Innovations in End-of-Life Care
an international journal of leaders in end-of-life care

The Cancer Pain Algorithm is a step-by-step decision-tree model developed by the Du Pens and their colleagues at the Swedish Medical Center in Seattle, Washington, for use by physician-nurse oncology teams for treating patients suffering from cancer pain. This algorithm represents one attempt to operationalize the Guidelines for Cancer Pain Management developed by the U.S. Agency for Health Care Policy and Research (AHCPR).¹ This comprehensive pain assessment and evidence-based analgesic algorithm offers clinicians the opportunity to systematize their responses to patients suffering from cancer pain. The Du Pens and their colleagues have embarked on a series of educational interventions and research studies to implement and evaluate the efficacy of the ACHPR guidelines as they are activated through this algorithm. So far, these researchers have tested the algorithm in one randomized, case-controlled study, which demonstrated that the algorithm does enhance patient pain outcomes.² Specifically, the researchers found that the oncology outpatients who were randomized to the treatment group did experience improved "usual" pain scores. In this Phase 1 study, researcher teams of physicians and nurses implemented the algorithm for the treatment group of patients. In a second, as-yet-unpublished phase of this work, the researchers went on to train physician-nurse oncology teams to use the algorithm. In the following edited interview with Innovations, the Du Pens offer an overview of how the algorithm works and discuss what they have learned about the process of designing and implementing this pharmacologic decision-making process, including the challenges and barriers encountered in training others to implement it in outpatient oncology clinics. The Du Pens also discuss what they have learned about barriers to patient adherence to medications recommended by the algorithm, as well as the cross-cultural and international applicability of the algorithm. [Citation: Du Pen A, Du Pen S. Designing and Implementing a Cancer Pain Algorithm: An interview with Anna and Stuart Du Pen, by AL Romer, Innovations in End-of-Life Care, 1999;1(3), www.edc.org/lastacts]

Problems in Practice That Led to Development of the Cancer Pain Algorithm

Stuart Du Pen: Why do we even have a protocol? Why did we set out to do this? It's not that nurses and physicians were doing a bad job per se. However, we made two important observations about practice and knowledge early on as we developed the algorithm: (1) Physicians would tend to simply grab the most recent drug they had been using, and/or the one they were most familiar with as their next choice to treat patients' pain. Clinicians—physicians and nurses—weren't tending to use the adjuvant drugs with their patients as extensively as the guidelines suggest is appropriate. Clinicians didn't really know when to use adjuvants, and, with opioids, when they should titrate up the opioid or move on to a different opioid, or which opioid would make a next-best choice.

There was no logical decision-making process that they could pull out and look at, and this is what led us to develop this decision-tree model from the very beginning, and to apply to the National Institutes of Health (NIH), to say, "We really think that there should be a logical way that a clinician approaches a patient who has cancer-related pain and works through the system . . . an efficient method versus just shooting in the dark." Treating pain gets confusing for clinicians, including oncologists. For example, when Mrs. Jones comes in with breast cancer and the tumor has spread to bone, and now we have movement-related pain, you think, "My God! The pain's not controlled! Where do we go?"

The algorithm is a useful tool, even for clinicians experienced in pain management, like myself. In my clinical practice, I may forget to apply certain segments, and having a written algorithm I can look at quickly reminds me of my choices: "Oh, I didn't ask about the neuropathic part of it, or the somatic part of it. I really should use a nonsteroidal anti-inflammatory; I haven't tried it." And all of a sudden, the patient has dramatic pain relief because of these adjuvants we've added, and we don't have to fight through the high-dose narcotics as much and deal with narcotic barriers as much as we would had we just tried to cover it with a narcotic alone. So having something to which you can refer back to remind you, the same way you do with antibiotics is a real boon. [See Cancer Pain Algorithm chart for a global overview of this decision-tree approach to cancer pain management.]

How the Algorithm Works: Key Steps and Links to Tools

What triggers the algorithm and how does it work?

