Features predicting the success of computerized decision support for prescribing: a systematic review of randomized controlled trials.

From Clinfowiki
Revision as of 06:16, 21 November 2015 by Hsong (Talk | contribs)

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search


Clinical Decision Support (CDSS) have the potential to enhance the quality of healthcare with the main goal of providing high quality services that can improve patients' outcomes. Resources used by clinicians should always be evidence-based medicine in order to provide transformational health care services. A systematic review was conducted to evaluate features predicting successful implementation of CDSS, change in clinicians behavior, and improvement in patient outcomes.


The databases Medline, EMBASE, CINAHL, and INSPEC were searched for articles published since the earliest entry to June 2008; Titles were reviewed and researchers kept those matching their definition of CDSS. In order to determine if studies met the inclusion criteria, abstracts were assess independently by different reviewers. Three different domains of focus were assessed to measure study outcomes: Implementation, change in clinician behavior and change in patient outcomes. Researchers considered an implementation a successful one if the CDSS was successfully introduced and implemented by the clinical staff. To consider a change in clinicians behavior it was necessary to have a reduction in inappropriate prescribing or implement a more cost-effective therapy. Finally, to consider a successful improvement in patient outcomes researchers just took in consideration those studies that reported improvements in patients' health such as decreases in mortality and morbidity. [1]


The selected 41 RCTs involved a total of 612,556 patients and 2963 providers. According to results systems were successfully implemented in 37 trials (provider behavior changed in 25 and patient outcomes improved just in 5). The most common features predicting success across studies were: support of the user's task at hand, access to decision support systems at the time and place of decision-making, automatic provision of decision support as part of clinician workflow, integration with CPOE and adequate locations for the computers. Regarding patient outcomes, features that all successful studies shared and the majority of unsuccessful studies did not, were: provision of recommendations (rather than just an assessment) and justification of decision support via provision of research evidence.


According to results CDSS have the potential to change clinicians' behavior in greater extent; however, it generates concern to learn that just few studies show improvement in patient outcomes. This is not the first study that reports a need to further research for features that can potentially improve patient outcomes.


There are unmet needs for improving effectiveness and usability with assisted by CDS system due to at least the following reasons:

  1. Implementation and integration issues.
  2. Technical design and display.
  3. No sufficient focus on dealing with human factors related medical errors.
  4. Newly introduced unintended consequences.
  5. Insufficient and inappropriate training to end users.

Related Articles


  1. Mollon, B., Chong, J., Holbrook, A. M., Sung, M., Thabane, L., & Foster, G. (2009). Features predicting the success of computerized decision support for prescribing: a systematic review of randomized controlled trials. BMC Medical Informatics and Decision Making, 9, 11. http://ca3cx5qj7w.search.serialssolutions.com/OpenURL_local?sid=Entrez:PubMed&id=pmid:19210782