Strengths and Difficulties Questionnaire (SDQ) as a mental health screening tool in general and clinical adolescent populations in Latvia.

Objectives. Mental health screening instruments are crucial for measuring the risks of psychopathology both in population mental health studies and in clinical psychiatric settings. Our study aimed to compare the psychometric performance of the self-report version of the Strengths and Difficulties Questionnaire (SDQ) in a community and clinical samples of Latvian adolescents.
Materials and Methods. The population arm of the study was conducted using data from the international Health Behaviour in School-aged Children (HBSC) study year 2017/2018 Latvian database. The analysed sample comprised 4004 11-, 13- and 15-y.o. adolescents. The clinical sample consisted of 207 adolescent outpatients aged 11 to 17 years attending the Children’s clinical university hospital, Child psychiatry clinic.
Results. The proportion of girls was higher in the clinical group (60.9% VS 50.4%). The adolescents in the clinical group were slightly older (mean age 13.9 VS 12.9). Adolescents in the clinical group scored significantly higher in emotional, peer problems and hyperactivity sub-scales then adolescents form the general population group, but there were no significant differences between the groups in conduct problems and prosocial behaviour subscales. When adjusted to age and gender, adolescents from the clinical group had 5.4x higher odds of reaching the “abnormality” threshold (above the 90th percentile in the population) in emotional problems, 3.0x in hyperactivity, 2.8x in peer problems and 4.5x in total difficulties. The SDQ subscales reached a higher level of internal consistency in the clinical population, but in both populations, the externalizing subscales (conduct problems, hyperactivity) demonstrated poor to outright unacceptable internal consistency, indicating that the factor structure of the SDQ in the Latvian adolescent population is substantially different from the structure of the scale in the original UK population.
Conclusions. This analysis illustrates the importance of proper psychometric evaluation of the language adaptations of mental health screening tools when applied in public health research and in clinical practice.