21847
Can Primary Health Professionals' Input to Parent Reports Improve the Ability to Detect Children with ASD?

Thursday, May 12, 2016: 5:30 PM-7:00 PM
Hall A (Baltimore Convention Center)
K. Ebishima, A. Stickley and Y. Kamio, National Center of Neurology and Psychiatry, Japan, National Institute of Mental Health, Tokyo 187-8553, Japan
Background:  Prior studies have demonstrated the effectiveness of detecting autism spectrum disorder (ASD) using parent-report questionnaires such as the M-CHAT at community-based health check-ups at age 18-24 months. However, one of the concerns about ASD screening based on parental reports is a high false negative rate. 

Objectives:  The aim of this study was to examine whether primary health professionals’ input to parent reports would improve the ability to detect children who are later diagnosed with ASD.

Methods:  Our sample comprised 1220 children who received a 24-month health check-up that included the Japanese version of the M-CHAT (M-CHAT-JV) at a health center in Tokyo in 2008-2010. The M-CHAT-JV rating was threefold: “parent report only (M only)”, “parent report under public health nurse instruction (M+PHNi)”, and “parent report combined with PHN observation (M+PHNo)”. Diagnostic evaluation was done according to DSM-IV-TR together with ADI-R, ADOS, CARS, SRS and other developmental assessments repeatedly during age 2-5.

Results:  Seventeen children were diagnosed with ASD. The number of screen-positives was 58, 34, and 55 using M only, M+PHNi, and M+PHNo, respectively. In addition to 13 children with ASD identified by M only, 2 children were newly identified by M+PHNi and 4 children by M+PHNo. A comparison between M only screen-positives (N=13) and screen-negatives (N=4) showed no significant differences in autism measurements at age 2, while M only screen-negatives scored significantly higher on the ADI-R RRB domain at age 3. M only screen-negatives tended to have higher DQs.

Conclusions:  This study suggests that face-to-face instruction by public health nurses may decrease screen-positives and detect more children who were not identified by M only, while their quick behavioral observation may also help identify a greater number of children with ASD without decreasing screen-positives. Optimal ASD screening should be chosen considering a community’s resources and needs.