Anna Du Pen: The nurses are trained that decision making flows from the assessment of the patient's pain. The day-to-day implementation of the algorithm is driven by a routine pain assessment of pain intensity and pain character. We have since added pain pattern, because there was such a divergence of effect between usual and worst pain in our first controlled study of the use of the algorithm. (Usual pain is "your pain level most of the time;" worst pain is "the worst your pain ever gets.") The practical implementation during the second study was that when the nurses had contact with the patients, whether it was a clinic visit, a phone triage, or a home care visit, the first and most basic part of the interaction was to get a pain score on a scale of 0 to 10 as a measure of pain intensity. To measure pain character, we made a concise word list of nine words (burning, aching, stabbing, shooting, tender, dull, sharp, throbbing and electric-like), just a bare-bones minimum number of words that patients could say, "Yes I have this one," or "It's shooting, and it's also aching." Our idea was that although we have some extensive tools like the BPI (Brief Pain Inventory) and the McGill-Melzack Scale, they're not feasible for practical, everyday use. So we really had to get it down to, "Is your pain controlled or not?" and "What kind of pain is it?" And that answer leads us to the choice of co-analgesics. [See the Pain Assessment: Routine chart.] A step in this routine pain assessment decision tree is to assess for the presence of side effects, such as constipation and oversedation. For each of the most common side effects, we've created a separate decision-tree tool.

Making the analgesic drug choices is a longitudinal process; there is no single intervention. We try out a particular medication, based on the algorithm, and then do a reassessment to check how it is working and cycle back through the process of drug choice and side-effect evaluation after each reassessment, or revisit the basic pain assessment if new kinds of pain occur. [See the Reassessment chart.]

Was this initial contact driven by the patient or by the study?

Anna Du Pen: In the Phase 1 study, which was published in the Journal of Clinical Oncology, the contact was driven by the study, because we were really trying to implement and validate the algorithm. And as you know, when researcher physician-nurse teams implemented the algorithm, it did improve pain management for usual pain, as reported by patients. In the second phase, when we trained the providers, the doctors and nurses, then I think the algorithm was just driven by patient contact. This systematic pain assessment was generally not part of standard treatment.

Modeling the Intervention on Existing Practice Patterns

Anna Du Pen: We were aware of several barriers that had to be overcome in order for the algorithm to be implemented efficiently into already-busy oncology outpatient clinics. We designed our training and tools to map onto existing systems so that clinicians, nurse and physician teams, would recognize them as similar routines to what they already do, but with a different content focus: pain. We used the analogy of chemotherapy treatment, because we felt that this would be comfortable for them.

We knew that for the implementation to work, nurses had to play a key role in operationalizing the algorithm. We were aware of process issues between nurses and doctors that can sometimes get in the way of best patient care, and we see these tools as a way to encourage teamwork for the benefit of the patient. These are not nurse documents or doctor documents; they are just short tools that could be used to tie together the work of the physicians and nurses.

During the one-day training in the Phase 2 study, one of the ways the role-model consultants tried to bring that point home was by making an analogy to the chemotherapy protocol. In that protocol, the patients are seen in the laboratory for their blood work; they come to the nurse first, the evaluation of the lab work is done, and then the nurse goes to the doctor and says, "The white count is such and such. It looks like the protocol calls for a drop in the next dose of chemo." The doctor basically signs off on it. The nurse runs the medication order to Pharmacy, and the dose reduction is carried out, essentially all facilitated, in an operational sense, by the nurse. So, what the consultants tried to do in talking to the doctor-nurse teams was to say, "Look, this is just another component of care that is best optimized by having the nurse operationalize it, and here's the way it's done: The patient comes in, the patient gives a pain score that's very similar to a lab test, and the nurse comes to the doctor and says, 'Look, the patient's pain level is an 8, and MSContin^TM is at 60, and the protocol calls for a 50 percent increase, which is the next dose that it would be, whatever that dose is. Is that all right with you? Can we go ahead and implement that?'" And the doctor signs off on it, because we still don't want to go so far in this automation of the process that the doctor is unable to use his or her judgment to make decisions based on the individual needs of the patient. But in a general sense, the care is optimized by having the protocol in the hands of the nurse.

So, the algorithm is a tool that encourages nurses to operationalize cancer pain management. In training the clinic-based physician-nurse teams, what other strategies did you use that were specifically geared to promoting the use of the algorithm and the cooperation of the physician-nurse team?

