e-Prescribing Adoption Lags

Earlier this month the Center for Studying Health System Change, based in Washington, D.C., released its issue brief, “Even When Physicians Adopt E-Prescribing, Use of Advanced Features Lags” that showed that about 1 in 3 office-based physicians routinely e-prescribed in 2008 and even fewer used advanced features like transmitting prescriptions electronically, identifying potential drug interactions, and checking formulary information. HSC senior researcher Joy Grossman, Ph.D. spoke with HCI associate editor Jennifer Prestigiacomo about these interesting findings.

Healthcare Informatics: In your opinion, what were some of the big takeaways from this study?

Joy Grossman, Ph.D.: Even when physicians adopt e-prescribing, it doesn’t mean they use the system routinely. The use of advanced features certainly lags, and that’s [because] of a number of factors. They might not have that functionality, and even if they do, it doesn’t mean that it’s used routinely by physicians. Those are the two key takeaways. The last point is there’s some variation, when you look into the different subgroups of physicians, whether that’s users of electronic medical records or physicians in larger groups, you see that some types of physicians are more likely to be routine users of the system, with respect to the basic functionality and the more advanced features.

HCI: What are your thoughts about the findings showing higher electronic prescribing rates among physician using EMRs exclusively than those with stand-alone systems?

Grossman: For one, physicians with EMRs are more likely to use the [e-prescribing] feature routinely. It is more integrated into their workflow. The computer’s open, they’re inputting the data, their staff is using the system and e-prescribing is often the first module that is implemented in an EMR because it’s one of the things that give physicians pretty quick payback. For practices where it’s easier to integrate into workflow, physicians are likely to get up to speed more quickly and use it more routinely.
Standalone products have the same issues around uptake and part of that is because it’s not integrated into the workflow. They might have a computer in their office, but not in the exam room, or they might be sharing computers with other people, or they might be using a handheld and there are problems with the wireless. The barriers to integrate these standalone systems into workflow are higher, and because of that, there’s potential for physicians to get frustrated and not use it.

In some practices you might have physicians with various specialties. Some physicians tend to write standard prescriptions, like in an ophthalmology practice, often their prescription pad has all of the medications they typically order, typed into the prescription pad, and they just check off what it is [they need to prescribe] because 90 percent of the time they prescribe 10 medications. Or surgeons who prescribe generic painkillers—they may find their threshold for using this technology is higher to connect it into their workflow because the payback is lower for them, relative to an internist or a rheumatologist.

HCI: Why do you think the physicians who do have e-prescribing in their practices aren’t using advanced e-prescribing features?

Grossman: The reason why doctors might not use advanced features is that, first of all, the systems may not have the functionality. The functionality even if it is available, might not be apparent to the doctor, so they may have not had adequate training, or they might not have absorbed that training. Second thing is that the practice may have decided the functionality does not work well and for that reason, they’ve actually turned it off. Even for the physicians who know it’s there and it’s on, they may find it hard to use, and it takes five clicks to get the information, or they may find the information is not particularly helpful.

For example, with the drug interaction alerts, there’s what’s called alert fatigue, and this is a well-written-about phenomenon. Alerts may trigger often for things physicians do any way or for patients they’ve treated for a long time, so they ignore the alerts, and then at some point there’s potential for an alert to come by that would be important for them to see, and they might not notice it.

HCI: Can these alerts be customizable?

Grossman: The more sophisticated the system, the more customizable it is. Sometimes it can be customizable at the practice-level, and sometimes individual physicians have control over that, but most times physicians express dissatisfaction with that. No one has really figured out the optimal design for your average physician and practice to take out of the box and go with.

HCI: Why do you think formulary checks were used less often than the other two advanced features [transmitting prescriptions electronically and identifying potential drug interactions] cited?

Grossman: There are other examples for the formulary information. Even when the information is provided, there are barriers to matching the patient to the formulary data. The formulary data may be missing for their health plan. Their health plan may be out of date.

HCI: The study showed a higher adoption rate in physicians in larger practices than smaller ones. Can you talk a little about why that is?

Grossman: The major issues are having the resources to set up the technology and incorporate it into workflow and then make it part of the practice culture. The large practices with employed physicians are more likely to have EMRs. Then, they have staff that is responsible for setting up the EMR and testing it. And often in those types of practices, when they bring physicians on board, they may get rid of prescription pads. The workflow is there. Then physicians have the expectation that their colleagues are going to use the systems in larger practices, so there’s an expectation that everyone is putting information in the medication lists and using the tool.

You certainly have examples in the smaller practices of the super user. It’s not to say that physicians in smaller practices can’t do that; it’s just the resources they need to set it up, test it, and the troubleshooting to make sure all the functionality works, there’s a huge resource investment that goes into that. Many of the standalone systems are Web based so all the upgrading is invisible to the physician because that’s all being done by the vendor. Some of that is mitigated by some of the vendors to use some of those ASP [application service provider] systems, but there are still challenges to use all the new functionality.

HCI: How much do you think meaningful use will incentivize physicians to use advanced e-prescribing features?

Grossman: Clearly with respect to transmission electronically, meaningful use is likely to have a direct impact on the use of that functionality for physicians that have that IT. So, right now they say 40 percent of prescriptions that are allowed, so the scheduled drugs the DEA allows, are to be sent electronically. I think this will definitely have an impact for those physicians who are responding to the incentives. Once physicians get that functionality in place and working, they’re going to use it frequently for all their prescriptions because it can be actually much easier.

I think for the other types of functionality right now, the requirements for the drug to drug interaction functionality has to be there and the drug formulary functionality is an option under that menu. There is also medication history that we didn’t talk about in this study, that’s a functionality provided by vendors like SureScripts that [allows] access to adjudicated claims data from health plans that patients are covered under. It shows what drugs they’ve bought under insurance plans. So that was actually not included in meaningful use, although it’s included under the MIPPA [Medicare Improvements for Patients and Providers Act] requirements as a functionality.

HCI: What do you think CIOs can do to encourage usage of e-prescribing?

Grossman: One thing is training, and that means making sure the physicians are aware of the functionality, and as systems get upgraded and changed, if access to that functionality changes, the physicians are made aware of that and don’t assume that they know.

The second thing is being able to work with the IT staff and the practice administrators to make sure the functionality is working well—testing to make sure the patient match is working, and that the formulary information is actually available to physicians at the point of prescribing. Also, that it’s up to date so that if there’s a major local health plan missing, that they’re going back to their vendor. Also, having someone troubleshooting across the practice. While incentives are useful , I think that when physicians find this functionality useful, it’s a kind of cost-benefit. If they find it valuable, and the value is worth the time it takes them to do it, they will do it. But if it turns out that it’s not worth the time to them, they’re not going to use it. For example, with formularies if that does it make it easier to make a decision with patients upfront and improve efficiency [they will use it].

There are a lot issues outside of the scope of the practice itself. There’s getting good formulary information and making it available. Getting all the local pharmacies onboard and having them send all their prescriptions electronically and making sure the connectivity works smoothly between them. EMR designers [need] to design systems that trigger alerts in a way that’s helpful to the practice and make it easy for physicians to set up. So there are a lot of other entities that need to be engaged, so its health plans, mail order, retail pharmacies, pharmacy benefit managers, and EMR vendors. So the point that we ended with in the issue brief was that in many ways e-prescribing is a more advanced IT functionality in terms of interoperability and in terms of health information exchange of data. There are still many challenges. It’s a complex task to setting up that connectivity and that interoperability and the exchange of that data and the quality of that data. It’s sort of a harbinger for the challenges that people face in implementing EHRs more fully because they’re much more complex.