Le deuil (perte par décès d'une personne significative) est une expérience universelle qui a suscité de nombreux travaux. Deux types de méthodes ont été utilisés : qualitatives et quantitatives. Outre la recherche des symptômes de la psychopathologie classique (dépression, troubles anxieux, stress et psychotraumatisme, etc.), les phénomènes de deuil font désormais l'objet d'investigations employant des échelles d'évaluation supposées spécifiques de l'expérience et du comportement de deuil, normal ou pathologique (par exemple les échelles TRIG, CBI, ICG, etc.). Un inventaire méthodologique critique est proposé.

Bereavement, grief and mourning (due to the loss of a significant person) are universal experiences. They have emerged as an important topic for scientific studies focused on losses and their consequences, studies of the attachment that precedes losses, and studies of other types of psychological trauma. Defined as a non-mental disorder (DSM IV, ICD 10), normal grief in International Classifications of Mental Diseases is now distinguished from pathological grief, i.e. complicated or traumatic (which amounts to 20 % of bereavements ?). Intensive empirical studies have been conducted in English-speaking countries as opposed to the psychodynamic Freudian and intuitive approaches in French literature. Scientific studies require several types of methods, usually quantitative approaches with the use of rating scales specially designed to measure the many symptoms of grief. There are many of these instruments, but almost none of them have been validated and compared to others. Only the TRIG (Texas Revised Inventory of Grief, Faschingbauer et al, 1987) has been translated and validated in French. The ICG (Inventory of Complicated Grief, Prigerson et al, 1995) is supposed to be a dimensional scale to discriminate complicated (traumatic) grief with a threshold (the cut-off point separating normal and traumatic grief). The ICG has been translated in French, validated and compared to the TRIG. Based essentially on the work of Prigerson, Jacobs and the Yale school ; Niemeyer and Hogan ; and the Japanese Tomita T., Kitamura T. ; we have reviewed the most useful and studied scals (TRIG, GEI, CBI, ICG). There are also qualitative studies of bereavement, based on the pure description of the mourning person, on content analysis, on the focus group method, on ethnography, on studies for cultural interpretation (cultural patterning) and finally on case studies. Bereavement, grief and mourning, although often very painful, are not pathological processes. The distress and the suffering, the depressive and anxious symptoms and behaviour specific of the process are usually transient. The quantitative psychopathology required for research can help the clinician to follow the course of the mourning process and, ideally, pinpoint the patient at risk of complicated (traumatic) grief. The scales should be reliable, validated and user-friendly. Many studies, in this clinical area, have used two types of methodology, qualitative and quantitative. Clinical research looks for the classical psychopathological disorders (depression, anxiety disorders, stress, psychological trauma etc.) and more recently symptoms and behaviour supposedly specific of grief and mourning. Several rating scales have been designed in order to assess the presence and intensity of the specific components of grief. This is a critical review of the methodologies used in this domain.