I wrote recently about a less than optimal shift where we had multiple intermittent issues with our EHR that impacted electronic prescribing and several other key functionalities. Although the issue was fairly serious, it turns out that a good portion of the lack of communication our users experienced was the fault of our practice leadership rather than the vendor. In trying to “protect” the clinicians from what they felt would be distracting, they erred in providing too little information. It turns out the vendor was communicating pretty well, but they weren’t passing on the information to those of us on the front lines.
This is a challenge for any healthcare organization – figuring out the right way to communicate about serious issues and knowing how often to provide updates, even if the update is “no update.” In organizations where clinical leaders work closely with the IT team, there can be good conversations around a service disruption and how to handle it. There can be discussions about user culture – whether they would benefit from more information or less and whether the disruption is more of an annoyance or a disaster. When clinical and technology leaders don’t work together, there’s an increased risk of erring too far in one direction or another.
I still encounter organizations that don’t believe it’s important to have dedicated clinician support for health information technology systems. When I put on my consulting hat, one of my tasks is to try to convince these practices that they can’t afford not to have clinician leadership where EHRs and other systems are concerned.
I recently pitched to a convenient care practice that has almost 100 providers. They have one nurse who spends a couple of hours each week working with the EHR, mostly responding to specific end user questions rather than working on global strategy. I spent a few days in the practice to identify potential opportunities for workflow improvement as well as long-term strategies.
The first thing I identified while watching physicians document was that none of the physicians were using medication favorites. Apparently they aren’t allowed to build their own favorites because the organization is concerned about upkeep. Instead, there are some global favorites that everyone can use, but that might not be aligned with current treatment guidelines. Providers can delete the global favorites, but can’t put anything in their place, leading to the opportunity for errors when physicians try to use the global favorites and edit them as they go.
There were some order sets available, but they weren’t very complete, and many common diagnoses didn’t have associated order sets. As a result of providers not being able to build their own order sets, I observed several clinicians using Word documents that contained their most common patient instructions that they would copy and paste into the free-text plan field. Some of these makeshift order sets didn’t seem terribly evidence-based and they varied dramatically from person to person. The diagnosis screens were cluttered with diagnoses that didn’t seem to be commonly used, while providers were having to search for conditions that they treated several times each day.
When I perform a practice analysis, I also dig into how the practice handles upgrades and changes to payer requirements or federal programs. It turns out that that when it’s time for an upgrade, the physician CEO and the EHR nurse evaluate the release notes and decide which features they will implement and how the end users will be trained. They don’t seek input from any of the users or even the physician group’s medical directors. The two of them personally deliver most of the training in a one-on-one fashion, which means that some users might get trained as much as six weeks prior to the upgrade. Others might just receive a PDF that they are supposed to review before launching into the new workflows. There’s not a lot of satisfaction around that process.
Understanding that process explained some of the issues I saw in the system, including a workflow for in-office medications that borders on dangerous. There are fewer than two dozen medications available in the office, many of which are in specific unit doses. Rather than configuring an order screen with those medications and defaulting in their strength, form, and administration instructions, providers are required to individually select every parameter for every order. Some medications can be ordered multiple ways.
For example, one drug can be ordered either as 3 ml or 2.5 mg. Since the medication is 2.5 mg/3 ml, either order is appropriate, but I saw several physicians click for 3 mg or 2.5 ml, neither of which were correct. The system didn’t flag these, but instead the clinical support staff was responsible for changing the orders. Incorrectly ordering albuterol at that scale isn’t going to cause significant harm, but for other drugs, those types of mistakes are far more serious. Beyond the safety issue, there’s the matter of the numerous clicks required to even order a single drug.
I identified all kinds of operational issues in the practice as well. Although they have a time clock system for both clinical and business office staff, they don’t have hourly providers use it. Instead, providers have to email their “stop time” every night and it takes a manual process to document the time in both the payroll system and the scheduling system. For the latter, they use one system for providers to request their schedules and another system to actually publish the schedules. They’ve switched payroll systems three times in five years, which makes me wonder whether it was really a software issue or something much more challenging to fix.
There were plenty of other issues to tackle, enough to keep several consultants busy for many weeks. I knew there were some internal disagreements on whether to bring in outside help, so I prepared a conservative proposal in multiple phases to allow them to get used to the idea of letting someone help them. The return on investment was easy to demonstrate, but as I presented to their leadership I could tell they weren’t interested. It was clear that the CEO believes his way is not only the best way but the only way to do things.
Although many of their technology struggles could be made better through the application of skilled assistance, they’re not ready for change. Given the challenges that will be upon them as healthcare continues to evolve, it will be interesting to see where they are in three to five years.
What’s the scariest CPOE system you’ve seen? Leave a comment or email me.
Email Dr. Jayne.