Validation of a French version of the Breakthrough Pain Assessment Tool in cancer patients: factorial structure, reliability and responsiveness

Abstract

Objective: Breakthrough cancer pain should be properly assessed for better-personalized treatment plan. The Breakthrough Pain Assessment Tool is a 14-item tool validated in English developed for this purpose; no French version is currently available and validated. This study aimed to translate it in French and assess the psychometric properties of a French version of the Breakthrough Pain Assessment Tool (BAT-FR).

Methods: First, translation and cross-cultural adaptation of the 14 items (9 ordinal and 5 nominal) of the original BAT tool in French language was made. Second, assessments of validity (convergent, divergent and discriminant validity), factorial structure (exploratory factor analysis) and test-retest reliability of the 9 ordinal items were done with data of 130 adult cancer patients suffering from breakthrough pain in a hospital-academic palliative care center. Test-retest reliability and responsiveness of total and dimension scores derived from these 9 items were also assessed. Acceptability of the 14 items was also assessed on the 130 patients.

Results: The 14 items had good content and face validity. Convergent and divergent validity, discriminant validity and test-retest reliability of the ordinal items were acceptable. Test-retest reliability and responsiveness of total and dimensions derived from ordinal items were also acceptable. The factorial structure of the ordinal items had two dimensions similar to the original version: "1-pain severity and impact" and "2-pain duration and medication". Items 2 and 8 had a low contribution to the dimension 1 they were assigned and item 14 clearly changed of dimension compared with the original tool. The acceptability of the 14 items was good.

Conclusion: The BAT-FR has shown acceptable validity, reliability and responsiveness supporting its use for assessing breakthrough cancer pain in French-speaking populations. Its structure needs nevertheless further confirmation.