Anna Du Pen: We have a progress note and a flow chart. The flow chart was modeled after a chemotherapy flow chart. In other words, in most oncology clinics there's a flow chart that shows what chemotherapy drugs the patient is on, what their white count is, when they got their last dose. It's a regimented form that oncology clinicians are used to filling out. It's a kind of tracking form.

We tried to integrate a pain flow chart into that same form. In other words, we have a model flow chart, but we wanted to give each individual clinic the option of whether to use it as a separate form or to integrate it into their existing flow chart. Basically, what it showed was that on the same date, right underneath the lab value, there's a pain score of 8, this is the medicine that the patient was on, this is the modification in medication that was made, and this is where we checked to see if there were any of the seven core side effects that we follow in the algorithm. So, we basically tried to integrate the use of the algorithm into the way the clinicians were already operating.

The progress note, which was designed as sort of clinic checkoff, is a one-page tool, very concise, and it says, "Is the pain controlled or not? Is the patient having side effects or not?"

So, you have some questions that they have to complete?

Anna Du Pen: Yes, and those were done by the nurse, for the most part, because that's the assessment phase, where the two primary questions are, "Is the pain controlled or not?" and "Are they having side effects from their analgesics, yes/no?" At the bottom of that piece of the form, there is a treatment recommendation that has to be followed through by the doctor. Again, it's just a checkbox, that is, "Optimize co-analgesics (check), Titrate opioids (check)." So it's another communication device that ties the doctor and nurse together.

Some nurses told us they didn't feel comfortable going to the doctor and saying that we need to increase the MSContin^TM from 60 to 75 or from 60 to 90 or whatever the titration protocol suggested. They weren't comfortable in their own discipline recommending something to the doctor. And so we tried to take that into account by having the recommendation come from the document, the Cancer Pain Algorithm.

Potential Barriers to Consider When Training Others to Implement the Algorithm

Physician-Nurse Team Issues

Can you speak more about the physician-nurse teams? What makes those teams work?

Stuart Du Pen: Well, let me just say that the closest relationship between a physician and a nurse seems to be in oncology, much more so than in, say, urology or general surgery, where a nurse practitioner or a nurse partner is more in the office, dealing with new patients as they come in, are treated, and then leave. Whereas with oncology, the patient comes into the practice, and is a long-term occupant, you might say, of the practice. And so the physician and nurse play a much closer role, both with each other and with the patient and family. The oncology nurses have individual patients within that physician-nurse practice they're following, and they know exactly where each patient is . . . the patient calls and talks to the nurse on a first-name basis, and they become very close. The nurses know who the relatives are; they become an integral part of that relationship . . . the physician, the nurse, the patient, the patient's family. So it's a much closer set of relationships than within the other types of practices, and it's very clearly related to the nature of the practice.

What specific things have you observed? Can you flesh out what that closeness means?

Anna Du Pen: Unfortunately, in my experience, that would be a really nice template if it were true for every oncology doctor and nurse, but I don't think it is. I think that the physician definitely drives the temperament of the interaction with the patient, and even if you have a wonderful nurse who is really into pain management, if the physician, for whatever reason, is one who believes that you can't treat the pain optimally because it will interfere with the antitumor treatment or diagnostic stuff, it doesn't happen.

During our one-day training in the Phase 2 study, Michael Levy and Pam Kedziera and Vivian Scheidler and Stuart Grossman—the consultants who served as role-model physician-nurse teams and trained the physician-nurse pairs—spent a lot of time trying to reinforce the idea that it was OK for the nurses to do some of the decision making. The algorithm is driven by assessment, even to the level of how much to titrate the opioid, and nurses are perfectly capable, and perhaps more in tune with the patient, in order to drive some of these protocols. So I think there was some background in their minds, and probably also in my mind, that the physician has to be committed to the concepts of palliative medicine, and the physician has to be able to delegate some of the responsibility for moving the protocol along to the nurse.

What did you find in that regard among the trained group of physicians and nurses who were implementing the algorithm?

