The Structured Observation of Motor Performance in Infants can detect cerebral palsy early in neonatal intensive care recipients

Abstract

BackgroundThe detection of motor problems in infancy requires a detailed assessment method that measures both the infants' level of motor development and movement quality. AimsTo evaluate the ability of the Structured Observation of Motor Performance in Infants (SOMP-I) to detect cerebral palsy (CP) in neonatal intensive care recipients. Study designProspective cohort study analyzed retrospectively. Subjects212 (girls: 96) neonatal intensive care recipients (mean gestational age 34weeks, range: 23–43). Twenty infants were diagnosed with CP. Outcome measuresThe infants were assessed using SOMP-I at 2, 4, 6 and 10months' corrected age. Accuracy measures were calculated for level of motor development, quality of motor performance and a combination of the two to detect CP at single and repeated assessments. ResultsAt 2months, 17 of 20 infants with CP were detected, giving a sensitivity of 85% (95% CI 62–97%) and a specificity of 48% (95% CI 40–55%), while the negative likelihood ratio was 0.3 (95% CI 0.1–0.9) and the positive likelihood ratio was 1.6 (95% CI 1.3–2.0). At 6months all infants with CP were detected using SOMP-I, and all infants had repeatedly been assessed outside the cut-offs. Specificity was generally lower for all assessment ages, however, for repeated assessments sensitivity reached 90% (95% CI 68–99%) and specificity 85% (95% CI 79–90%). ConclusionsSOMP-I is sensitive for detecting CP early, but using the chosen cut-off can lead to false positives for CP. Assessing level and quality in combination and at repeated assessments improved predictive ability.