Hi. Thanks for sticking around. My name's Mike McCrory. I'm at Wake Forest School of Medicine in Winston-Salem, North Carolina, and I'm going to talk about integrating social needs assessment into the care of critically ill children. I have no conflicts of interest. So my objectives are to list key areas of social drivers of health that should be assessed in critically ill children and discuss strategies to integrate that assessment into pediatric critical care. So how can we balance providing critical care, which can be very, of course, intense, time-consuming, along with this important social driver of screening? So some things I'm going to go over are when should we screen, is it feasible to do on PICU admission, should it be later, how to screen, should it be all patients or some kind of subset, and what kind of tool should we use, and how should we administer that tool? And then how can we integrate the screening into provider work and information flow? And this is a sign that our nurses put up in the break room to remind us that just adding more things for people to do doesn't always work. We need to make sure it's integrated and seen as valuable. So what do some major organizations tell us of what we should do and how we should do it? So the AAP has a statement on poverty and child health that says we should screen for risk factors during all patient encounters. And they tell us we can use a brief written screener or verbally ask families questions. CMS, as you may know, now requires screening of adults, adult inpatients, starting a few weeks ago. And they require the five core domains of food insecurity, safety, housing, transportation, and utilities. They do not currently have a requirement for pediatric inpatients. And then the guidelines for family-centered care in the neonatal pediatric and adult ICU from the ACCM says that we should include social workers and family navigators in an interdisciplinary team as part of supporting the family and making sure we're getting ready for discharge, but doesn't give any more guidance than that. So getting back to some of the main questions I wanted to address, when should we screen? Some studies or folks may think that PICU admission is not the right time because we have to focus on resuscitation as providers and the parents are particularly stressed. We do know that there is worse severity of illness and organ dysfunction in patients with lower SES. And then there's this qualitative study that was done out of Boston interviewing stakeholders and parents. And one of the parent quotes was, it definitely should not be done when patients immediately come in, especially for sick or hurt children, because the parents are very anxious and they just want to rush through it to make sure they can go see their child. However, despite those factors, the social drivers may be immediately relevant to providing family-centered care. In ICU families and multiple studies have been shown to be more likely to have social needs than the general population. The guardian may be more likely to be available on admission as compared to later. And in VPS, the median PICU stay is only about one and a half days. And depending on your unit, many patients may be discharged home from the PICU. So if we wait too long, we could miss important needs and opportunities. So what about how to screen? There have been some studies that have described screening selected patient populations for efficiency. One study described just screening government insurance patients that was in adults. There was a study that described a machine learning algorithm to help predict who would need social needs and direct services. And you could also imagine using an area-level metric such as Child Opportunity Index to determine which families might be most likely to have needs. However, universal screening is generally recommended. It minimizes bias and stigma. It likely will be required by CMS at some point. And it allows you to streamline a process of care and avoid missing families in need. What about what kind of questions or what screening tool to use? So CMS for adults just requires a standardized health-related social needs screening tool. So it must include those five key areas I mentioned earlier. And they give a list of examples on their documentation. I'm going to talk a little bit more about this first example, the Accountable Health Communities Health-Related Social Needs Screening Tool, or AHC-HRSN. The other ones I've listed there, CMS lists in their documentation, but they haven't really been used in children. So I'm not going to go into those in more detail. So this is a summary of the CMS AHC-HRSN. So the original version is 10 questions in the five domains. So you can see the domains here and the questions just to get a general idea. It's available in multiple languages. There's also actually an extended version of this with even more questions. This has been incorporated into our EPIC at our institution and probably at a lot of other places as well, since it's kind of the standard CMS one that's listed first in their documentation. And since it's required for adults now, a lot of folks have used this one, or a lot of EMRs have incorporated this one already. And in their documentation, they also say that they expect that clinicians and staff can easily use this short tool as part of their busy clinical workflows with people of all ages, backgrounds, and settings. And in most cases, they suggest that patients will self-administer it with little or no assistance required. So I think that's a little optimistic, and clearly it needs to be fleshed out more in our particular setting of the ICU and the pediatric ICU. What about tools that have actually been described and studied in children? So this was a systematic review, Sokol et al. and Pediatrics in 2019, and they looked at 11 tools and 17 studies. And you may have heard or be familiar with some of these acronyms, SEEK, PSQ, WeCare, iScreen. There are more, Thrive. But generally, the tools took between 3 and 20 minutes and were between 6 and 15 questions. In this systematic review, only one of them was used in the inpatient setting, the pediatric inpatient setting. Sokol et al. and Pediatrics, Colvin, 2016, using iHELP, which I'm going to come back to in a second. In that study, the pediatric interns administered the screening tool as part of their admission H&P, and they did find that there was three