Escalation Injuries: Pediatric Trauma Series 2017

Um, our experience in the development of a systematic protocol to identify victims of NAT. And so, my talk, um, is going to focus on, uh, 3 different studies that we've had the opportunity to publish. Um, the only disclosure I have is that we did have some intramural funding for, um, the 1st 2 studies to be able to be completed. And the objectives are listed here, and I'm gonna go into a little bit of the background behind those, um, right now. So, As you might have seen on the video that just introduced the talks, uh, trauma is the number one killer for children and 48% of, uh, child fatalities each year are a result of physical abuse. Um, one statistic that I found from the US Department of Health and Human Services revealed that approximately 12% of these fatalities involve families that had had some sort of CPS intervention in the last 5 years prior to the fatality. And the majority of these children were less than 4 years of age. There's a significant cost to society and, um, in at least 1 level 1 trauma center, only about a little over half of the kids being evaluated for NAT were seen by a pediatric surgeon. That's by no means to suggest that every single NAT patient needs to be evaluated by a pediatric surgeon. It's more of a um hallmark of how NAT is being managed across the system, even if those resources are available. And identification of risk factors and systematic screening programs may help avoid escalation injuries. But what are escalation injuries? So there are some classic studies that show that NAT is not necessarily even considered when a young child is being evaluated in an ER for an injury. And there's a classic study that Kristen's gonna discuss a little bit more in depth, um, by Carol Jenny and her group, um, that revealed approximately 30% of children with abusive head trauma, position and NAT was not originally considered or recognized. And um from uh Columbus, from Kate Dean's group, um they demonstrated that patients with recurrent NAT had a significantly higher mortality than during their first presentation. Um, and there was also concerns of missed abdominal injury from UK registries. So at Mary Bridge, um, we became interested in looking at NAT and how we screen for those patients when we had a Sentinel event. And I think a lot of hospitals have had this experience. And so, following the Sentinel, um, event, An ad hoc committee, um, initially led by one of my colleagues, Doctor Elizabeth Paulson, uh, and I got together and we got stakeholders from different, um, services and got together to see how could we improve, um, our screening for NAT. And so we subsequently, uh, incorporated this little ad hoc committee under the trauma Committee because we felt that the trauma committee and trauma presence gave us an across the system. Um, view of patients, starting in the ER, moving through the PICU, going through med surge, and, um, and because trauma is in the name of non-accidenal trauma, um. We all agreed that this is where it should live. We subsequently were able to obtain funding um to look at how we were doing before we implemented this and then how did we do after we implemented the standardized screen. And so as we went through, we adapted and then modified and developed um a screening algorithm that we actually found from Pittsburgh um and applied it for, to Mary Bridge. Um, and then in 2014, we went live with a standardized approach for screening for NAT. The screening tools based on three red flags, a historical component, uh, physical exam component, and then radiographic findings components. Um, and these were adapted and modified as literature came out and as I kind of continued my journey and learning more about NAT through the pediatric trauma Society. For history, the high level, uh, bullet points are, uh, families that present with no history of the injury or event or an inconsistent history, especially with the patient's developmental status. Changing history, somebody is being told something, one provider is told something, another provider is told something else, the staff is told something else, um, and delay in seeking care. And you can see some of the other, uh, historical red flags. We're gonna talk a little bit more about, um, the validity of these, uh, historical red flags as we talk more about the pediatric trauma Society work, um, that subsequently followed. In terms of physical exam findings, um, some very big, uh, catch-all, uh, findings that, uh, should point you in the direction of NAT include, um, bruising, especially in particular patterns which we will discuss as well. Um, there is something called the 10-4, uh, pattern group that Mary Clyde Pierce's group, um, uh, has reported and we're gonna discuss that. Um, there are still certain bruising patterns that are very much pathonomonic for, um, non-accidental trauma. Uh, torn frenulum was a classic example of something that has, uh, been associated with NAT for a very long time, but we're going to discuss whether or not that still holds true. Um. And in regards to radiographic findings, there are certain skeletal injuries that are very suggestive of non-accidenal trauma, especially depending on the patient's, um, development. As well as, um, rib fractures in infants, uh, without a history of a significant major trauma event like a motor vehicle collision or something along those lines, um, and undiagnosed healing fractures. Certain, uh, traumatic brain injuries are suggestive of abusive head trauma in the right setting. So, we first