Carlos G Forero1,2,3, Elena Olariu4,5, Pilar Álvarez6, José-Ignacio Castro-Rodriguez4,5,6, Maria Jesús Blasco7,4,5, Gemma Vilagut7,4,5, Víctor Pérez6,8, Jordi Alonso7,4,5. 1. CIBER en Epidemiología y Salud Pública (CIBERESP), Madrid, Spain. cgarcia@imim.es. 2. Health Services Research Unit, IMIM (Institut Hospital del Mar d'Investigacions Mèdiques), Carrer del Doctor Aiguader, 88, Edifici PRBB, 08003, Barcelona, Spain. cgarcia@imim.es. 3. Department of Experimental and Health Sciences, Universitat Pompeu Fabra (UPF), Barcelona, Spain. cgarcia@imim.es. 4. Health Services Research Unit, IMIM (Institut Hospital del Mar d'Investigacions Mèdiques), Carrer del Doctor Aiguader, 88, Edifici PRBB, 08003, Barcelona, Spain. 5. Department of Experimental and Health Sciences, Universitat Pompeu Fabra (UPF), Barcelona, Spain. 6. Institute of Neuropsychiatry and Addictions (INAD), Parc de Salut Mar, Barcelona, Spain. 7. CIBER en Epidemiología y Salud Pública (CIBERESP), Madrid, Spain. 8. CIBER en Salud Mental (CIBERSAM), Madrid, Spain.
Abstract
PURPOSE: Functioning is a necessary diagnostic criterion for depression, and thus routinely assessed in depressive patients. While it is highly informative of disorder severity, its change has not been tested for prognostic purposes. Our study aimed to analyze to what extent early functioning changes predict depression in the mid-term. METHODS: Longitudinal study (four occasions: baseline, 1, 3, and 12 months) of 243 patients with depressive symptomatology at three different services (primary care, outpatients, and hospital). Functioning was assessed on the first three occasions using the Global Assessment of Functioning (GAF), the WHODAS-2.0, and a self-reported functioning (SRF) rating scale. Growth mixture modeling of initial assessments served to estimate individual person-change parameters of each outcome. Person-growth parameters were used as predictors of major depressive episode at 12 months in a logistic regression model, adjusted by sex, age, healthcare level, and depression clinical status at third month. Predictive accuracy of all measures was assessed with area under the receiver operating curve (AUC). RESULTS: Of the 179 patients who completed all assessments, 58% had an active depression episode at baseline and 20% at 12 months (64% non-recoveries and 36% new onsets). Individual trends of change in functioning significantly predicted patient depression status a year later (AUCWHODAS = 0.76; AUCGAF = 0.92; AUCSRF = 0.93). CONCLUSIONS: Longitudinal modeling of functioning was highly predictive of patients' clinical status after 1 year. Although clinical and patient-reported assessment had high prognostic value, the use of very simple patient-reported outcome measures could improve case management outside specialized psychiatric services.
PURPOSE: Functioning is a necessary diagnostic criterion for depression, and thus routinely assessed in depressivepatients. While it is highly informative of disorder severity, its change has not been tested for prognostic purposes. Our study aimed to analyze to what extent early functioning changes predict depression in the mid-term. METHODS: Longitudinal study (four occasions: baseline, 1, 3, and 12 months) of 243 patients with depressive symptomatology at three different services (primary care, outpatients, and hospital). Functioning was assessed on the first three occasions using the Global Assessment of Functioning (GAF), the WHODAS-2.0, and a self-reported functioning (SRF) rating scale. Growth mixture modeling of initial assessments served to estimate individual person-change parameters of each outcome. Person-growth parameters were used as predictors of major depressive episode at 12 months in a logistic regression model, adjusted by sex, age, healthcare level, and depression clinical status at third month. Predictive accuracy of all measures was assessed with area under the receiver operating curve (AUC). RESULTS: Of the 179 patients who completed all assessments, 58% had an active depression episode at baseline and 20% at 12 months (64% non-recoveries and 36% new onsets). Individual trends of change in functioning significantly predicted patientdepression status a year later (AUCWHODAS = 0.76; AUCGAF = 0.92; AUCSRF = 0.93). CONCLUSIONS: Longitudinal modeling of functioning was highly predictive of patients' clinical status after 1 year. Although clinical and patient-reported assessment had high prognostic value, the use of very simple patient-reported outcome measures could improve case management outside specialized psychiatric services.
Entities:
Keywords:
Affective disorders; Early intervention; Functional disability; Questionnaire; WHODAS
Authors: André Hajek; Christian Brettschneider; Marion Eisele; Dagmar Lühmann; Silke Mamone; Birgitt Wiese; Siegfried Weyerer; Jochen Werle; Angela Fuchs; Michael Pentzek; Janine Stein; Tobias Luck; Horst Bickel; Edelgard Mösch; Kathrin Heser; Michael Wagner; Wolfgang Maier; Martin Scherer; Steffi G Riedel-Heller; Hans-Helmut König Journal: Int Psychogeriatr Date: 2017-01-30 Impact factor: 3.878
Authors: Klaas J Wardenaar; Erik J Giltay; Tineke van Veen; Frans G Zitman; Brenda W J H Penninx Journal: J Affect Disord Date: 2011-12-15 Impact factor: 4.839
Authors: Elena Olariu; José-Ignacio Castro-Rodriguez; Pilar Álvarez; Carolina Garnier; Marta Reinoso; Luis Miguel Martín-López; Jordi Alonso; Carlos G Forero Journal: Qual Life Res Date: 2014-10-04 Impact factor: 4.147