利用Rasch分析評估醫院層級間的護理服務滿意度缺口

Rasch Analysis of Patient- and Nurseassessed Satisfaction Gaps in Hospital Nursing Services

目的：探討護理服務滿意度之缺口。方法：評量3家醫院（醫學中心25間病房、區域醫院24間病房，及地區醫院7間病房），共計1,763位病人及1,270位護理人員參與調查。擬定護理服務滿意度調查表，以（1）Winsteps軟體進行Rasch（1960）分析量表單向度，平行模擬法決定因素個數，（2）病人量表分數進行單因子變異數分析，檢定醫院間病人滿意度異同，（3）Racsh模式估計部門滿意度、試題難度，及其標準誤，計算95％信賴區間，判定部門及題目滿意度缺口。結果：（1）此兩份量表皆符單向度測量；（2）不同層級醫院間服務滿意度，未呈顯著性的差異；（3）5個部門（8.9％）高估滿意度缺口；（4）3題（30％）高估缺口，3題（30％）低估缺口。結論：圖示分析部門及題目滿意度缺口，做為未來護理服務認知差距的改善參考。

Objective: To investigate the satisfaction gaps in nursing services assessed by patients and nurses in three different types of hospitals. Methods: Secondary data were collected from databanks of nursing service records and surveys conducted by a hospital group in southern Taiwan. These data were used for both scales to which nurses and patients responded in 2013. Fifty-six nursing units (medical center, 25; regional hospital, 24; and local hospital, 7) were assessed by 1763 patients and 1270 nurses. Two same-item questionnaires using 4-point Likert-type scales were developed which represented the perspectives of patients and nurses. (1) Rasch (1960) analysis was performed using Winsteps software to determine whether or not the scaled measurement was unidimensional. Parallel analysis was also used to determine the number of factors. (2) One-way ANOVA was used to examine satisfaction differences among the hospitals. (3) A graphical presentation was used to present satisfaction gaps identified with the nursing unit means and the 95% confidence intervals of person and item parameters estimated using Rasch analysis. Results: (1) Both scales were uni-dimensional and fit the Rasch model expectations. (2) There were no significant differences in service satisfaction amongst the three hospitals. (3) Five (8.9%) nurse units expressed satisfaction gaps. (4) Nurses expressed higher and lower average satisfaction gaps with respect to patient views on 3 (30%) items each. Conclusion: A graphical presentation was created to display satisfaction gaps in nursing units and items in the healthcare process, which can serve as sources of improvement in the future.