Satisfaction with care among patients with non-metastatic breast cancer: Development and first steps of validation of the REPERES-60 questionnaire

Abstract

Background: The care itinerary for cancer involves difficulties that occur in several different areas, whether in the diagnostic procedures, in surgery, or in adjuvant treatment. The aim of this work was to obtain a valid instrument measuring satisfaction among patients with breast cancer and exploring their care itinerary overall. Methods: Development phase: Patient focus groups were implemented in two French regions in order to identify areas of satisfaction in relation to the different phases of care provision in breast cancer. On the basis of the literature and the themes and wordings derived from the focus groups, the patients identified several areas of satisfaction, which they found to be partially covered in an American satisfaction measure that has been validated in the French general population (the Consumer Satisfaction Survey in its French version, CSS-VF, 39 items). The patient focus groups suggested adaptation of certain dimensions of this instrument to the potential care providers (37 items) and produced 45 new items in six areas. Validation phase: Using a large sample of patients (cohort of 820 women with invasive non-metastatic breast cancer) approached one month after treatment, this phase selected items that were comprehensible (non-response rate < 10%), non-redundant (r < 0.80) and reproducible (test-retest conducted on a sub-sample of 166 patients). The dimensions were identified by factor analysis on the selected items. Divergent and discriminant validity were assessed (relationships with quality of life questionnaire, comparisons between extreme groups). Results: Results were in favour of not inserting additional broken-down items into the CSS-VF and retaining 21 new items. The factor analysis found the initial structure of the CSS-VF (39 items in 9 dimensions) and the 21 new items divide up into four dimensions (listening abilities and information provided by doctors, organisation and follow-up of medical care provision, psychological support, material environment). No redundancy was observed between new items and CSS-VF items. Internal consistency was high. Divergent and discriminant validity were satisfactory. Conclusion: Adding four new dimensions to the CSS-VF yielded a valid 60-item instrument for assessment of care provided in breast cancer. These promising results now require further investigations of its responsiveness and its robustness in other linguistic, cultural and healthcare settings. © 2007 Defossez et al; licensee BioMed Central Ltd.