Busting the Myths about EMR Implementation

I am still crazy over the ridiculous EMR cost study recently published by CDW.  This regrettable study was obviously put together by a bunch of newbies that had no idea what they were doing.  A few weeks ago, during a rant on a previous post on this topic I promised I would review our practice’s implementation strategy.  It’s time to live up to that promise.

The CDW group assumed, as I imagine many still do, that EMR should be implemented as quickly as possible minimize the financial impact of decreased patient volume.  In our practice we took the opposite approach, realizing that if we brought in EMR slowly enough we wouldn’t have to decrease patient volume at all.  We overcame both the cultural and the financial barriers to EMR by creating an approach that was different from the prevailing wisdom at the time. Conventional project management principles don’t work in the physician practice setting.

1. We recognized that EMR was not a project with a defined end point – it would instead be an ongoing “work in progress.”
2. We avoided big “go-live” dates and hard deadlines for abandoning paper charts.  Paper charts were eliminated gradually, via attrition, over 2-3 years.   Docs are already uptight and uneasy about EMR; deadlines only make it worse.
3. We rejected the notion that we would have to decrease patient volume and lose revenue, even temporarily, to get EMR implemented.  Don’t even think about suggesting to a doc that (s)he will have to decrease patient volume.  We can’t tolerate it financially.
4. Every office and every physician was allowed to progress along its own timeline.  Every office has its own set of assets and liabilities – its own subculture.  It made no sense to force the same timeline on everyone.  We also offered (and continue to offer) each office / physician a fair amount of latitude on exactly how the EMR is used.  Some docs use speech recognition, some don’t.  Some offices didn’t scan outside records at first.  In the early days we didn’t care if docs wrote paper prescriptions.  The script pad is one of the hardest cultural icons for the physician to let go.

We had one physician who resisted EMR for almost a year.  I was approached several times to pressure this doc to “give in.”  I declined.  Then one day he discovered Dragon Speech and started EMR almost overnight.  We docs are self-selected fiercely independent souls; our training reinforces those characteristics.  I know this physician well; he had to do EMR on his timing and his terms.  If I had pressured him it would have backfired badly.  I probably would have behaved similarly.

To accommodate the physician’s need for independence the EMR adoption process was broken down into a large number of incremental steps.  After a short teaching session each physician had a training version of the EMR, complete with fictitious doctors and patients, installed on his/her laptop.  Over a few evenings the physician would work with the program to get used to the basic operations and functions.  Once the physician was comfortable we put the server communication software on the doc’s PC and showed him/her how to log in and use the same training EMR program on the server.  The training EMR on the server was configured with our custom templates.   The physician was then instructed how to create chart notes using our templates.  Then he/she could spend more time practicing at home, logging onto the server from there.

Then it was time to use the “real” EMR program on real patients.  But not all at once. Start with only one patient, the last patient of the day. Those first few notes took forever to complete.  But with our approach that was no problem.  For a while many of the docs printed out the completed EMR note and put it in the paper chart.  Why bother doing that?  It was a cultural trust issue.  With time, trust in the EMR increased and the practice disappeared naturally.

We advanced each doctor at his/her own pace.  Do EMR for the last 2 patients of the day, then the last 3, etc.  When ready, take on a half day of patients, then an entire day.  If there is a problem, back off.  Get the issues resolved and try again.   There were no deadlines and no pressure.  After getting settled with documentation move on to workflows such as prescriptions, ordering tests and imaging.  Then finally learn CPT/ ICD-9 charge code entry.

This process serves 2 goals.  First it allows the cultural change to an IT setting to progress at an acceptable, sustainable rate.  It also allows EMR to come in without decreasing patient volume.  It took almost a year to get 20 physicians in 15 offices implemented with basic EMR functions – but there was no panic, only modest chaos and no loss of patient volume.  We had our frustrating moments, but I am convinced that they would have been far worse with a conventional implementation plan.