All right. Thank you, everybody. It's nice to see you all. I literally thought there were going to be three people in the audience. So wonderful to see you all here. Thank you for sticking it out to the end. And as Liz said, I'm Erin Carlton from the University of Michigan. And I'm going to be talking about new and progressive medical conditions after sepsis. Just to report, this work was supported both by my KL2 and K12 award. And so I'll take a little bit of a step back just to remind us why we're all here talking about this today. So sepsis affects just over a million children and their families each year. In the US, this equates to about 70,000 children with severe sepsis, or per our new Phoenix guidelines, sepsis. And fortunately now, most of those children are surviving the hospitalization, with more than 80% surviving to hospital discharge. And with that, and as we've seen throughout critical care in general, there's been an increasing recognition of the challenges facing not only survivors, but their families as well. And so with that, if we think about survivorship at all, we think about PICS-P and the PICS-P framework created by some of the wonderful people here in this room. And we can think about the child experiencing hospitalization or sepsis, in our instance here today, coming in with their baseline status. They experience the hospitalization and then have these longitudinal outcomes in terms of social health, emotional health, cognitive health, as well as physical health. And it's really that physical health space that I'm going to be thinking about and talking about today. And that can be evaluated by different ways, looking at your functional status, looking at chronic organ failure, technology dependence, sleep, as well as fatigue. And so what we know from some of the prior literature is that children who survive sepsis experience increased hospital readmission rates, have increased emergency department use, as well as new increased rates of outpatient visits, both in talking about primary care visits as well as subspecialties. And similarly, as we think a little bit more about that chronic organ failure, we know that children who survive sepsis often are discharged with new technology dependence like gastrostomy tubes or tracheostomy tubes, may go home with dialysis, and as well as has new oxygen dependence. And this has been both shown within administrative data as well as with the LAPS study as mentioned today. And so one of the questions I had in thinking about all of this background data and just knowing what is happening anecdotally, what happens to a lot of the children who survive sepsis is, are the kids who survive their sepsis hospitalization more likely to develop new medical conditions after discharge? Or for those who had pre-existing medical conditions, are they more likely to experience a progression or a worsening of this particular condition? And so we looked at this last year and really trying to understand, do children, critically ill children who are hospitalized with sepsis, experience higher rates of chronic respiratory failure, seizure disorder, supplemental nutritional dependence, or chronic kidney disease? And we chose these four conditions specifically as there had been prior literature that had suggest increased rates of these after sepsis. It's important to note, however, that those studies were often done in adults or in pre-clinical data as well. And so there really hadn't been an examination of this in a pediatric population. And so while Liz mentioned the LAPS study, which had described some of these quality of life issues as well as increased rates of outpatient utilization, the study that we're going to be talking about today actually did not look at LAPS and said we looked at a large administrative database called MarketScan. So this database contains a convenient sample of about 170 employer-based and 40 health plan commercial insurance-based administrative claims, as well as Medicaid from about nine to 12 states. They don't tell us the exact number, nor do they tell us which states are included. We specifically chose this database for a couple of reasons. One, it was large, and so we could get the sample size that we needed to look at these rare outcomes. And two, it allowed us to follow children longitudinally over time, not only within the inpatient hospitalization period, but also in outpatient medical encounters as well. And this really allowed us to look at not only just relying on that inpatient hospitalization to determine what new morbidities that they were developing. And so in our cohort, we included children with non-neonatal, non-pregnancy-related hospitalizations and really enriched this to kind of a, quote-unquote, sicker sepsis cohort. And so we were looking at children that had sepsis or septic shock, ICD-9 or 10 codes, as well as sepsis or the addition of organ dysfunction. And then this was limited to children who were hospitalized in the ICU, which we identified through CPT codes. And then finally, we did limit the cohort to continuous insurance six months before and six months after the hospitalization, so we could really try to understand what was their pre-existing medical or did the child have any pre-existing medical conditions, and then looking at the development of these conditions following the sepsis hospitalization. Our comparison cohort were children that were hospitalized in the ICU who did not experience sepsis during that hospitalization. And so when we think about our outcomes, as I mentioned, chronic