started off trying to see where we were when we started. Um, and so we did a retrospective review looking at how good were we at screening patients for NAT before we implemented our tool. Um, and as you can see here, the demographics of that particular cohort. What we found was that we had an inconsistent or missing history um in our patients that presented with NAT in about 89% of the cases. Um, additionally, there was unwitnessed injuries in 79% of the cases and as you can see, these numbers don't add up to our total cohort. That's because oftentimes the stories would change and that's what we had available in our retrospective review in the chart. Um, interestingly though, almost 40% of patients had a prior ED visit and 60% of those were less than a year of age. So if you think about it, a less than 1 year old with multiple ED visits is a concerning finding. Now, interestingly, 41 up to 73%, and this is a range of patients were missing social history data. That meant that we were not recording it in the chart, and it's something as important as NAT and how social history impacts that, that was a very telling finding for us. Bruising was found in almost 2/3 of our patients. Um, and you can see that the, uh, The, the breakdown of those patterns there. I'd like to draw your attention to the final bullet point, which is our perineal bruising or injury. This was a small number of patients. However, what we did find was the majority were less than 4 years of age and there was a significantly increased risk of mortality if we had this finding on um physical examination as well as increased morbidity. Radiographically, we had 60% of our patients less than a year of age that had fractures, as well as a fraction of those that did have undiagnosed healing fractures in other sites. Um, and of the patients that had subdural or subarachnoid hemorrhages found, the vast majority were less than a year of age. Also, a third of our patients had rib fractures, and those were significantly injured, um, and associated with other undiagnosed healing fractures. And I'll draw your attention to that half of them had an injury severity score, which is somehow, is how we measure the severity of the injured patient of greater than or equal to 16, which is severe. 9 patients died in this in retrospective cohort and I've been asked at meetings, if we'd had this particular screening guideline, would we have been able to intervene sooner? Would we have been able to prevent this? Um, unfortunately, this is a retrospective study, so the answer is I don't know the answer to that. However, there is some telling data, and that is if you look. 2/3 of those patients had had a prior ED visit, so theoretically a time that we could have intervened. CPS history, 5 of the patients did have a prior contact with CPS. And in terms of domestic violence, this was something that was disturbing to me when we found was that although 3 patients did have a history of domestic violence recorded in the chart. The majority actually did not have data at all, which means we didn't record it and most likely didn't ask it. So now we're gonna move on to the prospective data. Oh, pardon me. One more slide. The, this just is to give a little bit of a sense of how injured our NAT patients are by looking at their severe ISS comparing it to our overall accidental trauma patients in our registry. So there's a significantly increased number of severely injured patients when they come in with NAT as a mechanism. There were concerns. So, we're talking about implementing a standardized uh NAT screening and some of the questions that arose, rightly so, were, what does that mean in terms of ED lengths of stay? What does that mean for the number of admissions? What does that mean for the number of HET CTs that we're gonna be doing? Um, are we gonna start getting unnecessary studies? Um, and so, uh, we were very fortunate in that we had partnered with all the different departments to, um, To implement and then study this. So, we did implement this on January 1st, 2014 and we were very interested in looking at our utilization of resources. As you can see here, the demographics were very similar. Before and after The implementation of our tool. And we didn't see a significant increase in the number of NATs that we actually did, uh, diagnose. We did see an increase in referral in our pediatric surgeons. And there was no significant difference between our severe or ISS. However, we did have a, although we did have a roughly similar mortality, we started to see some changes. We had less prior ED visits. We had less CPS history and at least 2 spots where we couldn't find documentation. And there was 1 episode of domestic violence in the home, but now we were down to 2 times when it wasn't asked as opposed to before when the majority of the times it wasn't asked. The use of HET CT actually did not change. So, we were still following our PCA guidelines. We're combining that with our NAT guidelines, um, and we did not see a significant increase in number of use of HET CT or the diagnosis of subdural subarachnoids. Our ED length of stay, and I think this is a really important slide to point out, is that they did not change. So, this was before and after implementation of our screening tool. The number of admissions went down, and that was very interesting. Um, and so we posit that because we were doing a more standardized screening in the ER and we were having a safer disposition planned from the get-go, that we were seeing less medical holds. Now I'm gonna move on to the