Stepped Palliative Care: A Podcast with Jennifer Temel, Chris Jones, and Pallavi Kumar

Eric 00:27

Welcome to the GeriPal podcast. This is Eric Widera.

Alex

This is Alex Smith

Eric

And Alex, who do we have with us today?

Alex 00:34

We are delighted to welcome back Jennifer Temel, who is a thoracic oncologist at Massachusetts General Hospital in Boston. Jennifer, welcome back to the GeriPal podcast.

Jennifer 00:44

Thanks.

Alex 00:46

And we’re delighted to welcome Chris Jones, who’s a palliative care doc at Duke in North Carolina. Chris, welcome to the GeriPal podcast.

Chris 00:54

Thank you.

Alex 00:55

And we’re delighted to welcome Pallavi Kumar, who is a palliative care doc and oncologist at the University of Pennsylvania. Pallavi, welcome to GeriPal.

Pallavi 01:04

Thank you.

Eric 01:06

So we’re going to be talking about a model of outpatient palliative care, a stepped model for palliative care. But before we jump into that topic, Jennifer, do you have a song request for Alex?

Jennifer 01:20

I do have a song request for Alex. I have requested Galileo by the Indigo girls. And the reason I requested it is because it is one of the two most famous Indigo girls song, with the other being closer to fine. And most people think closer to fine is better, but they are wrong.

Eric 01:36

Galileo, Alex, did you do closer to fine already? I feel like we’ve done that one.

Alex 01:42

Probably. Probably. I think I did. I was in an Indigo girls cover band in college at the University of Michigan. Saw them play a couple times. Actually got a pick from one of them. I think it was Emily. At some point. I carried around my wallet for years until it fell out. And then we got in. Our band, strangely, got in. We played like a talent show. And then we got invited to play at the Elks. Or was it the Masons? I think it was the Masons lodge. And boy, did they not know what to make of us. So this is a fun song. Thank you for requesting it. It’s taken me back quite a few years. Here we go.

Alex 02:29

(singing)

Eric 03:36

I have a trial that’s going on that’s showing that there’s association between the length that Alex plays a song on a podcast intro and his enjoyment of that song. [laughter]

Alex 03:48

That was a lot of fun. Thank you for taking me back.

Eric 03:58

So I’m really excited because I am very eager to talk about this study. So there’s a study published in JAMA just a couple months ago this year on stepped palliative care and a stepped model. And I wonder if we can, before we talk about, like, why is a step model important? Can someone describe for me what a stepped model is?

Jennifer 04:25

I can take that on. So, basically, with a stepped care model, the goal is to tailor care delivery to the patient’s needs while at the same time utilizing less clinician resources. How it works is that all patients will have access or encounters with the specialty trained clinician, a psychologist, a social worker, a palliative care clinician. But they’ll have infrequent contact, and they’ll be closely monitored for the condition that the specialist is treating them for. And if that low frequency contact sort of solves the problem, that is, their depression gets better or their depression doesn’t get worse, they stay with that low frequency contact.

But if their depression, for example, worsens, then they step up to more frequent contact with, for example, the psychiatrist. So it’s really trying to, again, solve two problems, be a more patient centered approach by having the care tailored to what the patient needs. But also when there’s limited resources, like psychiatrists or palliative care clinicians trying to sort of triage care to patients who need it most.

Eric 05:32

Yeah. And so let’s just, let me get this. So you, Jennifer, probably have, like, the most cited, probably should be the most cited randomized controlled trial of palliative care and lung cancer patients that everybody, every palliative care article I’ve seen, whether or not it works, always includes, it was outpatient palliative care in lung cancer patients. I know there’ve been studies afterwards, too. That was not a stepped model of care. Right. What was that?

Jennifer 06:02

That’s a great question. So when I started this research 20 years ago, I’m aging myself in national radio, tv, whatever you want to call this, the prognosis of lung cancer was pretty uniform. So, for example, everyone who was diagnosed with an advanced or metastatic lung cancer had a prognosis on the order of months. When we started this work, we made up an interval, but the interval we made up was monthly encounters with palliative care, because many patients didn’t survive more than four or six months.

Thank goodness. In the 20 years since then, lung cancer therapies have improved, and lung cancer is a cancer population where there’s really varying prognosis based on cancer characteristics. So, unfortunately, some people still do live for six, eight months. But there are subgroups of patients that have better responses to therapies and can live a year or two years, or in some cases, many years. That was really the impetus for our team thinking about, how do you take a population who has high palliative care needs, but has really varying palliative care and supportive care needs based on their treatment, their different prognosis, et cetera? That’s why we thought about applying a step care model to patients with advanced lung cancer.

Eric 07:25

I guess, Chris, from your perspective.

Jennifer 07:31

From.

Eric 07:31

A clinical perspective, why is it important to test this stepped care model versus implementing more? What we know was evidence based before? All of this is.

