Friday, September 18, 2009

#med2 - usability and human factors on the web

Live Blog from Medicine 2.0

Emily Seto, U of Toronto

Heart failure is a massive problem. 5M people in USA. Expensive

1/3 to 1/2 of hospitalizations are preventable but need to detect worsening health early.

Clinicians have too many patients to follow constantly. Patients don't adhere to care programs well.

Home monitoring reduces cost.


  • Bluetooth weight scale, ECG monitor and Blood pressure + Blackberry Pearl Smartphone.
  • Uploaded daily to central server at hospital.
  • Reminders sent if they don't report in.
  • Secure Web site to see logs and reports.


  • Patient reports and alerts
  • Alerts can be sent to blackberry

Decision making is very complex. Differing perspectives on rules and preferences.

There is a tough balance between flexibility and simplicity.

RHM must be individualized for each patient.

Clinician Concern: Clarity of messages

Patient Concern: False alerts. what to do with information.

Iterative development is needed with input from all system users.

Objective is less hospitalizations. Number of visits to ER is not necessarily a metric. System may send people - because it is appropriate. Issue or quality measure is severity of the issue being dealt with in the ER visit.

Tara McCurdie - Mobile Devices for Nursing

Focus of study was on features of mobile devices and applicability for nurses.

Nurses are on the move. Multi-tasking.

EMRs risk tying nurses to a desktop.

Nursing PDA Pilot project:

Devices: Nokia N810, HP iPaq 210, BB 8820, Palm Treo 680.


  • Lexi-Complete.
  • RN Assoc of Ontario - Nursing Best Practice Guidelines
  • McMaster PLUS

Research also looked at more devices: Nokia N810, HP iPaq 210, iPhone 3G, BB Bold, Palm Treo 680.

Nokia/Ipaq are wifi PDAs. Apple/BB are 3G

N810 - Heaviest biggest screen

iPhone was lightest

Human Factors:

  • Heuristics
  • Shadowing
  • Questionnaires
  • Usability

Weight is critical. Don't want scrubs falling down due to weight.

Devices are wanted by Nurses for communication with each other.

Testing had minimal training to test how intuitive they are.

4 tasks:

  • Email
  • BMI calculation
  • MedLine enquiry (did i get that right?)
  • Wikipedia

Nurses had problem with both BB and iPhone in finding create email task.

BMI task not much difference across tasks.

On iPhone Safari icon not understood to be the browser - typically took 72 seconds to find it.

On Medline - Nokia Tablet better than pocket Pc. Page loads are slow on IPAQ over wifi (1min 30sec)

Nokia then iPhone fastest. Nokia size of display helped.

Wikipedia query

Nokia a lot better than Palm. Took 1M 20s for page to render on Palm over 2G. Visual rendering is poor.

Nokia and iPhone fared best. Nokia 1st.

Task completion performance didn't correlate to user preference.

After period of use the Nokias (8/24 selected) weren't being used. Paper weight due to weight.

Iphone selected by 6/24

Nurses wanted tactile feedback on keyboard.

Quick data access wanted. 3G beats Wifi due to general availability.

Large Screen wanted.

BlackBerry Bold and Apple iPhone were most preferred devices.

Pierre Pluye - Information Assessment method to integrate constructive feedback

1 complainer represents 250 silent complainers.

There have been no studies on the integration of feedback.

Constructive feedback - should trigger action.

Incorporating Push feedback model increases level of feedback