Testing in hospitals
'Hi, Jonathan--
There are a few general principles that you can proceed from; for specific
issues on particular machines, I'm afraid there's still not a lot of info
available.
First, you absolutely need your Clinical Engineering department in on the
Y2k plans. They have a lot of equipment that you will need, and they have
mandates for testing equipment based on criticality (defibrillators are
tested more often than electronic thermometers, for example). You need to
know what they test, how and how often they test, etc., and build your
test plan around that foundation. Leverage off Clinical Engineering's test
schedule as much as you possibly can.
Second, you probably will not be able to get at the chip to change the
date in it. Most of this equipment has monitors and input devices, and for
most of them, altering the date through those input devices will be the
way you test that date transition. It is important to understand and
document that this is not a thorough test. On the real date, not only will
the date be rolling over, but it will be getting input from electrodes
attached to the infant; a nurse might be adjusting a setting; any number
of combinations of inputs can be going on. A device that tests compliant
for rolling over in isolation can fail in an ICU real-life setting. Watch
for false compliance, and plan what you will do in case devices you think
are compliant really fail. Testing can tell you which ones will really
fail, and which ones are less likely to, but watch out for false
positives.
Third, make a contingency plan and get buy-in from whomever you need to
make it effective. The transition is going to happen on a holiday weekend
at the biggest celebration in recent memory. Hospitals will have skeleton
crews and temps/floaters working unless they specifically plan to do
differently. It takes two hands to bag a patient if a ventilator fails;
one nurse can't do that *and* call tech support at the same time. Plan to
be more than staffed, in case you need the extra hands, and plan for your
clinical engineers to be available immediately. Make sure they know how to
reset any device you use, and are ready to do so. This is the most
important step; there is no testing you can do to eliminate it. And paying
floaters for a weekend's shift is a lot cheaper than the lawsuits that
will happen.
That's it off the top of my head; I'm still working on this problem with
local (Seattle) hospitals, and will be happy to share any further
information I have as I get it.
--Siobhan Harper-Jones, ComputingSafe 2000 LLC
__________
Date: Fri, 13 Mar 1998 14:35:07 -0800 (PST)
From: Siobhan Harper-Jones <siobhanh@blarg.net>
To: year2000-discuss@year2000.com
Subject: Re: Embedded Systems: Healthcare and other Y2K Issues |
