EARLY DIAGNOSIS OF CEREBRAL PALSY: MODERN METHODS OF NEURODEVELOPMENTAL ASSESSMENT AND A CLINICAL IDENTIFICATION ALGORITHM

Abstract

Cerebral palsy (CP) requires timely diagnosis to ensure early intervention; however, in clinical practice, the diagnosis is often established late. This study aimed to evaluate the diagnostic accuracy of standardized neurodevelopmental assessment methods (GMA, HINE, and clinical “red flags”) for the early identification of CP and other neurodevelopmental disorders (NDDs) in high-risk infants and to develop a clinical care pathway algorithm.A prospective cohort study of diagnostic accuracy was conducted at a specialized tertiary-level center in Tashkent. A total of 118 infants with perinatal risk factors (prematurity, low birth weight, and hypoxic-ischemic encephalopathy [HIE]) were assessed at a corrected age of 4–24 months. The index tests included the General Movements Assessment (GMA), the Hammersmith Infant Neurological Examination (HINE), and the identification of clinical “red flags” indicating suspected CP (sCP) or suspected NDDs (sNDDs). The reference standard was a consensus clinical diagnosis of CP/NDD established at 12–24 months of age. Sensitivity, specificity, predictive values, and the area under the receiver operating characteristic curve (AUC) were calculated.For predicting CP, the clinical assessment of “red flags” (sCP) demonstrated a sensitivity of 92% (95% CI, 63–99%) and a specificity of 84% (95% CI, 76–90%). The absence of fidgety movements on GMA yielded a sensitivity of 76% and a specificity of 82%. The HINE demonstrated good discriminatory performance, with an AUC of 0.88 (95% CI, 0.79–0.97). For predicting moderate-to-severe CP, the sensitivity of both sCP assessment and GMA reached 100%. However, the predictive performance of these methods for cognitive impairments was limited (HINE AUC = 0.62). The combined use of GMA, HINE, and clinical markers provides high diagnostic accuracy for the early identification of CP, particularly its more severe forms. Nevertheless, given the limited ability of these methods to predict cognitive and sensory impairments, children at risk require prolonged, multidisciplinary developmental follow-up. The proposed seven-stage algorithm may optimize patient referral and clinical management pathways in routine clinical practice.