Guerra On Healthcare: Electronic Medication Administration Records Work
Policy makers should push eMARs while letting computerized physician order entry systems mature.
Computers are very good at some things and not so good at others. When we automate the things they do well, businesses see the value and move quickly to adopt the new tools.
When we try to automate things computers don't do well, we need massive government-funded incentive programs followed by penalties to move the market. But sometimes, even coercion isn't enough to make good people use weak systems.
Computers are really good at accepting medication orders and routing them to pharmacies. They're excellent at letting nurses know when the pharmacist has approved those orders, and super-fantastic at matching barcodes on medication vials with patient wristbands. In short, they're great at the process known as bedside bar-coded electronic medication administration record, or eMAR.
Computers aren't yet very good at learning. They're not adept at understanding the preferences of individual physicians for alerts, for accepting the fact that physician A doesn't need to see "that" alert, but does need to see "this" one.
It's because of this weakness that computers aren't up to the task that the Office of the National Coordinator has prematurely assigned to them in Stage 1 of Meaningful Use--the first phase in which healthcare providers can qualify for HITECH funds under the American Recovery and Reinvestment Act.
A study released last week by Brigham and Women's Hospital in Boston compared 6,723 medication administrations in hospital units before barcode eMAR was introduced with 7,318 medication administrations afterwards. The barcode eMARs were associated with reductions in errors related to the timing of medications, such as giving a medicine at the wrong time, and non-timing medication administration, such as giving a patient the wrong dose. Put simply, eMAR worked.
Many health systems have naturally chosen to go with this improvement in medication safety before computerized physician order entry because it's a clearer win. eMAR users are hospital employees--nurses and pharmacists. CPOE users are usually independent physicians who bristle at the thought of being slowed down. They're independent contractors bringing revenue into the hospital who don't look kindly on being imposed upon.
But that doesn't mean eMAR is easy--not by a long shot. A huge amount of workflow change and training is necessary to make the process successful but, as the above study proves, the rewards can be huge.
For some reason, ONC has fallen in love with CPOE, putting it in Stage 1 of Meaningful Use while leaving eMAR for later stages. It's time for policy makers to switch things around, bring eMAR up front where it belongs and give CPOE a few more years of much-needed development time.
When the alert-fatigue issue has been solved and CPOE systems can better tune their warnings to the needs of each user, no one will have to force physicians to use it, they'll refuse to practice anywhere without it.