To Curb Prescription Opioid Abuse, Physicians Need to Get with the Program
Jeanmarie Perrone, MD
What’s the “big picture” currently when it comes to prescription drug monitoring programs (PDMP) around the country? There’s a large disparity between states in terms of what’s been funded, what’s been implemented, and how they’re being used.
They are a heterogeneous group of state programs. I’ve heard about some very functional programs from the physicians who use them and from the lawmakers in those states. Some programs are up and running and are as effective as we have hoped they could be. Then there are other programs such as the one in Pennsylvania, which is where I live. Whenever one of the PDMP organizations create those maps showing which states have functional programs, they always depict Pennsylvania as having one. [eg, this map http://bit.ly/ORKt0k from the National Alliance for Model State Drug Laws, or this one http://bit.ly/MVHAfa from the Alliance of States with Prescription Monitoring Programs] We do have a program in Pennsylvania, but it is not accessible by physicians or pharmacists. Only law enforcement can access it, and they don’t even really use it. If you think about the way a PDMP is supposed to work, the Pennsylvania program doesn’t fulfill any of those functions, and yet it gets counted as one of the states that have an operational program. If you look at our drug overdose death rate, you’d say that PDMPs don’t work because Pennsylvania’s drug overdose rate is quite high and hasn’t changed at all. But, in reality, if you look closer, maybe it’s because our PDMP doesn’t exist.
The programs are all designed very differently, with some uniting factors. Most states contract out their program design, and several states that have done that have used Health Information Design (HID; http://bit.ly/Ml0dpu). A lot of the programs that are run by HID are similar—they’re online, they’re pretty up-to-date, and they enable efficient access to the program and the information it provides. The states that are using that kind of program are able to provide pretty prompt function for providers.
Many states are using PDMPs that are actually working very well. There are a couple of things that could be adapted to make it a little bit easier, which is what we tried to point out in the NEJM article (http://bit.ly/OX8w9N), such as making it simpler to log in to the program. I recently spoke with people from Maine who said their PDMP has been up and running for a couple of years. Not only does it require an elaborate log-on process, but at one point the program just kicked them all off. Now they have to reapply for access. With busy physicians, it’s hard to get people to use a program again if they’ve already been through this process and found it wasn’t helpful, or they were kicked off the system, or they were never able to find any valuable information.
Do any state programs stand out above the others in terms of functionality or user feedback?
I’ve spoken on this topic a lot around the country in the past year and, anecdotally, I’ve heard about several successful programs. I know North Carolina has done a nice job with its program, Colorado, too. I’ve spoken to a lot of physicians in Florida who are very happy with the program there. I’ve had people from Minnesota come up to me and say, “Our program works great. Whenever I look someone up, I make a much better prescribing decision because of it.” It’s been encouraging to hear feedback from these users.
–Jeanmarie Perrone, MD
Would you say ease of use is a major obstacle to adoption and implementation?
I believe that ease of use is a big problem, but other people have said that actually computer availability is a bigger issue. For me, working in an emergency department, with computer access so readily available, I don’t think this is as much an issue as it might have been in the past. I think it basically comes down to having enough time and knowledge to access the system. If you can make it easy for physicians to log on and access the information, then I think people will be much more willing to use it.
If the log-on process is too cumbersome, people won’t use it?
My co-author and I went to the Harold Rogers Prescription Drug Monitoring Program meeting in DC, about a week after the article came out. The national lawmakers were convening to talk about how to streamline and coordinate research efforts among multiple states with PDMPs. We went there to give them the clinician’s perspective on why we do or don’t use the programs. My colleague made a slide showing every step in the process to access a program. It had a picture of about 10 different computer screens that a doctor who didn’t know the PDMP site URL would have to go through in order to gain access to the PDMP site. If the doctor also forgot his log-on, or once he gained access was told by the program “your log-in has expired, make a new log-in,” you can see why he would be discouraged. The audience thought it was amazingly illustrative to see it spelled out like that. Ten or 12 steps is too much. You have to be overly committed to think that somebody is going to go through that many steps to maybe gain some information, especially in an emergency department.
