Presentation on theme: "Parallel Session 2 2.3 Neglected Conditions : Depression and Violence THE SHORT FORM HEALTH SURVEY AS AN INSTRUMENT FOR THE SCREENING OF DEPRESSIVE SYMPTOMS."— Presentation transcript:
Parallel Session 2 2.3 Neglected Conditions : Depression and Violence THE SHORT FORM HEALTH SURVEY AS AN INSTRUMENT FOR THE SCREENING OF DEPRESSIVE SYMPTOMS IN THE ELDERLY POPULATION. Jair Licio Ferreira SANTOS (firstname.lastname@example.org) Ana Teresa de Abreu RAMOS-CERQUEIRA Antonia Regina Ferreira FUREGATO Maria Lucia LEBRÃO Yeda Aparecida de Oliveira DUARTE
Parallel Session 2 2.3 Neglected Conditions : Depression and Violence THE SHORT FORM HEALTH SURVEY AS AN INSTRUMENT FOR THE SCREENING OF DEPRESSIVE SYMPTOMS IN THE ELDERLY POPULATION. THIS PRESENTATION MAY BE CONSULTED AND DOWNLOADED IN THE SITE OF THE SUPERCOURSE : http://www.pitt.edu/~super1/ http://www.pitt.edu/~super1/ SUPERCOURSE is a global repository of lectures on public health and prevention.
THE SHORT FORM HEALTH SURVEY AS AN INSTRUMENT FOR THE SCREENING OF DEPRESSIVE SYMPTOMS IN THE ELDERLY POPULATION. Jair Licio Ferreira SANTOS Ana Teresa de Abreu RAMOS-CERQUEIRA Antonia Regina Ferreira FUREGATO Maria Lucia LEBRÃO Yeda Aparecida de Oliveira DUARTE SA ÚDE, B EM ESTAR E E NVELHECIMENTO: Estudo longitudinal sobre as condições de vida e saúde dos idosos do município de São Paulo Health, well-being and aging : A longitudinal study of health and living conditions of elderly in the city of São Paulo.
BACKGROUND Mental Illness - in particular Depression - has a broad impact on a patient's life, setting his perception of life, of himself, his health and his quality of life. (FLECK et al, 2002) The proximity of the worldview of the depressed patient, of his dismay, and self rated health to the possibility that, in trying to measure one of the entities, one is also measuring the other. LEADS the subjectivity implicit in the measurements of quality of life TO
BACKGROUND When measuring quality of life emphasis is given to subjective evaluation: responses refer to questions about one's perception of one's quality of life. Even the axes that would account for multiple dimensions are influenced by the individual's perception about himself, his state of anxiety and excitement. Thus, it follows from the subjective concept of quality of life its close relationship to mental health, particularly depression.
MOTIVATION Instruments designed to assess quality of life have been applied to trace mental conditions: -> anxiety (SANDERSON et al 2001), -> panic disorder, agoraphobia, social phobia ( SANDERSON, ANDREWS, 2002 ) ->depression (GILL et al 2007)
MOTIVATION The Short Form Health Survey (SF-12) with one of its two areas relating to the perception of mental health (MCS-12) may be useful to trace conditions such as depressive symptoms. But its validity as a measure of mental problems in general populations has received little attention (GILL et al 2007).
MOTIVATION Gill and colleagues (2007) evaluated the utility of the MCS-12 as screening instrument for depression and anxiety disorders diagnosed in a sample of the general population from the National Survey of Mental Health in Australia. The results confirmed the validity of using the MCS-12 in epidemiological research in Mental Health and its suitability as a screening tool for depression.
MOTIVATION These findings lead to a widening of interest in MCS-12 as a resource to assess indicators of depression, particularly in more susceptible individuals such as the elders.
OBJECTIVE To select a cutoff point of the MCS-12 scale that leads to an adequate discrimination of severe depressive symptoms with good sensitivity and specificity.
METHODS: SAMPLE Data was gathered by the SABE Study - Health Welfare and Ageing: in 2006, 1115 survivors of the first round (2000) were interviewed. A two-stage sample drawn in 2000 with clusters selected through probabilities proportional to the size was adjusted in 2006 to accurately represent the population of elders living in the city of São Paulo, Brasil.
