SCCM is performing maintenance on its websites. For the best browsing experience, please use Microsoft Edge or Safari. Those using Chrome or Firefox may experience access issues at this time.

SCCM Pod-470 PCCM: Adverse Events and Mortality in the PICU

visual bubble
visual bubble

The contribution of adverse events to the deaths of patients in the pediatric ICU (PICU) who die despite a low predicted mortality risk is unknown. Elizabeth H. Mack, MD, MS, FCCM, is joined by Carin W. Verlaat, MD, to discuss adverse events in low-risk nonsurvivors compared with low-risk survivors and high-risk PICU survivors and nonsurvivors and the contribution of adverse events to mortality. The podcast centers around the article, “Adverse Events in Pediatric Critical Care Nonsurvivors With a Low Predicted Mortality Risk: A Multicenter Case Control Study (Verlaat C, et al. Pediatr Crit Care Med. 2023;24:4-16). Dr. Verlaat is a pediatric intensivist at Radboud University Medical Center in Nijmegen, Netherlands.

*If you are unable to play the podcast please click here to download the file.

Category: PCCM Podcast


Dr. Mack: Hello and welcome to the Society of Critical Care Medicine Podcast. I’m your host, Dr. Elizabeth Mack. Today, I’ll be speaking with Dr. Carin W. Verlaat, and we will be talking about her article, “Adverse Events in Pediatric Critical Care Nonsurvivors With a Low Predicted Mortality Risk: A Multicenter Case Control Study,” published in the January 2023 issue of Pediatric Critical Care Medicine. To access the full article, visit Dr. Verlaat is a pediatric intensivist at Radboud University Medical Center in Nijmegen, Netherlands. Welcome, Dr. Verlaat. Before we start, do you have any disclosures to report?

Dr. Verlaat: Hello. No, I don’t have any disclosures. Thank you.

Dr. Mack: Thank you so much for being with us and thank you for this really important contribution to the literature. I’m curious, how did your team come up with the idea to do this study? What was the backstory there?

Dr. Verlaat: Thank you first for your kind words and for the opportunity to talk about my paper. Both as a pediatric intensivist and as a researcher, I’m interested in quality and safety in the PICU and also in PICU outcomes. Five years ago, we published a retrospective cohort study on PICU patients with a low predicted mortality risk; we call them low-risk nonsurvivors. It was published in PCCM. What we found was that children with complex chronic conditions and unplanned admissions were associated with death. But what happened with these patients during their PICU stay? Could adverse events play a role?

It’s probably known to the auditors, but if a patient is harmed by healthcare management rather than by the underlying disease, that’s what we call adverse events. We know that if you look in the general PICU population, between a quarter and three-quarters of all PICU patients experience at least one adverse event during their PICU stay. The numbers depend a bit on the method that is being used to detect adverse events. Of course, over the last years, many safety programs have been implemented, like line bundles, crew resource management, etc., but still adverse events occur, and they can lead to temporary harm but also to more severe harm, like permanent harm, or they might even contribute to death.

Gitte Larsen also studied adverse events in 2007, and she found that, among medical patients, the sicker patients, the patients with a higher PRISM score, had more adverse events. But among surgical patients, it was the opposite. Children with a lower PRISM score had more adverse events, so there might be differences in occurrence of adverse events among different patient groups. So we looked again in this group of PICU low-risk nonsurvivors; we wanted to know what happened with these patients, and might adverse events play a role in the mortality?

Because if you have a low-risk PICU patient, you don’t expect this patient to die. But, of course, if you have a big group of them, some patients will die. But they don’t die because the model says so, they die for some underlying medical reason. We wanted to know if adverse events contributed to that, and if this was the case, then we might have opportunities to improve the quality of care and to improve safety among this group of PICU patients. So, the primary aim of our study was to study the occurrence of adverse events in low-risk nonsurivors, and secondary aims were to study the nature, the severity, and the preventability of these adverse events and to establish the contribution of adverse events to mortality.

