Universal versus Targeted Q-CHAT-10 Autism Screening: Economic Evaluation in Chile and England.

Abstract Background: Autism screening may increase early diagnostic identification but can also generate false-positive referrals and specialist workload. Population predictions require explicit integration of autism prevalence with screening sensitivity and specificity. Objectives: To compare the true diagnostic yield, specialist assessment burden, false-positive assessments, and health-system costs of Universal versus developmentally Targeted Q-CHAT-10 screening in Chile and England. Methods: We constructed prevalence-anchored decision-analytic cost-consequence models for hypothetical cohorts of 100,000 children in Chile and England. Underlying autism status, Q-CHAT-10 accuracy, developmental-gateway performance, routine-review uptake, and diagnostic completion determined true diagnoses, false-positive specialist assessments, and health-system costs. The horizon extended from routine review to diagnostic disposition, using 2023 local costs. For Chile, gateway sensitivity for ASD was structurally unidentified and evaluated through equal-yield thresholds and prespecified scenarios. Parameter uncertainty was assessed in 10,000 Monte Carlo simulations conditional on specified gateway sensitivities. Diagnostic identification was treated as an intermediate outcome. Results: In Chile, Targeted screening matched Universal diagnostic yield at gateway sensitivity of 91.4% under reference follow-up, falling to 77.5% under alternative completion assumptions. With equal diagnostic completion, no feasible gateway-sensitivity threshold existed. In England, Universal versus Targeted screening produced 1,067.3 versus 946.0 true diagnoses, 8,268.7 versus 2,172.1 specialist assessments, and 7,201.4 versus 1,226.1 false-positive assessments per 100,000 eligible children. Universal screening cost an additional £64.75 per eligible child for 121.3 additional true diagnoses per 100,000. Derived positive predictive values were 12.9% and 43.6%, respectively. Conclusions: Prevalence-anchored evaluation reveals screening decisions as trade-offs between case identification and specialist workload. Universal screening increased modeled diagnoses in England but required substantially greater diagnostic capacity. In Chile, the relative diagnostic yield of Targeted screening depended on high developmental-gateway sensitivity. Local gateway performance, diagnostic completion, and capacity should inform screening design. Keywords: autism spectrum disorder, screening, early identification, decision analysis, health economics, diagnostic capacity, Q-CHAT-10