Chris 07:41

Yeah, it’s a really interesting thing, because the 2010 article was solving the problem of, hey, send us patients, we promise we won’t kill them. Right? That was the initial. And now when I think about the clinic at Duke, when we’re really running on all cylinders, we can get ourselves up to a five, six, seven week wait for a new palliative care referral. And there’s a mortality number of people who just can’t make it to get to our clinic. So the problem of, hey, look at us, we promise we’ll be helpful. We won’t kill your patient.

We’ve solved that problem, and we’ve solved it so well at some institutions that we now have this tremendous capacity problem problem. So that’s the thing that I’ve really liked about the stepped model. I’ve been the implementation arm of a couple of different studies, and I was in one where it was basically read this script, and we’re going to call it a palliative care intervention. I was like, oh, this is so hard to do, but this.

Eric 08:44

Did this happen in an ICU, by any chance?

Chris 08:48

There might have been a GeriPal podcast or a Christian Sinclair article about this. Yes, but the idea in the positive here is that this model says you do palliative care. We’ll keep an eye on the patients, and we’ll let you know if they need more of you. Now, the rate limiting step in my mind is I don’t have anybody to keep an eye on those patients in the clinic. So implementing stepped wedge would be hard. Now we can talk later on about the different models of paying for care coordination. That would actually be a way to solve this problem. But I love this because of the capacity problem that we’re suffering with.

Pallavi 09:26

Yeah.

Eric 09:26

Any other thoughts? Pahlavi?

Pallavi 09:28

I think it’s also just patient centered. Right. So time is a precious resource and I don’t know, in your respective cities, but parking at our beautiful parking garage costs like $30 for the day. So saving people time and money and just the burden of being here for a visit that they may not actually need that day, I think it’s hard to put a value on that.

Eric 09:51

I love. I don’t think a lot of doctors quite understand that. Sometimes patients feel like, I don’t really need to see my doctor, but they’re telling me to come, so I’m going to do this for them. Yeah, I have a question. Go ahead, Alex.

Alex 10:09

I have a lot of questions, but first question is, when I first looked at this, I was like, wait, there’s no usual care arm. There’s no usual care arm. And then I got an invitation to speak at the, what is the Montreal Congress to palliative care. And they want me to debate Vicki Jackson about this issue of we no longer should have usual care arms and palliative care trials. Like, have we moved beyond the point where it’s ethical to have a usual care arm for patients with advanced lung cancer? I don’t know who wants to take that one?

Eric 10:48

And Alex, when you’re saying usual care, do you mean usual care not being palliative care or like.

Alex 10:53

Yeah, or like palliative care if they get a consult for whatever reason, but not trn consult, like standardized, you know, randomized to palliative care versus whatever they’re doing before. If they get palliative care consult, they get one. If they don’t, they don’t.

Jennifer 11:07

First of all, Alex, I just want to compliment you on that, um, really heroic effort on a french accent. Great.

Eric 11:12

He loves doing a french accent, by the way.

Jennifer 11:14

Oh, really?

Eric 11:15

Get him started.

Jennifer 11:16

I mean, it wasn’t.

Eric 11:18

Don’t. I just started him. Will the rest of the podcast, he’ll.

Pallavi 11:23

Be doing a frank, not palliative. I’m just putting it out there.

Jennifer 11:32

I would be shocked if anyone on this call said anything other than, of course, usual care cannot be the control arm that just, it’s not ethical. We know that patients with serious cancer, like advanced lung cancer, need to be seeing palliative care. And I’m hinting at something we talked about before the recording started, but now that we’re going with it, Alex had asked about non inferiority and why that was the design and the reason non inferiority was the design.

If we accept that frequent contact with palliative care is the standard of care, and we’re trying to do something that entails less contact or less palliative care, that’s the rationale for it being a non inferiority design, because we have to make the argument it’s no worse. It would have been hard to hypothesize that less contact with palliative care would have been better. So again, that’s the rationale for non inferiority. But I definitely do not think it is ethical to have non palliative care. As Eric said, usual care, be a control arm when we’re trying to improve on early longitudinal palliative care models.

Alex 12:36

And most of our listeners are like clinician, practicing clinicians caring for older adults, people with serious illness. Can you, one of you, explain in lay terms what a non inferiority trial is and how it differs from. I guess we would say that most trials are superiority trials. Is this superior to that? But we don’t actually put the word superior in there. So what’s a non inferiority trial?

Jennifer 13:01

You guys want me to take this one?

Chris 13:02

Desperately, yes.

Jennifer 13:04

I feel like I want to make my statisticians proud. So it’s basically. So there are really. This is so unpleasant for a palliative care conversation, but sort of the three designs are superiority, non inferiority, and equivalence. Probably at some point in the next couple of months, we’ll talk about equivalence palliative care trial, and that one’s probably a little more easy or easier to understand. So, superiority, you’re just saying one arm is statistically better than the other. For an equivalence trial, you basically have the sample size and the power to look at whether one is better or one is worse. Basically that either one is better or worse.

An equivalence trial is kind of in the middle. You don’t need quite as much patience and quite as many patients and quite as much power as equivalents, but you need more patience and more power than superiority. So you’re basically saying that the experimental arm is no better or no worse, that it is equivalent, that within a margin of sort of variation in your p value, that the two arms are identical slash similar, and the difference in.

