Thank you for having me. I'm at the Oregon Health and Science University. And I'll apologize, because I'm an epidemiologist in my heart. So I guess I sort of really took this opportunity as a way to kind of delve back into the literature. And I was having so much fun that you'll see that I'll skip some slides, just because it's really just a way for me to kind of recognize sort of some of the things that we'll be talking about today. Now, how do I go forward? Let's see. I have no commercial interests. And these were the objectives that were given to me. We'll be really talking about the outcomes that families have after the death in the ICU. And then we'll talk about some strategies that we can glean from the literature about how to mitigate some of these outcomes. And I think in general, we should just ground ourselves in the notion that when we're talking about bereavement, we're talking really about, although we don't necessarily like to think of it as a medical condition, I think we can really leverage the disability justice movement to really understand it through the lens of the biopsychosocial sort of factors. And so the grief process really can create health problems, which can create difficulties with function. And can really limit person's social engagement in their world. And these challenges really can be sort of sustained or enhanced by sort of personal factors, whether it be sort of coping skills or psychological factors and psychosocial factors, environmental factors as well. And the other thing I think that is a more modern concept is that, you know, we have this sort of more Freudian notion of bereavement being letting go of the person that has died. I think a lot of the modern theories of bereavement really are centered in this notion that it's not about, you are losing the person, but you're also learning how to re-engage and restore the relationship in a new way. And so it's not about necessarily losing completely the story. And so that will be relevant to how we interpret some of the literature that you'll be seeing for mitigation. And I think we've sort of convinced you that, you know, bereavement care in the ICU is relevant. I think relevant here is that, you know, I think Judy Davidson's work certainly highlights the notion that, you know, the ICU is a crisis environment, right? So patients are coming in and crisis is a sort of biopsychosocial event as well, right? It can certainly instigate the organism into certain fight, flight, freeze responses. And I think that is going to be relevant to how we then interpret a lot of the data that's going to be coming forth. Looking back, families, we sort of are always beholden to the survivors in cross-sectional samples when we get data about long-term outcomes, right? And so families, this is a broad samples of families outside of the ICU, asking them about what they need after their death, after the death of their loved ones. And what you'll see is that there are some patient unmet needs, meaning families talk about what the patient suffered with, the dignity of the patient, the emotional and spiritual support that the patient might have wanted. And then they talk about their own sort of needs, whether it be their emotional support or their needs for better decision making. When families in the ICU look back, the struggle that we have is that the response rates tend to be very variable. And I'll highlight that in the next slide. But it tends to be very variable and very low. And so the people that we're actually sampling tends to be a very small percentage of the actual ICU patients or families that are surviving. And they talk about the perceived conflict, which really tends to be rooted in communication and staff behavior. They talk about the lack or presence of support. That's about pastoral care, understanding patient preferences, and the sort of social environment that they could be in in the ICU setting as well. Right. Like how permissive is the environment for them to be able to get the psychosocial support that they really need to be able to make the decisions that we're asking them to make. Right. And there's this whole literature with family satisfaction. And we don't have to review all of the studies here. But I think the bottom line is that families, on one hand, are very satisfied. The ceilings go through the roof and a lot of the nursing domains for satisfaction. And yet, they tend to rate the satisfaction and communication lower than all of the other domains. And there is one cross-sectional survey that suggested in a U.S. sample that the survivors' families of patients who had died were more satisfied than families who the patient had survived. Now, this was in a U.S. sample, in a Canadian sample that was not present. And so, again, I think like with all of the literature in the ICU outcomes, we'll have to really kind of really interpret a lot of this stuff into the local context. And so what are the symptoms that families are struggling with? Right. Early in the ICU, right away, you notice that there's lots of cross-sectional data to suggest that families are struggling with anxiety, depression, and post-traumatic stress symptoms, which I guess at this point we can think of as acute stress response. Again, the ICU is a crisis environment. So you can look at sort of the large French. And the French groups have really forged the way in this particular area of research. Right. And so this is a multi-center study looking at sort of the rates of anxiety and depression for patients who are at high risk for death in the ICU. In a smaller sample, you see similar rates as well. And what you start to see is that there are these sort of factors that are related to symptoms, whether it be the relationship that the family member has with the patient, whether it's a spouse or a child, those become factors that might be relevant. The characteristics of the patient themselves, whether the patient was chronically ill or quickly dying, those