METHODS: INSTRUMENTS Depressive symptoms were assessed using the Geriatric Depression Scale - GDS – (SHEIKH AND YESAVAGE, 1986). Cutoff score of 11 was used for severe symptoms. Mental Health component of the SF-12 (MSC- 12). (WARE, KOSINSKY & KELLER, 1996).
METHODS: ANALYSIS Indications of cutoff points for the MSC12 were found by Receiver Operating Characteristic curve (ROC) analysis. The area under the ROC curve (AUC) is an indication of the diagnostic power of the scale. This ROC analysis does not support weighting for effects of the Sample Design. Thus, final cutoff points were selected by examining the suggested points in weighted 2x2 tables.
RESULTS - SAMPLE TOTAL SAMPLE N0 = 1115 REFUSED TO ANSWER GDS N2 = 15 INCOMPLETE ANSWERS TO SF-12 N3 = 28 TOTAL NUMBER ANSWERING BOTHS SCALES N = 905 PROBABLE DEMENTIA (MMSE + PFAQ ) N1 = 167
RESULTS: SELECTION OF A CUTTING POINT. The ROC curve analysis indicated the points from 41 to 46 as possible good choices for cutoff. AUC was estimated as 89,4%. According to SWETS (1988) an AUC value of 90% or higher indicates a highly accurate scale, and between 80% and 90% indicates a useful scale for screening.
RESULTS In the analysis of double entry tables with weighting for sampling effects the score of 43 was chosen as the cutoff for severe depressive symptoms in both sexes. This score provided a good balance between the desirable values for sensitivity and specificity. We did focus on specificity, since the objective is to use the instrument for screening.
RESULTS: VALIDATION RESULTING FROM THE CHOICE OF 43 AS CUTOFF
RESULTS: VALIDATION BY SEX SEX Severe depressive symptoms MCS-12 Severe depressive symptoms - GDS PARAMETERS YESNOTOTAL FEMALES YES2,312,514,8 SENSIBILITY = 0,72 SPECIFICITY = 0,87 ACCURACY = 0,86 AREA UNDER CURVE = 0,84 CONFIDENCE INTERVAL (95%) : (0,76 – 0,92) NO1,084,285,2 TOTAL3,296,8100,0 MALES YES1.26,17,3 SENSIBILITY = 0,75 SPECIFICITY = 0,94 ACCURACY = 0,93 AREA UNDER CURVE = 0,97 CONFIDENCE INTERVAL (95%) : (0,93 – 0,99) NO0,492,392,7 TOTAL1,798,4100,0 Percentage of individuals classified as having severe depressive symptoms in MCS-12 and GDS and validation parameters.
DISCUSSION Gill and colleagues (2007) found similar values for the MCS-12 when compared to the CIDI-2.1: The area under the ROC curve was slightly higher (0.92), showed better sensitivity (87%) but lower specificity (83%). Their cutoff point was 45, which can be considered a similar result, since the population involved was older than 18 years, mean age 45, and in our sample we dealt with seniors aged 65 and over, mean age 73.7years.
CONCLUSIONS The mental component of SF-12 showed good performance for the screening of depression in an elderly population. It properly allocated 86% of females and 93% of men. The specificity was higher than sensitivity, which is convenient for screening instruments, and ranged from 87% to 94%.
CONCLUSIONS The SF-12 scale is widely used in epidemiologic research. These results indicate that its mental component MCS-12 is a useful screening instrument for depression among the elderly, and thus, a useful tool in epidemiological research in mental health.
LIMITATIONS MCS-12 was studied in comparison to only one other scale, also used for screening, the GDS. It would be important to evaluate the discriminatory properties of the MCS-12 in a population of elderly people diagnosed by the CIDI 2.1. This study refers to a population of elderly people in São Paulo - Brazil, and generalization of its application to other populations should take into account any cultural differences. The weights used in the analysis aimed to ensure representation of the senior community in general and were designed to apply in epidemiological studies. Thus, clinical studies or researches in primary care may achieve incomplete benefits from these results.