Limits...
Development and validation of a questionnaire on 'Satisfaction with dermatological treatment of hand eczema' (DermaSat).

Ruiz MA, Heras F, Alomar A, Conde-Salazar L, de la Cuadra J, Serra E, Regalado F, Halbach R, GEIDAC Gro - Health Qual Life Outcomes (2010)

Bottom Line: The reduction and validation sample was composed of 54% women and 46% men, of various educational levels with an average age of 43 years (SD = 13.7).The questionnaire was shortened to a version containing 17 items grouped in six dimensions: effectiveness, convenience, impact on HRQL, medical follow-up, side effects, and general opinion.Treatment satisfaction scores attained correlations between 0.003 and 0.222 with the HRQL measures, and showed higher correlations with the effectiveness (r = 0.41) and tolerability (0.22) measures, but very low correlation with compliance (r = 0.015).

View Article: PubMed Central - HTML - PubMed

Affiliation: Department of Methodology, School of Psychology, Universidad Autónoma de Madrid, Madrid, Spain. miguel.ruiz@uam.es

ABSTRACT

Objective: To develop a self-administered short questionnaire to assess patient satisfaction with medical treatment for hand eczema (dermatitis) with good psychometric properties.

Method: The content of the questionnaire was determined on the basis of clinical consultation with groups of patients, from studying the existing instruments, and from discussions with a panel of seven experts. A first draft version containing 38 items organised in six dimensions was tested on a pilot sample of patients to assess its legibility. The extended version was then tested on a sample of 217 patients of both genders enrolled at 18 hospitals representative of the national distribution. The questionnaire was supplied together with the Morisky-Green compliance questionnaire, the health-related quality of life (HRQL) SF-12 questionnaire, and a visual analogue scale (VAS) of perceived health status to assess concurrent validity. The dimensionality was reduced by means of exploratory factor analysis, and reliability was evaluated on the basis of internal consistency and two halves reliability estimates. Item discriminant capability and questionnaire discriminant validity with respect to known groups of patients (by gender, principal diagnosis, age, disease severity and treatment) were also assessed.

Results: The reduction and validation sample was composed of 54% women and 46% men, of various educational levels with an average age of 43 years (SD = 13.7). Of those who responded, 26% were diagnosed with hyperkeratotic dermatitis of the palms and 27% of the fingertips, and 47% with recurring palmar dyshidrotic eczema. The questionnaire was shortened to a version containing 17 items grouped in six dimensions: effectiveness, convenience, impact on HRQL, medical follow-up, side effects, and general opinion. Cronbach's alpha coefficient reached a value of 0.9. The dimensions showed different degrees of correlation, and the scores had a normal distribution with an average of 58.4 points (SD = 18.01). Treatment satisfaction scores attained correlations between 0.003 and 0.222 with the HRQL measures, and showed higher correlations with the effectiveness (r = 0.41) and tolerability (0.22) measures, but very low correlation with compliance (r = 0.015). Significant differences were observed between some diagnoses and treatments.

Conclusions: The shortened questionnaire proved to have good psychometric properties, providing excellent reliability, satisfactorily reproducing the proposed structure and supplying evidence of validity.

Show MeSH

Related in: MedlinePlus

Confirmatory factor analysis, standardized estimates. TE = Treatment Effectiveness, GS = General Satisfaction, ID = Impact on Activities of Daily Living, CU = Convenience of Use, MC = Medical Care, UE = Undesired Side-Effects.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC2992504&req=5

Figure 1: Confirmatory factor analysis, standardized estimates. TE = Treatment Effectiveness, GS = General Satisfaction, ID = Impact on Activities of Daily Living, CU = Convenience of Use, MC = Medical Care, UE = Undesired Side-Effects.

Mentions: Although the factor structure resulting from the scaling analysis provides evidence on how the items cluster together in the concepts they are intended to measure, a confirmatory factor analysis was conducted, imposing the theoretical structure and re-estimating item loadings. Figure 1 shows the results of the confirmatory estimation of the theoretical structure proposed for the questionnaire, using the generalised least squares method. All loadings were significant (p < 0.001) as were all the correlations between factors (p < 0.05) except for the correlation between undesirable side-effects and satisfaction with medical care (p = 0.217). The goodness-of-fit statistics indicate a good or very good fit: GFI = 0.918; AGFI = 0.880; CFI = 0.887; RMR = 0.047, χ2/gl = 1.312 and RMSEA = 0.040. Even the chi-square goodness-of-fit test makes it possible not to reject the hypothesis (with α = 0.01) stating that the correlation matrix can be correctly reproduced with the theoretical structure (χ2 = 136.4; gl = 104; p = 0.018), a desirable but infrequent output with sample sizes like the one used. However, the estimated correlations among the three dimensions were extremely high, especially the correlation between effectiveness and general satisfaction (r = 0.92), impact on daily living and general satisfaction (r = 0.85), and effectiveness and impact on daily living (r = 0.85). By contrast, the dimension showing the greatest independence was undesirable side-effects, which presented low correlations with respect to the dimensions of medical care (r = -0,10), effectiveness (r = -0,19), and general satisfaction (r = -0,18).