might become relevant. And then some of the organizational care factors might become relevant as well. Right. Like whether the family was involved in actual medical decision making, end of life, those kinds of things. And similarly, when you look at sort of post-traumatic stress symptoms in the sort of survivorship zone, right, so patients had died and they reached out to family members within 90 days of the outcome. What you notice, first of all, is this is one of the studies where you start to really see the challenges in sort of reaching patients to ask about some of these things over the phone. The very nature of post-traumatic stress symptoms is that some of the domains that you'll see named in this impact of event scale is re-experiencing, intrusion, avoidance, and arousal. So you can imagine we're calling patients' families, and if they were having some of these symptoms, they would be precisely the kind of people I would worry may not want to talk on the phone about their experience in the ICU. And so you see a lot of dropout. So 432 people might have said, oh, sure, call me in 90 days. 60 of them refused right off the bat. They're like, I don't want to do this, right? 372 of them consent for follow-up. But then after 15 average attempts, 88 of them can't be reached, right? You actually have two or three telephone numbers for these people, and they can't be reached. And I think it just goes back to say, I think that people, there is a sort of, we should worry about whether families want to talk to us about the past, particularly past that might be traumatic. And anyway, so we get sort of a small sample, and they do the study. And what they find is that, as you might expect, families for whom the patient had died in the ICU have more rates of anxiety, depression, and post-traumatic stress, and the rates are a little higher in those families, because this is a French study, when the family was involved in end-of-life decision-making. And then they do sort of start to look at some of those organizational factors, right, like the perceived incomplete information, the number of beds in the ICU, although the effect estimate is not very big, right? So I think you get a sense that the communication in the ICU and the information that families are getting kind of plays a role in how they perceive their stress, at least at the 90 days after discharge. This construct of complicated grief really starts to get then at, you know, the prolonged grief that families might suffer, and it really starts to include not just the high level of symptoms, but then also how maybe these symptoms might affect their function, right? So the complicated grief is really kind of meant to be a very kind of impaired subgroup of patients. And in this particular study, they looked at post-traumatic stress, and then they looked at caregiver complicated grief six and 12 months out. And again, what you're seeing is that the rates are quite high in the 40s for post-traumatic stress symptoms, and then complicated grief as well, I think, is almost like so high that it makes you wonder. And then they look at some of these factors as well, like the organizational factors, like whether the communication with the intensivist was satisfactory and whether those things were associated with complicated grief. And what you see is that a lot of them are. What do you make of that? I don't know, right? And what's interesting is that these next studies look at the over the time these symptoms continue. And what you see is that we find more hints that maybe sort of the types of communication, the types of activities that families are doing in the ICU may be relevant to how they perceive their symptoms in the months ahead. This is in a Taiwanese study where the prevalence is significantly lower, and it does diminish over the first year of the outcomes. But yet, you know, these are still concerningly high. And I'm going to skip these and then just kind of go to this notion of that these psychological distress are really correlated, right? They're not just kind of working independently of each other. So I like this study for sort of looking at the notion of how are these symptoms, anxiety, depression, and post-traumatic stress, related to each other? And how do they maybe relate to the prolonged grief that we might see in families? And what you're seeing here is that in this same population, Taiwan, where the prevalence is not as high in Europe or in the U.S., one month out, there's three subgroups of families, right? There's families with no distress. They don't have any anxiety, depression, or post-traumatic stress symptoms. And a month out, it's about 56% of the families. And then over time, you see that number starts to increase. So people are mostly getting better over time. And there's a subgroup of patients who have severe depression, anxiety, and then there's a smaller subgroup of people with severe anxiety, depression, and post-traumatic stress. And what you see is that the risk of becoming a complicated griever or a prolonged grief disorder is higher if you're in that initial subgroup of severe depression or severe anxiety, depression, and post-traumatic stress. But thankfully, in this particular context, the patients are mostly moving in the right direction over time, over the first year. So what do we say then in terms of the outcomes for families? I think what we're suggesting then is that psychological symptoms of depression, anxiety, and post-traumatic stress and complicated grief are common. The studies are poor. We've got a lot of selection bias to struggle with, partly because of what we're studying and also because of the convenience samples. Like, for example, we don't have any studies in middle-income countries, or very few of these studies have underrepresented minorities, even in the US, which I think should make us worry that we're definitely missing people. And I think what's good about those studies, though, is that they've really