Reliability and predictive validity of the Standardized Infant NeuroDevelopmental Assessment neurological scale

Aim To assess reliability and predictive validity of the neurological scale of the Standardized Infant NeuroDevelopmental Assessment (SINDA), a recently developed assessment for infants aged 6 weeks to 12 months. Method To assess reliability, three assessors independently rated video-recorded neurological assessments of 24 infants twice. Item difficulty and discrimination were determined. To evaluate predictive validity, 181 infants (median gestational age 30wks [range 22-41wks]; 92 males, 89 females) attending a non-academic outpatient clinic were assessed with SINDA's neurological scale (28 dichotomized items). Atypical neurodevelopmental outcome at 24 months or older corrected age implied a Bayley Mental Developmental Index or Psychomotor Developmental Index lower than 70 or a diagnosis of cerebral palsy (CP). Predictive values were calculated from SINDA (2-12mo corrected age, median 3mo) and typical versus atypical outcome. Results Intraclass correlation coefficients of intrarater and interrater agreement of the neurological score varied between 0.923 and 0.965. Item difficulty and discrimination were satisfactory. At 24 months or older, 56 children (31%) had an atypical outcome (29 had CP). Atypical neurological scores (below 25th centile, <= 21) predicted atypical outcome and CP with sensitivities of 89% and 100%, and specificities of 94% and 81% respectively. Interpretation SINDA's neurological scale is reliable and in a non-academic outpatient setting has a satisfactory predictive validity for atypical developmental outcome, including CP, at 24 months or older.