Dr. Mack: Thank you so much for that, and thank you for your transparency and vulnerability as you walk through this process. It’s a lot to digest, for sure. Can you share a little bit about the Dutch registry? I’m curious, for those of us around the world, does it include all the patients in every Dutch PICU? Who enters the data? How detailed? Tell us a little bit more about that.

Dr. Verlaat: In the Netherlands, we have like 18 million inhabitants and we have almost 5000 PICU admissions each year. PICU care is centralized into seven tertiary PICUs. All these centers contribute to the data registry and the PICUs transport their patients. So, all PICU care is taking place within those centers, and all data are within that data registry. I’ve been involved in this data registry for more than 20 years. The data are entered by either pediatric intensivists or, in some centers, by trained medical students. Some data are also gathered by automatic extraction from the medical records. But the way the data are gathered depends a bit on the center. It might differ between the PICUs. Do you understand?

Dr. Mack: Yeah, that makes good sense. Sorry I interrupted you. Keep going.

Dr. Verlaat: No problem. In the registry, we collect general patient characteristics like type of admission, ventilator days, length of stay, and PICU outcome. We also collected diagnosis, for which we used the diagnosis list developed by the Australian and New Zealand Pediatric Intensive Care Society. We use mortality prediction models. We use both PRISM models and PIM models and their updates. But of course, some of the models are a bit old and we recalibrate them so the mortality prediction is correct because some models were developed 30 years ago and mortality has decreased substantially since then.

Dr. Mack: Yes, thankfully. Your aim, as I understand it, was to determine the contribution of adverse events to mortality and to estimate the occurrence of adverse events in each of your groups, the low-risk nonsurvivors, the low-risk survivors, the high-risk survivors, and the high-risk nonsurvivors. It sounds like the risk was determined by PIM and PRISM. If you don’t mind, just tell us a little bit more about how you conducted the study, how were those risk categories defined, the cutoffs, and such.

Dr. Verlaat: All right. It’s quite a complex design, and I’ll try to explain. Our cases were the low-risk nonsurvivors; they were PICU patients with a mortality risk of less than 1% according to either the recalibrated PRISM model or the recalibrated PIM2 model, and they died during the PICU stay. Those were the cases. They were extracted from an 11-year cohort of PICU admissions between 2006 and 2017, so a long period. Then we selected three control groups. First of all, low-risk survivors. They also had a low mortality risk, less than 1%, but they survived, obviously. Then, we also selected patients from the other side of the risk profile, the high-risk patients; they had a recalibrated mortality risk of 30% or more. They also were high-risk nonsurvivors and high-risk survivors. So, first, stratification, then we randomly selected 125 patients from each of these four groups.

Dr. Mack: Thank you for that. It is complex and I appreciate you explaining. As someone in the quality and safety realm, I’m really interested in your findings, and it sounds like, as in many things in medicine, it depends. So, the low-risk nonsurvivor group had more chronic complex conditions than the other categories. But it sounds like the occurrence of adverse events in the lowest nonsurvivors was higher than in the other categories as well. Can you tell us a little bit more about the most frequent adverse events in the low-risk nonsurvivors, in other words, the category we’re most concerned about?

Dr. Verlaat: First of all, it’s important that all four groups had different clinical characteristics, as you mentioned, like complex chronic conditions. They were very common among the low-risk nonsurvivors; more than 90% of them had an underlying complex chronic condition, higher compared to all other groups. They also had a long length of stay and, indeed, they had more adverse events compared to the low-risk survivors and also compared to the high-risk nonsurvivors. If we look at the type of adverse event, it was very diverse.

The most frequently found adverse events were infectious by nature, like ventilator-acquired pneumonia, central line infections, etc. They occurred in one-third of the adverse events happening in the low-risk nonsurvivors. Second most common were drug- or fluid-related adverse events. But also, in 23%, we had to classify the nature of the adverse event as the category “other,” because there were either multiple factors involved or it was hard to establish the nature of the adverse events. For example, patients with delirium, it’s often very multifactorial. Still, what was a bit of a pity is that I cannot say that, if you rule out all infections, then we do not have any adverse events anymore. It’s very diverse.

