Validation of the parent version of the Strengths and Difficulties Questionnaire (SDQ) to screen mental health problems among school-age children in Mongolia

Abstract

Abstract Background Child and adolescent mental health problems are urgent health issues in low- and middle-income countries. To promote child and adolescent mental health services, simple validated screening tools are helpful. In Mongolia, the Strengths and Difficulties Questionnaire (SDQ), an internationally used child and adolescent mental health screening tool for children aged 4–17, was translated but not yet validated. To use the questionnaire appropriately, validation is necessary. Methods Children at 4th year at elementary school (community sample) and children visited psychiatric outpatient service (clinical sample) were recruited and their parental version of the SDQ was compared. The discriminating ability of the parental version of the SDQ was examined using Receiver Operating Characteristics (ROC) analysis on the SDQ total difficulties score. The area under the ROC curve (AUC) was used as a measure. Cut-off score was determined by normative banding that categorizes children with the highest 10% score range as abnormal and the second highest 10% as borderline following the original method; this cut-off score was compared with the cut-off score candidates with good balance between sensitivity and specificity using ROC analysis. Results We included 2301 children in the community sample, and 429 children in the clinical sample. Mean age was 9.7 years (SD 0.4, range 8.3–12.0) among the community sample and 10.4 years (SD 3.8, range 4.0–17.8) among the clinical sample. The mean total difficulties score was 12.9 (SD 4.8) among the community sample and 20.4 (SD 6.2) among the clinical sample. A total of 88.8% of the community sample and 98.8% of the clinical sample answered the SDQ. Using ROC analysis, the AUC was 0.82 (95% confident interval 0.80–0.85), which meant moderate discriminating ability. Using normative banding, the borderline cut-off score was 16/17 and abnormal cut-off score was 19/20. For cut-off scores of 16/17 and 19/20, sensitivity was 71.9 and 53.8% and specificity was 78.5 and 90.5%, respectively. The cut-off score candidates by ROC analysis were 16/17 and 17/18. Conclusions The parental version of the SDQ had moderate discriminating ability among Mongolian school-age children. For the screening of mental health problems among community children, cut-off score of 16/17 is recommended.