Validação da escala South Oaks Gambling Screen em população brasileira / Validity of the South Oaks Gambling Screen in brazilian population

AUTOR(ES)
DATA DE PUBLICAÇÃO

2006

RESUMO

Pathological gambling is considered a behavioral addiction that shares similarities with drug addiction. The South Oaks Gambling Screen - SOGS is the most employed screening tool for pathological gambling. The SOGS investigates gambling games and behaviors associated to gambling. Answering at least 5 from 20 questions classifies the respondent as a probable pathological gambler. The aim of this study was to verify the performance of the SOGS in comparison to the DSM IV criteria for pathological gambling. The Total Sample comprised 54 gamblers under treatment at the Gamblers Treatment Unit of PROAD/UNIFESP (Clinical Group); 71 participants from the general population assessed in a metropolitan train station of the city of São Paulo (Train Group); and 116 subjects recruited in a bingo house (Bingo Group). Factorial structure, internal consistency and psychometric properties of SOGS (sensibility, specificity, positive predictive value, negative predictive value and misclassification rate) and correlation among both instruments were verified. In the Total Sample women were older, divorced or widowed, did not work; they had chased, had had more discussions about money and had gambled to escape from problems or to relieve dysforic mood. Women bets were mostly concentrated in bingo whereas men bets were more diversified. The most frequent forms of gambling in lifetime were, in decreasing order: bingo, lotteries, sports, cards and electronic games. In the last year and last month frametime, bingo, lotteries and electronic games were the most frequent games. The prevalence of bingo and electronic games among PG is discussed. Differences in the SOGS scores among three groups were statistically significant, attesting good construct validity. The correlation between SOGS and DSM IV in the Total Sample was high (r = 0,854, p <0.01) and the internal consistency measured by Cronbachs alpha was 0,75. Factor analysis resulted in a two-dimensional solution accounting for 56,7% of the total variance: a first factor composed by questions related to the gambling behavior of pathological gamblers and a second factor encompassing questions related to the sources of gambling money. Considering the Total Sample, validity measures were good: (sensibility = 100, specificity = 74,7, positive predictive value = 60,7, negative predictive value = 100 e misclassification rate =0,18). Elevating the cut score to 8 the number of false negatives is reduced (95,4, 89,8, 78,5 , 98, 0.09, respectively). The difference between the classification of pathological gamblers by SOGS and DSM IV was striking, as the proportion of pathological gamblers evaluated by SOGS, at the cut score 5, was much higher than the proportion evaluated by DSM IV in the Train Group (8,5% e 2,8%, respectively) and in Bingo Group (44% e 12%, respectively). It is argued that this difference may be an artifact of the cut score used. Thus, a suggestion is made that either the cut score should be changed from 5 to 8 or that a 12-month timeframe be employed. If the cut score 5 is maintained, then the result should be confirmed by clinical interview.

ASSUNTO(S)

sogs tracking pathological gambling jogo patológico test validity escalas rastreamento gender sogs validade do teste scaling (testing) gênero

ACESSO AO ARTIGO

Documentos Relacionados