Study of opioid epidemic provides new method for thinking about long-term therapy

Carrying implications not just for clinical prescribing practices, but also larger pharmacoepidemiology trends, Dr. Hilary Mosher’s recent research using data from the Veterans Health Administration (VHA) shows a decrease from 2004 to 2011 in the proportion of new users of opioids who begin long-term opioid therapy. The paper, published in April in the journal Pain Medicine, may indicate the success of recent efforts in the VHA to lessen high-risk prescription of opioids.

Dr. Hilary Mosher

The team’s research builds on previous work done by Dr. Mosher, Clinical Assistant Professor in the Division of General Internal Medicine, and her colleague Dr. Brian Lund, Clinical Assistant Professor in the College of Public Health – Department of Epidemiology and investigator with the VA Comprehensive Access & Delivery Research & Evaluation (CADRE). Previously, their research team looked at prevalence and incidence of drug use. “We then wanted to ask that question of how people start becoming long-term users,” said Dr. Mosher of the impetus for their latest project. “Once people have been on opioids for a year or so, they really come off them at fairly low rates.”

Dr. Mosher’s publication joins a growing conversation (in both the scientific community and the popular press) surrounding the use and overuse of opioids in particular demographics. Dr. Mosher said, “There is some sense that the pendulum has been swinging very quickly over the last few decades on pain and on the appropriateness of opioids. And so now we’re in a situation where we have no evidence that these are effective long-term medications for non-cancer pain for the vast majority of people.”

Asking and answering meaningful research questions in the face of such rapidly changing attitudes and assumptions proved to be one of the more challenging, and therefore interesting, aspects of Dr. Mosher’s research. “I think it takes a lot of creativity and awareness to try to think of what are meaningful ways that we can tease out what might be effective interventions, rather than just ride this wave of social changes that really are, in some ways, discourses about the nature of suffering and expectations to be free or not free from pain,” said Dr. Mosher.

The Cabinet Supply Approach


The issue that Dr. Mosher and Dr. Lund have tackled is in the differentiation and characterization of short-term and long-term use. Using a “cabinet supply approach” which is novel in the field of pain medicine, the team’s research has taken a step in the right direction, identifying a large group of “Intermediate” users that fall in between short-term and long-term use.

Prescription form close-up

The method uses VHA data to establish an index date on which the prescription was made, and a cabinet supply value is assigned based on how many days’ worth of opioids were prescribed. Every day thereafter, one day’s worth of the prescription is subtracted until the cabinet supply reaches zero. That number stays at zero until the prescription is refilled. The number of non-zero cabinet supply days reflects the number of days of opioid receipt. Through all of this, researchers can gain a description of opioid usage longitudinally, over a span of time at the individual level, rather than a single aggregate number that may not reflect the true nature of a patient’s opioid use.

“It all comes down to how you decide to build that history,” said Dr. Lund. “Since I’m a pharmacist and an epidemiologist I think, ‘What is in that bottle in the patient’s cabinet at the time?’ In trying to characterize a patient’s medication use over time, you’re trying to figure out how many pills they have in their cabinet, and then making some judgments about whether time spans between refills represent continuous therapy or if there’s a break in treatment.” Through translation into patterns of prescription use, such data can ideally be used to make changes on the front end of clinical practice to prevent harmful long-term opioid therapy.

Responses and Next Steps

So far, the paper has garnered favorable commentary from Dr. William C. Becker of the Yale University School of Medicine. Dr. Becker writes in an editorial in the same issue of Pain Medicine that the methodology “should be useful in future quality assessments and studies explicitly designed to measure the impact of safety-oriented interventions and policy changes.” Both Dr. Becker and Dr. Lund note the efficacy of the method in building histories of overlapping prescription drug use.

One area in which Dr. Mosher intends to pursue further research is in the differences in opioid use between rural and urban areas. To her, the service UIHC provides to surrounding rural areas is partly why Iowa is an ideal location to conduct this type of research. “The population is different, the resources are different, the solutions are going to be different,” Dr. Mosher said. “I think that if we look at this as a national public health issue, then we have to have people working focused both on the nature of this problem in rural America as well as the nature of the problem in New York, San Francisco, Los Angeles, and Seattle.”’

“Zero opioids is not the answer. The level of opioids we have now is not the answer,” said Dr. Mosher. “So, where is the safe range?” Though Dr. Mosher recognizes there is still much to be known about the development of problematic long-term opioid therapy, the cabinet supply approach developed and put into practice in Dr. Mosher and Dr. Lund’s paper should prove a useful tool for answering these questions.

Leave a Reply