Predictive value of the test of infant motor performance and the Hammersmith infant neurological examination for cerebral palsy in infants

Abstract

BackgroundCurrent recommendations for early detection tools for cerebral palsy (CP) include assessments that vary in feasibility and resource requirements. The predictive value of less resource-intensive tools has not been fully explored. AimsTo determine the predictive value of the Test of Infant Motor Performance (TIMP) at 3–4 months corrected age (CA) for CP, and whether administration of both the TIMP and the Hammersmith Infant Neurological Exam (HINE) improves early CP detection. Study designFive-year retrospective observational study of infants who received the TIMP and the HINE at 3–4 months CA in a high-risk follow-up clinic. TIMP and HINE cut-off scores (alone and in combination) were compared for CP discriminatory ability. SubjectsOf patients with HINE scores (n = 1389; 676 [48.7 %] female; median gestational age at birth 31 weeks [interquartile range 29–34 weeks]), 1343 had concurrent TIMP scores available. Outcome measuresClinical diagnosis of CP. ResultsHINE total score <57 had optimal CP predictive value (AUC = 0.815; 77 % sensitivity; 91 % specificity) compared to optimal TIMP cut-off (1 SD below the mean, AUC = 0.71; 52 % sensitivity; 94 % specificity) and all tested TIMP and HINE combinations (all p < 0.001). ConclusionsHINE total score <57 at 3–4 months CA had the best CP predictive value, confirming its value absent first-line detection tools. Concurrent administration of TIMP did not improve predictive value.