Eric 14:13

The design of those studies is it mainly just the sample size? The power, yes.

Jennifer 14:20

So the power has to be higher for a non inferiority and then higher for an equivalence trial than it does for a superiority trial, and that increases the sample size. So, for example, for a lot of palliative care studies looking at superiority, it’s usually say 200, 150 to 250 patients. This trial was obviously 450 patients. We have an equivalence trial which will be published soon. That was 1250 patients.

Alex 14:48

Oh, wow.

Eric 14:49

Wow. We are going to have to have you back.

Jennifer 14:52

I just misquoted my study. This study was 510 patients. Just ignore me. But you know what I mean.

Chris 14:58

You’re creating like the all hat study now of palliative care. Once you get to four digits of patients, like, we can never, we cannot aspire to five digits.

Jennifer 15:06

It’s just 400 or 500 patients.

Eric 15:11

Well, can we talk about this study? Let’s talk about what you did and who was included. Let’s start off with who was actually included in this study, this 410 ish patients?

Jennifer 15:22

It was 510.

Eric 15:23

I’m sorry, 500, 510. Yeah, you’re right. You said, you said 500.

Jennifer 15:27

But now someone else should answer.

Eric 15:30

Pallavi, give us a little background.

Pallavi 15:33

Yeah. So adults with lung cancer. So that includes people with non small cell lung cancer, small cell lung cancer and mesothelioma being treated with non curative intent. So I think that was one of the key eligibility criteria. English or spanish speakers. And they had to be within. Oh, I’m going to get this wrong. Within eight weeks of diagnosis of advanced disease. Jennifer.

Jennifer 16:00

So within twelve weeks.

Alex 16:02

So early-ish, is that similar to your knee? Jim, paper on early palliative care?

Jennifer 16:08

Yeah, so that is a great question. So in the New England Journal of Medicine paper, it was eight weeks. In these more recent studies, we’ve expanded it to twelve weeks, just because when we do multi site trials, there’s definitely a learning curve. Enrolling patients in palliative care studies. And our research coordinator teams at multiple sites were saying eight weeks is kind of tighten to enroll patients or, you know, have the time to approach and enroll patients. So for most of our multi site trials, we’ve expanded the eligibility window to twelve weeks from eight.

Eric 16:38

So everybody was diagnosed with an advanced lung cancer sometime in the last twelve weeks and then they were randomized. Either this stepped early, stepped palliative care versus early palliative care, which is the. Maybe not the usual care like Alex was saying, but it’s the evidence based care that we have from prior trials. Is that right?

Pallavi 17:03

We called it the early intensive palliative care.

Eric 17:06

Early intensive palliative care, because the usual care is the usual probably thing that happens is we wait for a palliative care consult. It’s not case finding. And then when they come in, we’ll see them, and there may be more intensive palliative care. We see them every month. Does that sound about right, Chris?

Chris 17:25

Yeah, I think a month is pretty uncommon, at least in our clinic. Like, yeah, if I could hang the stars, if I was taking care of 25 people, I’d see them every day and they’d all feel great, right?

Eric 17:37

Or overwhelmed.

Chris 17:39

Yeah, well. Or they’d have so much time, toxicity, they’d drop out of cancer care. Those would be the two options. You decide if you’d want to see me every day. But no, that you named it right. That four weeks. We probably do six to eight weeks on our typical care patients. If we have somebody who’s hit a level spot. Jennifer mentioned earlier this idea of prognostication and the people who are the pd l one expressors. I learned more about people’s dogs and about their grandkids. Cause they were fine. There were a couple of people on my patient panel who were early intensive palliative care, and they did wonderfully for a long time outside of this study. I wouldn’t have seen them that often. So that’s where that full time toxicity question comes up.

Eric 18:28

So they were fine until they were not fine eventually, and some of them.

Chris 18:32

Were fine until the study ended, because we have some people in our clinic who are 4567 years stable metastatic non small cell lung cancer, stable on immunotherapy. It’s amazing. Back at that capacity question, right? If we’re still getting people in, but instead of living six months, they’re now living six years. That’s a lot of appointments that we have to account for over time.

Eric 19:00

So let’s talk about what was in the syringe in both groups. When we’re thinking about a palliative care trial. I was thinking about, okay, what do they mean by palliative care? What was in this stepped wise or intensive palliative care group? What was the team made of? Because, I mean, I can think of it like a spectrum. On one hand, in palliative care, it’s like primary palliative care, people who’ve got a little bit of teaching. On the other hand, you got a full interprofessional team. You got palliative care psychologists and social workers and pharmacists and MP’s and physicians and nurses, you know, the chaplains. Yeah, the full model that we love to talk about, that actually rarely happens in the real world. Pahlavi, who was, what was the team made out of?

Pallavi 19:45

It’s interesting. I think it actually varies a little bit, even among our three sites in terms of the intensive arm. So at our cancer center, we have physicians, advanced practice providers. We do have access to palliative care social work in the outpatient setting, and we have a nurse. We don’t have pastoral care actually readily available for our ambulatory patients. And so that’s sort of our interpretive team or interdisciplinary team here. And then the sept arm, I think we were pretty clear that they could not be receiving home palliative care services. That was an exclusion criteria. And then they, I think that was an exclusion criteria in both arms, actually. We did not want them to already have seen home palliative care.

