By Michael Burger, Senior Consultant

A study was published recently in the JAMA (link here) titled, “Analysis of Prescribers’ Notes in Electronic Prescriptions in Ambulatory Practice.”

The study found a high number of instances where extraneous data was included in the free form notes section of electronic prescriptions (ePrescriptions) by prescribers using electronic prescribing (ePrescribing) software.

In the study, of more than 26,000 ePrescriptions which contained free form notes, 66% contained what the authors consider to be inappropriate. 19% of these free form notes included directions that conflicted with directions contained elsewhere in the ePrescription.

I agree that the inappropriate use of the free text prescriber notes field in an ePrescription is an issue that must be addressed, especially from the patient safety perspective.   I also concur in part with the authors’ conclusion that accelerated implementation of existing and new ePrescribing electronic data interchange (EDI) standards would address a sizable number of these issues.

The study also suggests the need for “better prerelease usability testing, rigorous post-marketing evaluation and surveillance of EHR or ePrescribing software applications.”  On this point, I believe that the authors have missed the mark.

In my days as an EHR product manager, one of my usability-tester clients used this analogy to describe using the EHR. He said “Imagine composing an email using only words and phrases from pull down menus. That’s what it’s like to write a prescription using the EHR.”  Sometimes there is information that needs to be relayed to the pharmacist.  And while some (or maybe most) of that information could be tediously selected from pulldown lists, one piece at a time, sometimes a prescriber decides that it’s just quicker and easier to type it out as free form text.  So to say that EHRs are poorly designed and not user friendly is not a fair criticism. The process of ePrescribing by selecting individual components of data is not as "user friendly" as the process of writing a free form prescription on a piece of paper, regardless of the design of the software.

It is fair to say that the current design of ePrescribing software doesn’t accommodate certain kinds of ePrescribing well. For oral solids like pills and capsules, ePrescribing works great.  For liquids, ointments, injectables and compound drugs, not as well. In this area, I agree with the authors’ conclusion that “consistent end user training and feedback” is needed. Existing software can be appropriately used to order these not-easy-to-ePrescribe product categories.  But appropriate ordering may require prescribers to rethink how they order.  Instead of ordering “one tube,” they’ll need to order an appropriate number of grams or milligrams.

One ePrescribing challenge today is that there isn’t a feedback loop from pharmacist to prescriber. If the pharmacist receives an ePrescription containing conflicting or extraneous information, the prescriber isn’t typically notified by the pharmacist that there is an ePrescribing-related issue. The clarifying phone call is most frequently fielded by a nurse, not the prescriber. And the question is one of clarification of the prescriber’s intent – not an attempt to address the root cause of the issue.

There also seems to be a misconception of EHR vendors’ role in this equation. Vendors aren’t (and shouldn’t be) placed in a position of enforcement. If users of the software are sending extraneous information in the prescription, the discussion should be between the pharmacist and the prescriber.

[tweetthis]It isn’t the #EHR vendors’ role to surveil their customers’ use of their products. [/tweetthis]

It isn’t a fair characterization of EHR vendors to assume that they don’t perform usability testing or post marketing evaluation of their products. Frankly, if they didn’t, they’d be out of business because no one would buy inferior software.  It isn’t the vendors’ role to surveil their customers’ use of their products. If I don’t like the contents of a letter that I receive, I don’t talk to the letter carrier.  Nor do I talk to the manufacturer of the vehicle that the letter carrier is driving. Instead I talk to the sender of the letter. In the same way, if the pharmacist is receiving incorrect or extraneous data from the prescriber, the pharmacist and the prescriber should work that out.

I also observe that physicians are taught how to write prescriptions, on paper, in medical school.  I’m not aware of a study in which paper prescriptions have been evaluated to determine if there was “inappropriate information” included. I would wager that most of the same types of extraneous information being transmitted electronically is also included on paper prescriptions. So these “old habits” may be hard to break.

I am an avid proponent of more accurate and safer ePrescribing. It defeats the efficiencies that are gained by ePrescribing if such a high number of ePrescriptions require follow up to clarify the prescriber’s intent. Rather than place the burden for process improvement primarily on the EHR vendors, I believe that a more balanced approach would be to involve the prescribers and the pharmacists.  A quick and to-the-point conversation is likely all it will take.