Campos Juliana A D B, Zucoloto Miriane L, Bonafé Fernanda S S, Maroco João
Department of Food Science and Nutrition, São Paulo State University (UNESP), School of Pharmaceutical Sciences, Campus (Araraquara), Brazil.
Dentistry, Departamento de Odontologia Social, School of Dentistry, São Paulo State University (UNESP) - Araraquara Campus, Araraquara, São Paulo, Brazil.
Gerodontology. 2017 Sep;34(3):334-342. doi: 10.1111/ger.12270. Epub 2017 May 10.
To validity the General Oral Health Assessment Index (GOHAI) among adults who sought dental care and to present a new proposal for calculating scores on self-perception of oral health.
There is no study that presents a GOHAI scores using weight of the items.
The one-factor model, the three-factor model (physical function, psychosocial/psychological function and pain/discomfort) and the second-order hierarchical model (SOHM) were evaluated from confirmatory factor analysis (λ, χ /df, CFI,GFI and RMSEA). The reliability (CR,α) was estimated. Concurrent validity was assessed using the Oral Health Impact Profile (OHIP-14). The invariance of the models was estimated in independent samples. The calculation of an overall score using the factor scores was proposed to obtain the overall weighted scores. These overall weighted scores were compared to the scores estimated as the simple arithmetic mean (overall unweighted scores) using a repeated measures analysis of variance.
A total of 1000 individuals participated (74.1% female; age: 40.7 (SD=14.3) years). Three items of the GOHAI were excluded (λ<0.40). The one-factor model (λ=0.40-0.77; χ /df=6.291; CFI=0.947; GFI=0.960; RMSEA=0.073) and the three-factor model (λ=0.40-0.78; χ /df=8.321; CFI=0.932; GFI=0.954; RMSEA=0.086) each presented an adequate fit. Reliability was adequate (one-factor: CR=0.83/α=0.83; three-factor: CR=0.53-0.76/α=0.53-0.73), with the exception of the pain/discomfort factor. The GOHAI was invariant in independent samples, and the concurrent validity was adequate. The overall unweighted scores overestimated self-perceptions of oral health when compared with the weighted scores.
Both the one-factor and three-factor models of the GOHAI were found to be valid, reliable and invariant for the sample after the exclusion of three items. The use of overall weighted scores is recommended for calculating the score of self-perception of oral health.
验证在寻求牙科护理的成年人中通用口腔健康评估指数（GOHAI）的有效性，并提出一种计算口腔健康自我认知得分的新方法。
尚无研究使用各项目权重来呈现GOHAI得分。
通过验证性因素分析（λ、χ²/df、CFI、GFI和RMSEA）对单因素模型、三因素模型（身体功能、心理社会/心理功能和疼痛/不适）和二阶层次模型（SOHM）进行评估。估计信度（CR、α）。使用口腔健康影响程度量表（OHIP-14）评估同时效度。在独立样本中估计模型的不变性。建议使用因素得分计算总分以获得总体加权分数。使用重复测量方差分析将这些总体加权分数与作为简单算术平均数估计的分数（总体未加权分数）进行比较。
共有1000人参与（女性占74.1%；年龄：40.7（标准差=14.3）岁）。GOHAI的三个项目被排除（λ<0.40）。单因素模型（λ=0.40 - 0.77；χ²/df = 6.291；CFI = 0.947；GFI = 0.960；RMSEA = 0.073）和三因素模型（λ=0.40 - 0.78；χ²/df = 8.321；CFI = 0.932；GFI = 0.954；RMSEA = 0.086）的拟合度均良好。信度足够（单因素：CR = 0.83/α = 0.83；三因素：CR = 0.53 - 0.76/α = 0.53 - 0.73），疼痛/不适因素除外。GOHAI在独立样本中具有不变性，同时效度足够。与加权分数相比，总体未加权分数高估了口腔健康自我认知。
在排除三个项目后，GOHAI的单因素和三因素模型对该样本均有效、可靠且具有不变性。建议使用总体加权分数来计算口腔健康自我认知得分。
