Episode 48: CONVINCE Study – Patient Reported Outcomes with Krister Cromm

Episode Transcript:

Dr. Ben Hippen: Including patient-reported outcomes, or PROs, in the CONVINCE clinical trial opens new and meaningful pathways to understanding medical care from the patient's point of view. By combining clinical and patient-reported endpoints, we can identify opportunities to improve patient health status and define treatment goals. 

Krister Cromm is Fresenius Medical Care's Director of Person-Centered Outcomes Research.  He is leading the company's evolution from disease-centered to person-centered integrated trial design. 

Krister Cromm, welcome to Global Medical Office Dialogues. 

Krister Cromm: Thank you very much, Dr. Hippen. It's great to be here. 

Dr. Ben Hippen: I'd like to begin by asking you to talk a little bit about what a patient-reported outcome is. 

Krister Cromm: So a patient-reported outcome is basically a way to measure health-related quality of life.

So health-related quality of life is really important for all of us, and it has been defined by the World Health Organization. In order to measure that, you need a so-called patient reported outcome measure, and that would be like a survey tool that would allow you to ask certain questions on specific parameters.

So for example, you have physiological parameters, physical health, such as if you can walk upstairs. There are mental health components, such as measures about feeling depressed or anxiety. And there are also social aspects of health. Can you have a meal with your family? Can you play with your grandchildren? Can you participate in a social event at night? All of these are aspects that are really important for your quality of life, and they are in a certain relationship with your health status, and if you have a chronic disease, we measure how they are related to your treatment. 

Dr. Ben Hippen: So in the CONVINCE trial, you actually created a survey that really built upon a survey that had been used in previous dialysis-related trials, the KDQOL or the Kidney Disease Quality of Life Survey. Can you talk a little bit about the improvements you made on KDQOL and how your surveys are different from KDQOL? 

Krister Cromm: So we actually went back to the patients and we started with what we call qualitative interviews. So we sat down with patients and asked them, what is important to you in your life, in your dialysis journey, what is important to you when you go into dialysis? And we compared this to what is the standard measure, right? So the current gold standard, as we call it, is the kidney disease quality of life survey. And then we thought, well,

There are a few aspects that have not really been addressed right here because hemodiafiltration can offer certain benefits. And we already had patients on hemodiafiltration in Europe, so we asked them, what's better for you? They told us basically that they feel improvements in the recovery time after dialysis. They felt that their cognitive function improved throughout their journey.

They told us about symptoms during dialysis that are really important to them, such as sudden blood pressure drops and how that makes them feel. And we included them in the surveys. But even more than that, we asked them, how do you feel about filling in the surveys? And a lot of them told us that they actually want to use their mobile phones, they want to use tablet computers, and they wanted to reduce the number of questions. So we went back to the drawing board and looked at that and thought, well, what we really need to do is we need to find interactive surveys. We need to develop those. And in fact, there are already surveys in use in the United States that were developed at Northwestern University. They're called PROMIS, and they actually contain outcome sets, a lot of questions. And they can be chosen specifically for patients. So the computer system would actually choose the questions. And then we said, well, that's great. We really need that for our patients in this study. And this never been done before in dialysis. So we just said, okay, here we go. We will try this out. And that's what we did. 

Dr. Ben Hippen: That's very interesting. So an interactive survey seems to me, methodologically, to be able to capture very specific patient responses to previous questions and then prompt new ones in a dynamic way. But it does strike me that there may be a methodological problem with comparing that across different patient groups in different locations. So how do you do that intra-geographic or intra-market comparison if the subsequent questions are dynamically chosen by a computer based on previous answers? 

Krister Cromm: First of all, you need a lot of patients to answer these original questions for you, right? So you need a good base. And we were very lucky because PROMIS had been sponsored by the US government. And they had been actually used in the US in a lot of disease areas, in cancer and cardiovascular disease. So we've got that basis, and we know how patients in North America would answer these questions, not in dialysis, but in other disease areas. And we also know that actually these questions have been given to anybody, like the regular population and in a large sample size. So a few thousand patients, or just people in North America, have actually filled those in. So we know that these questions work. And what we've done in Europe then is we said, OK, we will first show these questionnaires and these items to patients in Europe, to dialysis patients, and see if they can answer them. And they could. So we included them in the study. And in our study, in CONVINCE, we had 1,360 patients participating in eight European countries, so people with eight different mother tongues. They answer the questions over the course of the study, so at least two and a half years, every three months. So we know how they answer, what the patterns will be. We can recommend using those in regular clinical care in dialysis patients. And since we have now translated those surveys into these eight European languages, they can now be used in these languages as well.  

