|
Abstract
|
BACKGROUND: With the object to obtain a screening scale of suicidality, assesses the reliability and validity of the suicidality scale of Okasha et al and determine the cutoff point to identify a suicide attempt.
SUBJECTS AND METHODS: The study was conducted in high school students (n = 763, N = 13.825). The reliability was evaluated with analysis of internal consistency and homogeneity of the items. For the validity logistic regression and Pearson coefficient of correlation were used. The cutoff point was determined with Characteristic of Operation of the Receiver (ROC) curve and the area under the curve for the level of discriminatory exactitude. Result The scale obtained a Cronbach alpha coefficient of 0, 89 and adecuate homogeneity (item-total correlations corrected: r = 0.76 items 1 and 3; r = 0.82 ítem 2); discriminate between adolescents who have attempted suicide and tose who did not (Wald de ÷2 (1) = 1.53, p < 0.001; coefficient b = 0.66; odds ratio = 1.94 (IC 95%: 1.75-2.16)). The area under ROC curve was of 0.89. The cutoff point corresponded to 5 point, sensitivity = 90% and specificity = 79%. The scores of the scale correlated significantly with low self-esteem (r (761)= -0, 52, p < 0.001), lower perceived social support (r (761) = -0, 37, p < 0.001) and higher impulsivity (r (761) = 0, 29, p < 0.001).
CONCLUSIONS: The evaluated scale presents high internal consistency and homogeneity, withstatistical of validity strongly significant and cutoff point adecuate sensitivity and specificity.
Language: es |