Dr. Mack: I appreciate you walking us through that. I’m curious, what did you find in terms of preventability, specifically in the low-risk nonsurvivor group?

Dr. Verlaat: Let’s first say that, among the low-risk nonsurvivors, in 30% of them, an adverse event contributed to death. Looking at all the adverse events in this group, about a third were preventable. So we had, in almost 9% of the those low-risk nonsurvivors, there was a preventable adverse event that contributed to death. But then, contribution to death can also be divided a little bit. It’s very rare that death is completely caused by an adverse event. It was only at 3% of the cases. Often, it’s a mixture of the underlying condition of the child and also the adverse event. For example, we had a child with a brain tumor who had an obstruction of the intraventricular vein, but there was a delay in discovering this and in resolving this problem. So the child died, but he died partially because of the adverse event, which was the delay in the treatment of the obstruction of the drain and partly because of the brain tumor. This was an example of a preventable adverse event, but two-thirds of the adverse events are unpreventable. It’s bad luck.

Dr. Mack: Do you mind sharing with us maybe some other examples of preventable and nonpreventable adverse events and maybe how the preventability is determined or adjudicated by your team?

Dr. Verlaat: Preventability is always hard, especially in retrospect. We used a 6-point Likert scale grading from 1, which is not preventable at all, to 6, which is certainly preventable. If you scored grade 4 or higher, we considered it preventable. Examples of preventable adverse events might be like central line infections. Preventability is hard to judge retrospectively because our information from the medical record is sometimes incomplete and there is a risk of what we call hindsight bias. As an investigator, I’m aware of the outcome of the patient and it might be hard to judge preventability. So, what we tried to do is use the guidelines and protocols that were valid at the year of admission to judge preventability and be not too strict. We also had an expert panel to discuss problems, especially in preventability of the adverse events. Second, we also performed a reliability study, not only for the presence, but also for the preventability, of the adverse events, which showed reasonable outcome. Preventability is quite hard and it changes over time.

Dr. Mack: Absolutely. As we develop more capacity to handle something, it’s maybe more preventable. Thank you for that. What are your next steps as a pediatric critical care community in the Netherlands?

Dr. Verlaat: There are several steps we are working on at the moment. One thing is trying to extract diagnoses more easily from the electronic medical records in a uniform way. The second thing is we are developing more long-time outcomes and we want to incorporate that into our data registry. We are also working on a collaboration with other Dutch data registries, since we are quite small as a PICU data registry, and it’s very hard to improve and to facilitate a registry if you are so small. So it’s better to work together.

Dr. Mack: Thank you for that. It sounds like there’s been quite a lot of work that has gone into developing this collaboration and this registry around the country. I know many of us around the world can sympathize with your EHR extraction issues. Thank you. Is there anything else that you’d like to share?

Dr. Verlaat: Maybe to stress again that I think it’s important to be aware of adverse events, especially in low-risk nonsurvivors. Patients with complex chronic conditions form a growing part of the PICU population, and they may have increased mortality risk, as we know, and they are very prone to adverse events. There’s no easy solution, but awareness is the first step. Then, it takes a multisystem approach, of course, to diminish the number of adverse events. So this is a small step, but I hope it’s encouraging other people to be aware of adverse events.

Dr. Mack: Yes, your work has been very inspiring, and I am really grateful for what you all have done as a collaboration. This concludes another edition of the Society of Critical Care Medicine Podcast. For the Society of Critical Care Medicine podcast, I’m Elizabeth Mack. Thank you.

Elizabeth H. Mack, MD, 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, USA.

Join or renew your membership with SCCM, the only multiprofessional society dedicated exclusively to the advancement of critical care. Contact a customer service representative at +1 847 827-6888 or visit 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

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.

Some episodes of the SCCM Podcast include a transcript of the episode’s audio. Although the transcription is largely accurate, in some cases it is incomplete or inaccurate due to inaudible passages or transcription errors and should not be treated as an authoritative record.