In Conversation With… Michael Cohen, RPh, MS, ScD (hon)
Approach to Improving Safety
Setting of Care
Editor's note: Dr. Cohen is President of the Institute for Safe Medication Practices, a nonprofit organization that operates the voluntary and confidential ISMP Medication Errors Reporting Program. He is also coeditor of the ISMP consumer website, chairperson of the International Medication Safety Network, and a consultant to the Food and Drug Administration. We spoke with him about safety in the retail pharmacy
Dr. Robert M. Wachter: Tell us a little bit about your interest in patient safety in the community pharmacy. Where did that come from and what are your big-picture impressions?.
Dr. Michael Cohen: We get quite a few error reports to our national Medication Errors Reporting Program from community pharmacists. We also have a consumer website and a Consumer Medication Errors Reporting Program, so we hear from consumers as well. For years, I wrote a weekly blog for the Philadelphia Inquirer. All of it was meant for consumers, and I shared a lot of information on dispensing errors and problems that patients had with medications at home.
RW: In the old days, when your doctor gave you a prescription on paper or the hospital faxed a prescription to your local pharmacy, what were the patient safety problems that you saw?
MC: Well, handwriting and lookalike drug names were a big problem. With no clinical decision support at the doctor end, sometimes medications would be ordered and patients were on other medications, and in many cases drug interactions weren't captured at either the doctor end or the pharmacy end. Although the paper era had a lot of negatives, there were some good points too. Because the patient usually would have something in their hand, in most cases at least they could read the name of the drug and know what to expect when they got to the pharmacy. We don't necessarily do that in the digital age. We don't always give the patient a printout. A lot of it goes electronically right to the pharmacy queue, and the patient doesn't know what to expect when they get there.
RW: Did computerization seem to be the main way we would get ourselves out of some of the messes that you were seeing with paper?
MC: Well, it was a long time before we even thought of electronic prescribing. They had computerized processing even back in the 1980s and 1990s, but electronic prescribing lagged. Once we knew that could be accomplished, it was a very hopeful period. We thought for sure that many errors that might have happened in the past wouldn't happen. It wasn't just electronic prescribing. It was also other technologies like barcoding. Almost every pharmacy uses barcoding and receives prescriptions electronically today, so we don't have the same problems that we had years ago. Unfortunately, we do sometimes have people choosing the wrong item off a drop-down list, particularly when you have two drug names that look alike and have a similar strength.
RW: What were the biggest surprises that came with computerizing the process? Did other safety concerns emerge that you hadn't expected?
MC: We were very hopeful about electronic prescribing, and I still believe it has helped quite a bit. But the medication safety problem hasn't gone away, obviously. One issue is we still don't prescribe by the patient's diagnosis, known as indication-based prescribing. Lookalike drug names or soundalike drug names that are used for widely different purposes could have easily been picked up if the doctor simply would have written the indication on the prescription; the pharmacist would see that, and the patient would have the prescription in their hand and know what the medication was for. Then with a little bit of conversation between the pharmacist and the patient, a lot of errors that we saw in the past would not have happened.
RW: How has the life of a pharmacist in a community pharmacy changed over the last 10 years?
MC: They've gotten busier, and the staffing hasn't always kept up with the increase in the number of prescriptions. We've gotten to the point now where we're able to maybe not cure but suppress leukemia with oral therapies, and there are so many more drugs on the market today. We're seeing the advent of the monoclonal antibodies being available in community pharmacies or sometimes specialty pharmacies. And the numbers of prescriptions have gone up dramatically. We're now into billions of prescriptions being filled, something like 4 billion prescriptions each year. At the same time, the design of the pharmacy has changed—we have drive-up windows and new expectations by patients who are in the car (even, "Can you get me some diapers?"). Things like that are a problem today.
RW: Has automation helped their workflow and workload, or not quite as much as we would have expected?
MC: I think computer processing can help, because with refills, for example, you don't have to look through thousands of prescriptions on a spindle that were kept in a file somewhere, and you matched the prescription container prescription number with the proper prescription. That would take some time to do that with refills in particular. Now it's just a few seconds to call up that medication on the computer, so it's a lot faster. It has helped to some extent, but it's negated to a large extent by the increase in the number of prescriptions being filled.
