Frequency of and risk factors for med errors during order verification

error

A friend and colleague of mine and I were talking about pharmacy order verification and errors the other day.

Many (all?) acute care pharmacies use the number of orders entered/verified by pharmacists per unit of time as a performance metric. The theory being that the more orders you verify the busier you are and therefore the more work you do. Sounds logical. And it is. And it also isn’t. It is because more orders typically equates to a heavier patient load and/or sicker patients, which in turn results in more work. It isn’t because some orders – and the associated distribution of medications associated with those orders – are significantly more complex than others. Chemotherapy is a good example of this. The amount of work required to verify and dispense your average chemotherapy order is several orders of magnitude greater than an order for, oh I don’t know, a saline-lock.

So in theory, a pharmacist verifying more chemotherapy will certainly have lower numbers than a pharmacist verifying routine medication orders. However, when evaluating the number of orders verified over an entire year, one would expect the number of complicated orders handled per pharmacist to average out over the long haul. That’s not complicated math, just common sense.

Assuming that every pharmacist is on equal footing, and that the number of orders verified is a reasonable performance metric, what then is a reasonable number of orders to be verified per hour, per shift, or per year to be considered good, bad, or average? It’s impossible to say. One would think that the higher the number the better the productivity. Here’s the thing, during periods of high volume order verification, pharmacists make more mistakes. Not just mathematically more, i.e. 1% of 300 is more than 1% of 200, but a higher percentage of mistakes.

From a blog post at the American Pharmacist Association (APhA) website: “the number of medication errors increased with the number of orders verified per pharmacist per shift” … According to the findings, the verification of more than 400 orders per shift per pharmacist was associated with the highest risk of errors…“Once we got to the 400 mark, meaning 400 orders verified per pharmacist, [we] started to reach a higher number of errors,” said Christy Gorbach, PharmD, coauthor of the study“. The study referred to in the APhA post is this one.

So it would appear on the surface that using volume of orders verified as a performance metric is actually driving pharmacists to make more mistakes, thus leading to less productivity, not more. (1)

With the increased adoption of EHRs, the number of orders verified per pharmacist is only going to grow. EHRs have made order entry quick and easy. Physicians create lists – favorites, order sets, etc. – that allow them to simply check a box and put patients on multiple medications in a matter of seconds. This is especially true for what I refer to as “don’t-call-me” orders. Don’t-call-me orders consist of multiple PRN medications to cover everything from fever to constipation; all designed to prevent the physician from receiving a phone call at 3 o’clock in the morning, i.e. don’t call me. Most of these orders go unused and simply complicate the medication profile and medication administration record.(2) Verifying such a large number of redundant, benign orders creates alert fatigue and selective blindness which ultimately leads to something important getting missed.

All in all, the results are increased verification numbers, more work, and as it turns out, more errors. Unfortunately, I have no answer to the problem. And make no mistake, it is a problem. Orders have to get verified and released, and patients must receive their medications in a timely manner. With that said, safety must also be a top priority. Medication errors are unacceptable.

Overall, it appears that using volume of orders verified by a pharmacist as a performance metric is a bad idea. I think it’s time to slow down, pay attention, and create an environment that rewards pharmacists for the quality of their work, not for the speed at which they perform it.

———————————————————–

  1. Medication errors can lead to all kinds of problems, among them wasted time.
  2. All the pharmacists reading this are shaking their head and grinning because they know exactly what I’m talking about. What pharmacist hasn’t seen a post-op C-section order with 15 to 20 PRN medication orders, including half a dozen different pain meds that never get used? It happens all the time.

4 thoughts on “Frequency of and risk factors for med errors during order verification”

  1. What a great article, it came in a right time!
    For the past few weeks, I was working on an interactive dashboard for our pharmacy based on generated reports from our EMR to give us more insights about the workload and the performance indicators.
    We are community general hospital and have different specialities and ICUs and as well.
    I built the dashboard to have the average orders per shift as a performance indicator for our pharmacists, but looking to the type of orders each pharmacist has been working on during the shift, made me start thinking is that fair?
    I can see a pharmacist who is doing an average of 100 orders per shift, but looking at the order type, it shows that 60% of this orders are Paracetamol 100 for, and 70% of his/her order were from general surgery ward
    How can I compare this to another pharmacist who is doing average of 60 order per shift but 80% of his work are more complex medications for ICUs which need calculation and more time in verification
    Also, the quality can’t be measured clearly.
    Same goes to th

  2. It’s a dilemma to be sure. Pharmacists have always found it difficult to come up with appropriate performance metrics. If you figure it out, please don’t hesitate to let me know (smile)!

  3. This is a weighty subject. Or, better put, this is something that can be solved by weighting.

    Many years ago, I worked for a medical center that had more industrial/management engineers than they knew what to do with. Given that, and the fact that the hospital pharmacy department had so many metrics, they wanted to figure out if they could “help me” know if I was using my staff “effectively and efficiently”.

    The result was a large management engineering study which determined “time information” associated with different types of medication orders and medication doses. Ultimately, we were able to create a “weighting system” that put a higher weight on certain types of orders (e.g., chemotherapy, antibiotics), where the pharmacist’s time per order was greater. A similar type of weighting was applied to the different types of doses prepared (e.g., a TPN versus a unit-dose pill).

    I think this principal would still apply. Using appropriately determined “weights” for each order and dosage type, you can determine the theoretical amount of time that “should be necessary” to perform all of the tasks, either for an individual person and/or an entire department.

    All orders or dosage types are NOT the same and if you really want to understand if you are effectively and efficiently utilizing your labor resources, you need a system (e.g., weighting) to normalize all of your data.

    I know this is a “heavy” subject, but I’m sure you get it…

  4. It’s always been difficult to figure out how to justify the services of a pharmacist, whether clinical or operational. Weighted values are certainly an option.

Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.