SCCM Pod-532 PCCM: Understanding Quality-of-Life Risks in PARDS Survivors
25:00 Share

Hello, and welcome to the Society of Critical Care Medicine podcast. I'm your host, Dr. Elizabeth Mack, and today I'll be speaking with Dr. Elizabeth Killeen, MD, MPH, about the article, Pediatric Acute Respiratory Distress Syndrome, Severity and Health-Related Quality of Life Outcomes, Single Center Retrospective Cohort from 2011 to 2017, published in the September 2024 issue of Pediatric Critical Care Medicine.

To access the full article, visit pccmjournal.org. Dr. Killeen is an attending physician in the pediatric ICU and an assistant professor at Seattle Children's Hospital, University of Washington School of Medicine in Seattle, Washington. Welcome, Dr. Killeen.

Thank you so much for having me. I'm looking forward to chatting. Awesome. So before we start, do you have any disclosures to report? I receive NIH funding from NICHD and have a completed consultancy relationship that's not related to the current work.

Perfect. Thanks so much for this fabulous contribution to the literature. We'd just love to hear where the idea for this study came from. Is this a continuation of prior work? Have you all collaborated before? We'd just love to hear. Yeah, this is a data set that I first accessed as a pediatric critical care medicine fellow for my fellowship project. It was a data set that had been collected by Arne.

different group in the hospital to assess quality of life outcomes among hospitalized patients, both in the general pediatric wards as well as in our ICU. And their data set included baseline discharge and follow-up quality of life data on a number of children close to

800 pediatric ICU patients. And so as a fellow, what was advantageous to me about that data set is that I could investigate my interest in post-ICU outcomes without having to collect prospective outcomes data, which is very arduous and takes a long time.

which is difficult for a fellow project. And so I was able to merge their data set of the baseline and outcomes data with EHR data to evaluate outcomes after pediatric sepsis. And that experience using that data set led me to subsequently work with the team who had collected that data to develop a more comprehensive data set, an ongoing project to access these existing data and

evaluate all sorts of different exposures. And so this group that co-authored this manuscript has actually collaborated on a number of manuscripts using the same data set with different exposures, such as pain, delirium, sepsis, and now ARDS. Well, thanks so much for sharing this story. I think that's very relatable for many of us, thinking back to our fellowship project.

Please, would you summarize the aims and methods? Really, what were you all attempting to do with this study? The question that we had was, how does the severity of hypoxia in pediatric ARDS, how is that associated with non-mortality or in-hospital outcomes? I was a member of the Pediatric Acute Lung Injury Consensus Conference 2, the 2023 version of that, which was an updated collaboration program.

or consensus conference from the original 2015 Pallet Guidelines. And as part of that second iteration of the Pallet Guidelines, there was a real focus in all of our discussions as part of that group on the delineation of severity of hypoxemia. In the original 2015 guidelines, pediatric ARDS was defined as mild, moderate, or severe based on your severity of hypoxemia for invasively ventilated patients only. In reviewing the literature,

between the publication of those guidelines, or the development of those guidelines rather, and the development of the subsequent second iteration of the guidelines, several studies had come out demonstrating that there really was not a large mortality difference between

the mild and the moderate categories. And because of that, it didn't seem like there was very much discrimination between those two groups. And so the decision was made in the new iteration of the guidelines to actually combine the mild and moderate categories into a binary hypoxemia severity categorization compared to a three-level or trichotomous categorization.

categorization. And so one of the questions that I had is, we know that there's no substantial difference between mild ARDS and moderate ARDS, what was previously termed that, in mortality, but how does that compare to non-mortality outcomes? There had been some initial literature showing that

some of the non-mortality but still inpatient outcomes that had been assessed, such as ventilator duration among survivors, ventilator-free days, perhaps did have some better discrimination with a three-level severity classification. And so it's possible that there may have been some better discrimination for non-mortality outcomes than mortality. So that was one of my motivations in trying to look at this question for the ARDS population specifically, having been a

again, a part of those pallet guidelines. The second part is something that I have been interested in a long time and is really what motivates a lot of my work now is this idea that mortality is so low among our pediatric ICU population and steadily decreasing still that using mortality as an end point for most studies

studies, whether it's clinical trials or even observational data, is a very inefficient endpoint. It just doesn't occur very frequently, which is great for the patients, but challenging when you're trying to develop a study looking at differences in risk factors or interventions as they relate to that outcome. And for any type of

research, the more common the outcome, the more efficient it will be in terms of the number of patients you need to evaluate differences in exposures and how that relates to the outcome. And so I've been interested for a while in what non-mortality outcomes are both common enough that they could be used as endpoints in clinical trials or observational studies, but also meaningful to patients and families.

