AUC for prediction of domain-specific neurocognitive impairment (>1.5 SD below the normative mean) using clinical characteristics and caregiver signaling questions. Performance of clinical characteristics and caregiver signaling questions as predictors of domain-specific neurocognitive impairment (>1.5 SD below the normative mean). The x-axes and y-axes indicate specificity (true negative rate) and sensitivity (true positive rate), respectively. Three models are presented in each individual figure. Model 1 (Clinical) included only clinical characteristics as predictors (patient age at the time of the evaluation, genotype, use of hydroxyurea before evaluation, social vulnerability). Model 2 (CRS) included only an aggregate CRS for a child’s history of grade retention, history of learning difficulties, or poor grades (Cs, Ds, Fs, variable). Model 3 (Clinical + CRS) included clinical characteristics and an aggregate CRS for a child’s history of grade retention, history of learning difficulties, or poor grades (Cs, Ds, Fs, variable). (A) Verbal comprehension index (VCI) measured with Wechsler Abbreviated Scale of Intelligence, Second Edition (WASI-II). (B) Coding subtest measured with Wechsler Adult Intelligence Scale, Fourth Edition (WAIS-IV) or Wechsler Intelligence Scale for Children, Fourth or Fifth Editions (WISC-IV/V). (C) Letter-Word Identification subtest measured with Woodcock-Johnson Tests of Achievement, Third or Fourth Editions (WJ-III/IV). (D) Perceptual reasoning index (PRI) measured with WASI-II. (E) Digit span subtest measured with WAIS-IV or WISC-IV/V. (F) Math fluency subtest measured with WJ-III/IV.