respiratory failure, seizure disorder, nutritional dependence, and renal failure, this slide is very, very busy. But we used a combination of inpatient and outpatient claims, procedural codes, medications, as well as durable medical equipment codes as well to really try to understand from a holistic standpoint to the best of our ability what new conditions were these kids developing. And when we think about new conditions, we were defining those as conditions that did not, that the child did not seem to have leading up to that hospitalization, whereas as a progressive condition, the child had evidence of having either one of our four outcomes prior to the hospitalization, and then we defined this as a worsening over time. So for instance, for chronic respiratory failure, a new condition would be a child who did not have oxygen or ventilator dependence prior to the hospitalization, but then developed it in the six months afterwards, versus a progression of that condition would be a child who had oxygen dependence pre-hospitalization, but then in the six months following had evidence of ventilator dependence. And so overall, we identified just over 3 million pediatric hospitalizations, and after all of our exclusion criteria, had a cohort of just over 5,000 children hospitalized with sepsis, compared to just over 96,000 children hospitalized in the ICU without sepsis. And then this breaks down a little bit further for each of the individual conditions when we're looking at our new outcomes, because we wanted to make sure to exclude those individuals who had evidence of the particular respiratory failure, kidney disease, et cetera, prior to the hospitalization. So we then used a propensity score to match these two cohorts. I matched them on age, sex, insurance, type being commercial insurance versus Medicaid, and then each individual complex chronic condition as well. And what this figure is showing you is that the yellow dots are the pre-match cohort versus the blue dots are after the propensity score match, and you really want them to be in between this blue line there, and so showing us that we had a pretty robust match. And so overall, we found that just over 13% of children develop a new medical condition after their sepsis hospitalization, and that chronic respiratory failure, supplemental nutritional dependence, and chronic kidney disease were significantly higher in children with sepsis compared to their carefully matched controls of critically ill children without sepsis. When we think about progression, there was actually no difference in the proportion of children who had progression of these conditions with the exception of supplemental nutritional dependence, which again was higher in the sepsis cohort. Interestingly enough, and this was actually a bit of a surprise to me, we found that seizure disorders were actually more common in the non-sepsis cohort compared to the sepsis cohort. We had done a sensitivity analysis as well, excluding those children who had come in for status epilepticus or some other condition as that, and it still held true. And so we did a lot of other subgroup analyses, and again found that children with sepsis were more likely to develop new medical conditions compared to children with non-sepsis infections. All come are hospitalized children, and then also after accounting for length of stay and organ dysfunction, which was our proxy for severity of illness. Additionally, we did a subgroup analysis looking at stratifying by age groups, and similar to a lot of the other studies looking at outcomes after sepsis, younger children were more likely to develop these new conditions compared to older children. As you can see with the dark blue here, had much higher rates of each of the conditions compared to the lighter blue bars as well. And then finally, we stratified by looking at children with pre-existing complex chronic conditions versus not. Again, because we know from prior literature that children with pre-existing comorbidity often have worse outcomes following their sepsis hospitalization, and again, this holds true with 16% of children with a complex chronic condition developing a new condition compared to only 5.5% of children without a baseline CCC. And so, I'd be remiss if I didn't talk about the limitations of this. And so, one, this is an administrative database, and so we're relying on individuals coding for sepsis, though we used a more stringent sepsis criteria to really try to enrich our cohort. Secondly, there was no real standardized screening of these target conditions, and so this is not a prospective cohort study where we'll be able to methodologically screen children for the development of these conditions. Again, we relied on these codes, and similarly from the pre-hospital state, we're relying on codes to ascertain whether or not they had any of these conditions prior to the hospitalization. And then finally, we don't know when these developed. And so, in summary, we found that one in five children have a new or progressive condition after their sepsis hospitalization, and that these rates specifically for respiratory failure, nutritional dependence, and kidney disease are higher than children, critically ill children hospitalized for other reasons. And this is important to consider as we think about screening and follow-up, as well as prevention of the development of these morbidities going forward. Thank you.

pediatric sepsis

new medical conditions

MarketScan database

respiratory failure

chronic kidney disease