...which is strongly suggestive, that the VA's problem with Cerner implementation? It's coming a lot more from the VA, than…
I was sorry to read of Mr. H’s COVID-like symptoms, which as of his last post hadn’t yet resulted in a positive test. Especially in vaccinated patients, we’re seeing a pattern where antigen testing doesn’t become positive until five days or more after symptoms have started. Although vaccinations are still very good for preventing hospitalizations and death, I’m seeing a fair number of people with some pretty debilitating symptoms as well as a large number of patients with long-COVID symptoms. Best wishes for a speedy recovery for Mr. H.
As a physician who has spent thousands of hours staffing emergency departments and urgent care facilities, I fully support the idea of having interoperable systems to better understand patients’ medication histories. My first informatics employer stood up a private health information exchange (HIE) to help its employed physicians better share data with community physicians who also used the same EHR. It was an exciting time to learn how the connectivity needed to work and to navigate through decisions such as whether we would be an opt-in or opt-out platform in the face of changing privacy laws and state requirements that weren’t keeping up with the times. As the first HIE in the state, if was something to be proud of, but we knew it could be done better and were ready to support broader statewide efforts when they came along.
One of the major benefits of even our limited regional system was being able to see prescription histories and to identify patients who were visiting multiple physicians for controlled substances, or who were trying to refill their medications sooner than they should. Having the data made it easier to have those uncomfortable conversations with patients so that we could try to arrange appropriate follow up. From those experiences, it was natural to be a champion for Physician Drug Monitoring Programs (PDMPs) when they were first proposed. Some states were faster in creating these programs than others, and at least one state still doesn’t have a true statewide platform. As soon as one became available to me, I signed up, and before long I was using it nearly every day.
My last urgent care employer had a large number of midlevel providers seeing patients – nurse practitioners and physician assistants outnumbered the physicians nearly three to one. In my location, those providers weren’t allowed to sign up for the PDMP on their own, but required a sponsoring physician to allow them to be a delegate on their account. I thought that was unusual since those providers can prescribe controlled substances in my state as long as they have the name of their collaborating physician on the prescription, but I’m not the one who makes the rules. Several of the nurse practitioners asked me to sponsor them even though I wasn’t their collaborating physician. Apparently, the physician owners (who were the collaborating physicians for many of the midlevels) refused to sponsor them for the PDMP because they didn’t want it to create inefficiency in the workflow.
I found that approach to be short-sighted as well as non-collegial, so I agreed to sponsor a few of them who I had worked with for years and knew very well. They continued to use the system under my sponsorship until I left the practice and terminated the linkage between our accounts. It looks like their collaborating physicians still won’t sponsor them, because every two weeks I get an email from the PDMP noting that I still have open requests for sponsorship. In thinking about the fears of inefficiency, it’s likely more complex than that since the practice dispensed medications on a cash basis (including controlled substances) and was well known for making sure providers were “treating the heck out of symptoms.” Patient satisfaction was also a big push and of course it takes more time to counsel a patient about something you might find in the PDMP versus just prescribing medication. The practice also refused to install technology that would allow e-prescribing of controlled substances, so you get the picture. Attitudes like that are part of why I no longer work there, but one does still have to think about the impact of any new systems on efficiency.
An article in JAMA Health Forum addressed the topic this week, with findings that having a PDMP that’s integrated with the EHR led to increased use of prescribing recommendations by primary care clinicians. It shouldn’t be surprising to anyone that if you have a solution that makes it easy to do the right thing, people will be more likely to actually do the right thing. The researchers looked at 43 clinics and how they used the Minnesota PDMP. There were 21 clinics that had the PDMP integrated with the EHR, which reduced the need for a separate login to the outside system while also lowering the need to intentionally think about using the PDMP.
Monthly query rates for the non-integrated sites were 6.6 per clinician before the intervention versus 6.9 after. Monthly query rates for the integrated sites were 8.8 before the intervention and 14.8 after. Additionally, the findings showed a greater impact of the EHR integration on less-experienced clinicians, which they noted “may reflect a less inflexible practice style and/or faster uptake of new features in the EHR.” While the study was underway, Minnesota initiated a mandate that clinicians query the PDMP before prescribing opioid drugs, which led to an uptick in query counts in both groups, although that increase was more significant for prescribers in the intervention group.
There are several potential limitations to the study, including the fact that it didn’t directly assess opioid prescribing. The study was also limited to a single health system and its affiliated prescribers within a limited geographic range, although there was inclusion of urban, suburban, and rural areas. The study also could not identify whether queries were clinically appropriate or not or whether the PDMP queries led to providers deciding to not prescribe controlled substances when they otherwise might have. The authors do plan to look at this in a further analysis using information from the EHR.
I do hope that those who are trying to increase the adoption of EHR-integrated solutions use information like this to show that they can have a positive impact on provider workflow. They can also have a significant impact on patient outcomes when it comes to identifying patients who might benefit from an intervention with regard to controlled substances. I’m interested to see what results other organizations may have had with EHR integration and how their clinicians responded to it.
Have you integrated your state’s PDMP with your EHR? Leave a comment or email me.
Email Dr. Jayne.