Validation of brief screening instruments for internalizing and externalizing disorders in Mozambican adolescents

Abstract

Abstract Background Mental disorders are the leading cause of disability for youth worldwide. However, there is a dearth of validated, brief instruments to assess mental health in low- and middle-income countries (LMIC). We aimed to facilitate identification of mental disorders in LMIC contexts by adapting and validating measures of internalizing and externalizing disorders for adolescents in Mozambique, an LMIC in southeastern Africa. Methods We selected instruments with good support for validity in high-income and other LMIC settings: the Patient Health Questionnaire Adolescent (PHQ-A), Generalized Anxiety Disorders 7 (GAD-7), and Strengths and Difficulties Questionnaire (SDQ). Instruments were adapted by local and international mental health specialists followed by cognitive interviews (n = 48) with Mozambican adolescents. We administered the instruments along with the Miniature International Neuropsychiatric Interview for Children and Adolescents (MINI-KID)to 485 adolescents aged 12–19 years attending two secondary schools in Maputo City, Mozambique. One week later, we re-administered instruments to a randomly selected sample of 49 adolescents. Results Participants were 66.2% (n = 321) female and the average age was 15.9 (S.D = 1.7).Internal consistency (alpha = 0.80, PHQ-A; 0.84, GAD-7; 0.80, SDQ) and test–retest reliabilty (ICC = 0.74, PHQ-A; 0.70, GAD-7; 0.77, SDQ) were acceptabe for the PHQ-A, GAD-7, and the full SDQ. The SDQ internalizing subscale showed poor test–retest reliability (ICC = 0.63) and the SDQ externalizing subscale showed poor internal consistency (alpha = 0.65). All instruments demonstrated good sensitivity and specificity (> 0.70). Youden’s index identified optimal cutoff scores of 8 for the PHQ-A, 5 for the GAD-7, 10 for the SDQ internalizing and 9 for the SDQ externalizing subscales, though a range of scores provided acceptable sensitivity and specificity. Conclusions Our data supports reliability and validity of the PHQ-A, GAD-7, and SDQ instruments for rapidly assessing mental health problems in Mozambican adolescents. Use of these tools in other contexts with limited specialists may asist with expanding mental health assessment. Specific instrument and cutoff selection should be based on screening goals, treatment resources, and program objectives.