Types of unintended consequences related to computerized provider order entry

From Clinfowiki
Revision as of 07:13, 21 February 2015 by Annathehybrid (Talk | contribs)

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

This is a review of Campbell, Sittig, Ash, Guappone, and Dykstra's 2006 article, Types of Unintended Consequences Related to Computerized Provider Order Entry.[1]

Review 1


Recently, many health care organizations have implemented computerized provider order entry (CPOE) in hopes of improving the quality and safety of health care and reducing the costs of providing medical care. While many advocate for the benefits of CPOE (i.e. Leapfrog Group) in this regard, there have been studies which have suggested that in fact it is not unusual for unexpected and negative effects to occur when CPOE is implemented.

The authors introduce the reader to Diffusion of Innovations (DOI) theory, which suggests that consequences from an innovation like CPOE can be anticipated or unanticipated (unintended), and desirable or undesirable (adverse). This paper describes a study which explored the unintended and adverse consequences of CPOE. The specific goal was to identify and describe the major types of these unanticipated adverse consequences (UACs).


Data was acquired by making site visits with direct observation of the daily provider activities at five hospitals within three different organizations; size, geography and academic affiliation of the hospitals varied. First, 79 UACs were identified by an expert panel. Categories of UACs were determined from this data. Then, 245 additional UACs that were identified during fieldwork were analyzed and categorized. The authors are careful to point out that “the categories emerged from the data, rather than from preconceived expectations.”


As a result of the data and 36 team meetings, nine major categories of UACs were identified. These were as follows (in order of decreasing frequency)

  1. More or New Work for Clinicians (like having to enter new information, respond to excessive alerts, complete complex orders)
  2. Unfavorable Work Flow Issues (such as duplications or contradictions among orders)
  3. Never-Ending Demands for System Changes (including demands for advanced hardware platforms and software application, 24 hour help desks, back-up systems)
  4. Problems Related to Paper Persistence (pointed out that while paper-based record storage should disappear, use of paper in the clinical setting probably will not)
  5. Untoward Changes in Communication Patterns and Practices (such as reduction in face-to-face communication, with possible increase in likelihood of errors as a result)
  6. Negative Emotions (like resentment in response to required change, though level of negative emotions noted to decrease over time)
  7. Generation of New Kinds of Errors (for example, juxtaposition errors)
  8. Unexpected and Unintended Changes in Institutional Power Structure (like physicians loss of professional autonomy, IT department becoming the new enforcer of standards, administration and QA gaining power by forcing compliance)
  9. Overdependence on Technology (for example, increasingly difficult for organization to work without it).

On a separate note, the researchers reported that (CPOE-based) clinical decision support (CDS) generated over 25% of the UACs.


The authors reported three major findings

  1. UACs occurred commonly, though not all institutions experienced all types of UACs
  2. the nine types of UACs occurred in a widespread fashion
  3. UACs pose significant consequences for providers, IT personnel and their organizations.

They also provide some further commentary on how workers in the field might be able to interpret these UACs and use this information in their future work.


This mostly qualitative study takes a large volume of data about unanticipated consequences (UACs) of CPOE, drawn from site visits, and uses an expert panel to devise a practical categorization scheme. The authors then take this scheme and identify lessons they believe are demonstrated within each type of unanticipated consequence. Ultimately, the authors drew a reasonable conclusion, which is that this “typology” should provide a framework for CPOE developers and implementers to improve their execution in the future. Indeed, software developers, vendors and hospital clinical information system leaders should be mindful of the details of this study as they move forward with their particular future responsibilities related to CPOE development and implementation.

Review 2

What are the types of unintended adverse consequences (UACs) that arise from computerized physician order entry (CPOE) implementation?


First, an expert panel established an initial set of UACs of CPOE, and an initial set of UAC categories was developed from these. Next, a large number of additional UACs were identified over a nine-month period through observations and/or interviews of staff at five diverse hospitals with successful CPOE systems. At each of these hospitals, the staff involved consisted of three groups of individuals: clinical end-users (such as physicians, nurses, pharmacists and unit secretaries), IT staff (the people who implement, configure, maintain and support CPOE systems), and administrative staff (the people who allocate the resources for, and established the policies, procedures and regulatory compliance of, the CPOE systems). The additional field-identified UACs were used to iteratively modify and refine the initial set of UAC categories.

Main Results

79 expert-panel UACs were initially established, and 245 additional field UACs were identified. Processing of all of these UACs ultimately resulted in the development of the following nine UAC categories:

  1. more/new work for clinicians
  2. unfavorable workflow issues
  3. never-ending system demands
  4. problems related to paper persistence
  5. unfavorable changes in patterns and practices of communication
  6. negative emotions
  7. the generation of new types of errors
  8. unexpected changes in power structure
  9. over-dependence on the technology

These nine categories of UACs occurred regularly at the five hospitals evaluated, but not all categories occurred at every hospital.


UACs occur during all CPOE implementations, and fall consistently into nine categories. Only through the careful evaluation of UACs in CPOE systems will it be possible to ameliorate the consequences of those UACs that are controllable.

Related articles

Also see Effects of CPOE on provider cognitive workload: a randomized crossover trial


  1. Campbell, E. M., Sittig, D. F., Ash, J. S., Guappone, K. P., & Dykstra, R. H. (2006). Types of unintended consequences related to computerized provider order entry doi: 10.1197/jamia.M2042. http://jamia.oxfordjournals.org.ezproxyhost.library.tmc.edu/content/13/5/547.long