times greater social work referrals. So in other words, they were identifying needs that were there, and they may not have otherwise identified. But there were some issues with implementation. So the average H&P only addressed half the domains, and 70% of the interns stopped using it once they weren't forced to anymore. So we want to make sure it's something that's streamlined and that can be sustainable. This is a great resource, SIREN, the Social Interventions Resource and Evaluation Network. It's from UCSF. And they compare different screening tools in both adult and pediatrics, including these PEDS tools that I list here. One really interesting thing they do is also list the reading level of each one. So like, for example, SEEK is a fifth grade reading level, iHELP is seventh. The AHC HRSN that I just mentioned is eighth grade, but the WE CARE is ninth grade, and this SWIC, Screening for Wellbeing in Young Children, is actually a tenth grade reading level. So that's important to consider if you're going to be asking caregivers to actually do the survey themselves, that literacy issue. And then another thing that's interesting here is they talk about whether desire for assistance or urgent needs is built into the screener. So that's not part of the AHC HRSN. But several studies have shown that even when you identify needs, families may not always want, for one reason or another, help with those needs at that time. And so it's really important to identify what the family's strengths and needs and desires are right at that point in time. So this is just the iHELP, the one that I mentioned was studied in the pediatric inpatient setting. And you can see it includes some of the same things as the AHC HRSN, although it doesn't have utilities or transportation. But here are some things that are additional that might be important in the pediatric population. Legal status or immigration, of course, this could be a sensitive question. But they recommend a follow-up question to make sure that families know they could still be eligible for benefits even if their child was not born in the United States. And then power of attorney guardianship, so making sure you know who the child lives with and who can make decisions on their behalf. And this could obviously be important immediately in the ICU for consenting for various things. OK, so what about how to do that screening? I alluded to how CMS said that the family or patient could just fill it out themselves. And that's actually been shown in several studies to increase disclosure. In other words, people are more likely to disclose needs if they're not being interviewed. So there's that privacy aspect of perhaps not feeling judged or interviewed and just being able to write it or fill it out on a tablet. So there are two studies in the PDD setting and one in the outpatient setting. And each of them showed higher disclosure of various combinations of stress at home due to violence, drugs and alcohol, or food insecurity. On the other hand, the qualitative study I mentioned earlier of pediatric inpatient caregivers and stakeholders, both groups in those studies preferred face-to-face interviews for social determinant screening. And they noted that it needs to be presented as caring, helpful, and routine because families can very easily feel shame and blame. And they worry about judgment of their parenting and about CPS referral if they disclose problems in the home, for example. And there was another study that indicated that families were particularly uncomfortable with that being asked their income. So it needs to feel not like a judgment or assessment and more so of identifying needs that the family wants help with. And one quote, interesting quote from a family member or a parent was, if you have mold in your apartment, it's because you have a slumlord and that's all you can afford. It doesn't mean your kids need to be taken away. So you have to realize there's a significant amount of worry on parents and many parents when they're being asked these questions. So getting back to the last question, so when, how, and then how do we integrate it into provider work and information flow in our busy ICUs? So there are many barriers, of course, in the ICU setting to social driver screening. We're task-saturated. We have many priorities. There isn't really a PICU standardized validated tool, although you heard Dr. Paquette and others are looking into this. There's discomfort by providers and families with the questions. There are issues with privacy, both if you're doing an interview, there's the child there who may have some understanding, there may be other family members there. And then also with documenting, if it goes into the patient portal and there's something about intimate partner violence, for example, that could lead to an unsafe situation. There's inconsistent documentation and then there's limited resources to address issues. But we do also actually have some facilitators. We do have longer than a 15-minute outpatient clinic visit to often interact with the families. We have a more favorable nurse to patient ratio than on the floor, for example. There are more and more tools available and these tools have been increasingly integrated into EHR products, which I'll talk about in a second. There are family sources of resilience that if we ask about, we can also often find strengths that the families already have. And then there's increasing evidence of recognition and recognition of the importance of this. So we have EPIC at our institution. This may look familiar to those of you who have EPIC as well, although in other EHR products over the past six years or so, it's been integrated. So the storyboard over on the left side for us has these various aspects of social drivers of health. You can see some of them were asked in this case and some weren't. If it was asked, it shows up as a color based on the risk, green, yellow, or red. We have it integrated into our HMP, but you can see in this example, it wasn't asked. And that's pretty common since we don't necessarily require it or track it right now. And then there's a wheel, the SDOH wheel, they call it an EPIC, which also has a color-based indicator to try to give you an alert of if something was asked and whether it's green positive. And then as far as community resource referral, we often, of course, rely on our social workers, but sometimes on nights and weekends, they may