Administrative databases to monitor the prevalence of cerebral palsy

Abstract

Cerebral palsy (CP) is the most common cause of motor impairment in children. Monitoring its prevalence is essential to improve the understanding of its determinants and to assess the impact of advances in obstetric and neonatal care. Numerous registers of CP have been established worldwide for such purposes. In South Carolina, where there is no such register, Li et al. used three administrative databases to analyse the trend in prevalence of CP. The use of administrative databases in health care is likely to gain more and more attention given the fact that generally they are population-based and store a lot of information. However, there is a need to validate such sources of information when using them to estimate prevalence rate. To provide accurate measures of CP prevalence, researchers have developed CP registers or CP surveillance programmes. The aim of a CP register or surveillance programme is to identify all cases of children with CP within a defined geographical area, regardless of the particular need or severity of the disorder. Population denominators used to calculate prevalence rates must refer to the population from which the cases are identified. To ensure completeness of ascertainment, a register must collect data from several sources. Professionals working in registers rely on a clear definition of the disorder and include the children using specified inclusion/exclusion criteria. To allow comparative studies between different areas, collaborations between CP registers or surveillance programmes have been formed in the USA (Autism and Developmental Disabilities Monitoring Network), Europe (Surveillance of Cerebral Palsy in Europe), and Australia (Australia Cerebral Palsy Register). The first achievement of these networks was the harmonization of definitions and inclusion/exclusion criteria – the prerequisite for reliable data. Twenty years ago, Johnson and King concluded that it was not possible to monitor trends in the prevalence rate of disabling disorders using the child health information systems in the UK. Indeed, in administrative databases there is no clear inclusion/exclusion criteria and the professionals coding the disorder may not be skilled enough in the domain of motor impairments. Also, in such databases there is a failure to update the records of children when a diagnosis is considered erroneous. In the present study of Li et al., such false positive cases are likely as the authors did not exclude cases with a diagnosis of progressive neurological disorder (a condition that is excluded in most CP registers). Additionally, the authors included all cases between 0 years and 4 years old, although the minimal age for confirmation of diagnosis is recommended to be 4 to 5 years old. The lack of an independent data source to verify or exclude the cases found with CP is also a limitation. In a more recent study, Hollung et al. compared the cases registered in the CP Register of Norway and in the Norwegian Patient Register. They found more cases in the Norwegian Patient Register, but with around 10% to 15% of false positive. The authors concluded that multiple sources and critical review of single cases were needed to obtain reliable figures. The study of Li et al. can serve as a basis for further exploration of administrative databases in the epidemiology of CP. Indeed, the use of administrative database is costeffective and may represent a valuable tool if proved to be reliable.