At my clinical practice, many of my partners have been out for spring break. Since the local school districts have staggered break schedules, nearly everyone wanted the overlap weekend off, so I was happy to work the whole thing.
Although Friday’s shift had more than its share of patients bearing the complaint of, “I just started getting sick and I’m going to Cancun and can’t be sick for break,” Saturday trended more towards, “I just got back from Cancun and am sick / hung over/sunburned.” I was starting to question my sanity until Sunday, when some “typical” patients started coming in.
I’ve mentioned this before, but with the shift to high-deductible health coverage, we’re seeing a tremendous increase in volume. Our pricing is transparent and we’re conveniently located and provide quick service, so the business is experiencing exponential growth.
With that comes some growing pains, however, which for me has been felt in the number of new staff members working on the teams. We have a really great training program – new staff members have formal training shifts and each shift has a different focus. One day may be clinical interview skills, another may be labs, and another may be procedures, etc. They work directly with a trainer whose only focus for the day is to train them – it’s not someone already on the care team who is training on the side.
Even after the formal training, some staff may be more green than others. I ran across a scenario yesterday where the staff failed to notice some nonsensical entries in the EHR. Although it should have been reviewed before addition to the chart, the patient care tech missed the errors:
- Pulse of 13270
- Respirations of 99/minute
- Temp of 15
It turned out that the tech had entered the data quickly, was just tabbing through the data entry fields, and was off by one field. The blood pressure field (which should have shown 132/70) was blank and he entered those numbers without a slash in the pulse field. The error then compounded as he tabbed. He was apologetic and immediately fixed the error.
Being in the health IT industry, I quickly flagged it as not only a human error, but also a software problem. Most of the EHRs I’ve worked with have restrictions on various data fields to prevent these kinds of errors. For example, a pulse field might only be able to hold three digits. Active or passive alerts might display for values outside the normal range.
Although the tech should have caught it, my bigger concern is that this happened in a Meaningful Use Certified EHR. I’ve asked the practice’s technology liaison to open a ticket with the vendor and see if it’s functioning as designed or whether there is a defect. If it’s functioning as designed, I have to wonder about the certification standards. I don’t beg to have a command of the details and I know there are hundreds of pages of requirements that must be met.
Knowing that some of the elements that are requirement for certification may not be something that physicians need or want, I’m surprised if there isn’t something in there to require safety checks for straightforward data entry like this.
I first dealt with an EHR that handled data like this in a conversion project more than a decade ago. We had vast amounts of data that couldn’t easily be brought into our new system because the blood pressure field was a single field that would accept numbers, letters, and symbols. Assuming a sample BP of 140/90, users had entered it as:
- 140/90 sit (meaning taken seated)
- 140/90 R (meaning taken on the right)
- 140/90 RA (meaning taken on the right arm)
- 140/90 RALC (meaning taken on the right arm with a large cuff)
And so on. Our new system had separate fields for the systolic BP (top number) and diastolic BP (bottom number) as well as discrete fields for position, side, site, and cuff size. Due to the work needed in trying to cleanse the data, we quickly decided that we would just not bring any values into the new system and would start from scratch.
Since that conversion project was so long ago and I haven’t run across the issue since, I assumed that such handling of data had gone the way of the dinosaurs. I guess it hasn’t, or I’ve just been spoiled by more sophisticated systems. But I would have hoped that with all the focus on patient safety and regulations, that we would have moved past this and that consistent handling of essential data such as vital signs would be a requirement for vendors seeking certification. How in the world can you be truly interoperable with data like this?
We’ll see what happens with the vendor ticket and what my practice decides to do about it otherwise. If I was the CMIO, CMO, or medical director and this was my system, I’d be tracing it all the way through to find out what is being sent to the patient portal and what appears on transition of care documents and how extensive the problem might be.
Although this particular scenario was a pretty significant and obvious error, I’m sure I could have missed less significant errors during the last couple of years. Since I’m wearing my hourly staff physician hat in this scenario, though, I’ve notified our leadership and have to let them work it as they see fit. I’ll be spending extra seconds reviewing my vitals going forward, however.
This should be basic functionality, but I guess it’s not. I’m interested in hearing how other certified systems handle this type of data – whether they have field restrictions that would have prevented these errors, and whether they have active or passive alerts to create additional patient safety support. Consider adding a comment and sharing what you’re seeing in the trenches.
Got screenshots? Email me.
Email Dr. Jayne.