A 2014 medical informatics perspective on clinical decision support systems: do we hit the ceiling of effectiveness

From Clinfowiki
Jump to: navigation, search


Bouaud, J and Lamy, J.B. recently published a paper in which they summarized over 1400 peer-reviewed articles associated with Clinical Decision Support Systems (CDSSs ) published in 2013 [1]. Two literature databases, namely Pubmed.Medline (NCBI) and Web of Science (Thomson Reuters), and the editors from each database site were employed for reviewing the literature respectively. The three articles which were the best cited and represented trends related to CDSSs in 2013 were selected based on the designated criteria.


Based on the three “best papers” related to CDSSs published in 2013, three major trends associated with CDSSs were identified as follows: 1. Generally speaking, CDSSs seldom demonstrated efficacy and effectiveness on clinical outcomes. The review results suggested that the critical factors for CDSS implementation have not been resolved yet and/or may be the usability ceiling was reached. 2. Secondly, the major focuses are to identify the obstacles of CDSSs implementation and knowledge formation. 3. The third trend is to expand new theoretical approaches, standards, terminologies, and technologies to develop new CDSSs which could overcome existing barriers related to CDSS implementation and usability.

The first trend represented by Gay, P. et. al, was focusing on theoretical works to expand CDSS capabilities as exemplified by providing a theoretical framework for integrating genetic information and knowledge, patient medical history, and family pedigree into CDSSs for cancer diagnosis [2]. However, this type of theoretical framework has not been tested whether or not it is effective in real-world application. More work needs to be done before it can be used for real application and play its role as expected. The second paper reflecting the trend related to reduce adverse drug effects (ADEs) by CDSS adoption. The authors systematically evaluated a classic CDSS termed ADE score card, which originally designed for reducing medical errors and potential ADEs [3]. Although this system was broadly accepted by healthcare professionals, however, ironically, it hardly demonstrated a significant impact on clinical outcomes by rigorous test using established criteria. As of today, it still remains to be answered whether or not such CDSSs such as ADE scorecard can improve clinical outcomes and promote patient safety and quality care. The third trend was related to the interactions among different stakeholders such as clinicians, staff, and patients [4]. An example provided for CDSSs in this aspect was e-Prescribing, which supposedly provided alerts for drug-drug, drug-allergy, and drug-food interactions. However, the authors observed that alerts could not be provided in a timely manner, very often “too late”. Therefore, they suggested that the alerts should be presented at appropriate time in order to provide useful information for healthcare providers. In addition, the acceptance of CDSSs between physicians and patients also affected effectiveness and efficacy of CDSSs and could potentially create barriers for CDSS implementation.


In summary, based on literature reports, CDSS is still an active research area for clinicians and health informaticians. However, with respect to the effectiveness and efficacy of CDSSs in healthcare applications, they seemed to come across bottlenecks in 2013. The authors finally suggested that paradigm shifts were needed “in the design, the development, or the implementation of CDSSs” in order to enhance their effectiveness and usability [1]. Furthermore, the following facts such as new CDSS decision models, commitment for the best integration of evidence based medicine, technology advance, including human-computer interface and display structure and technology, and better collaboration among stakeholders should be considered for enhancement of CDSS effectiveness and usability.


  1. 1.0 1.1 Bouaud, J. & Lamy, J.B. A 2014 medical informatics perspective on clinical decision support systems: do we hit the ceiling of effectiveness? Yearbook of Medical Informatics, 9 (1), 163-166. http://www.ncbi.nlm.nih.gov/pubmed/?term=A+2014+medical+informatics+perspective+on+clinical+decision++support+systems%3A+do+we+hit+the+ceiling+of+effectiveness%3F+Yearbook+of+Medical+Informatics
  2. Gay, P., Lopez, B., Pla, A., et al. (2013) Enabling the use of hereditary information from pedigree tools in medical knowledge-based systems. Journal of Biomedical Informatics, 46 (4), 710-720. http://www.ncbi.nlm.nih.gov/pubmed/?term=nabling+the+use+of+hereditary+information++from+pedigree+tools+in+medical+knowledge-based+systems.+Journal+of+Biomedical+Informatics
  3. Hackl, W.O., Ammenwerth, E., Marcilly, R., et al. (2013) Clinical evaluation of the ADE scorecards as a decision support tool for adverse drug event analysis and medication safety management. British Journal of Clinical Pharmacology, 76 Suppl 1 78-90. http://www.ncbi.nlm.nih.gov/pubmed/?term=Clinical+evaluation+of+the++ADE+scorecards+as+a+decision+support+tool+for+adverse+drug+event+analysis+and+medication+safety+management
  4. Hayward, J., Thomson, F., Milne, H., et al. (2013) 'Too much, too late': mixed methods multi-channel video recording study of computerized decision support systems and GP prescribing. Journal of American Medical Informatics Association, 20 (e1), e76-84. http://www.ncbi.nlm.nih.gov/pubmed/?term=Too+much%2C+too+late%27%3A+mixed+methods++multi-channel+video+recording+study+of+computerized+decision+support+systems+and+GP+prescribing