Dr. James Xie - Sharing Perioperative Records with Patients and Families: An Opportunity to Embrace Transparency

Good morning, everyone. I think we're gonna go ahead and get started. Welcome to grand rounds, uh, for the departments of anesthesia and Surgery. It is my great pleasure to introduce our speaker today, Doctor James Shi. Uh, he was a board certified uh pediatrician and pediatric anesthesiologist at Stanford University School of Medicine. His goal is to improve patient care and promote health equity with health information technologies. And currently, he is a, uh, clinical informatics physician and epic physician builder at Stanford Children's Health. Doctor Shi, uh, studied computer science and medicine at Stanford, uh, followed by a combined residency in general pediatrics here at Boston Children's Hospital and Boston Medical Center. And anesthesiology at Brigham and Metteman's Hospital. After residency, he completed a fellowship in pediatric anesthesiology at Stanford Children's Health, where he subsequently joined the faculty, and he's here to speak to us today on um transparency with regard to sharing peroperative records with patients and families. Um, so, thank you so much, uh, Doctor Schaffer, for joining us today and I'll turn things over to him. Thank you so much for the introduction and thank you for having me here. Uh, just catching up with Doctor Jackson, one of your colleagues, about many of the memories that we shared here on, in Falkman Auditorium and on the stage here. Um, as, oh, hang on. Doctor Ree mentioned, uh, I'm a pediatrician and pediatric anesthesiologist. I occasionally practice some obstetric anesthesia. No relevant conflicts of interest, and I will be discussing OpenNotes, which was actually an initiative that started here at the Harvard Medical School, and I don't have formal ties with them yet, but I'm meeting with some of their members later today, so that's kind of exciting. Um, and as mentioned, having trained here, it's, it's a real privilege to be back and see many old friends and make new ones too. So, what are we talking about today? So, sharing of perioperative records, I see some surgeons in the audience, so it's fantastic. Um, I have definitely geared some of the things that we're talking about to both surgeons and anesthesiologists. But today will be about talking about the implications of the 21st Century Cures Act and the information blocking rule that is part of that, uh, piece of legislation. And highlight some of the scenarios that we may run into as a result of being in the pediatric perioperative space. Stepping back a bit more, kind of review the history of the 21st Century Cures Act and its implementation over the last few years, and then I'll talk about some specific things that we can do, um, to embrace transparency and perioperative record sharing because it's already happened and we should really be focusing on how to grapple with how best to move forward with this piece of legislation. So, how might open record sharing impact the pediatric perioperative environment? Um, just a quick question for folks. Who's on their own healthcare providers, um, patient portal? Like if you get help, yeah, a lot of people have are experienced with the patient portal. Um, I was here in June when Epic went live, and Epic has its own patient portal, and you may have noticed that MyChart is now available to all of your patients. What does that mean for you, for them? Well, in our world, patients are anxious, parents are anxious enough as it is for their, um, child to undergo surgery. And what's unique to our world is that while the pediatric patient is asleep under anesthesia, their parent or guardians are awake. And so, if we are releasing things on an immediate basis, which is what the letter of the law says in the 21st Century Cures Act, then are we returning to a world where we're performing in an operating theater, which I know, uh, for our British colleagues, that's actually the norm for how we call the OR. But, um, This kind of concept of real-time data release perhaps wasn't the original intent of the law. And as we'll kind of illustrate today, the goal is not to be obfuscating or hiding stuff from our patients because ultimately, the medical record belongs to them. However, what we should be doing is being more thoughtful about how we release these records and ensuring that when we do release the records, it doesn't cause confusion and doesn't cause undue anxiety. So, could we uh implement kind of reasonable delays in record release? In the last 3 years since the Cures Act went live in 2021, this has been highly variable depending on which healthcare institution, which hospital you work in, um, depending on their respective informatics and legal teams, that, uh, kind of data release strategy is entirely dependent on how each legal team interpreted what does it mean to immediately release medical records to patients for free electronically. And so, this is kind of what we're grappling with. In addition to, you know, just the fact that we have to release records to patients, our adolescent, uh, our pediatric population is very unique. So, we take care of adolescents, right? And as many of the pediatricians in the audience know, adolescent confidentiality is a cornerstone of ensuring that they receive the care that they need to get. Because if we don't uphold confidentiality, we know that adolescents are far less likely to seek out appropriate care or disclose things that um their parents might not want uh to know, or the kid might not want their parents to know, right? And so, Uh, this is also kind of tricky because on a state by state basis, there are different things that adolescents can consent to. And so, this means that depending on what state you're in, there are