Lessons Learned from Anesthesia EMRs

Several years ago one of the hospitals where I operate spent 6 figures on an anesthesia EMR system.  After several months and a huge amount of money the whole thing was scrapped because it was so cumbersome to use.  They have not tried again.

A few weeks ago the anesthesia group that covers our surgery center got an EMR.  The product is called Anescan and apparently has many successful installs.  It runs on Windows 7 tablets that communicate with a central server.  Needless to say I was curious to see how this system differed from the failed system I had seen years ago.  What I learned was very interesting.

Medical record keeping in anesthesia is different from all other medical specialties.  The job includes monitoring vital signs constantly and documenting them in the anesthesia record every few minutes.  It is a task that begs to be automated.  Such technology would presumably free the anesthesiologist from mundane repetitive documentation, allowing more efficient and effective monitoring of the patient.   The necessary technology has been available for years and was used in the failed hospital system from years ago.

I was surprised to learn that Anescan avoids that technology.  A conversation with the Anescan rep revealed that is was precisely that technology which caused earlier systems to fail.  It’s easy to measure blood pressure, heart rate, respiratory rate, and blood oxygen level and push that data to an EMR.  The problem is that the data are often riddled with artifact.  If an EKG lead or pulse oximeter comes loose, or if the surgeon leans on the arm-mounted blood pressure cuff, it is not unusual to get an automated pulse or blood pressure of zero.  The anesthesiologist / anesthetist can easily recognize what is happening, fix the monitors and record accurate vital signs.  This often happens several times during a case and is no big deal.

The automated system makes it much worse.  By the time the bad data are recognized the automated system has already pushed that zero pulse and BP to your EMR.  Now the anesthesiologist / anesthetist has to open some kind of editing function in the EMR and delete, edit, or explain away the false readings…AND at the same time troubleshoot the monitors that sent the bad data in the first place…AND by the way your patient is still asleep and you can’t stop watching him.  AND you only have a couple of minutes to get caught up before the monitors send the next the next set of (? bad) vital signs to the EMR.  The potential downward spiral is easy to see.

Anescan avoids this problem.  The tablet PC presents an image of a standard anesthesia paper record with the patient demographics and other data already in place as structured data.  Vital signs are recorded with “digital ink.”   Use the stylus to record vital signs on the form, on the tablet.  When the case is complete the form images are sent to the server for centralized record keeping and billing.   A paper copy is printed for the surgery center chart.  This is an elegant solution that automates only those parts of record keeping where it is practical.

Someday the artifact problem will be solved either through better monitors or better error recognition within the EMR.  But today this serves as yet another example of too much IT and automation in health care causing more problems than it solves.