Improving Diagnosis in Healthcare

Improving Diagnosis in Healthcare – Report from the IOM released last year.
81% of physicians have experienced at least 1 significant delay in the past two months in getting timely diagnostic results.
Error in the diagnostic process OR “No fault” causes (silent disease, too early, atypical, patient misleads us, failure to follow-up) –> DIAGNOSTIC ERROR –> Inconsequential OR Harm
Diagnostic errors are common (around 10% of our diagnoses), making it an interesting topic for us both as clinicians and patients.
  • Most occur in the ambulatory setting – 58% (most medical care is delivered there)
  • Most do not occur with rare diseases!  Most involve common diseases e.g. asthma, T2DM…
  • 29% occur in the inpatient setting
The toll:
  • US:  40K – 80K deaths/year
  • The problem is – we have little mechanism to measure or get feedback on diagnostic error.  Autopsy rates have fallen precipitously.
WHY?  DIAGNOSIS IS THE HARDEST THING WE DO.
Root cause analysis:  Graber, et al.  Arch Int Med 2005; 165:1493ff.
  • about half the time, the error is attributable both to system errors and individual error
  • most common causes:  communication (e.g., critical lab lost); coordination of care; expertise unavailable; lack of trainee supervision; workload/stress/distraction; reliability of labs/imaging; lack of system to find dx error.
LACK OF KNOWLEDGE IS A RARE CAUSE – only 3%.  
Instead, most diagnostic errors are a failure of SYNTHESIS – putting it all together correctly.
 
HOW DO DOCTORS THINK?
  • In most cases, we use HEURISTICS – rules of thumb, illness scripts. (System 1)
  • Deliberate, conscious thought is only engaged if we don’t recognize the diagnosis!  Through repetition, this becomes recognized and thereby a heuristic. (System 2)
Croskerry 2003 – cognitive biases and heuristics.
Availability heuristic
  • fast, effortless, approximates the base rate of disease, often correct
  • pitfalls:  fails to consider DDx, experience is limited, available does not equal correct, we preferentially remember the “big case.”
Kathyrn Schultz:  “What does it feel like to be wrong?  Exactly what it feels like to be right.”
Search satisficing or premature closure – falling in love with the first puppy, “good enough” error.
  • the opposite of optimizing
HOW CAN WE PREVENT DIAGNOSTIC ERROR ON A SYSTEM LEVEL
  • Work in teams – use nurses, pathologists, radiologists…
  • Make the patient a partner in the process
  • Pay attention to the work environment:  reduce stress, allow enough time for diagnosis, provide enough support (including good EMR)
  • Improve communication
HOW CAN WE PREVENT DIAGNOSTIC ERROR ON A PERSONAL LEVEL?
  • What else could it be?
  • The Differential Diagnosis! (VINDICATE or VITAMIN CC&D
  • Be comprehensive
  • Practice reflectively
  • Consider the opposite.
Dxplain (
Isabel (isabelhealthcare.com) – Technol Health Care 2008;16:103ff – improved dx from 89% to 93%
Derm (visualdx.com)
What about Google?  Sens 58%, Spec 0%  – BMJ 2006 Dec 2;333(7579)
 – from Rocky Mountain Hospital Medicine conference, 2016, Graber
Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s