certain things that adolescents can consent to. For example, receiving contraception without the parents' permission or um getting treatment for drug or alcohol abuse, or mental health services. Each of these things varies by state. And as a result, your electronic health record has to navigate this, OK, should I release this to the parents or should I not? And as a result, uh, configuring this carefully is very tricky. And even if we are able to sub-segment all of the things that are in an adolescent's record, well, it turns out if the adolescent prox like a portal account, if the adolescent portal account, is not, uh, being used appropriately by the adolescent themselves, and in fact, a parent is messaging from the child's account, then this whole kind of sub-segmenting the chart doesn't make any sense either. Um, for people who receive clinical messages like our surgeons in clinic, you may receive messages from the child's account saying, Dear doctor, uh, this is the mother of Johnny, please help me with this problem, and you're like, huh, it's the child's account, but it's the parent messaging from that account, right? And has any, has that happened to anyone? Yeah, I see some nods in the audience. And so it turns out that across 3 large um children's hospitals across the nation, over 50% of patient messages. That were supposed to be coming from a child's account were very likely to be from a parent. And so we used a natural language processing technique, analyzing the messages from parent, uh, from child accounts, but they were clearly indicative that the parent was using the account. Whether or not the parent was doing this nefariously or whether the child had given the password to the parent, we don't know, but what we do know is that over half of the child's accounts were being used by the parents. So this is a problem if you're saying, oh, let's try and sub-segment these confidential parts of the adolescent chart in confidential notes, and then release the confidential notes to the adolescent account. That's a problem if the adolescent account is accessible to the parent. Interestingly, the situation is no better in adult patient portals. Um, I met with a friend who works at Dana-Farber, and because there isn't a very robust way of sub-segmenting confidential information in the adult chart, and because they really want to have family involved when they discuss chemotherapy and what to expect as they're undergoing treatment, They literally have to tell the family members, just log into your loved one's account using their password because the other thing for telehealth visits, they have to do that um through their portal account. And so, these workflows have not been well developed or well thought out. So, interestingly, yes, this applies to children, but also to adults. And You know, only 20% of hospitals offered a way for electronic health information to be restricted. So, let's say you didn't want every single result in my chart to be released to my spouse. There isn't really a good way to do that in current, um, patient portal accounts in electronic health records. The other interesting thing in our world, um, there are many maternal confidentiality concerns, right? So, oftentimes we take care of newborns, we sometimes will put in a lot of the maternal history, whether it's STI testing, drug exposures, or, um, different things like that. And so you can imagine scenarios where perhaps the other guardians of the child who uh will do have a legal right to access the child's record, perhaps they are not, uh, no longer having a You know, happy family situation, and the mother would like to have their Health history, uh, kept confidential from the other parent. And so, in this scenario, um, our newborn nursery has really rethought, oh, we shouldn't just be, you know, copy pasting the entire maternal record into the child's record. In addition, the child will, you know, as they get older, they'll also have access to their own account and then see all sorts of stuff about their mother. Lots of interesting scenarios here. Um, I was at dinner, Doctor Chuck Birdy, one of our pain medicine physicians here at Boston Children's, and relaying some of the things that we've run into in our own pediatric pain clinic at Stanford. Interestingly, you know, pain medicine notes aren't treated like psychotherapy or psychiatry notes. What do I mean by that? Well, the reason why I draw that distinction is because the 21st Century Cures Act information-blocking rule has certain exceptions, specifically carved out for psychotherapy. However, pain medicine notes don't count. And therefore, you know, the sensitivity of some of the topics that are discussed in pain medicine, such as psychosocial dynamics and um things that are going on with either the child or family or beyond, are now kind of more vaguely documented by some of my pain medicine colleagues, and they're worried that You know, they can't be as fully transparent as they normally would have been in the past when things were not shared, uh, completely with their families. Another example, uh, shortly after the Cures Act went live, um, this was a letter to the editor in anesthesiology. This was an example of a patient who had undergone a routine anesthetic, a young lady who had initially a laryngeal mask airway placed that wasn't seating well, so it was replaced with an endotracheal tube. But because the, only the intubation record was released to the patient, Several days later, the patient called the anesthesia practice back and was extremely distraught and thought something terrible had happened because there were two notes, an LMA placement and an endotracheal tube placement. Thought that she had had something really bad happen. Obviously, she was awake enough to make the