Evelyn Corsini of the PainEDU team interviews John Poikonen, Medication Safety Pharmacist in the Office of Patient Safety at Partners HealthCare System in Boston. He has spent 20 years in clinical applications and pharmacy informatics.

--------------------------------------------------------------------------------------------------------------------------

Questions

--------------------------------------------------------------------------------------------------------------------------

Evelyn Corsini:

Is the overuse of opioids really a problem in the US?

John Poikonen, Pharm.D.: When you look at the evidence based practices for prescribing opioids, and then look at the prescriptions that are written, I suspect you would see that the use of these medications follows a bell shaped curve. That is, on one end of the curve there are a few standard deviations of numbers of patients who are under-prescribed, and at the other end of the curve there are a certain number of patients who are over-prescribed. When I was a consultant for a state Medicaid program we searched their database for patients who were given prescriptions for 5 or more narcotics by 5 or more different prescribers. We found a number of these people. There is no rationale behind this kind of overuse.

EC: What is DUR or drug utilization review?

JP: DUR stands for drug utilization review. There are two kinds of DUR, retrospective DUR and concurrent DUR. Most people are familiar with the retrospective reviews because for many years they have been undertaken electronically by organizations like health insurance companies, health care providers, and pharmacy benefit managers. They are required of all state Medicaid programs.

EC: How are retrospective DURs used?

JP: The goal of these reviews is to reduce the rate of potential prescribing errors and gather information that will lead to more cost effective and better clinical outcomes. Personally, I believe that for many reasons, the use of retrospective DUR has not met these goals. The “exceptions” (that is the deviation from criteria on the use of the drug) are reviewed in retrospect to determine if a physician “alert” should be issued about the prescription. This may be by mail or telephone. Pharmacy benefit managers have effectively used this tool to change prescriptions to lower cost alternatives and generics. However, there is not a lot of good clinical evidence that retrospective DUR has improved outcomes. In addition, these reviews focus heavily on inappropriate prescribing or over-prescribing of drugs, but not under-prescribing which can be an equally serious issue. Concurrent DURs pose a different set of issues.

EC: How are concurrent DURs used?

JP: Concurrent DUR is an electronic system that typically checks for problems such as drug allergy, drug-drug interaction problems, high dosage, or duplications, at the time the prescription is being filled. The goal is to prevent adverse events from happening. Unfortunately, using the technology that is currently available, concurrent DUR is notorious for “false positives”. These are electronic alerts which require the user to stop the flow of work to get the drug to the patient, but do not actually provide valuable information. For example, the prescriber might receive an alert with material that they already know. Or the alert may need more specific patient information to be entered into the system. And worst yet, because of so many false positive alerts clinicians ignore all of the alerts, due to a ‘cry wolf’ syndrome. The end result is that many of the alerts are considered an annoyance and are just dismissed. I do not believe there is good evidence that concurrent DUR, especially as it has been used in pharmacy systems and pharmacy benefit management, has achieved its goal of better patient outcomes and error prevention.

EC: What efforts are being made to improve the prescribing processes?

JP: Throughout my career I have been heavily immersed in technology and I believe technology has a role in fixing these programs. This is far from easy. First, it requires a universal electronic prescribing system. There are many initiatives and pilot programs around the country trying to do this, with organizations and communities putting in the necessary infrastructure. At a national level, the CMS (Centers for Medicare and Medicaid Services) consider this an important priority. A national electronic prescribing system would be a logical first step in developing a system of national electronic medical records. The ideal of national electronic medical records has a great deal of support in this country. Second, it requires clinical decision support that has lower false positives and greater specificity than those gotten ‘out of the box’. I am currently involved at Partners Healthcare, in developing clinical decision support programs.

EC: What is a Clinical Decision Support program?

JP: Concurrent DURs have traditionally been canned messages or alerts that are provided by vendors. Right now what is available is one size fits all. Clinical Decision Support to be effective is much more specific and customized to give messages and alerts during the prescribing process. It has to work in concert with applications or electronic prescribing systems that have to be easy to use if it is going to be successful. This takes a great deal of special expertise to accomplish. It takes a lot of work to get clinical alerts to be effective in preventing errors. Here at Partners we have a whole division working on drug decision support to provide high quality messaging at the time of prescribing.

EC: Can you give an example of electronic messages that you are already using at Partners?

JP: Currently we have a system in place to generate alerts when opioids are prescribed for two specific populations of patients: the elderly and those with renal problems. For renal patients, there are known excretion issues with some opioids and these should be avoided. For the elderly, there are some drugs that just should not be used in this population. We will present this information to the physician, with recommendations for switching to a better drug choice.

EC: What do you predict will be the future of Clinical Decision Support programs?

JP: I think we are definitely moving in the direction of implementation of more of these programs. This will not be the cure-all for problems with prescribing drugs, but it will be a major step forward that should shorten the two ends of the bell shaped curve to assure that more patients get the proper drug, in the proper dose, at the proper time.