BACKGROUND
Clinical decision support systems (CDSS) generate excessive alerts that lead to alert fatigue and override. Alert overrides have resulted in patient death and medical errors. CDSS with its alert function can also disrupt clinical workflow. Therefore, inefficient clinical processes that contribute to the misfit between CDSS alert and workflows must be evaluated. Evaluation findings can serve as input to the process redesign. Redesigning clinical processes can enhance CDSS alert appropriateness and subsequently improve patient safety.
OBJECTIVE
The paper presents a proposed framework for evaluating CDSS appropriateness in supporting clinical workflow. The paper also discusses the preliminary results of the framework validation.
METHODS
A subjectivist, qualitative case study evaluation was conducted at a 620-bed public teaching hospital using semi-structured interview, observation, and document analysis methods to investigate the features and functions of alert appropriateness and workflow related issues. The current state map for medication prescription process was also modelled to identify problems pertinent to alert appropriateness.
RESULTS
The main findings showed that CDSS is not well designed to fit into clinical workflow due to several influencing factors including technology (system design and implementation), human (information analysis and acquisition), organization (clinical tasks, organizational policies & procedures) and process (process analysis, redesign, implementation, monitoring and improvement) impeding the use of CDSS with its alert function.
CONCLUSIONS
CDSS alerts should be integrated into clinical workflows due to their potentials in enhancing patient safety. Process improvement methods such as Lean can be used to enhance the appropriateness of CDSS alerts by identifying inefficient clinical processes that impede the fit of CDSS alerts with clinical workflow. The validated framework can be used to address alert and workflow related problems in any healthcare setting.