final part of my talk, which is the pediatric trauma Society, um, review article that I was very fortunate to be a part of. About 2 to 3 years ago, um, the Pediatric Trauma Society identified screening for NAT as a key objective for the development of a guideline. And so I was lucky to co-chair this with uh Mark Auerbach uh from Yale, uh, who's a Ped EM physician, and we got together a multidisciplinary, multi-institutional, international group to review the literature and try to come up with high-quality recommendations for NAT screening. And what we found was, um, we were hoping to use, uh, the approach of a systematic review and then create grade criteria guidelines. Um, and what we found was that the literature wasn't quite up to where we needed it to be, to go down that route. Nevertheless, there was a significant amount of high-quality literature that we could evaluate and so we turned the project around into a more generalized review article with the concept that this could Be something that an general ED physician, a traumatologist, somebody could essentially rip out our one big table and put up on the board so that it could be looked at every time a patient comes in that should prompt a red flag and what should prompt a child abuse workup. And so, it was broken down into 7 categories. We had bruising, we had, um, abusive head trauma, fractures, burns, abdominal injuries, um, oral injuries, etc. And then based on the high quality literature, we were able to make recommendations about when to pull the trigger on a child abuse workup. Um, the first was bruising, and some of the best literature is, uh, exists on this and is currently, um, To the last of my knowledge is being worked on to be validated, and Kristen can probably make a comment on this as well. So, we use the 10-4 uh clinical prediction rule, uh, which was developed by Mary Clyde Pierce, as I mentioned earlier. And the idea is that you have torso, ears, and neck bruising that are highly suspicious for NAT in children less than 4 years of age, or any bruising in an infant less than 4 months of age. The sensitivity and specificity are listed there. And if there is bruising found in this group or characteristic patterns of bruising such as whip marks or cigarette burns, then we do recommend, um, a child abuse workup. And this is just an example of what that looks like, um, of what low-risk and high-risk areas are. Anybody who has a toddler knows, you're gonna see bruises on the shins and on the knees and elbows all the time. Um, but areas of the diaper area and on the chest and back and the buttocks are highly suspicious. In regards to burns, there are two, large systematic reviews, um, looking at, um, burns, most recently from Parkland, um, that found that these associations with non-accidenal trauma. And interestingly, in the most recent review, up to 25% of children that were admitted to the burn center had been abused. Um, and so there are characteristic patterns to intentional burns. Um, they are listed there. There's also characteristic high risk ages, um, where a significant number of kids less than they, pardon me. Of the patients that presented with intentional burns, greater than 95% were aged less than 5 years. So, based on that, the characteristics of the burns as well as the age, we do recommend a child abuse workup. Now, less than 5 is a big spectrum and we recognize that. But this is geared more towards the general ED that may not have the experience or the child abuse experts to be able to intervene and say yes or no on the risk of uh factor or risk assessment for NAT. And this is an example of a of a uh skull burn, uh, in which the patient was dipped into a scalding uh tub. In regards to intracerebral injury or um abusive head trauma. So, the two clinical prediction rules that have the, the largest amount of evidence behind them are the PD Burn uh clinical prediction rule as well as the, um, we call the PREDAHT, uh, clinical prediction rule. Um, and both of these have associated other findings. Some are very specific to having been admitted to the PICU and having these findings, but nevertheless, we generalize these concepts. Um, and so if there is. Uh, a, a traumatic brain injury associated with an additional feature that you can see here, then the sensitivity and specificity is as listed, and then we do recommend a, uh, NAT workup. Additionally, um, the Pred AHT rule has a slightly lower sensitivity and specificity, and it does have these, um, associated injuries, but nevertheless, if it is, it trips positive on that, we also recommend a non-extendal trauma workup. And it's just an example of an abusive head trauma that I've taken care of. Abdominal injury. This is where the literature starts to not be quite as robust, um, but of the existing literature, um, it's consistent that hollow viscous injuries, specifically duodenal injuries in age less than 4, are highly suggestive of, um, abuse. And so if a patient presents in this particular age group. Or if there is an associated hollow viscous injury and solid organ injury in this age group, then we do recommend a non-external trauma workup. This is less common, but the mortality is significantly higher if it's present. One thing I do want to point out, and that is that you can have abusive abdominal injury without external signs such as bruising. And so, we do recommend screening or the general consensus is to screen for um elevated liver function tests. Um, and I can refer you to the reference in, in terms of that. And if those are elevated, then further