Early detection of infants with neurodevelopmental concerns indicative of cerebral palsy in a lower middle-income country (India).

Follow-up was not complete for all infants; only 176 of the screened cohort had 18-month outcome data. The sample was enriched for high-risk infants, so results do not apply to the general population. Diagnosis was made via video assessment, which may have missed subtle tone differences, potentially reducing reported accuracy. Specificity of the GMA was notably low (44.4%), leading to many false positives. The study was nested within an RCT, so infants identified as high-risk received early intervention, which could influence outcomes, though the paper states this does not change diagnosis. Community sites had lower rates of predictive birth risk factors, which may explain the variability in diagnostic accuracy compared to hospital sites.

Funded by the Cerebral Palsy Alliance Australia, QEII Diamond Jubilee Endeavour scholarship, Commonwealth Government Australia, Merchant Charitable Funding, and the Australian National Health and Medical Research Council. Authors declared no conflicts of interest.

Do not interpret the high sensitivity of the GMA as evidence that it is a standalone diagnostic tool. Its specificity was only 44.4%, meaning it incorrectly flagged more than half of the infants who did not have neuromotor concerns as high-risk. Use it as part of a screening pathway, not a definitive diagnosis.