Can family members reliably assess mood and suicidal ideation in individuals with epilepsy?

BACKGROUND

Routine detection of depression, anxiety and suicidal ideation in people with epilepsy (PWE) remains suboptimal. We investigated the level of agreement between PWE and their proxies when evaluating these psychiatric symptoms.

METHODS

From October 2021 to March 2022, we conducted a cross-sectional anonymous survey at Vilnius University Hospital Santaros Klinikos (Vilnius, Lithuania). Persons accompanying PWE completed different scales measuring symptoms of depression and anxiety and evaluated suicidal ideation among PWE (defined as a non-zero score of the suicide item of the Beck depression inventory (BDI) or as a score > 1 of the item 4 of the Neurological Disorders Depression Inventory for Epilepsy, NDDI-E). Agreement between PWE and their proxies was measured using the Wilcoxon test for paired samples and Pearson's correlation analysis.

RESULTS

The study included 140 PWE (mean age 41.0 ± 18.1, 71 (50.7 %) female) and 140 proxies (mean age 49.0 ± 15.01, 105 (75.0 %) female, 57 (40.7 %) were parents, 51 (36.4 %) - spouses, 15 (10.7 %) - partners, 11 (7.9 %) - children and 6 (4.3 %) - siblings of PWE). There were no statistical differences in paired scores of the BDI, Hospital Anxiety and Depression scale, Generalized Anxiety Disorder scale-7, and the NDDI-E (p > 0.05). Proxies reported statistically higher scores on the Geriatric Depression Scale (Z =  - 2.026, p = 0.043) than PWE. The correlation between PWE and proxy evaluations for all instruments was moderate (varying from r = 0.500 to r = 0.688, p < 0.001). For most psychometric scales proxies became less accurate with higher scale scores for anxiety and depression and tended to underestimate the level of such symptoms (correlation between scale results and signed (PWE - proxy) mismatch varied from r = 0.368 to r = 0.641, p < 0.001). The measure of proxy-PWE agreement on suicidal ideation was low (Cohen's Κ = 0.192, p = 0.025); proxies missed 63 % (NDDI-E) to 70 % (BDI) of cases of suicidal ideation.

CONCLUSION

Companions of PWE provided comparable reports of participant anxiety and depression on a group level but tended to underestimate psychiatric symptoms as their scale scores became higher. Proxies significantly underreported suicidal ideation in those people who did express it. While proxy reporting should not be used interchangeably on an individual level, its use in group studies may be useful and ought to be investigated further.