[French translation and validation of a short version of the Suicide Intervention Response Inventory-2 (SIRI-2-VF)]

Training health professionals about suicidal crisis is one major line of suicide prevention worldwide. France has one of the highest suicide rates in Europe, and although crisis intervention training has been set up since 2001, there presently is no training assessment tool in the French language for health professionals trained in suicide prevention. In the four levels of Kirpatrick's education pyramid, training that takes place in France today solely assesses level one status, that is to say relative to the level of satisfaction of participants (self-report). This study proposes a validated short French version of the Suicide Intervention Response Inventory-2 (SIRI-2) of Neimeyer & Mc Innes. The SIRI-2 questionnaire assesses the ability of first line intervention in dealing with suicidal individuals.

The translation methodology was inspired from Vallerand's model of cross-cultural back translation. This method is regularly used for translating from the English language to a French version. In order to translate the English version, we used an extensive 7-step methodology implicating several bilingual translators, expert reviewers (psychologists and psychiatrists) and a scientific committee. Participants were 107 students from different French universities and study programs: psychology, medicine and nursing (17 were men; average age was 26.6). Fifteen of these participants answered the SIRI-2-VF on two occasions (separated by a 15-day interval) in order to estimate the temporal stability of the instrument. The scores of the students were compared to six French experts in suicide prevention and with the original expert group who worked on the development of SIRI-2 (n=7). We used Student t Test for construct validity, Cronbach's Alpha for internal consistency and Pearson's correlation coefficient for temporal stability.

Following a fidelity comparison of the results of the French experts with those of the American experts, ten items presenting the least good fidelity were suppressed in order to obtain a short version of the SIRI-2 containing 15 questions (SIRI-2-VF). Statistical analyses of the short version (15 questions, SIRI-2-VF) showed good validity (difference between experts and subjects is significant: t=31.5, P<0.001) and reliability (good internal consistency: α=0.850 for positive statements and α=0.830 for negative statements, and a temporal stability: r=0.827, bilateral test, P<0.001).

This tool should improve the range of specific instruments in French suicidology adapted for French culture of suicide intervention. It is the first tool in France that reaches level 2a of Kirkpatrick's pyramid to assess clinical skills after training in suicidology.