IJCS | Volume 31, Nº4, July / August 2018

DOI: 10.5935/2359-4802.20180037 414 ORIGINAL ARTICLE International Journal of Cardiovascular Sciences. 2018;31(4)414-421 Mailing Address: Thais de Rezende Bessa Guerra Rua dos Cristais, 22, quadra 92. Postal Code: 24342-685, Itaipu, Niterói, RJ - Brazil. E-mail: tcrezende34@gmail.com , thaisbessa@id.uff.br Methods of Screening for Depression in Outpatients with Heart Failure Thaís de Rezende Bessa Guerra, 1 Isabella Cristina Diniz Venancio, 1 Daniel Mählmann de Moura Pinheiro, 2 Mauro Vitor Mendlowicz, 1 Ana Carla Dantas Cavalcanti, 2 Evandro Tinoco Mesquita 1 Programa de Pós Graduação em Ciências Cardiovasculares, Universidade Federal Fluminense; 1 Niterói, RJ - Brazil Escola de Enfermagem Aurora de Afonso Costa, Universidade Federal Fluminense; 2 Niterói, RJ - Brazil Manuscript received August 03, 2017, revised manuscript November 07, 2017, accepted December 19, 2017. Abstract Background: Depression is a syndromic clinical condition underdiagnosed in patients with heart failure. Several instruments are currently applied to screen for depression. Objective: To determine the prevalence of depression and the agreement among screening methods for depression in patients with heart failure. Methods: Cross-sectional study conducted between March 2015 and January 2017 including 76 outpatients following up at a clinic specialized in heart failure. Depression was screened with the Hamilton Depression Rating Scale (HAM-D), Beck Depression Inventory-II (BDI-II), and Patient Health Questionnaire-9 (PHQ-9). The agreement among the three instruments was analyzed with Fleiss’ kappa coefficient (k F ), Krippendorff’s alpha coefficient (C k ) and Cronbach’s alpha coefficient. The accuracy, sensitivity, and specificity, as well as false-positive and false-negative results of the HAM-D and PHQ-9 were calculated considering the BDI-II as the gold-standard instrument in the diagnosis of depression. Results: The prevalence rates of depression were 72.4% (n = 55) with the HAM-D, 67.1% (n = 51) with the BDI-II, and 40.8% (n = 31) with the PHQ-9 scales. The prevalence of depression simultaneously identified by all three instruments was 28.9% (n = 22) and the diagnostic agreement (presence or absence of depression) was 47.4% (n = 36). The analysis revealed a superficial agreement (k F = C k = 0.27) and moderate consistency (( ↓ C = 0.602, significantly not null, p = 0.000). Sociodemographic and clinical variables were not risk factors for depression in the evaluated sample. Conclusion: The screening methods analyzed showed agreement and were useful in detecting depression among outpatients with heart failure. (Int J Cardiovasc Sci. 2018;31(4)414-421) Keywords: Heart Failure; Depression / diagnosis; Depression / prevalence; Medical Records; Surveys and Questionnaires; Cross-Sectional Studies. Introduction Depression is a disorder of multifactorial nature. 1 When associated with heart failure (HF), depression compromises the functional capacity, quality of life, and survival of the patient. 2-7 It is necessary to explore the screening methods used to diagnose depression because despite the variety of applied instruments, no studies have been conducted to evaluate the agreement of these methods in patients with HF. 1,8 The diagnosis of depression is established through the patient’s clinical history and duration of signs and symptoms, as well as the application of specific scales. 1,8-9 There are approximately 49 scales used in the multidimensional assessment of depression,10 including the Hamilton Depression Rating Scale (HAM-D), 11 Beck Depression Inventory-II, 12 and Patient Health Questionnaire-9 (PHQ-9), 13 among others. A study conducted by Matias et al. 1 comparing the screening of depression with the PHQ-9 and the