Development and validation of a questionnaire on 'Satisfaction with dermatological treatment of hand eczema' (DermaSat).

Ruiz MA, Heras F, Alomar A, Conde-Salazar L, de la Cuadra J, Serra E, Regalado F, Halbach R, GEIDAC Gro - Health Qual Life Outcomes (2010)

Confirmatory factor analysis, standardized estimates. TE = Treatment Effectiveness, GS = General Satisfaction, ID = Impact on Activities of Daily Living, CU = Convenience of Use, MC = Medical Care, UE = Undesired Side-Effects.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC2992504&req=5

Figure 1: Confirmatory factor analysis, standardized estimates. TE = Treatment Effectiveness, GS = General Satisfaction, ID = Impact on Activities of Daily Living, CU = Convenience of Use, MC = Medical Care, UE = Undesired Side-Effects.
Mentions: Although the factor structure resulting from the scaling analysis provides evidence on how the items cluster together in the concepts they are intended to measure, a confirmatory factor analysis was conducted, imposing the theoretical structure and re-estimating item loadings. Figure 1 shows the results of the confirmatory estimation of the theoretical structure proposed for the questionnaire, using the generalised least squares method. All loadings were significant (p < 0.001) as were all the correlations between factors (p < 0.05) except for the correlation between undesirable side-effects and satisfaction with medical care (p = 0.217). The goodness-of-fit statistics indicate a good or very good fit: GFI = 0.918; AGFI = 0.880; CFI = 0.887; RMR = 0.047, χ2/gl = 1.312 and RMSEA = 0.040. Even the chi-square goodness-of-fit test makes it possible not to reject the hypothesis (with α = 0.01) stating that the correlation matrix can be correctly reproduced with the theoretical structure (χ2 = 136.4; gl = 104; p = 0.018), a desirable but infrequent output with sample sizes like the one used. However, the estimated correlations among the three dimensions were extremely high, especially the correlation between effectiveness and general satisfaction (r = 0.92), impact on daily living and general satisfaction (r = 0.85), and effectiveness and impact on daily living (r = 0.85). By contrast, the dimension showing the greatest independence was undesirable side-effects, which presented low correlations with respect to the dimensions of medical care (r = -0,10), effectiveness (r = -0,19), and general satisfaction (r = -0,18).

Bottom Line: The reduction and validation sample was composed of 54% women and 46% men, of various educational levels with an average age of 43 years (SD = 13.7).The questionnaire was shortened to a version containing 17 items grouped in six dimensions: effectiveness, convenience, impact on HRQL, medical follow-up, side effects, and general opinion.Treatment satisfaction scores attained correlations between 0.003 and 0.222 with the HRQL measures, and showed higher correlations with the effectiveness (r = 0.41) and tolerability (0.22) measures, but very low correlation with compliance (r = 0.015).

View Article: PubMed Central - HTML - PubMed

Affiliation: Department of Methodology, School of Psychology, Universidad Autónoma de Madrid, Madrid, Spain. miguel.ruiz@uam.es

ABSTRACT

Objective: To develop a self-administered short questionnaire to assess patient satisfaction with medical treatment for hand eczema (dermatitis) with good psychometric properties.

Method: The content of the questionnaire was determined on the basis of clinical consultation with groups of patients, from studying the existing instruments, and from discussions with a panel of seven experts. A first draft version containing 38 items organised in six dimensions was tested on a pilot sample of patients to assess its legibility. The extended version was then tested on a sample of 217 patients of both genders enrolled at 18 hospitals representative of the national distribution. The questionnaire was supplied together with the Morisky-Green compliance questionnaire, the health-related quality of life (HRQL) SF-12 questionnaire, and a visual analogue scale (VAS) of perceived health status to assess concurrent validity. The dimensionality was reduced by means of exploratory factor analysis, and reliability was evaluated on the basis of internal consistency and two halves reliability estimates. Item discriminant capability and questionnaire discriminant validity with respect to known groups of patients (by gender, principal diagnosis, age, disease severity and treatment) were also assessed.

Results: The reduction and validation sample was composed of 54% women and 46% men, of various educational levels with an average age of 43 years (SD = 13.7). Of those who responded, 26% were diagnosed with hyperkeratotic dermatitis of the palms and 27% of the fingertips, and 47% with recurring palmar dyshidrotic eczema. The questionnaire was shortened to a version containing 17 items grouped in six dimensions: effectiveness, convenience, impact on HRQL, medical follow-up, side effects, and general opinion. Cronbach's alpha coefficient reached a value of 0.9. The dimensions showed different degrees of correlation, and the scores had a normal distribution with an average of 58.4 points (SD = 18.01). Treatment satisfaction scores attained correlations between 0.003 and 0.222 with the HRQL measures, and showed higher correlations with the effectiveness (r = 0.41) and tolerability (0.22) measures, but very low correlation with compliance (r = 0.015). Significant differences were observed between some diagnoses and treatments.

Conclusions: The shortened questionnaire proved to have good psychometric properties, providing excellent reliability, satisfactorily reproducing the proposed structure and supplying evidence of validity.

Show MeSH
Related in: MedlinePlus