Eric 20:26

And that includes home hospice care.

Pallavi 20:28

Yes, definitely. And then we also didn’t give them access to palliative care social work on the stepdarmouse. They had access to the cancer center resources, but they didn’t have access to sort of regular contact with palliative care mental health professionals. Was that the same at MGH for.

Jennifer 20:48

We don’t, you know, our palliative care team here in the outpatient setting is only physicians and advanced practice providers. I would say generally, though, to answer your question, Eric, you know that palliative care content and clinicians were identical in both study arms. So those were intentionally not varied, because then you’re not just comparing the cadence and the frequency, you’re comparing different interventions. So, for example, the intervention visits had to be delivered by a physician or advanced practice provider.

Eric 21:19

So I got a question for you, Chris. You brought up ICU study. It was titled like, had the word palliative care in the title, I think.

Alex 21:31

Blog post, if we’re talking about the same study, was like, fast food palliative care, in and out palliative care, something like that.

Jennifer 21:37

In and out.

Eric 21:38

Palliative care doesn’t work.

Chris 21:39

It was nationally representative. So it was not in and out because I can’t get those burgers here. It was just fast food generally. You’re right.

Eric 21:46

So this study showed that in people with chronic critical illness, having palliative care lead one and a half sessions in an ICU around prognosis and goals. I’m not even sure goals of care was a huge part of it, didn’t actually help, but actually worsen potentially some PTSD outcomes. And the entire community of palliative care said yes, not a palliative care study. Sure, it has palliative care in the title. Sure, there was a palliative care led prognosis, meaning.

But it was not a full interprofessional palliative care team. They didn’t include all these palliative care NCCN guidelines or. Sorry, I always forget that. Yeah. Not all the domains of palliative care were included. I wonder, when you think about that trial versus this, which you can also make the argument that this was not a full palliative care consult team. How do you think about that? I’m bringing it up since you brought up that trial.

Chris 22:52

I know, I know. I’m happy to. And I’m going to have to call the first author and tell them that I love them still. Cause I still do. But the idea of do this thing and we’re gonna call it palliative care versus do palliative care, and I. We’re what? We’re nine years after that original article. Right. Or the article that, the fast food article, to use the parlance from the blog post. So I think we’re in a different spot now. I think we’ve.

That article really helped me to realize if you’re overly prescriptive about what a palliative care intervention is, it’s less likely to work. Palliative care is zigging and zagging and following the patient where they are. So being told, you are an expert at this, we trust you do palliative care. It makes it hard to report what’s in the syringe, but it’s putting the stuff in the syringe that the patient needs.

Eric 23:45

Yeah, I love that, too, because I think that one for me really is, can our interventions need to target the patient population that we’re managing? So in chronically critically ill patients, the idea that one visit is going to do anything is just. No, I think it was a great trial. I think we learned a lot from it. Palliative care. I actually think it’s a great palliative care trial. And I wonder, Jennifer, from this study, were there other things that putting this together that was prescriptive that these palliative care physicians, advanced practice nurse that they were always going to do, or was it, we’re going to let you be do your thing, but this is who the team members are going to be.

Pallavi 24:29

Yeah.

Jennifer 24:30

So the team members were prescribed that at least intervention visits had to be physicians or advanced practice providers. As Pahlavi said, you could refer to other providers. The intervention, I would say, was somewhere in the middle of scripted versus not scripted. I mean, the pro of scripted is we have fidelity, we know exactly what happened. The con of scripted is exactly what Chris says. You come in with an agenda, that’s not what the patient’s agenda is. So we trained all the palliative care clinicians on the model that we mostly deliver here at MGH. That early visits, when people who often look and feel like all of us, even though they’re on chemotherapy or immunotherapy. It’s about establishing a relationship and rapport. It’s about instilling coping skills.

And if they have symptoms, you address symptoms, and at some point, you might elicit goals and values, and at some point, you might talk about end of life care. But it’s a very different model. It’s an outpatient model where establishing relationships, as Chris said, getting to know what their grandchildren are doing, getting to know what they like to do, so that when the hard times come 3612 months later, you have that trust and relationship to make those conversations happen, easier to facilitate them. What was very scripted was when palliative care visits occurred in step and how they occurred. So again, the step model is patient centered. We monitored patients to know when they should step up to more frequent palliative care, and we did that with a quality of life measure.

And that was very intentional. We did not just rely on a physical symptom measure, which many of our colleagues are doing. And the reason for that is the nature of our palliative care intervention. It’s not just about physical symptom management. It’s about coping. It’s about psychological support. And so we thought a quality of life measure, which incorporates physical well being, functional well being, emotional well being, social well being, and physical symptoms, was the right trigger to say someone needed more palliative care. Additionally, in case patients didn’t step up, we did have sort of events where they were scheduled to see palliative care. So if their cancer treatment changed or they were admitted to the hospital, they were always triggered to have a palliative care visit, just because we know those tend to be very physically and emotionally taxing events in a cancer patient’s life.

