The FDA has recognized the need to include the patient's voice in the drug development process. The question is: what is the best way to do this?
PATIENT-REPORTED OUTCOMES (PROs) are finding a significant place in healthcare quality metrics, and PROs are sometimes thought to be more reliable than clinician-reported data. The challenge continues to be adopting service workflows to collect this information from patients, and electronic PROs (ePROs) have definitely enhanced this process.
Evidence-Based Oncology (EBO) spoke with Thomas W. LeBlanc, MD, medical oncologist, Duke University School of medicine, Durham, North Carolina, about the real-world influence of PROs.
EBOTM: Could you provide a brief introduction to PROs, their influence on the cancer drug development process, and the subsequent influence on cancer care overall?
: PROs are validated tools that help provide additional information about what patients are going through and how they are feeling. Until relatively recently, these were mostly just tools used in research studies. Patient-reported outcome measures of things like health-related quality of life include different domains about how a person is functioning physically, about what symptoms they’re going through, their emotional state, their social well-being, and so on.
The FDA, too, has recognized that this aspect of measuring the patient experience was missing from the drug development process. They have increasingly been saying that PROs need to be part of what we measure when we test new drugs or compare drugs, and it’s an important part of what should be in the repertoire of clinical trial end points. While traditional clinical trial end points— remission rate, complete response to chemotherapy, overall survival [OS], and event-free survival—are all very important, other kinds of end points, such as what happens to someone’s overall health-related quality of life when they get treatment A versus the standard treatment, are important as well. If all things are equal, if people live about the same length of time, what is that time like? Is the time potentially better? Is it worse because of side effects?
PROs are an important way to amplify the patient voice as part of what’s going on in clinical trials.
EBOTM: Are PROs usually secondary or tertiary end points in a clinical trial? What is their influence on the outcome of a trial?
: There are several recent examples of PRO measures being important in the drug approval process, and [they are among] the things we review in our paper.1 There is a drug called ruxolitinib, indicated for the rare blood cancer myelo brosis, that was approved in large part because it dramatically improved a PRO measure called the total symptom score scale [TSS]. Patients with myelofibrosis have really debilitating symptoms: They have spleen enlargement that pushes on their stomach, so they feel full all the time; as a result they don’t eat well, they lose a lot of weight, and they experience other constitutional symptoms such as fever and dramatic fatigue. A lot of these things were improved with this new medication. So, although ruxolitinib met some traditional end points too, such as improved OS, some of the earlier, more exciting, and interesting findings really had to with TSS, and that was a big part of the story around the drug getting approved.
Similarly, another story we mention in the paper is the approval of a chemotherapy for pancreatic cancer, gemcitabine, which was compared with the standard treatment at that time, 5-fluorouracil. While gemcitabine improved overall survival, it also improved a composite measure that evaluated patients’ experiences, including pain, maintenance of physical function, and weight loss, among other things—our paper describes this in detail. These are just a couple of examples where in the last few decades the patient voice, through PROs, has become so much more important in how we count what is a meaningful di erence with a new treatment.
EBOTM: In your paper, you discuss how ePROs could make it easier for patients or their family caregivers to participate in the process of reporting the patient experience. Can you tell us more?
: Electronic PROs are the next frontier in the field. The big distinction here is what happens on research studies and clinical trials versus what can just happen in how we take care of people. For me, what’s most exciting about ePROs is that they can be a part of good cancer care, or medical care, and not simply a clinical trial tool.
A few recent examples have shown that using ePRO methods for people with advanced cancer enhances our ability to take care of them. One such landmark study, which we refer to in our paper, was done by Dr Ethan Basch and published in the Journal of Clinical Oncology,2 regarding its quality-of-life end points, and then more recently in JAMA, regarding the survival measures from that study.3
What the authors did was electronic symptom monitoring on a weekly basis. So, when people were home, they were filling out a questionnaire about 12 common symptoms; the results were relayed back to the care team and to nurses who could then intervene and not have to wait for the person to call or show up in the emergency department [ED] to get help. It’s almost like an early warning system for people with cancer, whereby reporting their experiences and symptom burden between visits led to improvements in overall quality of life, fewer ED visits, and then eventually about a 5-month survival benefit. In patients with advanced solid tumors, it’s quite remarkable.
