Institute for Medical Sociology, Health Services Research and Rehabilitation Science (IMVR), Faculty of Human Sciences and Medical Faculty, University of Cologne & Centre for Health Services Research Cologne (ZVFK), Eupener Strasse 129, Cologne 50933, Germany.
BMC Fam Pract. 2010 Feb 1;11:9. doi: 10.1186/1471-2296-11-9.
We developed an instrument assessing the extent of smoking cessation activities by general practitioners (GPs) within the Cologne Smoking Study (CoSmoS). The objective of the present study was to examine further psychometric quality of the "SmoCess-GP" instrument (Smoking Cessation by General Practitioners).
127 current smokers who had participated in the Cologne Smoking Study (CoSmoS) were included in our analyses. Confirmatory factor analysis (CFA) was conducted to examine the model fit and to retest the single-factor structure of the instrument using the Mplus software. Further construct validity was tested with bivariate analysis using an instrument which measures patients' trust in physicians.
CFA supported the unidimensional structure of the instrument. The factor loadings exceed the threshold of > or = 0.50. All indicator reliabilities were higher than 0.30. The composite reliability was 0.86 and the average variance extracted (AVE) resulted in a value of 0.50. The calculation of global fit indices identified a CFI value of 1.00 and for TLI a value of 1.02. The root mean square error of approximation (RMSEA) indicates that 0% of the information is not accounted for by the model. The chi-square value was chi2df = 6 = 4.63 (p = 0.59). Analysis of discriminant validity resulted in a non-significant correlation of r = 0.092 (p = 0.350).
Results indicate preliminary evidence for the construct validity of the "SmoCess-GP" instrument which therefore appears to be a promising tool for analyzing the extent of smoking cessation advice offered by GPs from the patients' perspective. Future research should examine the psychometric properties in a population based sample, further improvements of the instrument and should apply other methods of validation.
我们开发了一种评估科隆戒烟研究（CoSmoS）中全科医生（GP）戒烟活动程度的工具。本研究的目的是进一步检验“Smocess-GP”工具（GP 戒烟）的心理测量质量。
我们纳入了 127 名参与科隆戒烟研究（CoSmoS）的当前吸烟者进行分析。使用 Mplus 软件进行验证性因子分析（CFA），以检验模型拟合度，并重新测试工具的单因素结构。进一步通过测量患者对医生信任的工具进行二元分析来检验结构效度。
CFA 支持该工具的单维结构。因子负荷大于等于 0.50。所有指标的可靠性均高于 0.30。综合可靠性为 0.86，平均方差提取（AVE）值为 0.50。计算总体拟合指数得出 CFI 值为 1.00，TLI 值为 1.02。近似均方根误差（RMSEA）表明，模型未解释的信息量为 0%。卡方值为 chi2df = 6 = 4.63（p = 0.59）。判别有效性分析得出 r = 0.092（p = 0.350）的非显著相关性。
结果初步表明“Smocess-GP”工具具有结构有效性，因此它似乎是一种从患者角度分析 GP 提供的戒烟建议程度的有前途的工具。未来的研究应在基于人群的样本中检验心理测量特性，进一步改进工具，并应应用其他验证方法。