Eric 26:47

How soon after from a hospitalization, all.

Jennifer 26:50

The palliative care visits were triggered? Generally, we gave palliative care because of the reason, exactly what Chris said. Sometimes we can’t get people in with palliative care. You sort of had a month after all of those events. So once the team saw that you had cancer change in your cancer treatment or admitted to the hospital, you were scheduled to see palliative care within a.

Eric 27:07

Month of that event, did they also see inpatient palliative care?

Jennifer 27:11

Yes.

Eric 27:12

And that was, was that a trigger in the stepped group, that everybody who got admitted saw inpatient palliative care?

Jennifer 27:18

So that’s a great question. So in stepped palliative care, it was sort of up to the team in the, what Chris and Pallavi are calling intensive. I haven’t called it that before, but I might steal it. Then inpatient visits were mandated.

Eric 27:30

Oh, truly intensive. Like you’re seeing everybody. Every four weeks, they get admitted, they’re going to see inpatient palliative care. After admission, they also see palliative care.

Jennifer 27:43

I know we’re getting you confused. And the early integrated palliative care was every month. And if hospitalized, and then the step quality of life is monitored, if their quality of life dropped, they stepped up. But for patients who weren’t stepped up after a change in cancer treatment or hospitalization, they would have palliative care visits.

Eric 28:01

And, Ken, I asked Pallavi, maybe you know this, so you’re doing these quality of life surveys. I’m trying to figure out how would this be implemented in, outside of a research setting? How are the, how are the folks getting these quality of life surveys? How are you getting them back? Great question. Is this practical outside of a research setting?

Pallavi 28:23

It depends. Which is not a great answer, but. So it actually changed over the course of our study because, you know, before COVID everything was in person, and we’d have these lovely research assistants and coordinators running around sitting with patients while they did the surveys, getting them back in real time, or hounding them to send them back by mail or email. And then we started pushing a lot more surveys electronically to patients, which, of course, has its challenges, because not everyone’s as tech savvy as, you know, our highly educated patients who have access to technology. So it even changed over the course of the study.

Pragmatically, I think as more and more cancer centers and health systems move towards collection of pros, I think it is going to be more feasible. But pros, patient reported outcomes. I’m so sorry. Jargon. I think bigger cancer centers and health systems can probably implement these, but I think it’s tough in a community based practice to reliably collect these without boots on the ground.

Eric 29:20

Yeah. Chris, do you use them in your practice?

Chris 29:24

So it’s an interesting question. We had some, something that was required of all patients. And the thing, if they didn’t fill it out, they couldn’t join the video visit. So people would be three minutes late because they’d be feverously clicking these boxes and we didn’t ask for them and we didn’t read them and we didn’t look at them. So not what we’re talking about within this trial, but there was a patient reported outcome. I begged our tech people, I said, get rid of this. This is a barrier to getting the patient in the door.

Now, if I had somebody who could catch these things, similar to what the research assistants are doing, or if I had a dashboard that Mister Smith three months ago was this and now is this, that would be so useful to me. So I think that’s the future. Is being able to automate some stuff through the portal and then have a dashboard on the clinician side of look, these are the domains that the patient declined in between these two visits. Well, I’m not going to wait for them to bring it up. I’m going to bring it up in our next scheduled visit.

Eric 30:27

Yeah. Or if you have a team huddle. Oh, these ones are in the red. We need to.

Chris 30:32

We aspired within our group. Now we have two nurses for 8.5 ftes. So we’re like super tight in terms of our support staff, but we aspire to have like ten patients or so that we call the hot list. So these are the patients who are. We’re actively changing methadone. They’ve been in the hospital twice in the last six weeks. We want our nurses to keep up with those patients, but it’s just, it’s really hard to implement when the prior authorizations all need to get done.

Alex 31:00

It’s real world.

Eric 31:02

So let’s hear. I was going to say, did it work? But that’s not the question, right, Jennifer?

Alex 31:06

That’s not the question.

Eric 31:09

Was it, did it not work? I’m not even going to say, was it equivalent? Because that’s not the question either. I just learned.

Jennifer 31:15

I feel comfortable saying it worked, and I’ll tell you why. Because quality of life was non inferior. So quality of life was, if we want to talk about just the as good as the more intensive model, but we used significantly fewer palliative care resources. So basically, to achieve the same quality of life benefit as the intensive arm patients saw palliative care, significantly fewer.

Eric 31:38

Half is less, right?

Jennifer 31:40

About half. Is that good English? I’m not sure.

Eric 31:42

About half. Good enough for Jerry. That’s GeriPal grammar.

Alex 31:44

People understand, though.

Jennifer 31:46

Jerry Pal Grammar, that’s about half as much.

Eric 31:52

And the quality of life included symptoms, right? Their symptoms weren’t worse. Quality of life wasn’t worse.

