Premium
Treatment‐Resistant Depression and Risk of Suicide
Author(s) -
Pfeiffer Paul N.,
Kim Hyungjin M.,
Ganoczy Dara,
Zivin Kara,
Valenstein Marcia
Publication year - 2013
Publication title -
suicide and life‐threatening behavior
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.544
H-Index - 90
eISSN - 1943-278X
pISSN - 0363-0234
DOI - 10.1111/sltb.12022
Subject(s) - depression (economics) , medicine , antidepressant , logistic regression , conditional logistic regression , comorbidity , stage (stratigraphy) , antidepressant medication , medical record , psychiatry , case control study , paleontology , biology , hippocampus , economics , macroeconomics
We evaluated whether treatment‐resistant depression ( TRD ) as measured by the M assachusetts G eneral H ospital ( MGH ) staging method was associated with suicide in a large U . S . health system. Data from the V eterans H ealth A dministration and the N ational D eath I ndex were used to conduct a case–control study of patients newly diagnosed with depression who received antidepressant treatment between 2003 and 2006. Suicide cases ( N = 499) were matched with nonsuicide controls ( N = 1994). Conditional logistic regression was used to assess whether MGH stage at time of suicide (or matched date) was associated with case status, adjusting for patient demographic characteristics, comorbidity, and service use. Results indicated 11.6% of suicide cases had MGH stage 3 or greater (indicating at least two antidepressant trials) compared to 6.4% of controls ( p < .001). In adjusted analyses, suicide was not significantly more likely among patients with stage 3 or greater ( OR 1.52; 95% CI : 0.98, 2.37) or stages 1.5–2.5 ( OR 1.19; 95% CI : 0.91, 1.55) compared to patients with stage 1 or less (<10 weeks of antidepressant medication). Staging TRD using MGH criteria is unlikely to substantially improve suicide risk assessment of depressed patients beyond existing measures contained in health system records.