Difference between revisions of "Systematic review of clinical decision support interventions with potential for inpatient cost reduction"

From Clinfowiki
Jump to: navigation, search
Line 47: Line 47:
 
== Results ==
 
== Results ==
  
Findings from both studies raised issues with the amount and organization of information in the display, interference with workflow patterns of primary care physicians, and the availability of visual cues and feedback. These findings were then used to recommend user interface design changes.
+
After an extensive search, a total of 7,663 articles and 78 manuscripts were reviewed.  The types of studies were broken down into before-after studies with control groups which had 78.2%. and randomized controlled studies which encompassed 15.4%.  The largest studied area by far was pharmacotherapy which had 53.8% of the studies  70.5% of the studies resulted in statistically and clinically significant improvements in an explicit financial measure or a proxy financial measure. However, of all the studies, data on cost effectiveness was available for only one study.
  
=== Main results ===
 
  
The two qualitative studies showed that there were a lot of consistencies in issues with usability of LMR.  Deficiencies were identified specifically with regard to the following aspects:
 
 
'''Navigation'''
 
 
Both studies qualified the navigation aspect as "awkward" and subjectively used too many clicks to get data entered or retrieved.  Too many popup menus were offered which crowded the screen.  Physicians created workarounds by opening up multiple browsers which was not ideal as it was time-consuming and consumed the system's resources, slowing down the computer. 
 
 
* '''Information Design''' The presentation of the screens created issues.  The Results Manager’s usage of color and low contrast with data objects that are in their “selected” state made it difficult to read or identify the information quickly.  In addition, there was a poor balance of displaying what the provider needed with what was available in "one click away" from the current screen.
 
* '''Customization''' Comments regarding customization were targeted primarily toward the letter-writing feature in Results Manager. Many physicians often used their own letters and found the pre-defined letter templates of Results Manager to be inadequate for all their workflow needs.
 
* '''Workflow''' The participants came from a variety of workflow backgrounds.  Some blocked off time at the end of the day to enter notes, while others entered at the end of each patient visit.  The biggest complaint from workflow again came from navigational issues--specifically the popup menus which slowed productivity.
 
  
 
== Conclusion ==
 
== Conclusion ==

Revision as of 22:17, 2 March 2015

This is a review for Christopher L Fillmore, Bruce E. Bray and Kensaku Kawamoto's systematic review of other completed studies to assess if clinical decision support CDS interventions cut inpatient costs. [1]

Research question

In the inpatient setting, has there been a conclusion of whether CDS cut the cost of healthcare and if so by what average amount and in which specific area?

Methods

MEDLINE was searched up through July 2013 to find the studies which deal with the economic savings of CDS intervention. These studies were identified through titles and abstracts, and subsequently full text articles were accessed and reviewed to make a final determination on inclusion. Relevant details of the studies were extracted and summarized.

Environment

MEDLINE was used in searching from the beginning of the project to July 2013. The following search terms were used in combination:

  • Clinical decision support systems
  • Decision-making
  • Computer-assisted
  • Computerized decision support
  • Reminder systems
  • Guideline adherence
  • Medical informatics

All researched articles dealt with human beings and were written in the English language.

Design

Data was gathered by a single reviewer in order to prevent multiple biases using a standard form. If there was uncertainty any uncertainty, it was resolved through consensus. Each article was examined followed the exact same steps in the following order:

1. Evaluation for inclusion criteria 2. Data extraction 3. Analysis of trial design 4. Analysis of intervention 5. Analysis of trial results

The inpatient setting was defined by the wards, ER, ICU and surgical settings.

Measurement

The design of the study was such that data gathered were analyzed and summarized in a table. Themes, trends and patterns were identified and looked to see if identifiable patterns were evolving. A Fisher's exact test of independence was used. The independent variable was a statistically and clinically significant improvement in cost (reduction in cost). A p-value of < 0.05 was considered statistically significant.


Results

After an extensive search, a total of 7,663 articles and 78 manuscripts were reviewed. The types of studies were broken down into before-after studies with control groups which had 78.2%. and randomized controlled studies which encompassed 15.4%. The largest studied area by far was pharmacotherapy which had 53.8% of the studies 70.5% of the studies resulted in statistically and clinically significant improvements in an explicit financial measure or a proxy financial measure. However, of all the studies, data on cost effectiveness was available for only one study.


Conclusion

Through a thorough examination of two studies, it can be shown that qualitative research can help focus attention on user tasks and goals and identify patterns of care. Findings from both studies found consistency with regards to issues with the organization of information in the display, interference with workflow patterns of primary care physicians, and the availability of visual cues.

Commentary

Usability engineering can play a valuable role in assisting product design teams, unfortunately it has not been a routine part of designing clinical computing systems. Upon interviewing many of the test subjects who evaluated the EMR system, some identified system speed to be the primary determinant of user satisfaction, but most felt that usability principles and not speed or technology alone was necessary for the success of the EMR. Although usability is a broad term, it can be narrowed to two groups via various theories of human cognition and visual sensory perception.

First, it relates to navigation and system content. This is where efficiency of the EMR comes into play in defining usability. When efficiency slows down, the physician blames system speed, however technology is not the issue, rather often its the myriad of details and popup menus which create an overload in visual sensation. The fact that often the users created workarounds by opening multiple browsers reinforced this conclusion. Navigation and system content needs to be designed around user workflow--not a one size fits all. Different practice styles from various specialties and personalities come into play.re is

Second, information design is an important aspect of usability. Cognitive load theory defines the amount of "mental energy" needed to process the information or task in front of the user. There is a direct relationship between cognitive load and the amount of information present. Visualization also affects cognitive load as when too many screen elements come close together, the processing of information becomes slowed. Resolution can decrease efficiency if contrast or colors are not ideal and can mask the visual hierarchy.

Related papers


References

  1. . Systematic review of clinical decision support interventions with potential for inpatient cost reduction. BMC Medical Informatics and Decision Making 2013, 13:135. http://www.biomedcentral.com/1472-6947/13/135