Awalin Sopan, Catherine Plaisant, Seth Powsner, Ben Shneiderman Human-Computer Interaction Lab & Department of Computer Science, University of Maryland.

User Interface Techniques to Reduce Wrong Patient Errors

Awalin Sopan, Catherine Plaisant,Seth Powsner, Ben Shneiderman

Human-Computer Interaction Lab & Department of Computer Science,

University of Maryland

A Tale of Two Patients

http://www.nytimes.com/2002/06/18/health/oops-wrong-patient-journal-takes-on-medical-mistakes.html

A Tale of Two Patients

Mrs. Morris, 67 Mrs. Morrison, 77

They were in same hospital floor.

Mrs. Morris was taken to the operation room for the heart

surgery

Problems

• A drug administered to wrong patient

• Reading of wrong patients’ test results

• Patients miss needed treatment

• etc.

Is Computerized Patient Order System a Panacea for

These Problems?

interruption

fatigueurgenc

y

multitasking

long work-hours

Our ApproachError Classification

what are the error-scenarios clinicians face

Task Analysis which stage is more susceptible to a

particular type of error

27 Specific Techniques what to do, and then how to do it

Error Classification

Mistake

Slips

Failure to recogniz

e

Error Classification

Mistake

Slips

Failure to recogniz

e

Recalling the wrong patient due to short term memory failure, name similarity, unfamiliarity with the patient, fatigue.

Error Classification

Mistake

Slips

Failure to recogniz

e

Mechanical errors such as wrong key press, mouse slip, or errors due to unreadable fonts and too small button size.

Error Classification

Mistake

Slips

Failure to recogniz

e

Failures to detect errors due to interruptions, multitasking, absence of relevant information.

Task Analysis

Task Analysis

Task Analysis

Task Analysis

Task Analysis

UI Techniques:Reduce Mistakes Facilitate recall:▪ Provide more context: room number, photo,…

Avoid confusion:▪Emphasize the salient features: age, chief complaint,…

▪Use at least two sources of identification: name, medical record number,…

Facilitate Recall

Poor recall strategy, more mistakes

Other Techniques

Allow sorting Always show patient’s full name Scan RFID to retrieve the patient Use indoor location to retrieve the

patients

UI Techniques:Reduce Slips

Improve target-selection Improve text-readability Highlight target under cursor

Facilitate Selection

Poor selection mechanism, more slips

Highlight row under cursor Use an icon-based 2D grid instead of

list

Other Techniques

UI Techniques:Increase Recognition

Draw attention to patient information▪ Taieb-Maimon et al. : recognition increased from 7% to 43% with photo

Use decision support system

Increase RecognitionPoor verification, less error recognition

During Verification

Use visual summary of patient history

Avoid visual distraction Re-enter ID

Other Techniques

During Confirmation

What Guided Us

▪Human Error Classification▪Attention Theory▪Context Recovery Process▪Cognitive Task Analysis▪User Interface Design Principles▪Expert Feedback▪Medical Literature

Contributions

Categorization of the error-types, and sources

Suggestions of User Interface remedies

Prototype demonstrating the techniques

Take-away Messages

Small changes in the UI can make big difference in patient safety

Include Clinicians and HCI researchers in the design process

To err is human, the systems should make up for it

www.cs.umd.edu/hcil/WPE

www.youtube.com/watch?v=CrwOJIrnsg8

Awalin Sopan, Catherine Plaisant, Seth Powsner, Ben Shneiderman

@[email protected]

We thank the Patient-Centered Cognitive Support under the Strategic Health IT Advanced Research Projects Program (SHARP) from the Office of the National Coordinator for Health Information Technology (Grant No. 10510592).