Jennifer 31:59

The quality of life tool includes sort of the four domains that I mentioned, physical, functional, emotional and social well being, and that it also has a lung cancer specific subscale. So it’s the fact measurement system. So there’s like a fact lung effect, brain effect, colon. And so they always have those four domains and then a cancer specific. So, for example, the lung cancer specific one has questions about cough and shortness of breath anorexia. So the quality of life measure encompasses all of those domains.

Eric 32:29

And the secondary. Now, no worsening in depression symptoms, or I guess, equivalency as far as depression symptoms, use of coping strategies, prognosis, communication about serious illness or end of life care. Just no difference between any of those outcomes.

Jennifer 32:46

No difference in any of our patient reported outcomes. That is correct.

Eric 32:50

All right, I’m going to turn. This is the tough one. There was a difference, though, in hospice utilization, right?

Jennifer 33:00

Correct. There was a difference in length of stay in hospice. So the length of stay in hospice in both study groups was longer than, I think, any other study that’s ever been published, but officially, it was longer in the intensive arm than the step arm. And I have not asked Chris or Pallavi on their opinion as to why that is. So I’m very excited to hear their thoughts on why they think that was.

Eric 33:23

Pallavi, I want to hear from you first. Why do you think so? I’m looking at the data right now, adjusted mean days, 20 days versus 35.

Pallavi 33:33

Days, which hats off to the intensive arm. 35 days is amazing and many, many times the national average. Right. So my thoughts, and this is speculation on my part, but I do wonder if there was a slightly more sort of orderly transition to hospice on the intensive arm, because we had these social visits, these rapport building visits, we had a chance to address that sort of low level existential distress that I think we can pick up that may not be picked up in the fact l, but may be picked up in these very organic clinician patient interactions.

And so I wonder if the. And this is where I think qualitative data from caregiver bereaved caregivers would be super useful. On the lower touch arm, the stepped arm, was it a slightly more chaotic or sort of, you know, not as smooth transition to hospice? Maybe it happened a little bit later because it just wasn’t as smooth and we hadn’t had months and months of monthly visits to talk about it. The what ifs.

Eric 34:41

All right, Chris, my theory, and you can tell me if this is a dumb theory or not. My theory is that more people were seen by inpatient palliative care. And that’s where a lot of hospice referrals happen for these individuals when they’re really declining and they get discharged with, with hospice. So you’re seeing. Because everybody in the intensive care got inpatient palliative care or they were supposed to. That’s the reason. Stupid theory. Do we know of any data and I haven’t seen.

Chris 35:11

It’s an interesting question. I don’t know if we captured where the hospice referral came from. What my thought is, I like Paul Lavie’s thought. My thought is probably driving alongside hers, which is that patients sometimes need to be told, I love you, I know you. It’s time now for hospice. And if you wait for the hospitalization, for the inpatient team to be the one to say that you’re going to get hospice, but you’re going to get it for a shorter time because things have already decompensated, where if we can do that in the office without the hospitalization, that’s when people have longer times on hospice is when it can be controlled.

Eric 35:53

So you’re saying the exact, exact opposite. Jennifer, do we have any data to suggest which theory fits closer?

Jennifer 36:06

We are investigating it right now. I think one thing that’s a bit tricky about it is just to be careful to not make inferences about subgroups. So we’re looking at people who are sort of on the low touch versus who stepped up. And it gets a little confusing. You’re slicing and dicing the population. The data on hospitalizations by study group we are just starting to analyze right now. So you’ll have to have us all come back in a couple of months so we can tell you if there’s a difference in hospitalizations and palliative care exposure. My guess is it’s not that I don’t think that we’re going to see differences in hospitalizations between groups or exposure so much to palliative care during hospitalizations.

My guess is it’s more what Pahlavi was saying, that the sort of less intensive palliative care was able to do everything else except a super early referral to hospice. I don’t want us to forget that 20 days in hospice is still something that is still pretty amazing. And so even though there was a difference, it’s probably worth the discussion. Is that difference matter? Is that something we have to chase, or can we say the length of Stan hospice or step palliative care was 20 days? That’s amazing. And it’s good enough. I don’t know.

Eric 37:20

Pallavi. Good enough.

Pallavi 37:22

Good enough. In terms of the length of stay.

Eric 37:25

When you’re thinking about this study, I’m trying to put this together in my mind. We have this study that shows less intensive palliative care compared to intensive palliative care actually was non equivalent in pretty much everything we measured except length of stay in hospice. Like, how much weight do I put in that hospice use secondary outcome?

Pallavi 37:48

I don’t know that we know exactly the right dose of hospice to get max benefit. Like, when do you plateau? Right. We know that is usually better, but we’re looking at short times being hours to days. So I don’t know how clinically meaningful that week difference was to families. That’s really who we’re talking about, who got the support they needed to take care of their dying loved one. Here’s what I really want to know. I want to know how are we going to better predict who’s going to step up? That’s what. And someone’s got to say it this hour. AI, right?

This 2024 is the year of AI. How we can we predict who’s going to step up? So I can give those people that intensive care just, you know, from the get go. I don’t know. I don’t know that there. I don’t know we’re going to know the answer to that. But that’s really what I wanted to know after we looked at the results.