Other than English and Spanish and French that are obviously spoken a lot in North America. And we are confident that the surveys can actually be used in the North American dialysis population because the treatment as such is obviously similar. With the difference now that hemodiafiltration has shown to improve mortality and has also shown to improve patient reported outcomes, especially cognitive function as well as social participation and I believe this is a truly important outcome for patients. It's meaningful to know that if you go on hemodiafiltration you can participate more in social life and you will live longer, you will have less complications in dialysis if treated correctly on hemodiafiltration. And I think these are very, very good prospects and a light at the end of the tunnel for many patients.

Dr. Ben Hippen: I do want to come back to what you found in the CONVINCE trial, but I want to start by asking you about some findings that you and your team identified with regard to the relationship between self-reported self-efficacy and social support and health-related quality of life outcomes. 

Krister Cromm: Social determinants of health are fundamental for all of healthcare, no matter where you are globally. And we set out to really ask patients about, how they feel about that. Traditionally, a lot of allied health workforces, they would ask, how is your social support? Right? Do you have people that you can talk to? And so we also included that in CONVINCE.

But more so, we also looked at self-efficacy. And self-efficacy basically is the ability to adapt to new situations and take responsibility for your own healthcare and your life in general, if you face adversity. And what we have found out, that this is equally important, in particular in dialysis. We have employed a model. It's called the Wilson-Cleary model of health-related quality of life.

It basically tries to give a theoretical basis on how health, reality, quality of life emerges in a treatment situation. And self-efficacy is closer to the physiological basis because if you think about it, that makes a lot of sense. If you take your medication and you ask your doctor, why you should take a certain medication and at what time, and can you take it with a certain meal or not, then that means that obviously your outcomes will improve. And what we have learned is that people who have that feeling of self-efficacy will do better on dialysis. And that means that within this trial, we have learned that we need to actually reinforce the self-efficacy of patients. And there are actually trials going on also in North America, that look at cognitive behavioral changes and support for patients to increase just that, to basically give them the agency to take care of their disease, to speak to their doctors, and to get the care that they deserve. 

Dr. Ben Hippen: So let me ask you a real-world evidence question about self-efficacy, because you mentioned that one of the things you discovered is that not only is there impairments in self-efficacy, but in fact, a need for intervention.

So if you're assessing the relationship between a clinical intervention like hemodiafiltration and outcomes like self-efficacy, but you also come to the realization that you need to reinforce self-efficacy, is it the treatment or is it the intervention to improve self-efficacy that's really improving self-efficacy insofar as you do see an improvement? 

Krister Cromm: What we've seen in the results is that the effects go into the same direction. So we've seen that hemodiafiltration as such improves cognitive function, social support. But we've also seen that it's actually linked to self-efficacy. Now what this trial has not done, it has not looked at causal relationships, right? So we have not measured at various time points, how self-efficacy has improved. So it could be, of course, that because HDF is doing what it does, so it improves outcomes. Therefore, you're feeling that yourself efficacious, And then that reinforces, you know, that notion, you know, you feel, oh, I'm better, you know, I have more energy in the morning. So maybe that's why I want to take my medication, I want to go for a walk. And this is like a positive circle.

It could also be the other way around. It's because, okay, you're already taking your medication well, and then HDF comes in, and, woo, it improves your quality of life. We have not investigated these interactions throughout the trial, but as it stands, these two effects support each other. And I think it's good to see that even in patients who are on regular hemodialysis in this trial, also have seen improvements. When they took care of their health and they took charge and took the initiative. Now what we can also say is that hemodiafiltration as such is a powerful intervention, but in a clinical trial setting, you give people the opportunity to address their questions more. So just the fact that people are in a trial has shown that you can improve outcomes. In the trial, we have shown that patients can be on dialysis for four hours. They can feel better with this. And that, as such, is already a tremendous achievement, because it shows that being on dialysis doesn't need to necessarily make you feel bad, make you have symptoms during dialysis or after. It shows also that you can work towards it with the people who care for you and who are there with you in dialysis and outside of dialysis. 

Going forward, I think there is a lot to do to address these different questions to actually improve quality of life. It's not only about hemodiafiltration, but it is about the environment and what you make out of it. 

Dr. Ben Hippen: So you brought up patient symptoms and I was very excited to learn about the IDS survey that you and your team developed. Can you tell us a little bit about the IDS survey and how it's different from other quality of life measures? 