Anna Du Pen: I think that in some cases it did work that way. In other cases, the nurses would say things like, "Well, he doesn't want me to make a recommendation . . . he wants to make all the decisions himself." Some nurses reported that physicians seemed to be intimidated when they used tools associated with the algorithm, such as equianalgesic charts to determine what the dose of the new drug should be. It seemed in those instances that control issues got in the way, and that physicians who did not already have access to this knowledge found it intimidating in this form, that is, coming from the nurse. So the response in those cases, as reported by nurses, was that the doctor wanted to have control of the dosage decision on his own—even though he didn't have this knowledge at his fingertips—without using the equianalgesic chart. That does happen, and I'm not sure what the best way to deal with that tension is.

Limitations of One-Day Training

Stuart Du Pen: We found that when we trained clinicians to implement the algorithm, once we completed the teaching process, it only worked for about four months, that is, the patient outcomes only showed improved pain scores for the first four months after the training. Then the effects of the training seemed to wear off. The physician-nurse teams were back to doing the same things they were doing before the training.

What happened?

Stuart Du Pen: We taught them, but we didn't update the teaching. In Phase 2, we wanted to determine what impact that one intensive day of teaching would have over time. So, purposefully, we gave tools to the participants, and we, as the researchers, remained available to them, but we didn't go back in and reeducate.

Anna Du Pen: I think that some of what Stu is talking about was that we were hamstrung a bit by the clinical research format. But the benefits clearly are that the outcomes are improved when the process is implemented. The weaknesses are that unless you have a motivated learner, the one-shot teaching may not stick. We did notice the variability of interest and commitment among the clinicians we were training. One of the groups included a medical director of a hospice, who was obviously very interested in implementing this. Another group had some doctors that were totally uninterested in palliative care, yet the nurses in the clinic were very interested and had the reference guide with them at their desk. Some of those things are nearly impossible to capture in terms of a research variable.

Organizational Barriers to Implementation

Anna Du Pen: We noticed that implementation fell down in some individual clinics because of reorganization and systems issues. For example, in one clinic we had a doctor-nurse team that was very strong. Both were very interested in palliative care to begin with and were clearly working hard to implement the algorithm. The way we knew that is because the progress notes were consistently done, and in the chart, which we knew because we were doing chart audits. Then the managed care organization where their clinic was located underwent major shifts in personnel and oversight. As a result, they lost one of the doctors out of the clinic and then the main nurse who was working with one of the doctors got ill and things just sort of fell apart there for several months, mainly because of lack of resources to follow through on the implementation.

"I Feel Like I'm Buried in Paperwork"

Anna Du Pen: We surveyed the doctors and nurses at the end of participating in the study, and they gave us anecdotal feedback. One said, "You know, I just can't do the progress note, because I'm totally overwhelmed and at the end of the day I literally feel like I'm buried in paperwork." So, the other issue here about why it may or may not work has to do with whether or not the resources are there to allow them to do things like call back the patient and say, "Look, Mr. Jones, your pain was a 9 out of 10 two days ago or whatever and we made this change in your medicine . . . what's your pain score today?"

This brings us back to one of the key steps in the algorithm: the reassessment. The reassessment parameters say that if you have a patient with severe pain, they need to have some sort of checkback in 24 hours. So if a patient comes in with a 9 out of 10 pain score, you implement some sort of recommendation from the algorithm, and you have a checkback the next day. We tried to encourage the nurses either to ask the patient to call them or to call the patient themselves, depending on how the system was best going to work for that practice. But we found that although the checkback would happen, it wouldn't happen for a week. Actually, we were gratified that it happened in a week, because we were fearful that had it not been for the algorithm, that contact probably wouldn't have been made at all, because there was an overwhelming sense among the nurses of just needing to get through their day, to give the antitumor therapy, which was sort of the first priority.

Ways to Address These Barriers

Anna Du Pen: I think two things are going to help with that kind of problem, in other words, with implementing a system or a process in an institution, where people lack motivation in the area or feel they don't have time to do the paperwork and follow through. As you are probably well aware, the JCAHO (Joint Commission of Accreditation of Healthcare Organization) has been moving toward including pain assessment and management standards for accreditation. That will be the big hammer, obviously—to mandate that processes like this be instituted.