That’s about what we expect to see with new therapies for advanced solid tumors—and this wasn’t a treatment, it was just improved monitoring and better care. That’s why I like to say that ePROs really amplify the patient voice and help us to better see the patient experience to incorporate it into what we are doing for that person to provide excellent cancer care to them.
EBOTM: How do you ensure that the most important PROs are being collected but that the patient is not burdened?
: The issue of survey overload is really the big one that we are facing now in this part of the eld. We can measure anything and everything at this point—there is a PRO for so many di erent things—but you can’t ask every patient 250 questions on a weekly basis, daily basis, or every time they come to the clinic, because people quickly get tired of it.
Part of what we wrote about in our paper,1 though, is that the recipe to success with PROs seems to be not so much related to how many questions you ask people, but rather if you make it useful to them to answer those questions. What I mean by that is that we found when the answers to the questions impact the care that people receive, that is a reinforcing experience. So, if patients realize, “My doctor actually looked at this, and cares about the result, and cares about how I am feeling. This is a way for me to communicate things that we didn’t have time to talk about in a short visit; my doctor went over these things with me that I said were really bothering me during the next visit, and I can see that it actually changed the care that I received, it helped me be heard in ways that maybe I wouldn’t have been,” that’s important.
In 1 of our analyses, we evaluated the completeness of questionnaire batteries and people getting PRO assessments as part of routine care, here at Duke, and found a significant correlation, such that people who weren’t likely to come back for a consult were a lot less likely to fill out the questions. On the other hand, among people who filled out the questions and then realized that it changed the care they received, the completion rates were more than 90%. They always filled it out at subsequent visits because they realized it was an important part of coming to the clinic and us being able to take good care of people by knowing what’s going on with them and how they are feeling.
EBOTM: What are some of the challenges of collecting PROs in the clinic?
: One of the major challenges is the work ow. Everyone is really busy in oncology clinics, and if you do something where you create a new app, and the patient has to download and use it, and then the doctor has to use it as they are also trying to do everything else they are supposed to do in a 10-min- ute visit, which is already too much, then it just won’t work. They are struggling with documentation and the electronic health record [EHR]—if the app doesn’t integrate with the electronic record, then nobody is going to use it. It doesn’t matter how much money you spend developing an amazing app; if it doesn’t really make things easier for the clinicians and the patients, then it will not be used, or will not be viewed as being helpful.
The other issue is the EHR itself. Despite the advancements, EHRs are quite problematic in fundamental ways. Each is a bit different, but I have yet to see one that really does a good job of providing an off-the-shelf PRO module, or that does it in a way that works well in real-world practice. For example, with the system that we are using here at Duke, I can send out electronic questionnaires to patients automatically before a visit. However, I can’t necessarily control what the data look like when they come back to me—I can’t control where they go in the chart, and they don’t come back in ways that I find particularly useful for patient care. EHR vendors have not really stepped up to the plate on this issue, which has resulted in the development of homegrown PRO systems in the clinic, which creates other problems with sustainability. I hope to see this change in the coming years.
EBOTM: Is there a challenge in bringing providers on board to be a part of this process?
: It is very much a challenge. There is a lot of activation energy required to change what you do. Anyone who comes in and says, “Now we are going to start collecting this new questionnaire and we want you to use it in clinic,” is perceived as someone who is just telling you to do more work when you are already overworked. That’s a recipe for failure with busy clinicians.
Incorporating PROs in the clinic will require buy-in at every level—from not only the patients, but the clinicians who are there in the trenches. When I am in my clinic and I am running an hour late through no fault of my own—maybe the lab is running behind or I had a patient who was very sick and I needed to spend some extra time with them— the last thing that I am going to do then is spend extra time on something else that isn’t required for me to take care of a person. If the change does not save time or is not well-integrated into people’s work flows, then it will be perceived as difficult or not helpful, and people will rally against it. This is especially true if they don’t recognize its value.
An ideal example of this is symptom screening. A part of what clinicians do when they see patients is a “review of systems” assessment, which includes queries on di erent types of symptoms and issues and about whether they are bothering a patient. Well, PRO systems can, and should, do that, but that would mean that the PROs would have to be collected in the waiting room or before a patient’s clinic visit, and then be presented back to the clinician in a useful way that doesn’t create confusion, slowdowns, or extra work.
