Monitoring clinical performance: the role of software architecture.

Methods of assessing and monitoring the performance of clinicians have received a lot of publicity in recent years. We review the main methodologies concentrating on the distinction between monitoring individual performance and monitoring aggregated performance. We also highlight the importance and difficulties associated with incorporating and assessing risk factors into the process. We discuss how software architecture can be developed to implement these methodologies. We illustrate this development by a case study involving the creation of a software tool to produce funnel plots to analyse surgeon performance. We discuss how such tools are currently evaluated and propose that in future assessments of usability would benefit from an experimental study.

A software system for clinical monitoring.

Three graphical methods for monitoring clinician performance - the Funnel Plot, the CUSUM and the Variable Life Adjusted Display - are described. In addition the problems associated with incorporating and assessing risk factors into the process are discussed. The software system produced to implement these methodologies is presented; the creation of the user friendly interfaces, the underling algorithms and the structure and connectivity of the database are described. The methodology for piloting and evaluating the system is described. Drawing on developments in experimental usability we propose an experimental procedure aimed at evaluating possible changes in design so as to maximise the use of the system in terms of its accuracy of input and clarity of output.