The second thing is, we're trying to move more toward an automated system, which can be easily updated. We have another grant from the National Cancer Institute (NCI) to transfer the knowledge into a computerized decision support system that will take pain assessment data and generate treatment recommendations. We're hoping that this electronic tool will reinforce implementation of the algorithm, because if the algorithm is not used, then it obviously doesn't work!

Patient "Nonadherence" in the Face of Pain and No Medication Side Effects

In your first paper, you observed that some patients refused to take prescribed treatments even when pain was not well controlled and there were no side effects. What did the researcher physicians and nurses do in those cases?

Anna Du Pen: Well, we were actually surprised that it was such a problem. About 30 percent . . . or one out of three patients . . . refused to take the prescribed medications. We went into this with the idea that if we had the ability to educate patients and make sure they understood the medicines they were taking, made sure they understood things like fear of addiction, and made sure side effects were tracked and treated aggressively because we knew side effects were a major barrier, then we would be able to see much less nonadherence in the group treated using the algorithm versus those receiving standard treatment.

What we found is that there were some pretty major barriers to patient adherence that we didn't have a good grip on before we went into this. One of them definitely was the cost of the medication. Patients reported to us that they were actually rationing their pills because of the cost of the pills. We were saying to patients, "Make sure that if you feel the pain coming on, you take the medicine before the pain gets too bad, because it's easier to treat the pain beforehand than to treat it after it's out of control." That's one of our standard patient teaching messages.

What we discovered, however, in talking to some of these people was that their decision making about when to take their pills had less to do with our instructions— even when they understood our instruction to take their medicine preemptively—than with how much money they had to spend on their pills versus groceries or rent.

Patients Who Say, "I Don't Want To"

Anna Du Pen: Probably a second barrier to following the algorithm was a particular stance some patients took toward taking the recommended medicines, what we began to call the "I-don't-want-to" group. And this was a substantial group of patients. What we found is that people in this group are just antipill. When you talk to these people, you can explain the situation about the narcotic medicines not being dangerous or addictive and you can explain about the implications of unrelieved pain, which are pretty drastic. For example, we know a lot now about the fact that unrelieved pain compromises the immune system. So, we have a lot of powerful teaching tools, yet the patients would tell us things like, "Ever since I was a young kid, I've tried to stay away from pills. I just don't like pills. I don't want to take pills. It's OK, I'll just have some pain; I'd rather have the pain than take the pills."

Other clinicians and researchers would tell us, "Well, that's just an educational problem." But you know, in Phase 1, we had three full-time nurses who did nothing but interact with these patients and educate them, to an extent that would not be feasible in terms of what can normally occur in a clinic; I mean, it's just not feasible for three nurses to be working on nothing but pain management. So we're pretty confident that we covered the education piece pretty solidly. Even so, some patients, although fully informed about the medication, still said, "I don't want to."

A Patient-Centric Approach Addresses These Barriers

Anna Du Pen: The reality is that we need to address these issues in a very patient-centric way. You do that by asking patients what they do use for their pain, or what else they would like to pursue in pain management if they're antipill.

Integrating Nonpharmacologic Types of Pain Management

Anna Du Pen: There is room for the nonpharmacologic types of pain management, there is room for more complementary care, herbal remedies, and so on. Some patients are going to alternative care providers to get herbal remedies or something that is more "natural." Many, many times, patients would say to us, "We want something that's more natural." And some patients wanted to try other strategies, such as stress management, relaxation and imagery, physiotherapy, or heating pads. Some patients said, "If I have a choice of taking a pill or running to the store to buy a heating pad, I'm going to try the heating pad first."

Many times, it wasn't necessarily that they were antipill to the point where they weren't going to take any pills; it's just that they would prefer to do other things prior to doing that. And when we went to measure patient adherence, of course, we were measuring it against the algorithm, which said things like, "If you're beginning to feel an escalation of pain, take the pill early." This recommendation is supported by the evidence-based literature, which suggests that it may take more pain medicine to reduce their pain if they wait to take the pill. But for the patient, the choice did not follow that line of reasoning. Such a patient would be labeled as less adherent according to this algorithm, even if in fact they did eventually take their pill after trying the heating pad.

