Accurate data collection for head injury monitoring studies: a data validation methodology

Abstract

Background: BrainIT is a multi centre, European project, to collect high quality continuous data from severely head injured patients using a previously defined [6] core data set. This includes minute-by-minute physiological data and simultaneous treatment and management information. It is crucial that the data is correctly collected and validated.

Methods: Minute-by-minute physiological monitoring data is collected from the bedside monitors. Demographic and clinical information, intensive care management and secondary insult management data, are collected using a handheld computer. Data is transferred from the handheld device to a local computer where it is reviewed and anonymised before being sent electronically, with the physiological data, to the central database in Glasgow. Automated computer tools highlight missing or ambiguous data. A request is then sent to the contributing centre where the data is amended and returned to Glasgow. Of the required data elements 20% are randomly selected for validation against original documentation along with the actual number of specific episodic events during a known period. This will determine accuracy and the percentage of missing data for each record.

Conclusion: Advances in patient care require an improved evidence base. For accurate, consistent and repeatable data collection, robust mechanisms are required which should enhance the reliability of clinical trials, assessment of management protocols and equipment evaluations.