Thank you so much for that. I think this is the sort of thing that we all think about if we end up in the ICU or our family ends up in the ICU, not whether we live or die, but really it's our functional status. So really grateful for you digging down into the next layer.

Yeah, and what we know is that previous work has found that when you ask patients and families what they care about, it is quality of life and functional status. It's not necessarily these in-hospital non-mortality outcomes like ventilator-free days or length of stay. And those are much easier to assess after mortality, but they don't necessarily mean as much to the actual patients and families experiencing those outcomes than longer-term outcomes.

And so there's been certainly increasing interest in evaluating these longer-term non-mortality outcomes. What struck me when we were evaluating the outcomes that we do know about for pediatric ARDS as part of the PALIC guideline development for assessing outcomes after pediatric ARDS

ARDS is that there's actually been very few studies that have evaluated quality of life after ARDS, despite it becoming an increasingly common measure for people to investigate. There actually were only three studies published so far that looked at health-related quality of life after pediatric ARDS, and none of them evaluated hypoxemia as a risk factor, which is the factor that best discriminates mortality risk. And so

One of the motivations in this study was to essentially establish whether the risk factors for mortality are also similar for health-related quality of life decline. Because if you're thinking about eventually developing intervention that could improve mortality, which is certainly the

goal of most, if not all, of the interventions being developed, you also need to understand whether it could be expected to also contribute to improved quality of life among those who do not die in the ICU, but live long enough to have those non-mortality outcomes assessed.

And so what we did for this study was that we used the existing data that had been already collected with baseline and follow-up quality of life data. The follow-up was conducted a median of six weeks after hospital discharge and then used that as a retrospective design to go back into the electronic health record and extract oxygenation data and do a manual chart review to specifically look at x-ray data, diagnostic

timing of hypoxemia, and some of these other nuances to the diagnosis of pediatric ARDS that are easily extracted directly from the EHR. And then we evaluated how patients who were diagnosed with pediatric ARDS based on our definition of it using the PALIC guidelines, how ARDS itself was associated with quality of life decline compared to patients who did not have ARDS but were also in the ICU. And then if you subdivided that cohort by

a binary categorization of hypoxemia for the 2023 PALIC-2 guidelines or per the trichotomous severity stratification of mild, moderate, and severe from the original 2015 PALIC guidelines.

Thanks so much for that detail and really that history of kind of how this came about and really from a patient and family lens. Give us an overview just briefly of the various scales that you all use. We'd love to hear your thoughts. So the Pediatric Quality of Life Inventory and the Functional Status II are where the two outcome measures that were used. The Pediatric Quality of Life Inventory or PEDSQL has become

the most commonly used health-related quality of life measure used in pediatric critical care studies. There's a number of quality of life measures that are validated for use in this population, but the PEDSQL has really emerged as the most commonly used, and that was the measure used for many of our patients in this study. The functional status 2R is a measure that has been adopted by

to be used as a essentially surrogate quality of life measure for patients with severe developmental disabilities. The PTQL asks a lot of questions about school, running, interacting with peers, interacting with the family that may not be applicable

for many patients with severe cognitive or physical disabilities. And so Seattle Children's, as part of its program to assess outcomes across many inpatient ward and ICU patients, has adopted the use of the functional status 2R, FS2R, for patients for whom the PEDS QL is not necessarily as applicable.

And families are actually given the choice of which one is more applicable to their child when they complete it with some guidelines or recommendations based on the content of each of those. The functional status 2R is really more focused on functional status and includes things like smiling, grimacing, feeding that are potentially more applicable than school function, exercise activity, playing with friends for some of our patients.

So those were the two outcome measures. The PEDSQL is well validated in the pediatric ICU population. The FS2R has been used less frequently, but our group has used it in a way that has actually been incorporated alongside the PEDSQL in terms of the scoring for that. And so that's been at least established as an approach.

is to use the same scoring methodology and the definition of what a change from baseline might be. That's one of the potential limitations of this study is that we don't know what a clinically important change in the FS2R is. And again, it's just based on prior literature that we've adopted the same change in score as the PSQL, which is four and a half points from their baseline score.