This observation contributes to our constant revision of how we're measuring things. With respect to the algorithm—which, of course, is a pharmacologic tool—we can only comment on patients' adherence to the drug therapy. In fact, we've looked at and are still looking at complementary techniques and nonpharmacologic techniques and how they fit in. We did ask patients about this, and we have lots of data on whether or not they were using alternative or complementary techniques, but from a research perspective it's not possible to factor those things in and still call them "drug-adherent." So, it was more in terms of explaining the factors that patients use to determine whether or not they will take their medicine, and that will be the subject of another paper we plan to write on patient adherence, in which we will look at factors used by patients to determine whether or not they want to take their medicine.

For purposes of implementing the algorithm, I think this means that not only do we have to deal with institutional issues that create roadblocks to implementing the process, not only do we have to figure out a way to really help clinicians retain and continue to use the process, but we also have to recognize that the algorithm is just a tool to help bolster the patient-provider interface. We still have to come back and say, "We're in the business of patient-centric care." We recognize that patients will make decisions, and we absolutely want to make sure that they're educated about their medications, about pain and the implications of unrelieved pain; but ultimately (and I'm speaking for both of us, I think), we are in favor of patient-centric pain management.

Ongoing Negotiation Between Patient and Clinician Understandings

Stuart Du Pen: It truly becomes a negotiation. I find certain patients will accept, for example, Kadian, which is a long-acting, 24-hour morphine, when they don't want to take as many pills. Or they'll take a Duragesic^TM patch, because it isn't a pill and it's a different way of getting medication.

The other part of that is that the patient has a personal algorithm that may be different from the one we are using. We often try to get the patient to fit into our understanding of the process of prostate cancer, for example—what's going to happen to them in the long term, it's widely metastatic, they're going to die—but let's try to make them as comfortable as possible. Well, the patient's algorithm is, "I'm not going to die. This next chemotherapy is going to cure the problem. I'm going to be out skydiving like I've done before, or rock climbing, and this is where I'm going. I'm not going along your pattern that you're planning for me." So, it does become a real negotiation with the patient as to what medications the patient is willing to accept, because their plan for the future is totally different from ours.

Reassessment of Pain as a Means to Renegotiate Medication Options

Anna Du Pen: The idea of an algorithm is not that 100 percent of the time, if you follow this box by an arrow to another box, then you will have the golden answer to the problem. I think that from the perspective of the algorithm, the reassessment component probably speaks most to the idea of continuing to keep the patient in the loop. That is, you assess the patient, you elect a logical choice of pain medicine therapy, and then you always go back to the patient and say, "Is the pain controlled? Are you having side effects?" And generally speaking, the answer to one of those two questions will lead you into some renegotiation with the patient.

It's very flexible. You say to the patient, "I hear you're still having pain, but you're not having side effects and you've been taking your medicine only 30 percent of the time. My next step, according to the algorithm, would be to increase the medicine or change to a different medicine or add an adjuvant, but I need to know why you're not taking your pills so together we can take the next step in the algorithm and decide about either changing the drug or adding a second drug so that you don't need as much of the first drug."

The patient may say, "I'm only taking 30 percent of my medication because I want to use the heating pad and drink my chamomile tea, because I think that they're better for me." So, at some point you try to say, "Well, let's explore why you think that. Is it because you're afraid of becoming addicted?" If so, you try to deal with it.

Or they may say, "No, I don't have any trouble taking the pill. I understand that it's OK to take the pill. I just prefer to do these other things." That response isn't necessarily anathema to the algorithm, because you're still doing the reassessment component, you're still monitoring the patient's pain level, you're still continuing to offer options to the patient; but at some point you listen to the patient and say, "OK. You're going to do your heating pad, you're going to do your chamomile tea. On the occasions that you do choose to take the medication, let's see if there isn't a better way that you can gain more efficacy from the medicine during that 30 percent of the time that you're taking it." So, you try to utilize the logic and knowledge of the algorithm in the space where the patient is ready for you to do that.

Cultural Diversity and the Applicability of the Algorithm

Were there any issues connected to cultural diversity related to patient adherence to pain medication?

Stuart Du Pen: The big key is the assessment and reassessment process, which I think would work fine in any ethnic environment.

