Introduction

SERVQUAL methodology has been used to build service quality evaluation instruments in many countries, in private and public organizations as well as in different services, including health services.

This methodology was designed by Parasuraman, Zeithaml and Berry in 1985, with the purpose of developing a marketing construct, considering «service quality as an elusive and hard to measure subject, since it is the result of a comparison of consumer expectations or wishes versus a provider, and their perceptions with respect to received service».¹ Thus, the instrument is composed of two parts: one corresponding to «expectations» applied before the patient or user enters the service and the other corresponding to «perceptions» applied when service is completed. When expectations are surpassed by patients expectations during the course of treatment, service rendering is considered excellent, contrarily, when expectations are not met, service is considered deficient; when user’s expectations are met service is considered satisfactory. This can be calculated with the following formula:

Formula SQi represents global perceived quality of stimulus i; k indicates the number of attributes (22 in this case); Pij shows perception of stimulus i result with respect to attribute j and Eij is the expectation for service quality of attribute j for stimulus i. In this model, perceived service quality increases as difference between Eij and Pij increases along different assessed attributes.²

According to studies conducted by Grönroos in 1984, it has been identified for decades that technical and functional quality weigh heavily in service providing organizations.³

From this perspective, technical and functional quality are identified in health service quality, technical quality is for the exclusive use of health professionals and administrators; functional quality is perceived by the user and/or patient with respect to service final result and its development. Due to the fact that frequently patients cannot accurately assess technical quality of a medical care service, functional quality is normally the main determinant factor which can be identified through patient perception.⁴

Evaluation of provision of a service, including health services, has common dimensions, nevertheless, items of the model’s dimensions must be adapted to the special characteristics of this service and of the users, therefore, to be reliable, instruments must correspond to processes of each service, in addition to being statistically validated with the aim of obtaining objective results which might allow identification of opportunity areas in an organization so as to implement continuous improvement processes.

Reliability and validity are the essential characteristics of a measuring instrument in the field of social and behavioral sciences which provide «psychometric solidity» to the instrument.⁵

Material and methods

With previous authorization of the Research Committee of the Stomatology Faculty, Autonomous University of Puebla, a descriptive, cross-sectional, observational and prospective study was conducted in the service and teaching clinics (STC) at undergraduate level. At the first stage, a measurement instrument was designed in accordance to the five categories of the SERVQUAL: tangible elements, reliability, response capacity, security and empathy. The original model was composed of 22 items, in the present study two additional items were included due to the fact that a teacher participates in this process, who is responsible for supervising the development of the student’ s clinical skills, through treatment protocols established within normativity and ethical principles in order to provide stomatological services to patients, therefore the instrument was composed of 24 items, which were evaluated by means of a Likert scale with five response options.

According to SERVQUAL methodology, an instrument was designed; it was divided into two parts in order to conduct evaluation in two moments; the first corresponds to «expectations», and was used before patient attendance to STC, the second corresponds to «perceptions» and was applied at the end of the service.

Once the instrument was designed and adapted it was examined by a team of experts to assess questions’ relevance and clarity. A pilot study was additionally conducted in a stratified, random sample of 100 patients in different schedules in all seven service and teaching clinics: Pediatrics Clinic, Integrated Clinic, Comprehensive Clinic 1 and Endodontics Clinic with the aim of identifying opportunity areas and be able to perform improvements in the instrument.

Its application was later conducted in a stratified randomized sample of 400 patients who attended teaching and service clinics; to this end researchers were subjected to a standardization process, it was additionally considered that they should have no relationship with the clinics, therefore, students of first courses were selected.

Statistical treatment of the instrument consisted on a Cronbach’s alpha test, which measures a construct to evaluate its correlation; the closer to value 1 found in alpha Cronbach, greater will the internal consistency be,⁶ as observed in Table I.

Results

Instrument’s validity was assessed through a reliability analysis, conducted with alpha Cronbach coefficient; with it both parts of the instrument were individually analyzed using factorial analysis.

For the first time, corresponding to «expectations» three common factors were obtained, which explained the 70.18% of studied phenomenon’s total variability: the first factor explains 34.22%, the second factor 18.44% and the third factor explains 17.51%. Interpretation of each factor must be obtained interpreting saturations presented in Table II.

For the second part, corresponding to «perceptions» two common factors were obtained, which explain 65.5% of total variability in the studied phenomenon: the first factor explains 42.7% and the second factor 22.8%. Interpretation of each factor can be obtained by interpreting saturations as observed in Table III.

Reliability determined by alpha Cronbach coefficient exhibited a value of 0.967 for the section «expectations», whereas for the second section corresponding to «perceptions» it exhibited a value of 0.935, therefore, analysis conducted eliminating each question secures instrument’s pertinence.

Discussion

Both parts of the questionnaire, that corresponding to «expectations» and the one corresponding to «perceptions» remain unchanged, this shows the instrument’s pertinence, given the high reliability indexes calculated with the help of alpha Cronbach coefficient, the validation of the construct’s content was obtained through factorial analysis method; in the «expectations» aspect, three factors were identified: first one is facultative staff care, the second would be service facilities and administrative personnel, the third would be equipment availability and suitable orientation, whereas in the part corresponding to «perceptions» two factors were identified: the first one was medical care provided by the clinic’s personnel and the second one equipment and orientation. Results obtained corroborate validity of this instrument.

Conclusion

In health services there is a great number of validated instruments to identify the degree of satisfaction, nevertheless, each instrument must identify and locate each of the model’s dimensions; selecting undertaken processes in the services to be evaluated in order to obtain results. Obtained results for the SERVQUAL instrument show excellent reliability and validity. Methodology of Parasuraman, Zeithaml and Berry allows undertaking adaptations according to specificity and pertinence of processes of the service to be evaluated, this allows its successful, continuous use in different countries and languages.^7-10