Accuracy of Community General Pediatrician Diagnosis of Autism Spectrum Disorder (ASD) Compared to Multi-Disciplinary Team Assessment

Background: The prevalence of ASD has increased dramatically in recent years, placing a strain on access to diagnostic assessment. Early diagnosis is of significant importance to obtaining early intervention and optimizing outcomes. Encouraging community-based clinicians, such as general pediatricians (GPs), to make the diagnosis may increase diagnostic capacity and decrease wait times. If pediatricians are to be used as a diagnostic resource for ASD, it is important to evaluate whether they provide accurate diagnostic assessments compared to a multi-disciplinary team (MDT) assessment.

Objectives: The objective of this study is to determine agreement in ASD diagnosis between GPs and a subspecialist MDT assessment.

Methods: This study follows a prospective, cross-sectional design. There are two target populations for recruitment: 1) GPs in Ontario (projected n = 20), and 2) their patients under 5.5 years referred for developmental concerns (n = 200 total; 10/pediatrician). Participants are randomized 1:1 to have their MDT visits scheduled either before or after the pediatrician visit. MDT and general pediatrician assessments are conducted blinded to each other’s impressions. The MDT visits consist of clinical assessments by a developmental pediatrician and a psychologist. The MDT and GP independently record a forced decision on whether the child has ASD, and the GP indicates the actions they would have taken had the child not been enrolled in the study. The MDT team provides feedback on their assessment results to the family. Agreement between the two assessments is measured with a simple kappa statistic. Descriptive data around the breakdown of true and false positive and negative cases is provided.

Results: To date, 36 assessments have been completed (24 males, 12 females; mean age 40 months, range 17-63). Thirty of the children had an ASD diagnosis as determined by the MDT. GPs agreed with the MDT team on 66.7% of cases (24/36 cases), corresponding to a kappa of 0.2 (poor agreement). In 20 true positive cases, GPs indicated that they would have given the ASD diagnosis in 12 cases, referred to a specialist in 7 cases, and ‘watched and waited’ in 1 case. In 2 false positive cases, GPs indicated they would have referred to a specialist. In 4 true negative cases, the GPs indicated they would have given the non-ASD diagnosis in 3 cases, and referred to a specialist in 1 case. In 10 false negative cases, the GP would have incorrectly give a non-ASD diagnosis in 4 cases, ‘watched and waited’ in 2 cases, and referred to a specialist in 4 cases.

Conclusions: These results are preliminary but are illustrative of the importance of this work. GP participants are willing to diagnose ASD in their practices. The incidence of several false negative cases, some of which may have been missed diagnoses as the pediatricians would not have referred for further assessment, suggests that further GP education is needed on the varied clinical presentations of ASD. Further analysis with a larger sample size will permit identification of child features associated with ‘accurate’ ASD diagnostic determination by GPs.