Given the volume of computer transactions in hospitals, these errors should surprise no one. Humans are fallible. So are human designed machines. In the case of EHRs, add to these factors the fact that 50% of the time, patients omit information or lie when they know their information is being entered into the record.
Wednesday, April 10, 2013
Human
Entered Health Information Errors and Computer Malfunctions
Garbage
in, Garbage Out.
Computer
Users’ Aphorism, 1970
And
as the smart ship grew,
In
stature, grace, and hue
In
shadow silent distant grew
The
Iceberg too.
Thomas
Hardy (1840-1928), The Convergence of the
Hue (Line on the Loss of the Titanic), 1912
The April 5 issue of HealthLeaders Media ran an
article “HIT Errors ‘Tip of the Iceberg,’ says ECRI."
The ECRI (Economic Cycle
Research Institute) publishes and consults on patient safety, quality
improvement, risk management, medical devices, health care technology, and
health policy.
The iceberg analogy may be a little misleading. HIT is
improving, but given the billions of transactions in hospitals and doctors’
office, I have no doubt errors occur and
like icebergs, 90% of these errors may be beneath the surface.
In any case, ECRI has just issued a 40 page report ,
“Health IT and Patient Safety Building a Safer System for Better Care.” The report documents 171 errors in 36
hospitals. The errors were traced to human mistakes and to computer malfunctions, each of which might have caused patient
harm.
The 171 errors involved:
·
53%,Medication mismanagement (25%
computer entry. 15% EHRs, 13% lab information, 11% pharmacy system, 2% other)
·
17%, poor clinical documentation
·
13%, inaccurate lab information
·
9% faulty computer malfunction
·
8% misleading diagnostic information
·
1% inadequate clinical decision support
The errors were of two types:
·
Caused by humans in their
interaction with computers (47%)
·
Caused by computers that moved too slow,
couldn’t communicate, or crashed (53%)
Given the volume of computer transactions in hospitals, these errors should surprise no one. Humans are fallible. So are human designed machines. In the case of EHRs, add to these factors the fact that 50% of the time, patients omit information or lie when they know their information is being entered into the record.
Garbage in, Garbage out is too harsh a term to apply to these errors,
which may be decreasing with advances such dictation entries, standardized entry, computing in the cloud, and better training of
entry personnel. It is too early to tell if the iceberg analogy applies to HIT. We should not let the perfect drive out the good.
Tweet: In an HIT study in 23 hospitals,
the Economic Cycle Research Institute found 171 errors that might cause patient
harm.
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment