How valid are proxy assessment of mental health and sleep comorbidities of patients with epilepsy using standardized questionnaires?

• Published in: Seizure (London, England) England), 2023-10, Vol.111, p.151-157
• Main Authors: Serrand, Chris, Faucanié, Marie, Jaussent, Audrey, Crespel, Arielle, Denuelle, Marie, Bartolomei, Fabrice, Vercueil, Laurent, Derambure, Philippe, Tyvaert, Louise, Marchal, Cécile, Landre, Elisabeth, Szurhaj, William, Mura, Thibault, Navarro, Vincent, Rheims, Sylvain, Picot, Marie-Christine
• Format: Article
• Language: English
• Subjects: Human health and pathology; Inter observer agreement; Life Sciences; Psychiatric comorbidities; Psychiatrics and mental health; Sleep disorders; SUDEP; Verbal autopsy
• ISSN: 1059-1311; 1532-2688
• DOI: 10.1016/j.seizure.2023.08.011

•Good agreement between patients and proxies on psychiatric comorbidities.•Also good agreement when evaluating sleep disorders and treatment adherence.•Agreement was especially good for respondents living with the patient.•In case of SUDEP, level of agreement allows to gather reliable information. The study aimed to determine the level of agreement between patients with epilepsy and their proxies when assessing psychiatric comorbidities, sleep disorders, and medication adherence using standardized questionnaires. This agreement study is an ancillary analysis of the PRERIES study, a matched case-control study exploring SUDEP risk factors. Controls aged 15 years and older, with active epilepsy or in remission for less than 5 years were recruited between 01/01/2011 and 03/31/2019. An interview was carried out by a trained psychologist on both the patient and a proxy-respondent. During these independent interviews, the following comorbidities were explored: psychiatric comorbidities using the MINI, the STAI- Y2 and NDDI-E scales, sleep disorders with the SDQ-SA and Epworth scales and medication adherence. Level of agreement between patient and their proxy was estimated using Gwet's AC1&2. Among the 107 patient–proxy dyads recruited, proxy respondents were mainly family members (65.4%) or spouses (30.8%). Exploration of present major depression showed excellent agreement at 0.81 [0.65;0.97], as well as exploration of dysthymia at 0.96 [0.61;1]. Suicidal risk evaluation had a lesser agreement at 0.77 [0.60;0.94]. Agreement on anxiety was moderate 0.5 [0.38;0.62]. For sleep disorder, SDQ-SA presented a better agreement than the Epworth questionnaire with respectively 0.73 [0.51;0.95] and 0.45 [0.26;0.63]. For medication adherence, the overall agreement rate was excellent (0.90 [0.78;1]). Exploration of potential risk factors through families can give valuable and relatively robust information, especially if the respondent lives with the patient, and should be retrieved, when possible, in usual clinical setting.