RW: I have been told that pharmacies run by some of the big chains have completely automated the process. Is that a threat? Is that opportunity? Can you imagine that happening?
MC: Well, one of the things I'd like to talk about is deployment of pharmacists right now in the United States. It's poorly done. Maybe things are a little different in California than in Pennsylvania, but it's a pretty rare thing to see the pharmacist literally dispense the medication and speak to the patient directly. At ISMP, we work with PharmD students, and they're pretty well trained when they graduate. They then go to a community pharmacy where many of them have to work to pay off their loans for several years. They wind up behind that counter, and all they're doing all day long is dispensing prescriptions, taking doctor and pharmacy calls, the people driving up to the window, and a lot of interruptions. It's amazing there aren't more errors than there are.
I would much rather see well-trained technicians, whatever it takes—if it's a 2-year or 3-year program or a degree program of some type to train technicians to do that type of legwork and have the pharmacists out front at a desk. They look over every prescription before it's dispensed to the patient. They could compare it to the electronic prescription and it could come out in a basket with the medication bottle that it came from, then they could talk to the patient. In some states there is mandatory counseling. We've done research, we've been in pharmacies where they have mandatory counseling. Although they're doing it, it's little more than "take one tablet three times a day." They're not going through a checklist for high-alert drugs like oral anticoagulants or insulin or whatever they're dispensing that is potentially dangerous.
RW: California may be one of those states, because in my local pharmacy the pharmacist pops over and asks if you have any questions. It feels pretty rote, pro forma. Whether they have the time to actually do the counseling in a way that would be helpful, I'm not sure.
MC: First of all, what you just said is exactly what happens nationwide. It may not be a pharmacist that comes out and asks, it's a sales clerk or a pharmacy technician. They ask the patient, "Do you have any questions?" They don't have any questions. They don't know what to talk about or what to ask, and they're dismissed. That's not the question that should be asked. There should be mandatory movement of the pharmacist, when patients are on certain medications you want to make sure that they are educated about them. Can you imagine walking out of a pharmacy the first time you got a fentanyl patch? We have some really dangerous errors that happen. Or with methotrexate for a patient with psoriasis or rheumatoid arthritis, and the doctor accidentally prescribed it daily instead of weekly? Somebody needs to pick that up. A lot of that can happen at the counseling stage, where there's some communication between the pharmacist and the patient.
RW: So, you can imagine a pharmacy where the lower level, more logistical tasks are done by lesser trained people and/or technology, and the more highly trained pharmacist is freed up to counsel and to double check things that are particularly risky. Do you know of pharmacies that look like that today?
MC: Walgreens did try to do that. I'm not sure where they are; they may still have some of those pharmacies running. But they got a lot of pushback from state boards concerned about pharmacy technicians not having a pharmacist literally looking over their shoulder—the check was done at the desk. And they stopped it. But that is something we as pharmacists need to be getting very serious about. We need to be looking into ways to improve pharmacist activities.
RW: Can you talk about how ISMP works vis-à-vis community pharmacies? Let's say there's a bad prescription error or a pharmacist sees something that they think is error prone. How do they get to you? How do you get that to the FDA [Food and Drug Administration]?
MC: We have the Medication Errors Reporting Program. We also have a Vaccine Errors Reporting Program. (One of the big public health pushes in pharmacy has been vaccinations. Close to 25% of the vaccinations given in the United States are done by pharmacists.) If pharmacists are involved in an error or even if they have a concern about something that might lead to an error, usually that is a regulated product that looks like something else or has an ambiguous label or whatever, they can contact our program. We publish in multiple journals, so they're used to seeing our column. For example, the Journal of the American Pharmacists Association has an ISMP column. We're in Pharmacy Times. We have a community pharmacy newsletter, and more than 10,000 pharmacies subscribe to that. They can contact us about a problem. They can go to our website and tap on "Error Reporting" at the top and it guides you through. We receive many every day. We are very serious about them. They get an automatic acknowledgement, but in many cases we follow up with the pharmacist directly to learn more information. It's not a particularly difficult form where you must fill out multiple fields. We just want your story; it's a narrative. Where appropriate, we interact with the pharmaceutical manufacturer, The Joint Commission, and the FDA daily.