In a primary care office you can designate an admin or nurse or someone else to look people up in the system. If you are in a pain management office and you are always dealing with this issue, you could pull the patient’s chart and have a designee check the program when a person is scheduled to come in. But in emergency departments this is a big issue because we never know our patients. We have no history; they’re all strangers to us. They come in and say they lost their medicine, or they need a refill, or they’re traveling from out of town. We’re trying to make it easier to get our prescribing a little bit more under control. But for a program to require that amount of time, energy, and effort to access is asking a lot of the average ER doctor.
What is the evidence so far that these programs are helping to curb abuse and diversion?
Very limited data have been published. There is an article by Paulozzi and colleagues (http://1.usa.gov/NQjDW8), published in 2011, that looked at the number of prescriptions for opioids written and at the rate of overdose deaths in states that had a PDMP. They used data from 1999 to 2005, so it was old data. They found that PDMPs did not have much of an impact on overdose death rates. However, I think that if you were to repeat that study now, using the same endpoints, you would actually find that the programs are making a difference. There are also a couple of articles that have been published more recently that have found a little bit of a difference. In one article on Ohio’s PDMP, the authors looked at whether physicians with access to the program were changing their prescription habits (http://bit.ly/PUZDif). The authors found that physicians who reviewed PDMP data changed their prescription plan about 40% of the time.
Many of the states with PDMPs are going to publish data in the next few months showing that their overdose endpoints are actually going down now that more people are enrolled in the PDMP. The programs are working. Not all of them, but maybe half of the programs are working really well.
In addition to ease of use and access, what are some of the most important characteristics of an ideal PDMP?
In the NEJM article, we used the example of Oklahoma’s PDMP, which has instantaneous access—if you prescribed something two hours earlier, it would already be in the database because once a prescription is filled at a pharmacy in Oklahoma, it’s immediately scanned into the database. No other state has anything even close to that; most of the other states are one week to one month behind. The question is whether that time lag is good enough. Probably one month isn’t, but with a one-week delay you’re still getting a pretty current view of that patient’s history.
Simplifying the log-on and password process would be also very important. Another is the need for these programs to be able to communicate and collaborate on a research basis so that we can better study the impact of these things. Other considerations include what is the critical number of physicians who need to be using the program, and can we track that? Can we track how many of the physicians are emergency doctors, primary care doctors, pain management specialists, etc? Collecting demographic data in the PDMP so that you can reach out to those physicians who aren’t using it but are writing a lot of prescriptions would be helpful. We didn’t emphasize that so much in the article, but it’s become clearer to us that that would have been a good additional factor that should be addressed when these programs are designed.
New York recently passed a law that requires all physicians to access the state’s PDMP when they prescribe controlled substances (http://on.wsj.com/Pe0JYX). How do you feel about requiring prescribers to use these systems?
I don’t feel strongly about mandatory use. My understanding of the way the New York program is designed is that you also have to log in to the system to write your prescriptions. To me, that is going to cause some under-prescribing in some instances where people probably should be getting their opioids. Or it could lead some doctors to not prescribe opioids at all.
On the other hand, we’ve been quoting that Ohio study over and over again, and trying to make it clear that these programs can actually also help doctors write for more pain medicine. For example, when I use the electronic medical record in the emergency department and find evidence that a patient has had appropriate opioid prescribing in the past, I feel better about writing them another prescription.
Essentially, I’m using the equivalent of a PDMP to feel better about my prescribing decision so I don’t have to say to a patient, “No, my limit in the ER is four pills and that’s all you’re going to get, no matter what.”
What about concerns that PDMP systems could be used to monitor physicians’ prescribing habits and used as a tool against them if they’re being investigated?
I understand physicians are worried. When you first get your DEA number and you start writing prescriptions, you think there’s going to be a lot of scrutiny, and you think that every prescription is going to be analyzed for appropriate use of pain management. Now there’s so much over-prescribing that it’s crazy how little scrutiny there is. I understand people’s concern because I was coming from that same angle when I first started prescribing, but now I don’t know. I think they ought to have that same level of concern over a patient going home and dying because they got too much oxycodone in the setting of asthma or emphysema. I would think that would be a bigger problem than concerns about overprescribing and scrutiny by the DEA. If you’re really worried about scrutiny by the DEA, you’re probably overprescribing. I have to tell you, I’ve had many ER colleagues who’ve had their prescription pad stolen or DEA number stolen but when they reported it to the DEA they received absolutely no response. I’m not sure that the DEA is as proactive as maybe I thought when I first got my DEA number. I think I now have a more enlightened view that the DEA is kind of overworked. Plus, they say over and over again that’s not the goal of these programs.