focused on some of the practices in the ICU to give us a hint as to kind of what can we do better to kind of improve some of those symptoms. And a lot of it centers around our practices on communication and end-of-life. What's missing is that, like, we know that the body keeps the score, right? So a lot of these studies don't get into the details about other symptoms, like cognitive impairment, sleep, even health impairments, whether it be new medications or things like that. And that might be very important. We know that the caregiver literature suggests that people, when they're struggling with emotional stress, have a lot of body symptoms as well. And we don't have any good studies looking at sort of people's baseline psychiatric history, the social network in which they live, and those kinds of things that will be very relevant to sort of how they then mitigate some of these outcomes. So what can we do? The studies have been alluded to already. I like this study because it's just a simple brochure about bereavement, and then they do a structured communication, the value mnemonic, and they found important differences in the intervention compared to the control group. And Dr. Brown alluded to this particular project as well, where we're really just focusing on asking families what they would want us to focus on for the patient who is dying and really implementing three wishes. In this particular project, I'm looking at an implementation in California where they just focused on implementing it over six ICUs, over two academic medical centers. And they looked at sort of how families then perceive the satisfaction after the process, and they find significant improvement compared to the families who were not exposed to the three wishes program. So that's reassuring and helps us think about maybe ways that we can be better present for families. The notion of family presence, I think, is an important one. I think going back to this notion that it's not about, like, letting go of the patient or letting go of the memory of the person. Sometimes this might explain why this literature around families being present during events has been shown in some studies to actually improve the psychological outcomes. It might have to do with their ability to kind of integrate the memory a little bit better, have a sense of what's happened. So this is a clinical trial that looked at this and showed some important outcomes, again, from the French group. Intensive care diaries are very, very commonly used in Norway and Sweden. The studies are quite small, and so, again, we can't say that it's necessarily recommended. And yet when you look at it, it's kind of similar to what we're talking about in terms of, like, integrating memory. And so a lot of the literature is really qualitative, asking families what they perceive to be the benefit, those kinds of things. And so far, we don't see a lot of signal for harm. More interestingly, there's been a recent pilot study that's completely different from diaries, but also kind of like a narrative pilot. Again, like asking families to kind of speak and tell their stories about what they felt was going on, and that might be another avenue for kind of memory integration. The bottom line is I think that we probably still have to really think about integrating the multidisciplinary team, right? And we alluded to this already, the spiritual support, getting an ethics consult of their values, conflicts, allowing time. The families are under a lot of stress, and they're probably not processing a lot of what we're saying. And so the communication has to be really frequent. We have to really chunk it up into small sizes. We've got to try it in different ways. We've got to be willing to be present, like actually present, right? And let things play themselves out a little bit more slowly. And then we have to assess families for risk factors, because we know that there are some risk factors for bad outcomes, for complicated grief. And that could be helpful, I think, as well. The condolence letter intervention, I think, is so funny, partly because I used to, as a church member, I used to do pastoral care. And one of the three options we had as a leader in pastoral care in my church was you could send a card, you could pray, or you could be present with the person. And most of the time, you would do all three. And so I thought that it was very funny, the notion that a hospital would send a card, and that's it. And that would change what? And yet, that's what they did. They sent a card. It was a personalized card. The physician and the nurse wrote a card 15 days after the death. And lo and behold, they saw not just a small sort of signal at one month out, which was the primary outcome, but six months out, they still saw a worsening of the depression and anxiety scores. And so that was concerning. They did a subsequent qualitative analysis of the data, and they didn't see any real sort of – the families felt supported by the card and felt like – And so it's probably more complicated than just the outcomes that we're measuring. But I would argue that maybe that's not the right outcome for sending a card. I think maybe it's something else. And so I think the strategies need to get real, right? Like, we need to really ground ourselves in the bereavement research from psychology if we're really going to try to do complex bereavement interventions, really maybe integrate qualitative work into the intervention study so that we actually have a sense of why something may or may not be working. But in the meantime, I think we have to sort of root ourselves in the proactive, structured, and paced communication in the ICU setting so that we can really get families to kind of understand what's happening and really affect their emotions and integrate sort of all the multidisciplinary care so that they can improve. And with that, I'm ready for questions.

epidemiologist

bereavement

ICU

outcomes

bereavement care