90 Screening Test Accuracy and Reliability of the Parent-Report Preschool Strengths and Difficulties Questionnaire (P-SDQ) in Primary Care Settings

Abstract

Abstract Introduction/Background Previous research shows that mental health disorders can be identified in preschoolers. Detecting disorders early in life is key for early intervention and prevention of life-long behavioural and emotional issues. Individuals who experience mental health issues before the age of 14 years are at increased risk of a mental health disorder in adulthood, contrary to the idea that issues seen in young children are transient. Although the rate of mental health disorders in preschoolers is similar to the rate in older children, there is no agreed upon standard of care for routine screening or case identification of psychiatric problems in young children. Improving mental health and long-term outcomes for children requires early identification and monitoring of children at risk, including a screening tool that can be used in primary care settings. Objectives To evaluate the criterion validity (screening test accuracy) and test-retest reliability of the parent-report preschool Strengths and Difficulties Questionnaire (P-SDQ) in primary care settings for a sample of 2-4 year olds. Specifically, we were interested in evaluating the screening test accuracy of the P-SDQ for presence of DSM-5 diagnoses. Design/Methods Children 24-48 months were recruited at scheduled primary care visits in Toronto, Canada. Parents completed the P-SDQ at baseline, 2 and 12 weeks. At 12 weeks, parents participated in the Preschool Age Psychiatric Assessment (PAPA). We assessed the criterion validity of the baseline P-SDQ using three different criterion measures. We compared the Total Difficulties Score (TDS) with presence of any DSM-5 diagnoses on the PAPA. For the internalizing subscale score, the criterion was presence of any internalizing disorder. For the externalizing subscale score, the criterion was presence of any externalizing disorder. Criterion validity was evaluated using area under the curve (AUC) and calculating screening test properties for a range of threshold values of the TDS and two subscale scores. Optimal thresholds were selected based on the maximal value of both sensitivity and specificity. Test-retest reliability at baseline and 2 weeks was evaluated using intraclass correlation coefficient (ICC). Results 183 children were enrolled, mean age 39.3 (SD 7.4) months, 46.4% male, 120 (66%) completed the P-SDQ at 2 weeks, 107 (58%) completed the PAPA at 12 weeks. Of those with a PAPA, 26 (24%) had any psychiatric diagnosis, 22 (21%) had an internalizing disorder, 9 (8%) had an externalizing disorder, and 5 (5%) had both. TDS identified any diagnosis with AUC = 0.67 (95% CI: 0.55, 0.79); the internalizing subscale identified internalizing disorders with AUC = 0.61 (95% CI: 0.47, 0.74); the externalizing subscale identified externalizing disorders with AUC = 0.77 (95% CI: 0.60, 0.94). Sensitivity and specificity for TDS were 50% and 78%; for internalizing subscale were 18% and 99%; for externalizing subscale were 78% and 71%. Test-retest reliability (ICC) for the TDS was 0.72 (95% CI: 0.62, 0.80); internalizing subscale was 0.62 (95% CI: 0.49, 0.71); externalizing subscale was 0.79 (95% CI: 0.72, 0.86). Conclusion The parent report P-SDQ externalizing subscale has sufficient accuracy and reliability for identifying externalizing disorders in primary care. The TDS and internalizing subscale perform less well. Table 3: Screening test properties of the P-SDQ compared with the PAPA (n=107)