Some legislators who voted against PDMP bills claimed they did so because they were worried about patient privacy and confidentiality. Many patients have also expressed concerns about having their prescription information stored and tracked. How do you weigh those concerns against the public health need to curb overuse and diversion?
I’m a public health practitioner, so I would have to say that these privacy worries do not weigh highly on my list of concerns. If a physician is over-prescribing or an individual is over-receiving medications and is going to be scrutinized, I guess they probably are worried because they already have a problem. I can’t think through a scenario where appropriate prescribing will lead somebody to inappropriate scrutiny. I think we’re giving the government too much credit.
On the other hand, if you’re somebody with a high-profile job and are getting prescriptions for Percocet, if that were to become public information I can understand that there are difficult privacy issues that could come into play. I don’t know the way around that. That’s one of the things that of course we have to be protective of. But I think of this as a public health issue more than how it affects any one patient. If a PDMP thwarts one patient from getting too much medication, it’s a good thing. If it causes a high-profile individual to say, “You know what? I think this is bad but I take this all the time, I’m going to stop,” then that’s a good thing. Part of our overall goal is to think about every prescription written and think about every pill taken. It’s part of how we ended up here. Remember, it’s all doctors causing the problem. If we don’t think we need to look at our prescribing, then nothing is going to change.
In the next couple of years, what can we expect to see in terms of the evolution of these programs? What’s going to gain more attention for these programs?
I believe that the interstate exchanges will continue to expand. When people in one state talk to people in another about how happy they are with their program, it creates a certain amount of pressure to build a better system. Eventually that kind of peer pressure—at the governor’s level, at the state level—is going to get us in a position where programs will actually prioritize and increase their capabilities. What motivates politicians is when some incident occurs in their jurisdiction and gets a lot of press, like with Derek Boogaard, the hockey player who was getting multiple prescriptions from team doctors and other people. If we could spin that around and say PDMPs might have prevented this from happening, then that kind of high-profile event could actually lead to real changes, much more so than CDC reports that tell us that overdose deaths continue to climb. Unfortunately, even after some of these high-profile cases, I’ve found that most of the people in my community have no idea what I’m talking about when it comes to the prescription opioid abuse epidemic.
So all of the work to raise awareness of this issue that’s been done over the past few years has had little effect?
I think that’s true. I recently interviewed a scientist whose research is on nicotine and other addiction issues. I had to explain the whole prescription opioid epidemic to him. Here’s a scientist who actually does research on some of these issues and he still didn’t know about it. Even at Penn, I’ve spent a lot of time trying to educate people in the hospital and I still come across oncology pain specialists who say, “I don’t want you to spread the word that these drugs are dangerous or addictive because I see these little old ladies who will only give their husband a half a pill with his cancer metastasizing to his bones because she doesn’t want him to get addicted.”
Which physicians would you target for more education? What message would you convey?
I’ll tell you what I want to educate people about—I want to educate emergency physicians and primary care doctors and get them to understand that the time to be frugal with these drugs is not with the chronic pain patients, but rather with that acute fracture or acute trauma patient. I want them to say, “You broke your wrist. I’m going to give you Motrin, which is a great pain reliever for bone pain. I’m also going to give you some Percocet because you’re going to have times where your wrist kind of throbs. You can take one or two of these but I hope you don’t take any because they’re really addictive and you’re only 20 years-old, and I’m really worried about your future after being exposed to these drugs.” We should be warning the new acute users that we don’t want to turn them into chronic pain patients. That’s what happens to some fraction of people. Maybe they’re genetically predisposed. Maybe they have an addictive personality. But the kids who get their wisdom teeth pulled when they’re 18 and they get 20 Vicodin, how many of them result in experimental drug use after that and what fraction of them go on to have real problems? It just comes down to a little bit of exposure.
My message to primary care physicians would be, “Let’s be more careful with new prescriptions, and let’s promote a public health message that helps everybody understand that these drugs are potentially dangerous; they have a role in treatment but they come with caveats.”