The change of four and a half points from baseline was our primary outcome measure as a binary yes or no outcome. We also looked at the continuous score change as a secondary outcome. The other measures were used as covariates in our multivariable modeling. So the Pediatric Overall Performance Category, or POPC, is a measure that assesses children's level of general performance

functional status on a relatively non-granular basis. So from normal to coma on a scale of essentially one to five. And it gives a very broad sense without getting into details of the person's specific functional status in terms of their baseline functional status. The pediatric cerebral performance category score is a measure of cognitive function. So we use that as a covariant. The other

metric that we used as a covariate is the Pediatric Logistic Organ Dysfunction score or the PILAD score and that's the measure of organ dysfunction. We elected to use their non-respiratory PILAD score and essentially all that means is that the PILAD score is made up of a number of different

organ systems. And if you remove the points allotted towards your respiratory status, that becomes your non-respiratory PELAD score. And so it's really a measure of how severe the organ dysfunction is across different organ systems, not the lungs. And

This was intended to, when we compare patients with and without ARDS and with different levels of ARDS severity, to try to as closely as possible approximate similar conditions in terms of baseline health status, other non-respiratory organ dysfunction. We also adjusted for age, diagnosis, length of stay, and time to follow up, which were all factors that we thought could conceivably be associated with both ARDS and their outcome.

Awesome. Thank you for that review and overview. I'm curious, why classify pediatric ARDS severity both in binary and trichotomous ways? The question that we really have is whether the discrimination seen with different levels of

Hypoxemia severity is better or worse if you look at it in two categories versus three. Those groupings are drawn directly from the two different iterations of the Pallet Guidelines with the 2015 original Pallet Guidelines suggesting a three-level or trichotomous severity stratification and the subsequent iteration, the second Pallet Guidelines in 2023, revising that to instead suggest a binomial

categorization based on lack of discrimination for mortality. Our first question with this study was, is hypoxemia related at all to quality of life? There have been many studies across different patient populations and disease categories that suggest that there's

many factors that are associated with post-discharge quality of life after pediatric critical illness. And illness severity is not consistently one of those factors. And so we consistently see that patients with relatively low illness severity may have a risk of quality of life decline that is similar to patients with

higher severity of illness and that the contribution of environmental, patient, family, etc. factors really contributes quite a bit to their post-discharge status. And so question number one was, what are the risk factors associated with quality of life after pediatric ARDS? We found just in our initial evaluation that severity of hypoxemia was one of those factors that was associated and was in fact the only factor that we assessed that was associated with quality of life outcome.

Again, we did not do a detailed analysis of other patient family and environmental factors. These were all demographic characteristics and clinical characteristics of the patient. And so once we had established that

There was an association between severity of hypoxemia and quality of life. We wanted to know, are the new guidelines that had just come out at that time appropriate for assessing quality of life? Or is a more granular evaluation with a three-level severity stratification going to give us better discrimination?

Anything more that you'd like to tell us about the results and how they have changed your practice in any way, whether that's earlier rehabilitative efforts or developmentally appropriate supports or anything along those lines?

Yeah. So what we found is that before adjusting for any covariates, patients with pediatric ARDS had a much higher prevalence of quality of life decline compared to their baseline. 41.5% of all of the ARDS patients in our cohort had a clinically important quality of life decline compared to 23% of patients with an unplanned ICU admission who did not have ARDS. So that I think is important just to acknowledge that the

prevalence of quality of life decline among patients with ARDS of any severity is extremely common and suggests that that could be, if used in a conventional trial, a very efficient outcome. The other part is that we did identify a essentially stepwise increase in risk of

of quality of life decline based on your severity of hypoxemia from mild to moderate to severe when evaluated in a trichotomous fashion and from mild moderate to severe when evaluated in a binary fashion. And that is both done overall prevalence and when adjusted for multiple covariates.

given relatively low sample sizes, we're not able to say definitively that there is a difference between, for example, mild and moderate for every category, but there is an increase in prevalence and there certainly seems to be a notable and statistically significant increase in risk for severe compared to the other categories. The other component that I

I think is very interesting, and this is consistent with other studies, is that we were able to, as a secondary analysis, stratify our groups based on invasive versus non-invasive ventilation. I had mentioned that the updated pallet guidelines included severity stratification not just for invasive mechanically ventilated patients, but for non-invasively ventilated patients, which was new compared to the original guidelines.

