Psicometric properties of an instrument for measuring patient’s satisfaction with physical therapy (medrisk) in spanish population: cultural diversity

Abstract

The Spanish-language versión of the MedRisk Insttrument for Mesauring Partient Satisfaction with Physical therapy Care was validated in a simple of 203 Spanish-speakers patients in New York City. The purpose of this investigation is to adapt this instrument to Spanish population (patients who assist to physical therapy services in Spain); to corroborate if exist a psychometric equivalence and to identify if Spanish patients consider relevant the same dimensions as the preliminary validation. Relevance: At time to used a measurement instrument that has been validated in another country, it is necessary to assure that it is not only reliable and valid to the country where was developed, but also it needs to be appropriated to the patients subjects of the current investigation. Participants: A total of 300 patients of the Hospital of San Juan de Dios del Aljarafe (in Sevilla, Spain) participated in this study (44% men and 56% women between 13 and 93 years old).

Methods: It has been done a cross-sectional descriptive study. Data were collected between April 2008 and November 2009.To the patients involved in the research, it was applied the original 20-item version of the MRPS (18 items and two global measures) and not the validated 12-item MRPS (10 items and two global measures). The main investigator randomly selected days during which the data were collected. The sample was taken finally on by convenience: all patients, in these selected days, who assist to medical care once they were done with their prescribed physiotherapy treatment.

Analysis: Like Spanish version in New York, factor structure was investigated using item-correlation and exploratory factor analysis. Group-level reliability for single test administration was assessed using the Cronbach alpha.

Results: Factor structure: the inter-item correlation matrix revealed that 3 items were not significantly correlated to both of the global measures . These 3 items were not included in the subsequent analysis. Principal components analysis revealed 4 eigenvalues greater than 2.0 explaining 67.93% of the cumulative variance. This suggested a 4-factor solutions. Following varimax rotation, a total of 14 items loaded on four factors were retained. Reliability: The Cronbach alpha was .87

Conclusions: Spanish version of MRPS that was validated with Spanish-speakers patients in New York is similar but not identical to the validation obtained with Spanish Population. In both of them, 9 of the 12 and 14 items finally selected in each other, have a coincidence. In the first one, there is a conclusion of two factors, internal and external. And in our investigation there were obtained 4 dimensions, 2 of them refer to internal factors (respect of the healthcare team and professional behavior of the therapist) and 2 that does it to external factors (adaptation of the center and accessibility)