Evaluating a Training Model for the Use of Enhanced Diagnostic Screening Measures within Community Pediatric Practice

Background:  Early accurate diagnostic classification of young children with ASD is a critical public health issue. A growing body of research indicates that children who receive early ASD-specialized intervention services show significant gains in cognitive and adaptive functioning. While caregivers are able to report concerns to medical professionals by the age of 12 - 18 months and the diagnosis appears quite stable by 24 - 36 months, the average age of diagnosis is 4 years or older, representing a serious time lag. Since pediatricians and family physicians are often the first point of contact for families, they play an important role in early identification of ASD. The path from positive screening to appropriate diagnosis, however, is not well-established.

Objectives:  In the current report we describe results from an ongoing training program in Tennessee designed to help pediatricians diagnose young children with ASD in community settings within a time-limited framework.

Methods:  Six community health professionals (i.e., pediatricians, nurse practitioners) who had previously completed an office-based training program for introducing clinically-validated screening tools into primary care settings were invited to participate in the program. Phase 1 was a two-day evaluation workshop involving training on the STAT, MCHAT, and diagnostic interviewing. Phase 2 included evaluation practice and feedback, and Phase 3 involved independent evaluations and referrals of these cases to the university-based clinic for a separate evaluation.

Results:  13 cases were referred to the university-based clinic from this training. For the referred cases there was 84.6 % (11/13) agreement as to whether a child met criteria for an ASD diagnosis. Within the group flagged for ASD, there was 62.5 % (5/8) agreement on the specific ASD diagnosis made. For the two cases where there was disagreement on the presence of an ASD, the referral source was the same and there were no significant differences in calibrated severity, cognitive, or adaptive scores when compared to the cases in agreement.

Conclusions:  The current results support previous findings from this program, namely that training in standardized ASD assessments can help community practitioners accurately identify children with ASD within their local practice settings. Continued expansion of the model will need to systematically assess trainee characteristics associated with accuracy of diagnosis. We are currently evaluating the model's impact on families and analyzing feedback from participating professionals, in part to determine whether definitive diagnosis or risk designation is the most appropriate goal. Findings from this program are promising given the large impact even a small number of community medical practitioners could have in providing timely diagnostic services across the state and country.