However, if someone hands them an iPad and asks them to fill it out while they are in an exam room, then nobody can see them until they are done, and that holds up other patients from being put in the room and slows down clinicians—it’s not going to fly. But if PROs are collected when the patient is sitting in the waiting room, and then integrated into the EHR in a way that also pulls it into their clinical documen- tation so the clinician has all of those data at their ngertips during the examination, and saves time when writing their notes, then you have streamlined care. It would improve e ciency as well as the quality of the information.
We know from other studies that patients are much more accurate about reporting their experience of illness than doctors or other clinicians are, who to some degree are guessing at what they think people are going through. It’s important to recognize that. Patients are the experts on their own experiences.
EBOTM: Do you think the physician on-boarding with respect to PROs can be spearheaded by payers?
: That’s a tough question. There has been some discussion about including PRO measures as part of value-based care in pay-for-performance initiatives. I’ve heard more about these things in surgical settings, such as postoperative clinics, looking at different PRO measures around functional status and quality of life as hallmarks of whether you provided good care. But this could go wrong, in some unfair ways. Imagine you see a patient who has a joint replacement and then they just don’t do their rehabilitation program, and then end up with poor mobility and maybe pain; that doesn’t mean that you provided bad care, it means the patient didn’t do their part. That’s an instance where you can see the folly of a PRO measure being related to reimburse- ment, or perceptions of value of care. Is it reason- able to do that kind of assessment and then hold physicians accountable? I’m concerned about that possibility and what that might look like. Similarly, we might expect PROs to worsen steadily over time in people with advanced cancer, and it doesn’t mean we provided bad care but rather that the disease and its treatments take a toll over time. So, what might be the right PROs to track in a pay-for-performance kind of way? I’m not sure.
On the other hand, there may be instances, say in the cancer care setting, where we might think it’s more agreeable to use more specific PRO measures around what a person is going through—for example, a measure of nausea or vomiting in people who are getting chemotherapy. There are standard protocols for preventing chemotherapy-induced nausea and vomiting, plus we now have really effective drugs. While we probably won’t get to a zero rate where nobody has either of those side effects, sometimes we may not do as well as we could or should be doing to prevent them. If we actually measured the rate of these side e ects and if practices were held accountable to that, I can guarantee you they would pay more attention to it, and get much better at addressing the issue than they are now. That is 1 example where a very focused measure could be quite helpful and very appropriate.
EBOTM: What are your predictions for the growth of this field?
: Overall, PRO measures are very useful tools for amplifying the patient’s voice as part of their cancer care. That’s a voice that often gets muted and sti ed in how we take care of people. We know that our healthcare system is not set up to be patient-friendly—it’s built around clinicians and nurses rather than patients, ultimately.
Inclusion of PROs into routine care will improve patient care by increasing emphasis on the patient experience—what they go through, how they feel, and how they live when they are not in the clinic. Most of these people’s lives, when they are dealing with something like cancer, is spent mostly outside of the clinic; we only see them a few days a month. What’s going on during these other times, that’s where PROs can be very helpful in amplifying the patient voice and making us pay attention to the patient experience in ways we unfortunately tend not to in routine cancer care today.References
1. LeBlanc TW, Abernethy AP. Patient-reported outcomes in cancer care— hearing the patient voice at greater volume [published online October 4, 2017]. Nat Rev Clin Oncol. doi: 10.1038/nrclinonc.2017.153.
2. Basch EM, Deal AM, Dueck AC, Bennett AV, Atkinson TM, Scher HI. Overall survival results of a randomized trial assessing patient-reported outcomes for symptom monitoring during routine cancer treatment. J Clin Oncol. 2017;35(18 suppl). doi: 10.1200/JCO.2017.35.18_suppl.LBA2.
3. Basch E, Deal AM, Dueck AC, et al. Overall survival results of a trial assessing patient-reported outcomes for symptom monitoring during routine cancer treatment. JAMA. 2017;318(2):197-198. doi: 10.1001/jama.2017.7156.