Research

Construction of a questionnaire measuring outpatients' opinion of quality of hospital consultation departments

Isabelle Gasquet^1,2 , Sylvie Villeminot² , Carla Estaquio² , Pierre Durieux³ , Philippe Ravaud⁴ and Bruno Falissard²

¹  Delegation of innovation, health security and assessment, Direction of medical policy – Assistance Publique-Hôpitaux de Paris, 3 avenue Victoria 75184 Paris cedex, France

²  UPRES JE 2360 (PSIGIM : Paris South Innovation Group in Mental health), Paul Brousse hospital, Villejuif, France

³  Department of Public Health and Medical Informatics, Faculté de Médecine Broussais Hôtel Dieu, Paris, France

⁴  Department of Epidemiology, Biostatistics et and clinical research, Hôpital Bichat Assistance Publique-Hôpitaux de Paris, Paris, France

author email corresponding author email

Health and Quality of Life Outcomes 2004, 2:43doi:10.1186/1477-7525-2-43

Published:	4 August 2004

Abstract

Background

Few questionnaires on outpatients' satisfaction with hospital exist. All have been constructed without giving enough room for the patient's point of view in the validation procedure. The main objective was to develop, according to psychometric standards, a self-administered generic outpatient questionnaire exploring opinion on quality of hospital care.

Method

First, a qualitative phase was conducted to generate items and identify domains using critical analysis incident technique and literature review. A list of easily comprehensible non-redundant items was defined using Delphi technique and a pilot study on outpatients. This phase involved outpatients, patient association representatives and experts. The second step was a quantitative validation phase comprised a multicenter study in 3 hospitals, 10 departments and 1007 outpatients. It was designed to select items, identify dimensions, measure reliability, internal and concurrent validity. Patients were randomized according to the place of questionnaire completion (hospital v. home) (participation rate = 65%). Third, a mail-back study on 2 departments and 248 outpatients was conducted to replicate the validation (participation rate = 57%).

Results

A 27-item questionnaire comprising 4 subscales (appointment making, reception facilities, waiting time and consultation with the doctor). The factorial structure was satisfactory (loading >0.50 on each subscale for all items, except one item). Interscale correlations ranged from 0.42 to 0.59, Cronbach α coefficients ranged from 0.79 to 0.94. All Item-scale correlations were higher than 0.40. Test-retest intraclass coefficients ranged from 0.69 to 0.85. A unidimensional 9-item version was produced by selection of one third of the items within each subscale with the strongest loading on the principal component and the best item-scale correlation corrected for overlap. Factors related to satisfaction level independent from departments were age, previous consultations in the department and satisfaction with life. Completion at hospital immediately after consultation led to an overestimation of satisfaction. No satisfaction score differences existed between spontaneous respondents and patients responding after reminder(s).

Conclusion

Good estimation of patient opinion on hospital consultation performance was obtained with these questionnaires. When comparing performances between departments or the same department over time scores need to be adjusted on 3 variables that influence satisfaction independently from department. Completion of the questionnaire at home is preferable to completion in the consultation facility and reminders are not necessary to produce non-biased data.