For example, we had a surge in mix-ups between lamotrigine and labetalol. We've been interacting about and trying to find out the reason for that. It will result in an article in our newsletter. We've been working with FDA directly since the 1980s. Although we didn't form as a nonprofit until 1994, our reporting program started in 1975 as a journal column. And FDA caught on around the mid-1980s, and it has resulted in a pharmacist or a doctor or nurse that contacts us, they can be sure that FDA gets a copy of every report. They're allowed to select whether they want to be identified or not. It has resulted in many hundreds if not thousands of changes in drug names, labeling, and packaging. If you look at The Joint Commission, there are a lot of practice issues, high-alert drugs, and abbreviations that shouldn't be used. Some of these have wound up as National Patient Safety Goals. A lot of good has come from someone's willingness to turn information around through our channels.
RW: Most observers in the safety field see your organization as a successful model of reporting. It has made a difference in getting the information out to people and connecting with regulators and accreditors. But they would see other efforts at reporting in patient safety as not being as happy a story. Do you agree with that, and if so what do you think was different: Is it the organization or is it something about medication safety that's different?
MC: I do agree with you by and large. Although we are a Patient Safety Organization, we are not gathering thousands of reports in a database and having the hospital (or pharmacists for that matter) fill out a bunch of fields. We just get the story. If you go to our website, it's very easy—just tell us what happened. All these people want really is acknowledgement that you sent them something and that you changed something that they thought was really serious. You followed up, published it in your newsletter or in the journals, and years later they look back and see several things that they may have reported or a colleague has and it has been taken care of. The focus is on the narrative and two-way communication. We contact a lot of people on a daily basis, and it's more about quality of the report than quantity. I went to the IOM committee that was publishing To Err Is Human, and they asked us how many reports we get a year—this was in 1999. We said we're getting about maybe 1100 reports a year. They looked at me and they said, "Are you serious. That's all you get?" In other words, how important could our program be?
A large database is important too. We're very fortunate that we have groups like ECRI, Pennsylvania Patient Safety Reporting System, and some others that we've been able to interact with that have given us that, and FDA, the MedWatch program. When I need reports regarding how many of this happened, they do that for me. That helps me to make a case for the need for change. But the individual provider who has shared the narrative, that's where it really starts. I was a clinician for 28 years, and almost my entire career was spent on the nursing unit side. I started at Temple University Hospital and wound up at a community hospital as a director, and it was on the third floor outside the ICU. As clinically experienced pharmacists and nurses, along with our medical director, an anesthesiologist at Children's Hospital in Philadelphia—we understand. We have walked in their shoes, and that's what we try to communicate. That has helped to make this program so successful.
RW: It seems like there are fewer and fewer Mom-and-Pop community pharmacies and more international and national chains. Is that accurate and do you think is that a good thing or a bad thing?
MC: Oh, it's definitely accurate. The Mom-and-Pop operations just aren't there anymore. I worry to some extent about the large chains. I don't care what they say or what their public relations people tell you—again and again we've seen how many prescriptions an individual pharmacist is filling, how quickly. There was one large chain that had a 15-minute promise. If they didn't fill your prescriptions within 15 minutes, they would give you a coupon that you could use elsewhere in the store to buy something.
Right now, every pharmacist that I've talked to that works in community confirms that if they don't administer so many vaccines per period of time, it affects their bonus, it can affect their salary, it can even affect their job. That's concerning because you know what speed does in a situation like that, and the time pressures that people are under. And I've yet to see supervisors that are trained that go into a pharmacy and say "Have you read the FDA MedWatch report? Have you read the ISMP report? What are you doing to assure that methotrexate isn't given out to a patient on a daily basis when they have psoriasis? And what is the pharmacy chain doing to make sure that something like that doesn't happen?" We could stand some improvement in that area.
RW: If you were going to design a retail pharmacy of the future, what would it look like?
