LJ: You’re listening to “Leading Voices in Health Care Policy”, a podcast brought to you by the department of health care policy at Harvard Medical School. I’m Lauren Jett. Today we’re speaking with Dr. Nicole Maestas, an associate professor of health care policy at Harvard Medical School. Thank you, Dr. Maestas, for speaking with us today.
NM: You’re welcome.
LJ: Research shows that a lot of people that are addicted to opioids are getting them legally through prescriptions from their physicians. What’s a condition that a doctor might prescribe an opioid for?
NM: Well back pain might be a very common condition for which a doctor would prescribe an opioid. Not all kinds of back pain, but certainly more severe forms of back pain. Arthritis would be another one, other forms of muscle pain. You often find that musculoskeletal conditions are most common reasons for an opioid.
LJ: Your recent study in Annals of Internal Medicine shows that these prescriptions aren’t always written for conditions like this. Some people are getting opioid prescriptions without any documented pain. Do undocumented conditions make up a lot of opioid prescriptions?
NM: We were surprised by just how many are undocumented. For example, we found that about 5% of opioid prescriptions were for cancer, cancer pain. About 66% were for non-cancer pain, and then there was this twenty eight and a half percent for which there was no identifiable pain condition documented on the record. By pain condition I mean not only a condition that might be described as pain, but also conditions that are known to cause pain. That is pain that is severe enough that a doctor might need to prescribe prescription pain relief.
LJ: Why are there so many undocumented prescriptions? Does it have to do with the way this data is actually documented?
NM: Well I think that’s the big question here of why are there so many undocumented prescriptions? Now we analyzed this surveyed dataset called the National Ambulatory Medical Care survey and this is a survey, it’s not electronic health record data, it’s not claims data. Because it’s a survey it often collects additional information that you might not find in a medical claim. So it certainly is possible that if we looked across different types of data collection vehicles- health records, claims, survey data- we might see some differences here. We did some tests of robustness to make sure that the way the survey form was recording information wasn’t writing results, that didn’t appear to be the case at all, which then leads us to the bigger questions which is: what’s going on here? Why are these not documented? We think that that many of these likely are justifiable prescriptions for justifiable conditions, so this is not to say that these are all inappropriate. But the point we’re making in this paper is that we should be able to determine which opioid prescriptions are appropriate and which are inappropriate. In this instance, without the documentation you can’t even begin to look at that question.
LJ: So some of these patients may have a valid pain condition, but they’re just not getting documented. What would you suggest they do to increase the number of physicians that report this information?
NM: Well I think that we need somewhere to require that when an opioid is prescribed that an appropriate documentation it entered into the record. Many times insurers will require that this type of information is included in requests for novel drugs. Usually this is driven by cost considerations rather than public health considerations. But it certainly could be done by insurers. It could be done by practices- a practice could adopt a policy that opioid prescriptions will be documented and justified. The thing that we want to be careful about is that we have to recognize that physicians already have extraordinary demands on their time for record keeping, documenting, and compliance activities. So by bringing attention to the problem I don’t think we’ve solved in a sense where that we can say how you do it, I think that there are some considerations to think deeply about. How can we make it easier for physicians to document the indication when they give an opioid prescription without being overly burdensome?
LJ: Are opioids the only place we see a lack of condition documentation?
MN: We only examined opioids. Opioids are certainly not the only type of prescription where there are concerns about inappropriate prescribing. Antibiotics is an area where people have used data just like this to detect inappropriate and appropriate prescribing of antibiotics as a way again to serve the public health by bringing attention to antibiotics prescribing that doesn’t comply with guidelines. You could also look at other kinds of benzodiazepines for example as another kind of drug that you might worry about lack of documentation and inappropriate prescribing. Our analysis is a framework that could be used to look across other kinds of conditions and examine indication rates and lack thereof.
LJ: You’ve touched on this a little bit, but why is it so important for physicians to report what they’re prescribing and why they’re prescribing it?
NM: Because in the US, the use of opioid therapy has risen far faster than the instance of pain. That really brings us all to ask “Why? Why are people being given opioids more and more if pain is not any different than it used to be?” This raises questions about if we’re overprescribing opioids or even erroneously prescribing opioids. A first step there is to be able to identify it when it’s actually happening. That’s when we uncovered this limitation that when the clinical rationale is not documented you really can’t say whether it was okay or not okay, appropriate or not appropriate. If you could identify which were appropriate and which are inappropriate you could detect what kind of errors are being made by physicians and in principle deliver feedback about those kinds of common errors. I think it’s an important part of opioid control policy that hasn’t really been emphasized much to date. On that note, I wanted to point out that most of our policy efforts to date have been quantity-focused. That is we ask “Is this particular patient obtaining too many opioid prescriptions? Is this doctor providing too many opioid prescriptions overall?” But this is really more of a question of quality. Was the prescription appropriate or inappropriate? Doctors need to ask “Does this person, who might be receiving an opioid for the first time, have a medical indication that’s appropriate to treat with an opioid, or is there some other therapy that could be tried first?” For someone who has been started on opioids, would it be appropriate to continue that patient given this medical indication on opioid therapy if they came and asked for more? I think that this study sheds light on another type of opioid control policy that isn’t just about looking for behavior in the extreme but asking about the average person who goes into a doctor’s office experiencing pain. Do they or do they not get an opioid prescription? And if they do, should they have?
LJ: Do you think that more thorough documentation could help curb the opioid epidemic?
NM: It’s a really good question. The opioid epidemic of course is a multifaceted epidemic. We have excessive prescribing by legitimate doctors to legitimate patients all the way to people obtaining street drugs, heroin, and winding up overdosing and even dying. Often when we talk about the epidemic we’re talking about overdose deaths. Do I think requiring physicians to document what they’re prescribing and why would stop overdose deaths? Probably not. Many things happen between those first legitimate medical prescriptions on that path to an overdose death. But everybody starts somewhere, and even though elicit use of fentanyl as an important driver that you hear about, it is the case that many people that use and abuse these drugs obtained illegally began with prescription medication. In many ways dialing the practice back to its beginnings- starts with doctors and to whom and for what they give pain medication for. I think in the longer run arc of the epidemic public policies or greater efforts by physicians to justify their prescribing could help stem the tide of future use and misuse.
LJ: Thank you so much, Dr. Maestas, for joining us today and discussing opioid prescriptions. You can read more about Dr. Maestas’ research on the department of health care policy’s website at hcp.med.harvard.edu.
That’s all for this episode of Leading Voices in Health Care Policy. From the department of health care policy at Harvard Medical School, I’m Lauren Jett.
For more of the top news and updates on health care policy, be sure to follow us on twitter at HMSHCP.