The Extent and Importance of Unintended Consequences Related to Computerized Provider Order Entry

From Clinfowiki
Revision as of 22:12, 4 November 2007 by Shapiromr (Talk | contribs)

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

Ash JS, Sittig DF, Poon EG, Guappone K, Campbell E, Dykstra RH J Am Med Inform Assoc. 2007 Jul-Aug;14(4):415-23

Introduction: This article presents qualitative research reporting the range of unintended consequences of CPOE in a large group of health care organizations.

Disclosures: There are no explicit disclosures of conflicts of interest. The study received funding from the National Institutes of Health and the National Library of Medicine.

Background: The deployment of CPOE causes dramatic changes in workflow for all members of the health care system. Many changes are intended and can be planned for but there are a large number of unintended consequences which can emerge during and after implementation.

Question: A research question is not explicitly stated but the authors seek to determine the types, frequencies and impact of unintended consequences of CPOE systems.

Objective: The research goals are explicitly stated as "...to identify the types of unintended consequences, so that they can be monitored and managed…"

Methods: The authors conducted two activities in this study. First they directly observed the work flow of clinicians in five sites experienced in the use of CPOE for a total of 390 observer-hours. They then then aggregated their observations using qualitative research methodologies into 9 clusters of consequence types. They eliminated one, "persistant use of paper", as it was felt not to be an unintended consequence, but rather an alternative display modality. They then provided summary statistics of the frequency of the various unintended consequences.

Key Findings: The authors drew their respondents from an initial group of 561 hospitals listed in HIMSS Analytics as users of CPOE. CPOE had been in place from 6 months to 25 years with a median of 5 years. Their final group size was 176. There were 6 clusters of unintended consequences and associated frequencies of moderate to high importance (indicated in parentheses): New or More Work (72%), Workflow(88%), System Demands(82%), Communication(84%), Emotions(80%), Dependence on Technology(83%), and two with low ratings: New Kinds of Errors and Power Shifts. There was no correlation with responses and duration of use of CPOE.

Discussion: The authors note that unintended consequences are very frequent, often unanticipated, and can have a large effect on the acceptance and usability of CPOE. They suggest that those contemplating implementation of CPOE will benefit by either planning strategies to avoid these unintended consequences or at least plan, in an informed manner, for the trade offs that result. They further suggest that developers of CPOE systems carefully attend to these unintended consequences and make efforts to mitigate them in their development strategies.

Reviewer Comments: This article is a very valuable resource for health systems which are planning to implement CPOE. It offers the opportunity to normalize the disruptions and stressors of system deployment and mitigate the worst of the unintended consequences. The inclusion of 176 bona fide users of CPOE provides a very broad survey of the current CPOE environment, lending great credibility to the findings. This is a must read for the leaders of any health system that is planning to deploy CPOE.

Retrieved from "http://www.informatics-review.com/wiki/index.php/Ash_JS%2C_Sittig_DF%2C_Poon_EG%2C_Guappone_K%2C_Campbell_E%2C_Dykstra_RH_J_Am_Med_Inform_Assoc._2007_Jul-Aug%3B14%284%29:415-23"