Anna Du Pen: We don't have huge blocks of minority folks in this part of the country, [Seattle] so we did not have enough patients from culturally diverse backgrounds to explore issues related to ethnicity. However, we do have a fair number of Pacific Islanders, and a large population of Scandinavian folks, Norwegians and Swedes, in a subgroup in Seattle. The Scandinavians can be incredibly stubborn and stoic, and comprise a big part of the "I-don't-want-to" group of patients. Among the Scandinavians is an attitude that pain is an obligatory part of the experience: "I can tough it out; I can get through this and it's a part of what I have to go through."

After the first part of the study, we did begin to ask patients what an acceptable level of pain was, because we heard so many patients saying, "No, I don't take the pain medicine until the pain gets to a certain level." We were surprised by how high the pain had to get before some patients would consider medication. I'd have to go back and look specifically at this, but some of these patients would say 7 out of 10: "I don't take a pill until it gets to 7 out of 10 because I'm just determined to win over this pain." Or "I am tough enough to put up with it." In my opinion, that definitely is a cultural kind of decision-making process.

Developing New Tools for Patient Education

Anna Du Pen: That leads into some of the other things we're currently doing with the algorithm, which includes trying to develop good handouts for patient education and support so that in the event—as is the case many times—that the doctors or even the nurses are not skilled at being able to work through differences with their patients about pain medication, they will have some tools to use. I'm thinking specifically of the stuff that's coming out right now about unrelieved pain and the immune system, because it speaks directly to that "I'm-going-to-be-tough, I'm-going-to-fight-this-cancer" attitude. As a matter of fact, unless your pain is relieved, your body is not able to do its best work at fighting the cancer.

Did you find that when a nurse actually explained this to patients that it affected their adherence to the treatment recommendations of the algorithm?

Anna Du Pen: In this study, we weren't measuring the patient-level outcomes of the educational intervention component. Basically, our feeling is that patient education is a component of what the nurse does as a part of the process of implementing the algorithm. But just anecdotally, I can tell you that many times it does work. For example, in the general category of fear of addiction, people are misinformed, sometimes by their own doctors, which is kind of alarming. But the fear of addiction can and often is worked through fairly quickly; it doesn't have to be a 45-minute discussion.

International Applicability of the Cancer Pain Algorithm

Is your algorithm capable of adoption in countries where different drugs are used or available or where morphine is difficult to obtain?

Stuart Du Pen: Yes, because we give options. We've got to give options to the doctor-nurse teams because they have different drugs, and they're going to tend to use different drugs. In Australia, we found, there is no Dilaudid^TM or hydromorphone available, and they were just getting the Duragesic^TM patch, the fentanyl patch. So yes, there are differences, yet you can still use the algorithm. At the reassessment step, in which you are talking about titration of the opioid, clinicians can use whatever sort of opioid they have available.

We found that when we were lecturing in Australia, we had very little problem, because of the availability of alternative drugs. In China, on the other hand, where morphine is very restricted and Demerol^TM is available but still restricted, that would pose a major problem for implementing the algorithm there.

Anna Du Pen: Sometimes clinicians in other countries have more and sometimes they have fewer options than we have. In Australia, for example, they have nasal fentanyl, which we don't have. So in some cases, they may have drugs that are better than the drugs we have, or they may add to the list of the ones we have. In some cases, the list of available agents will change. I think what we're emphasizing here is that it is not necessarily the availability of one or more of the drugs by themselves; it's really the process that's important. In a country where they don't have access to opioids at all, I think it would be much more difficult to implement the algorithm, but as long as there's some access to drugs in these categories, they can be used and combined in the process.

References

1. Agency for Health Care Policy and Research. Clinical Practice Guideline: Management of Cancer Pain (AHCPR Publication No. 94-0592). Rockville, MD: U.S. Department of Health and Human Services, Public Health Services, Agency for Health Care Policy and Research, 1994.[Return to Featured Innovation]

2. Du Pen S, Du Pen AR, Polissar N, Hansberry J, Kraybill BM, Stillman M, Panke J, Everly R, Syrjala K. Implementing guidelines for cancer pain management: Results of a randomized controlled clinical trial. Journal of Clinical Oncology. 1999;17(1):361-370.[Return to Featured Innovation]