Posted on November 8, 2016 by Dr. Michael Hamilton
It’s important not to think of e-prescribing as an isolated end in itself. We really need a whole system and process built around getting medicines to patients — from drug regulation and oversight, to drug manufacturing, to packaging and labelling, through pharmacists and pharmacy software, dispensing systems, and even a patient’s own medicine cabinets. All of these aspects of the system need to interact at a high degree of accuracy and integrity to ensure safety.
Even in its simplest form, e-prescribing certainly solves the legibility problem. Handwriting remains one of the important contributing factors in prescribing errors. Perhaps we think that EMR-printed prescriptions are better — they are, but those printed prescriptions need to be transcribed, by a human, to the pharmacy software system. This again is a significant source of medication error. A service that communicates seamlessly between the EMR and the pharmacy system reduces these vulnerabilities.
The development of an e-prescribing service is also an opportunity to engineer out some of the problematic ways we, as physicians, write prescriptions — we use short forms that can be misinterpreted and we use acronyms that carry different meanings. Is “HCT” hydrochlorothiazide or hydrocortisone? We use the symbol “U” to mean units and it is easily mistaken for 0 — so 4U of insulin is mistaken for 40 of insulin. We use trailing zeroes after decimal points — so that 4.0 looks like 40 (we should use 4). QOD is confused with QID is confused with QD. This, again, continues to occur even with computer printed EMR prescriptions.
In e-prescribing there is a unique opportunity to standardize the elements we need for prescribing and to force input of appropriate parameters, reducing the ambiguity and the need for interpretation which lead to errors — what we prescribe, how we prescribe, and the way in which we communicate instructions.
The transition to new technologies and new ways of doing things is almost always accompanied by new types of errors, and e-prescribing is no different. EMRs addressed some safety concerns inherent in paper charts, but created new ones.
We know of errors in electronic systems’ drop down box or pick list selections; selecting dilantin instead of diltiazem. We know of errors in legacy prescriptions; medications that were stopped years ago being re-activated accidentally because computers are so good at storing information and they remain on the EMR or in the pharmacy software in perpetuity. We also know that, once in there, for whatever reason, electronic systems are very good at propagating and perpetuating erroneous information. A drug prescribed in error through an e-prescribing service, in particular if it automatically populates other services (provincial drug information systems, insurance claim systems, hospital systems, pharmacy groups), can be very difficult to rectify.
And finally, in automating this process, we remove an underappreciated safety mechanism — we remove the patient’s assessment of the prescription. I think we have all had a patient, after receiving a prescription, turn around to us and point out an error in a drug name, or a dose, or even the fact that we documented all of this in the wrong chart. We lose this safety double check, or at least we delay it — and that’s an important consideration.
Despite these challenges, a thoughtfully designed and robust e-prescribing system, particularly if left open to future enhancements, will contribute enormously to the enhancement of patient safety.
Have a comment about this post? We’d love to hear from you.
Michael is a member of the medication incident analysis team at ISMP Canada, contributing to the investigation of medication errors and collaborating in research projects with the Institute. He has been privileged to be able to advise health care providers, administrators, policymakers, and legislators about issues in medication safety.