職場員工疲勞、病假、離職與假性出席之相關性研究

Abstract

前言：假性出席的定義是當工作者健康不佳時，抱病出席工作，但未能有如正常健康情形下的工作表現。近年來，職場健康之研究逐漸留意到假性出席的問題，且發展出評估工具。然而，台灣目前並沒有對假性出席之評估工具建立良好的信效度，來測量假性出席，進而預測之相關性。因此，本研究目的有兩個：第一、建立中文版假性出席量表並評估其信效度；第二、評估假性出席與疲勞、離職、病假之間的相關性。 方法：本研究為一年之世代研究，以台灣10家工廠員工為對象，利用兩次之健康問卷檢查進行調查。我們利用「史丹佛工作表現量表」 (Stanford Presenteeism Scale, SPS)建立中文版「假性出席量表」評估假性出席。其效度包含建構效度和區別效度，使用的統計分析方法為因素分析和相關性分析。信度以Cronbach’s alpha表示題目之間的內在一致性。使用項目分析、因素分析、信度分析刪減23題「假性出席量表」至6題版本。使用多變項邏輯斯回歸分析第一年假性出席狀況對於之後的離職與職場個人疲勞；另與一年中的請病假時數進行相關性分析。利用ROC曲線下面積推估假性出席對於之後的離職與職場個人疲勞以及一年中的請病假時數之預測能力。 結果：研究對象依所述問卷中「假性出席量表」分數可區分為三組：因自覺無症狀而不需填寫的健康對照組、分數高於15分低假性出席組、分數低於15分高假性出席組。總計20,957人進入分析，其中1,334位為高假性出席組、11,383位低假性出席組，剩下8,240位為健康對照組。在調整性別、年齡、年資、輪班、抽菸、喝酒後，表現不佳組較健康對照組對於之後的離職有很大的影響 (OR=1.94，其ROC曲線下面積為0.69)。同樣地，假性出席也是疲勞和請病假的危險因子。 結論：我們發展出之中文版假性出席量表，具有良好的信效度。此工具應可以在職場之中使用，協助假性出席之評估。綜合問卷調查結果可知，在控制了可能的干擾因子後，假性出席可以有效地預測一年後的離職。我們研究利用簡短的中文版假性出席量表突顯了假性出席為預測因子的重要性。

Background and Objective: The definition of presenteeism is that the worker goes to work while feeling ill; therefore, the worker cannot perform as well as usual. In recent years, several studies have conducted on workers’ presenteeism in the industries. However, to the best of our knowledge, no study on the reliability and validity of presenteeism scale has been conducted in Taiwan. This study aimed to e develop a presenteeism scale and to evaluate its reliability and validity. Furthermore, we examined the association between burnout, sick leave, resignation, and presenteeism. Method: This was a longitudinal follow-up study. Workers in Taiwan were recruited to complete a questionnaire. The English version of Stanford Presenteeism Scale, developed by Koopman in 2002, was translated into Chinese. Factor analysis and Spearman’s rank-order correlation analysis were conducted to assess the construct validity and discriminant validity of SPS. The value of Cronbach’s alpha was computed to evaluate the reliability of SPS. Factor analysis, item analyses, reliability analysis were used for item reduction. Multiple logistic regressions were used to estimate the association between sick leave, resignation, and burnout. Thus, the receiver operating characteristic (ROC) curve was used to assess the effect of the models. Result: 20,957 respondents with the completed questionnaire were divided into three groups: “high prsenteeism” (n=1,334), “low presenteeism” (n=11,383), and “healthy control” (n=8,240). After adjustment for age, gender, work tenure, shift work, medical history, drug history, surgical history, smoking, alcohol consumption, exercise frequency, and stress, the group of high prsenteeism has a significantly higher risk of resignation (OR=1.94, AUC of ROC=0.69). Also, presenteeism is one of the risk factors of burnout and sick leave. Conclusion: Results from this study showed that the Chinese version of Stanford Presenteeism Scale was reliable and valid. Further use of this instrument could be applied to the industries. Also, the findings of this study indicated that after adjustment for potential confounders, presenteeism could predict resignation in one year. Our research stressed the importance of presenteeism as a predictor measuring by simple and short SPS-6.