High practice variability in cerebral palsy diagnosis: Need for clarification of the consensus definitions?.

Abstract

Background and Objective(s): Cerebral palsy (CP) is the most common cause of motor disability in childhood. The increasing emergence of genetic etiologies for the CP phenotype combined with the variable involvement of different medical disciplines in CP diagnosis can promote practice variability. This diagnostic practice variability has broad implications for patient understanding of their symptoms and access to care. Our objectives were to determine whether CP diagnostic practice variability exists and, if so, to determine what factors might contribute to CP diagnostic practice variability. Study Design: Cross-sectional survey. Study Participants & Setting: We surveyed physician views on CP diagnosis via an online link distributed by email between 1st September and 15th December 2019. Surveyed physicians practiced in the United States or Canada and demonstrated an interest in CP or related disorders as evidenced by membership in the American Academy of Cerebral Palsy and Developmental Medicine or the Child Neurology Society Neonatal Neurology, Movement Disorders, or Neurodevelopmental Disabilities Special Interest Groups. Materials/Methods: The survey measure was developed with iterative input from physicians and non-physician parents of children with CP who are members of the Cerebral Palsy Research Network. It included questions on physician training and experience, the 2007 international consensus definition of CP (Rosenbaum et al. 2007), and four hypothetical case scenarios for which respondents were asked whether they would or would not diagnose CP. Logistic regression analysis was used to assess which factors contributed to the odds of diagnosing CP in each case scenario. Result(s): Of 695 contacted physicians, 330 physicians (47%) completed the survey. Two case scenarios yielded diagnostic consensus: (1) non-progressive spastic diplegia following premature birth with MRI findings consistent with periventricular leukomalacia (96% would diagnose CP), and (2) progressive spastic diplegia (92% would not diagnose CP). Scenarios featuring genetic etiologies for non-progressive motor disability yielded diagnostic variability (67% would diagnose CP in the setting of spastic diplegia and 46% would diagnose CP in the setting of generalized hypotonia). Adjusting for training and experience factors, neurologists were more likely than other specialties to diagnose CP in the setting of non-progressive generalized hypotonia with a genetic etiology. Conclusions/Significance: There is diagnostic variability in whether a child with a non-progressive motor disability due to generalized hypotonia or due to a genetic etiology will be diagnosed with CP. This diagnostic variability occurred despite provision of the 2007 international consensus definition of CP. To reduce diagnostic variability, it may be useful to emphasize or clarify key aspects of this consensus definition.