Alex 38:37

Yeah, speaking of, like, a practical approach, one thing that was mentioned in discussion, I’m not sure if this was like, the reviewers asked you to do this using prognosis is a trigger.

And, you know, we’re kind of like, we do e prognosis. We do a lot of prognosis stuff. And I completely agree, like, quality of life should be the trigger. Like a decrement in quality of life is the ideal trigger, not prognosis, not a shorter prognosis. And it’s a lot easier to calculate prognosis likely from, you know, the labs and the medical record than it is to, like, get patients to stomp onto a survey about quality of life. So there’s some trade offs there. I don’t know if any of you.

Eric 39:14

Have thoughts about that or are both important. Like, quality of life is really important, but also, like, when somebody’s prognosis suddenly changes through the much worse, is. Shouldn’t that be a trigger as well?

Jennifer 39:26

Yeah, I mean, I think if I think about the next step, which is actually what I’m going to call the next study, isn’t that cute? Next step I think exactly what you are saying is what we need to do. We need to combine stepped care with AI prognostic models. Polybe knows this. I’m not sure if Chris knows it as well, but we have, and if anyone is on the study section is listening to this, you are meeting next week, so hopefully this will convince you. We have a proposal that’s been to NIH several times using an AI prognostic model to identify palliative care needs, and it has been trashed multiple times for a wide variety. I know this is so inappropriate, but I’m just going with it now. Maybe a reviewer is listening because of concerns about the ethics of using AI to identify palliative care needs.

Alex 40:15

Huh?

Eric 40:16

What’s the concern there? Because we’re using AI to do everything right now it feels like we sure are a reviewer.

Jennifer 40:23

We sure are.

Jennifer 40:31

I feel like Kamala Harris about drama. Next question. And what’s the risk of someone seeing palliative care earlier than they were supposed to? I don’t know, but I’m hoping the third tries of charm, but I make promises.

Alex 40:52

I have another question about relating to your prior study. One the landmark reason, let’s just call it, you know, speaking of, can we say it on this podcast? AI the need. Jim, paper on early palliative care. The big reason it made headlines wasn’t the improvement in quality of life, which was great. It was the length in survival, you know, in the patients randomized to early palliative care. Now, I notice survival wasn’t a component of this study. I wonder what the reasons for that are. And I might it be that my hypothesis is that those small gains that might be seen in a palliative care arm in terms of survival are dwarfed by the gigantic differences that we’ve seen in the intervening ten years in cancer directed therapy.

Jennifer 41:38

Yeah, that’s exactly right. You really could not look at survival in an unselected metastatic lung cancer population in this five to almost ten year period. Now, there are just too many sort of predictive and prognostic factors that we would not be able to control for. So that is exactly correct. But it goes to how you kicked off this podcast. We don’t need a usual care arm anymore, and we don’t need to show the value and benefits of early palliative care. It has been shown, done. Close that chapter.

Eric 42:09

Well, in wait, can I just add, what was the survival in this group? Do we know how long people lived?

Jennifer 42:17

We made a very thoughtful decision to not look at it, because if we do, people are going to ask for the differences between study groups, and that is a flawed analysis. And I don’t think we should be looking at data that we don’t feel like we have the rigor to make inferences about. So we are not looking at it.

Alex 42:36

You’re saying it’s a flawed analysis because as what you said earlier, there’s so many components, what goes into survival for these different groups that it would be hard to make comparisons that were fair. You couldn’t digest for all those factors because you didn’t collect them or you don’t have access to them or whatever.

Jennifer 42:52

Correct?

Eric 42:53

Yeah, sorry, Alex, you were going to say something?

Alex 42:56

No, I was just going to say that that case is shut, that palliative care should be usual care in advanced lung cancer and that for non cancer illnesses, for other conditions, there are still questions where usual care might be appropriate. Is that fair to say? I’m workshopping what I’m going to say at the “Congress du caretive” [speaking in French]. Thank you. Thank you.

Chris 43:26

You’re welcome.

Eric 43:27

Yeah, I mean, I think Alex is a great point. So we have this for lung cancer. Do we need to repeat all of this for different kind of cancers? Is it really open and shut for other conditions or do we need randomized controlled trial for different conditions, different stages of conditions? I think palsy. I sure hope not, Chris. No, hope not.

Chris 43:50

Because if you, you know, if we go back and you think about palliative care as cross cutting. Right. And if we go back to every single underlying condition that we can help with and if we’ve got to build that house from the dirt, we’re going to be talking about this in 2086.

Eric 44:10

Well, we have. Because every ten years we’re going to have to update it too, to keep up with.

Chris 44:13

We still haven’t gotten to Huntington’s disease, but you know, we’ll get there by 20, 116. Right. Like at some point we have to say we don’t hurt people. We know that the question of, as a limited resource, who are the right people for us to help, that’s in my mind probably still open. But I think when we have an opening like an ASCO guideline, we should drive the truck through there and then. I love the idea that really this trial was about is how much is enough? Because if we give somebody a lot, that means at least at my institution, some other people are not going to get any.