Krister Cromm: We actually started off with the KDQOL that includes a symptom survey and the traditional symptoms that you would measure such as cramps And what we did is we gave this to patients and we asked them what other symptoms are really important for you. And they addressed questions like blood pressure drops.

And I was actually surprised to see that this was not included in the KDQOL before. So we developed five additional questions that actually look at quality of life during the dialysis session. And what I've learned actually is that simultaneously other research groups have done the same thing. So I know that Jenny Flythe in North America has actually developed it, more or less exactly the same questions. So we have really learned from these patients that you need to go back to them doing qualitative research at regular intervals to understand how their journey has changed. And this could have had many reasons. It could have been through regular improvements in clinical care. It could have been about new therapy options that are available or simply a change in the environment.

The KDQOL was developed more than 20 years ago, I believe, and it's really been time to go back to patients and ask them. So one of the things that I have learned from developing, especially the symptom survey, is that you can develop questions, but every time you do a new clinical study, you'd need to go back to patients and ask them, is there something that has changed for you? Do we need to or want to address another question? So that we can measure it and improve your quality of life step by step. And that is also what large initiatives like the SONG initiative, that's a big outcome initiative in dialysis care, actually calls for. So it's looking at different aspects of quality of life like fatigue and mortality, and they should be measured every time you do a clinical study.

They are also proposed, of course, to be measured in regular care. And you should then go back and maybe develop a toolbox. And this is what we want to do with our COMVINCE surveys. We actually don't want to see them as static, but we want to see them as toolboxes. Or like individual tools in a toolbox, you can take it out, and you can say, OK, this is the fatigue measure.

Does it fit here? Is that the right approach for you right now? Do we need to add questions? Do we need to change it? Or are we ready to go? And because of the electronic survey methodology that is available to us now and that wasn't available 20 or 30 years ago because people simply didn't have mobile phones at that time, our smartphones and, you know, not everybody was feeling comfortable using them. We now have this opportunity, so we should seize this opportunity. Make people feel welcome and share their quality of life on a regular basis. And therefore, it's also not so much of a burden because if we pop in a new question into such a survey, because patients have not filled in that many questions because they're doing it on their mobile phone, maybe they've just answered three or four questions, then we can test along the way, okay, does an additional question actually make sense? and then we can introduce it into a new survey method and a new clinical study. 

Dr. Ben Hippen: So I gather this toolbox approach really focuses on making sure that the survey for any individual clinical trial is very much tailored to the patient themselves. And when comparing quality of life outcomes across different trials, I gather it's perhaps more important to make sure that the quality of life survey tools that you use are specifically tailored to those patients and those trials and less a concern about apples to apples comparisons between different trials having exactly the same questions. Is that right? 

Krister Cromm: Yeah, that's pretty much what it means. What we try to do is because we have the PROMIS surveys and they've been validated in so many people around the globe.

We obviously take that as a basis, so we can use that as a means to compare between treatment populations, between people that live in different countries, and then we can add other questions. So it's like, you know, you have a hammer, but you also have a precision tool for that particular patient population. And you can use both of them at the same time. And they give you just the information that you need so that you can move forward and improve care and really address the question that a patient or anybody for that matter may not have come up with. So that's the true advantage of electronic patient reported outcome measures and item response theories. So that's basically what we are using technically in these computer adaptive tests.

Dr. Ben Hippen: So I do want to come back to the CONVINCE trial and ask you to talk about what you found in terms of quality of life outcomes in patients who were enrolled in the hemodiafiltration arm versus the conventional high-flux hemodialysis arm. 

Krister Cromm: We have seen remarkable improvements in cognitive function, and we have also seen differences in social health of patients and that is in line exactly with the results of the mortality outcomes. We have not really expected that, honestly, because if you think about it you start an intervention like hemodiafiltration, and you would think that, patients will improve on all scales. So they could, you know, physical means like that's what you would expect, right? But you would not see an additional benefit in these areas. And I think what is important about that is that we can now lead discussions with researchers about the reasons behind that, the physiological reasons, the pathological reasons. And we can also use that in regular clinical care.

So these are the two aspects that we have really found out. So it's about social health and cognitive health. I think it's also a good sign for physicians to see that if you go on hemodiafiltration, you do something for your patients that is immediately improving their health status and that helps them to interact. And that's something that you can support. 

Make them interact more with other people or encourage that. And that will ultimately even lead to an improvement in mortality. It's definitely an unusual outcome. Or one that has made us happy, but that we did not expect. 