axial imaging is recommended to evaluate for intraabdominal injury. Um, and these are just examples of, uh, bruising that's associated with, um, non-accidenal trauma, um, as well as a, uh, ruptured duodenum. In terms of skeletal injury, the key take-home message of skeletal injury is it's developmentally um appropriate, is the history developmentally appropriate for the injury that you've seen. High risk, um, long bone injuries include, uh, proximal and mid uh humeral fractures, uh, femur fractures, um, and especially rib fractures in infants that are in without a preceding history of a, uh, major trauma. In this particular case, if the patterns are inconsistent with the degree of mobility, then we do recommend child abuse workup. Additionally, the American Academy of Pediatrics recommends a skeletal survey to screen for occult fractures in any child less than 2 with suspected non-accidenal trauma. Um, and, uh, it is recommended that a pediatric radiologist read those. And this is just an example of a bucket handle fracture that's his, um, classically been associated in smaller children with NAT uh due to the twisting motion that occurs, uh, in this particular kind of injury. And this is an example of posterior rib fractures, um, where there were sets of fractures here and then here you can see, I think, quite well, the posterior rib fracture. And that's from a squeezing type of, um, mechanism for, uh, infants. Frenulum tears are very interesting, and of course I'll be happy to hear Kristen's uh opinion on this, but classically, this was always taught to be associated with non-accidenal trauma. However, um, once you get into, uh, uh, children that are mobile, the incidence of lip injuries, uh, oral injuries, renum injuries are equivalent between accidental trauma and non-accidenal trauma mechanisms. So, in and of themselves, there's not sufficient evidence to recommend a child abuse workup. However, if you do have an infant with an oral injury, then it should at least raise the red flag. And if there is another finding, then absolutely, it is recommended to proceed with the workup. And this is an example of some of the um older literature uh associated with a torn frenum or frenulum uh in an infant. Finally, historical factors, um, historical factors are the least well studied of all of the categories that I just discussed. Um, usually you find that the historical factors that I discussed earlier in conjunction with a finding a suspicious. For non-accidal trauma is going to be what prompts the child abuse workup. Um, there are several studies that look at, and Kristen's gonna discuss, um, standardized implementation, um, of NAT, uh, screening guidelines, um, but there's still a lot of work to be done in this particular category. So, in conclusion, from our Mary Bridge studies, we found that the majority of patients less than a year of age um had had prior ED visits in our retrospective study. And in the absence of a standardized process, we did not obtain a thorough social history. We found that perineal bruising, though rare, was associated with mortality, and we had nearly 1/3 of patients who had undiagnosed healing fractures. Um, I'm not gonna talk about this aspect of our study, but we did feel that anecdotally, there was more comfort discussing NAT workup with parents, um, from both physicians and staff, uh, once we standardize the language. And then finally, our implementation of our NAT screening guideline did not lead to increased utilization of resources and in fact, we felt that they were better utilized. Um, and that ultimately led to the work I did with the Pediatric Trauma Society with the recommendations that we just discussed. And these are the, uh, people that I need to acknowledge, uh, for all their hard work. And the references uh to the studies I just discussed. Thank you. All right, thanks, Tony. That was, that was really great and uh a nice start to our discussion. Um, I have a couple of questions kind of for our, for our audience more than, more than me or, or at a children's hospital, but what if you don't have a pediatric radiologist? Um, and you're saying I need to do this child abuse workup and I don't have that or I don't have you or Kristen at my hospital to help evaluate these patients. So I've had the uh wonderful opportunity to go speak around our region, um, Olympia, other places to talk about this work, and that is the number one question I get. Um, so, My answer is that as a level 2 pediatric trauma center, as a level 1, we're the resource for our region. And so, um, and I certainly, we certainly capture probably a smaller area, uh, in Southern Washington, um, than perhaps in some of the uh greater Ohio areas. Um, but we recommend that, um, if there's really, if there's concern. We, we ask them to send them to us and we do the evaluation. Um, we also try to establish their, with their resources because in our, uh, state, every, um, the child abuse resources are, uh, county dependent. So we try to make sure that they're set up with the appropriate referrals, uh, in regards to CPS and then child abuse follow-up. Um, but we do ask them to come to us. I'm not suggesting that every hospital has to do that, um, but that's how we handle it. Great, no, I, it, it is a real challenge for hospitals that we talked to as well and, and those out there who are seeing kids and as you know, seeing the majority of kids before, before we do, and, and that's why of helping to figure out how to help them screen and know who to screen and also be connected to their regional trauma.