Practical aspects of conducting a prospective statewide incidence study: the incidence of serious inflicted traumatic brain injury in North Carolina.

The article describes practical problems encountered in setting up and maintaining an active statewide surveillance system for a low-frequency but high-impact injury, inflicted traumatic brain injury (inflicted TBI). A system was designed to identify prospectively all children aged <2 years with a traumatic brain injury (TBI) admitted to any of the nine pediatric intensive care units (ICUs) in North Carolina in 2000 and 2001. Children who died prior to admission to hospital were identified from the records of the Office of the Chief Medical Examiner. The study design had strengths and weaknesses for ongoing surveillance. Strengths included a clear definition of a case, mechanisms to jury undecided cases, and a high level of quality control. Difficulties included appropriately addressing investigators' ethical and legal obligations in the study of child abuse, differing requirements of multiple institutional review boards (IRBs), and the time-intensive nature of the system. The described surveillance system produced high-quality data, but may not be practical for ongoing multi-year injury surveillance. Incorporation of inflicted TBI into an existing surveillance system, such as a trauma database, may be a feasible solution to many of the problems encountered.