信度。期望WPGB廣泛被採用，共同探討健保總額預算對醫療工作影響的課題，研究結果也能互相比較。

Objectives: Since 2000, the payment system of global budget (GB) has been included in Taiwan National Health Insurance, but no study has been conducted to investigate work perception about global budget for medical employees (WPGB), and no questionnaire to measure WPBG has been available. Therefore, we aimed to design a questionnaire to measure medical employees' perceptions of their work tasks before and after implementation of GB so that further study of WPGB could subsequently be conducted. The validity and reliability of the WPGB questionnaire were evaluated. Methods: The WPGB questionnaire was based on literature review, interviews, and observation and consisted of nine questions on the following topics: workload, working hours, manpower, paperwork, salary, medical resources, working procedure, treatment quality, and relationship with patients before and after GB. Expert validity, hypothesis testing of construct validity, internal consistency, and test-retest reliability of the WPGB questionnaire were evaluated. Results: The mean content validity index of the WPGB questionnaire for 13 experts was 83%, with a range of 55% to 100%. A total of 384 medical employees (response rate, 48.3%) at one hospital in northern Taiwan participated in the study. Regarding hypothesis testing of construct validity, the correlation coefficient between WPGB and job satisfaction in the second version of the occupational stress index (OSI-2) was -0.35 and reached statistical significance. Another 24 people from the same hospital completed the WPGB questionnaire twice in 2 weeks, and the reliability intraclass correlation coefficient was 0.78. Conclusion: The content and hypothesis testing of construct validity and test-retest reliability of the WPGB questionnaire are acceptable. We welcome other researchers to use the WPGB questionnaire and evaluate the findings on the same basis.