MC: For one thing, I'd like to see pharmacists paid in a different way. Right now, they're paid for how many prescriptions they fill and how quickly they can get it to the patient rather than how many patients they educate or phone calls made to doctors where they feel something is unsafe and they get the doctor to change it. You could work for half an hour on a problem and you're not going to get reimbursed one cent. Clinical efforts should be mandated for high-alert drugs. We have a list of high-alert drugs for community pharmacy on our website just like we do for acute care. We also have one for long-term care.
Years ago, we had a grant from AHRQ, and we put together checklists for community pharmacists to use in dispensing medications so that they would not overlook something that was important, for example, when you dispense an opioid if that's what you have the prescription for. They've worked very well for some pharmacists. They come back and tell us that they really like using them and they've reordered them—there's no charge, you just print them out from the website. We'd like to see those deployed.
There are some other issues. For example, safety culture. Remember the Eric Cropp case? Fortunately, we haven't had any more arrests and prison sentences, but we still have punishment doled out to pharmacists who made an error—an error that didn't involve reckless behavior. They simply tripped up on something that was pretty well known in the past. It was not as much their fault as it was the pharmacy's fault or the supervisor's fault for not addressing the issue to make it much more unlikely or even impossible to happen. We still have that problem, and the Boards don't necessarily send their surveyors out to look at what the pharmacists are doing to address major errors that have been reported to FDA or ISMP. That could make things quite a bit safer. That's what I would do if I was redesigning. I probably would want to see a medication safety officer on the Board of Pharmacy in each state. These are people that eat, live, and breathe safety; follow the literature; and could help to guide the Boards of Pharmacy on what to look for when they send their surveyors out.
RW: Interestingly, you answered that question with not a mention of robots or artificial intelligence.
MC: No, we have that already. Robotics have been around for quite some time. The most common drugs are loaded into a dispensing machine, but it's automated and it puts a label on the prescription container. The pills, tablets, capsules are in a cell. There can be hundreds in a pharmacy depending on how large you want it to be. These are used in very busy pharmacies. When that prescription is being filled, a signal is sent to the dispensing equipment and it literally prepares the vial ready for dispensing.
Another thing, in research published in the Journal of the American Pharmacists Association, we did an ST-PRA [sociotechnical probabilistic risk assessment] project where we met with groups of pharmacists and technicians seven different times with 10–12 different people each time. We looked specifically at high-alert drugs. One of the most frequent types of errors was giving a patient a bag with someone else's medication. At least one container was for another patient, even though the name on the outside of the bag was correct. So, patients would go home, see their name on the outside of their bag, never look at the label itself, and start taking someone else's medication. We were able to show that something as simple as telling the patient to open the bag and look at the label before they leave the pharmacy immediately cut down on that type of error by more than 50%.
RW: I'm wondering about a future in which somebody prescribes a medication and there's an automated review of the chart around indication. Because it sounds like you believe that indication-based prescribing would help a little bit, but for whatever reason getting clinicians to write the indication on the prescription is too much. But you could probably have a computer figure out whether the patient has a given disease from their medical record.
MC: Did you ever see Gordy Schiff's indication-based model? For example, if your patient has diabetes it gives you a list of drugs that would be appropriate for a patient with diabetes. If you tried to prescribe something else, "We don't have this listed for diabetes. Do you want to add this to the list?" Then immediately they realize they've selected the wrong drug off the drop-down list. He had something in New England Journal of Medicine. There are many different conditions where this would work. It made sense, and I hope they continue to pursue that.
RW: You talked a little about culture in terms of punitive culture, you talked about production pressures. You talked about the importance of people being able to speak up. Any other cultural issues that you think are important in the community pharmacy setting that need to be addressed?
MC: A lot of pharmacies have automatic prescription renewal. In many cases, they have the patient sign a form that they want their chronic medication renewed. The issue right now is most people have trouble with electronically canceling the prescription. There is a software version, and many vendors are making it available now for the first time, called CancelRx, but that doesn't happen by and large. A lot of times automatic renewal can lead to duplication. Say you're taking metoprolol 50 mg and you go to a cardiologist and they decide to use a different beta-blocker. No one has ever discontinued the metoprolol, and the cardiologist may not know that the patient is on that. They prescribe it, and now the patient is taking both and the pharmacy keeps renewing it. Automated renewal systems are a major issue.
Previous interviews can be heard by subscribing to the Podcast