Data_Sheet_2_Factor structure of post-operative quality of recovery questionnaire (QoR-15): An Italian adaptation and validation.docx

BackgroundThe Quality of Recovery questionnaire (QoR-15) is an English instrument for measuring quality of recovery in surgical patients, not yet translated and validated in Italian when the Enhanced Recovery After Surgery (ERAS) Piemonte studies were planned.
ObjectiveTo produce the Italian version of the QoR-15 questionnaire, to evaluate its factorial structure and to assess the invariance between two types of surgery.
MethodsThe Italian version (QoR-15I) was obtained translating and adapting the original version to the Italian context. The validation was performed suppling the QoR-15I to 3,784 patients enrolled in two parallel stepped wedge cluster randomised trials (ERAS Colon-rectum Piemonte; ERAS Gyneco Piemonte). The factor structure and its invariance between types of surgery was tested using confirmatory bifactor model and multi-group analysis. Comparative fit index (CFI), root mean square error of approximation (RMSEA), and standardized root mean square residual (SRMR) fit indices and their changes between nested models were used to assess the factor structure and the invariance.
ResultsThe bifactor model showed good fit (RMSEA = 0.049, CFI =0.957, SRMR = 0.036) and provided a general recovery factor and two specific factors for physical and mental recovery. Eighty-four percent of the common variance is attributable to the general factor, and thus the QoR-15I is sufficiently ‘one-dimensional’ with an adequate reliability (ωh = 0.70). The ωs values for the physical and mental recovery factors were 0.01 and 0.13, respectively. Multigroup analysis supported configural (RMSEA = 0.053, CFI = 0.950, SRMR = 0.035) and metric invariance (ΔRMSEA = -0.004; ΔCFI = -0.002; ΔSRMR = 0.014), whereas the intercept constraint was removed from item 15 to obtain partial scalar invariance (ΔRMSEA = 0.002; ΔCFI = 0.007; ΔSRMR = 0.004). Construct validity was supported by a negative association of QoR-15I scores with all variables related to worse patient condition and more complex surgery.
ConclusionOur results support the use of the QoR-15I as a valid, reliable, and clinically feasible tool for measuring the quality of recovery after surgery. The results of the confirmatory factor analyses suggest that a unique recovery score can be calculated and support measurement invariance of the QOR-15I across the two type of surgery, suggesting that the questionnaire has the same meaning and the same measurement parameters in colorectal and gynaecologic patients.

背景：康复质量问卷（Quality of Recovery Questionnaire, QoR-15）是一款用于评估外科患者术后康复质量的英文量表。在规划皮埃蒙特加速康复外科（Enhanced Recovery After Surgery, ERAS）相关研究时，该问卷尚未有意大利语版本且未完成意大利语本土化验证。
研究目标：研制QoR-15问卷的意大利语本土化版本，评估其因子结构，并检验两种手术类型间的测量不变性。
研究方法：通过对原始问卷进行翻译并适配意大利医疗场景，得到QoR-15的意大利语版本（QoR-15I）。本研究共纳入两项平行阶梯式整群随机试验（皮埃蒙特结直肠加速康复外科研究、皮埃蒙特妇科加速康复外科研究）中的3784例患者，对QoR-15I进行效度检验。采用验证性双因子模型与多组分析，检验问卷的因子结构以及不同手术类型间的测量不变性；以比较拟合指数（Comparative fit index, CFI）、近似误差均方根（root mean square error of approximation, RMSEA）、标准化残差均方根（standardized root mean square residual, SRMR）及其嵌套模型间的变化量作为评估指标。
研究结果：验证性双因子模型拟合良好（RMSEA=0.049，CFI=0.957，SRMR=0.036），可提取出一个通用康复因子以及躯体康复、心理康复两个特异性因子。总共同方差中有84%可归因于通用因子，因此QoR-15I具备良好的单维性，且具有足够的信度（层级Omega系数ωh=0.70）。躯体康复与心理康复因子的Omega系数（ωs）分别为0.01与0.13。多组分析结果支持构型不变性（RMSEA=0.053，CFI=0.950，SRMR=0.035）与度量不变性（ΔRMSEA=-0.004；ΔCFI=-0.002；ΔSRMR=0.014）；为获得部分标量不变性，移除了第15条目截距项的约束（ΔRMSEA=0.002；ΔCFI=0.007；ΔSRMR=0.004）。QoR-15I得分与患者不良状态及复杂手术相关变量均呈负相关，验证了其结构效度。
研究结论：本研究结果证实，QoR-15I可作为一款有效、可靠且临床可行的术后康复质量评估工具。验证性因子分析结果表明，可计算得到单一的康复总分，且QoR-15I在两种手术类型间具备测量不变性，即该问卷在结直肠手术与妇科手术患者中具有一致的测量意义与参数。