And so we were able to look at, for the non-invasively ventilated patients, what was their risk of quality of life decline based on mild, moderate, and severe hypoxemia. And what we found is that the stepwise increase in risk of quality of life decline remains true for both invasively and non-invasively ventilated patients, and that severe hypoxemia among patients who are non-invasively ventilated was associated with a higher risk of

And so I think one of the really important takeaways there is that patients who are non-invasively ventilated, and again, and these are patients who never were invasively ventilated during their entire ICU course, so who remained during their entire ICU course on non-invasive ventilation, have a higher chance of having a non-invasive life decline.

have a similar risk of poor quality of life outcomes. And this has been seen in other types of outcomes as well as patients who are invasively ventilated, that our management of patients does not necessarily change their original risk of that poor outcome. So other than curing pediatric ARDS and shouting from the mountaintops the need for health-related quality of life outcomes to be an endpoint in our studies, what's your next step?

The field of outcomes after pediatric critical care has focused for a while now on identifying the prevalence of abnormalities across various domains. And we're at a point where that's been pretty well established. There is emerging literature about risk factors for those declines. And this work, I hope, contributes to that. But really, the next

step is to actually intervene to change those outcomes. And I think that the first step to doing that is to really think about how our in-hospital outcomes affect not just our likelihood of survival or ventilator-free days or these short-term outcomes, but really our long-term outcomes.

And the reason why that's important is that an intervention that, for example, were to decrease the time on the ventilator, but in turn increases the number of days in the ICU, for example, that could conceivably have a result in which the

quality of life of patients a month, two months, six months down the road was actually worse for the patients who have spent a longer time in the ICU. We know that exposure to the ICU in and of itself is a risk factor for poor quality of life and functional outcomes.

not to mention emotional and family outcomes related to PTSD and reintegration into society. And so I think acknowledging that not all of our assumptions about what a good or bad outcome is on the inpatient side necessarily translates to what the patient status will be several months down the road.

The other piece of this is to really acknowledge that many of the risk factors for poor quality of life or other outcomes are not necessarily very modifiable. We looked at, in this study, their severity of hypoxemia four hours after diagnosis of ARDS. That's not a lot of time to intervene to change that trajectory. And so thinking about what interventions are

improve recovery is something that more and more people are starting to focus on. And whether that's reduction of sedation, earlier physical therapy, post-discharge support with the primary care team, and other specialty services to improve functional, cognitive, et cetera, recovery is going to be really important.

And one of the next projects that we're working on is to build on some of the work that's been done in identifying these post-discharge outcomes to really figure out what are the best ways that we can screen for these and then subsequently who's the right person to manage them, whether it's the primary care provider, whether it's post-ICU clinics, whether there's another model that's most effective. Hmm.

Thank you so much. Lots of food for thought there about how we might pivot going forward. And thanks to all that you've done in this field. This concludes another episode of the Society of Critical Care Medicine podcast. If you're listening on your favorite podcast app and you like what you heard, consider rating and leaving us a review. For the Society of Critical Care Medicine podcast, I'm Elizabeth Mack. Elizabeth H. Mack, MD, MDM.

MS. FCCM is a professor of pediatrics and chief of pediatric critical care at Medical University of South Carolina Children's Health in Charleston, South Carolina. Join or renew your membership with SCCM, the only multi-professional society dedicated exclusively to the advancement of critical care. Contact a customer service representative at 847-827-6888 or visit

sccm.org membership for more information. The SCCM podcast is the copyrighted material of the Society of Critical Care Medicine and all rights are reserved. Find more episodes at sccm.org podcast.

This podcast is for educational purposes only. The material presented is intended to represent an approach, view, statement, or opinion of the presenter that may be helpful to others. The views and opinions expressed herein are those of the presenters and do not necessarily reflect the opinions or views of SCCM. SCCM does not recommend or endorse any specific test, physician, product, procedure, opinion, or other information that may be mentioned.

♪♪♪