not be available or the family may want a resource to go to directly. And so this table on the left was from a nice review article by Ogurek et al. of different resources. And I'm going to talk a little bit about this Aunt Bertha because that's what's integrated into our EPIC and many others probably as well. So I think it comes from where Uncle Sam leaves off, Aunt Bertha picks up. So it's just a free tool. For some reason, that's how they came up with it. So it's a free tool where if you click on it directly from our EPIC, you can enter the zip code and it gives you various areas where you can find local resources. So what's been described in the literature about how this has actually been integrated into flow in the ICU setting or pediatric inpatient setting? So these first two studies up here were in PCCM. They were both online letters to the editors, so they were brief studies, but they described screening families in the PICU for social drivers. The first one was in Pittsburgh, LeCount et al. asking about the two-item hunger vital sign, which is those two questions that are part of the AHCHRS and the CMS questions. And they found that 20% of their families were food insecure, which is roughly double the US population, consistent with other studies where PICU families have increased needs compared to other families. And then the second study was out of, and that was direct interview by social workers. The second study here was using a tablet screening in Cincinnati. And what I want to point out here is that only 59% of the families were able to be screened. There were families who declined, caregivers unavailable, technical issues with the tablet, and then issues with patient acuity and language barriers. But the families who were screened were mostly comfortable with the screening, and many of them did have social needs. But again, as I mentioned, only about a third desired assistance. So I think that's an important part to determine, what do the families actually want? Where can we help them the most? And then the adult ICU, there was a big study with embedded tool into the EHR. The providers were administering it. But again, I just want to point out, only 45% of the patients were screened. Often they refused, were incapacitated, or were admitted in off hours. So we need things that can be broadly available and get streamlined into workflow. And then I'd like to mention this, I'm sure I'm running low on time, but I'd like to mention this study that was done in the inpatient setting by Fortin et al at CHOP. Again, they were only able to screen the minority of admissions because of doing it during social work hours, language, caregiver availability. But very few patients declined in this study. They also used the red cap or tablet. They didn't want the inserts to go straight into the patient portal, again, because of safety concerns, particularly with intimate partner violence and other things. Some really cool things they did here was have an automatic notification to the social worker when there were positive screens and the case manager. They also integrated that community resource link, the Aunt Bertha I just showed you. And they had the social worker indicate preferred resources that the social worker was familiar with and felt would be most helpful, which was really cool. And they also had more longitudinal or attempted longitudinal integration by automatically populating the discharge paperwork so that the follow-up physician, presumably the primary care doctor, would know, because you hate to just start from scratch and ask the same questions without knowing what the family has already indicated needs. And then how can we track this over time? These ICD-Z codes were introduced in 2016. This is probably the most well-known way to track these. There are 10 of them related to social drivers, and there's one that's fairly pediatric-specific problems related to upbringing, including custody of non-parental relative. These are not directly reimbursable, but I spoke with our coder, and it can definitely affect the DRG via comorbid condition. And we'll be able to help us over time in tracking community needs and outcomes. But it has to be documented in the medical documentation, the social driver, as a condition that affects health for it to be able to be coded in the first place. And there are a couple of studies that showed since these Z codes were introduced, there hasn't been a lot of use of them. And so there was an interesting quote from the second study out of UCSF, simply adding dedicated fields and tools to EHRs is insufficient to drive increased documentation of social needs. So we need to make sure that everybody knows about it, we track it, give feedback, and make sure we are actually using it. So in conclusion, based on review of the literature, we don't have necessarily a clear-cut best screening tool in the PICU right now. But I would say that when should we screen on admission with the option to defer if there are reasons that that needs to be done? How should we screen? It should be standardized. It should at least include those five CMS core domains, and possibly also guardianship, desire for assistance, and some would say immigration status. There should be considerations of privacy, language, and literacy with the screening. I think there is increased disclosure in the literature when the patient or caregiver proxied or completes it directly. But there should be assistance available. And it should definitely be presented to the family as something that's routine, supportive, and non-judgmental, and looking for their strengths they may have, and be helpful and not burdensome and streamlined into the provider's workflow. And then as far as integration, it has to be integrated into the EHR with privacy regarding what shows up in the portal. There needs to be judicious alerts to the social workers, case managers, and providers. There needs to be resources accessible and ideally longitudinal integration so that future providers can see what was screened positive and what was done. Finally, I'll close with a quote from a family. What if I haven't gone to my doctors in a while so they're unable to ask me, but now I'm here at the hospital and the hospital providers do ask? I'd be like, oh, thank God someone brought it up. Thank you.

pediatric critical care

social needs assessments

universal screening

electronic health records

systemic barriers