Eric 44:49

So I want to be mindful of the time. I want to ask kind of what’s next? What do we do with this? And I wonder from each of you. I’m going to start off with you, Chris, kind of what is next? What do you think the next thing we should either do in your own clinics with this data, or maybe from a research perspective, you can take either one of those.

Chris 45:08

Yeah. And as the non researcher in the group, if I could kind of hang the stars around this, what I would do is build some patient reported outcomes that allow me to know my patients, who I see on kind of the routine refill every ten weeks, every twelve weeks, who is suffering in that group, and I don’t know, so that I can pull them in in an earlier.

Eric 45:31

Slot, see, like a dashboard with how they’re doing.

Chris 45:34

Exactly.

Eric 45:35

So similar to this study, except kind of automated, no researcher research assistance needed, and then you’re looking at that dashboard.

Chris 45:44

I would love to build that. Yeah.

Eric 45:46

Pahlavi?

Pallavi 45:48

I think I would like to see us operationalize these critical clinical points and use them as triggers for sort of, kind of an automated palliative care consultation. So if a patient with advanced lung cancer needs palliative radiation, a change in therapy, or an unplanned hospitalization, to me is actually the number one sentinel event that always alarms me. I feel personally ashamed when my patients get admitted to the hospital. So when that happens, I think that should trigger sort of reconnecting with palliative care. So I think that’s actually fairly easy to operationalize and shouldn’t require a great deal of manpower. I think the quality of life assessment was really brilliant in this trial, and I, on my wish list, would be a way to operationalize it that doesn’t add to clicks or headaches or make clinicians want to pull their hair out. But I do think that’s really important.

And then my third wish is that if we were operating in a stepped care model, is giving folks access to people like palliative care social work, because we know so much of what’s in the syringe is coping support and that rapport building. We have found that when we partner with our palliative care social worker, even if we’re trading visits, patients feel cared for and they don’t have to have a financially toxic visit every time they can do a virtual visit with a social worker and still feel connected to the team. So those are three things I’d love to implement, like today.

Eric 47:12

And Jennifer, next steps.

Jennifer 47:15

I would like us to continue to maximize palliative and supportive care for patients with newly diagnosed advanced cancer and used fewer palliative care resources. So my next step is to try to integrate digital apps and videos and other things that deliver palliative and supportive care.

Eric 47:34

AI palliative care versus human palliative care.

Jennifer 47:37

There you go. That’s it. That’s the end of the podcast.

Pallavi 47:43

Very fitting.

Eric 47:45

Well, we could talk much more about everything. I just want to thank all three of you. But before we end, and Alex, I think a little bit more is it Galileo?

Speaker 6 48:01

Offer thanks to those before me. That’s all I’ve got to say. Cause maybe you squandered big bucks in your lifetime. Now I have to pay. But then again, it feels like some sort of inspiration to let the next life off the hook. And she’ll say, look what I had to overcome from my last life. I think I’ll ride above how long till my soul gets it right? Can any human being ever reach the highest light except for the resting soul? Galileo, king of night vision, king of inside.

Chris 49:07

Alex Smith, Indigo guy, everybody.

Eric 49:14

Follow me. Chris, Jennifer, thank you for being on this podcast. That was wonderful, wonderful study. We’ll have links to the study and to some of the other things we talked about, like the ICU study in our show notes today. Thank you all.

Pallavi 49:27

Thank you.

Jennifer 49:28

Thank you. Bye.

Eric 49:30

And thank you to all of our listeners for your continued support.

***** Claim your CME credit for this episode! *****

Claim your CME credit for EP327 “Stepped Palliative Care”
https://ww2.highmarksce.com/ucsf/index.cfm?do=ip.claimCreditApp&eventID=13848

Note:
If you have not already registered for the annual CME subscription (cost is $100 for a year’s worth of CME podcasts), you can register here https://cme-reg.configio.com/pd/3315?code=6PhHcL752r

For more info on the CME credit, go to https://geripal.org/cme/

Disclosures:
Moderators Drs. Widera and Smith have no relationships to disclose.  Guests Jennifer Temel, Chris Jones, and Pallavi Kumar have no relationships to disclose.

Accreditation
In support of improving patient care, UCSF Office of CME is jointly accredited by the Accreditation Council for Continuing Medical Education (ACCME), the Accreditation Council for Pharmacy Education (ACPE), and the American Nurses Credentialing Center (ANCC), to provide continuing education for the healthcare team.

Designation
University of California, San Francisco, designates this enduring material for a maximum of 0.75 AMA PRA Category 1 credit(s)™. Physicians should claim only the credit commensurate with the extent of their participation in the activity.

MOC
Successful completion of this CME activity, which includes participation in the evaluation component, enables the participant to earn up to 0.75 MOC points per podcast in the American Board of Internal Medicine’s (ABIM) Maintenance of Certification (MOC) program. It is the CME activity provider’s responsibility to submit participant completion information to ACCME for the purpose of granting ABIM MOC credit.

ABIM MOC credit will be offered to subscribers in November, 2024.  Subscribers will claim MOC credit by completing an evaluation with self-reflection questions. For any MOC questions, please email moc@ucsf.edu.