Dr. Ben Hippen: Did you identify any correlations or associations between specific patient reported outcomes and more traditional primary and secondary clinical endpoints in the CONVINCE trial? 

Krister Cromm: We have not done analyses that look at these interaction effects. And what we are going to do now is we will actually do predictive modeling to understand just that. So we have collected data throughout a course of about five years by so many patients. And what we actually need to do is go back and analyze this. And there will probably be hordes of PhD students doing just that.  This is a work that cannot be done within a few months. So we know these outcomes are true, they are reliable, but we don't know everything. So we will go back to this data set. We will definitely address some of the more urgent questions like how can we develop tools, screening tools that will bring physiological parameters, lab values, and patient reported outcomes together and show in what combinations will these now lead to improved outcomes in regular clinical care.

And then we will analyze that, we aim to publish that. And then hopefully there will be new clinical trials that will look at these interactions, how patients actually benefit from hemodiafiltration. Now there are a few spots that have not yet been addressed in our trials. For example, our studies has been conducted in Europe and that may or may not be transferable to a US setting. We believe that the patient reported outcomes are very likely to be achievable in the US setting. But of course it is important to address that and to run a trial here in the US to compare these outcomes.

We don't have to do another CONVINCE trial in North America, but what we'll need to do is we'll need to see if patients can actually achieve the convective volumes in North America, if there are differences based on participants' backgrounds, social backgrounds, also in terms of their genetic heritage, are there any differences that we can see in North America. For all that we can see now from the data, there don't seem to be any of these differences. We've got a remarkable diversity in Europe and obviously a patient in the east of Romania looks very different and receives quite a different healthcare than a patient in the south of Spain or in the north of England. So we are quite confident that these results are actually transferable to other settings. And what we know from, for example, Japan, where HDF is also a regular treatment option, that seems to be reasonable. 

Dr. Ben Hippen: Do you think that these patient-reported outcome research methods are going to become the norm for future dialysis-related clinical trials? And do you foresee any barriers to generalizing these approaches to prospective trials?

Krister Cromm: I think that they will become the norm eventually, because first of all, it's a lot of work to collect paper questionnaires. They are long. And eventually, people just want to use their private personal devices to fill in surveys. And it will also be a lot easier for staff at clinics to then review them. If we now use what we have learned in the study and create tools for regular clinical care so that the dialysis nurses, the social worker, the doctors can actually receive this information and discuss them in their regular care with patients then that will probably be the next step that we'll have to do so that it's easier for people to use them but I believe that yes they will become the norm.

And also as we collect more data in regular care from different countries, we will see if they are outliers, if there are certain questions that may not be appropriate in a certain country or population, so that they can be exchanged or they can be recalibrated, so to say, right? So that you might need to adjust the answer option and it may mean something else in Turkey than in South Africa, for example. And that's what we really need to do in the future. But I think it's exciting and it will definitely be something that will help patients every day. 

Dr. Ben Hippen: What do you see on the horizon for patient-reported outcomes-related research? Are there novel methods of data collection or data analysis that you're excited about? 

Krister Cromm: IRT, so item response theory and computer adaptive testing, have been around now for quite a while, a good 10 years, but they have not really made their ways into regular clinical care. Right now, I think it's on us to implement them. There are definitely new methods around that are exciting. 

There's voice recognition, there are other ways that would help you to do these things. But I think for now, if we manage to bring patient-reported outcomes to everybody's phones, then we've achieved a lot. I think looking forward, there is a lot to be excited about. So there may be more visual cues, for example, so you won't see a question, but maybe somebody's talking to you, or you would see a picture, and it would be easier to describe how you feel or what issue you have. But all of these things, yeah, will come, and they will develop over the years once people get used to answering questions on their health, on their devices in their daily routines.

Dr. Ben Hippen: I share your excitement with this novel vista of patient-centered research. Any final thoughts you'd like to offer?

Krister Cromm: I think that it's great that we are here now at this point. It's been eight years since we started the CONVINCE trial. And in the beginning, nobody really believed that we could run a trial with patients filling out surveys for four years on electronic devices. And all I can say is, Yes, they did it. They liked it. We liked it. And it shows that change is possible. So let's go for it. 

Dr. Benn Hippen: That's perfect. I've been joined today by Krister Cromm and we've been talking about patient reported outcomes findings in the CONVINCE Study. Krister, thanks again for joining us today on Global Medical Office Dialogues. 

Krister Cromm: Thank you very much.