phone call. But, again, only saw bits and pieces of the record, and as a result, was very anxious. The letter actually exposed some other things that were interesting about how people have responded to the CURES Act. They went through kind of risk management and legal channels because they're like, oh, I don't want to have to deal with any kind of legal fallout from, from this situation. But obviously, um, I think if we had some anticipatory guidance ahead of time, where we explained, oh, you know, here's what's actually being released. Here are the things that you should look out for, and, you know, we're going to take care of you during the, the anesthetic and we're happy to follow up about questions later. But because there was only part of the record that was released, it led to this spiral of anxiety and poor communication between the anesthesia team and the patient. So, why is it that we're releasing all this stuff to people and what is the Cures Act? How did we get here? I've kind of illustrated all these like, oh, these, these are highly inconvenient and kind of potentially aggravating situations clinically, uh, for why transparency isn't so good. Well, well, it's gonna turn around soon, don't worry. Um, what's the CureRES Act? Why should we care about compliance? So, as we know, healthcare information has become increasingly electronic and Because it's so easily accessible to us and others, it doesn't really make sense that we don't give it to patients and families too. It used to be when things were in paper, you'd have to go to medical records, go through this arduous request process to get a copy of your own record. But now that it's electronic, it doesn't really make sense not to. And interestingly, there have been a lot of studies, as I mentioned earlier, OpenNotes is a concept and kind of concept, framework, and team here at Harvard that started at Harvard that has grown and actually studied this in-depth. So, where did this all come from? So, all electronic health information um was mandated to be made available immediately and for free in October of 2022. So, it's already been two years since, since this has been live. But the first open note study uh out of Beth Israel was in 2010, 14 years ago. And since then, they have found over and over again, not just for outpatient notes, but that's the most well studied uh arena. But for outpatient, inpatient notes, there have been many studies that they've conducted that show better patient and physician engagement, better understanding of, um, the healthcare, the health plans for, for the patients. And improved, uh, adherence to kind of the plan of care, all sorts of positive benefits, and some of the feared things such as, oh, it's gonna take longer to document, patients are gonna freak out when they see things that they don't like. All of those things actually turned out to not be, uh, the case for outpatient stuff. How about for perioperative records? Well, that's the thing, we don't know. Um, so, the Cures Act is part of codifying this concept that We should be more transparent with our medical records, with our patients, because ultimately, they own the record. So, 2016, the 21st Century Cures Act was signed into law. And you may hear about the 21st Century Cures Act in other contexts with respect to pharmaceutical development and medical device development. But specifically, the information blocking rule is a part of the Cures Act. That originally went live, or at least the final rules were released in March of 2020, but I think you may remember that we were busy with something else in March of 2020. And so, that's why they delete, uh, delayed the initial go-live until April 2021, when the final rules went live. And so, I mentioned earlier, there are some exceptions to sharing records, but these are actually really narrow, and for many of these 9 things, they're more having to do with the technology, not quite, um, Being at the, being ready for every component of the medical record to be shared. And the one thing that we can use, so you may notice in Epic, there's a share with patient question mark button for certain notes or labs. And so you can uncheck that, but it, um, in most, or at least that, in my institution, when you uncheck the share with patient box, you have to give a reason for why. And preventing harm is one very narrow reason. And, uh, interestingly, the way harm is defined, emotional distress is not considered harm. So, but if there is, uh, a risk of like imminent harm because you are releasing information that might trigger someone to commit an act of violence, then yeah, that's, that's an, that's preventing harm. But if someone's just going to be anxious about something, that is not a valid reason to uncheck, share with, share with the patient. So, information blocking claims are costly. Uh, if a patient or anyone actually, you can go on to an online portal and submit a claim, and each claim can be, uh, up to a million dollars against the healthcare facility or institution that has committed information blocking. As of November 30th, there's been over 1000 of these information blocking claims since, uh, the Cures Act first went live in 2021. And so, most hospitals have Interpreted this quite conservatively and said, oh, I, we don't want to get fined, so let's just push everything out as soon as it gets out. But, you know, I've heard from colleagues even here that when things are not exactly configured to be, uh, that to workflows that makes sense in the perioperative environment, it can lead to unintended consequences, especially in the perioperative environment. So, interestingly, there are some state-level medical associations that have advocated for bills that allow a reasonable delay in release of information, especially if it pertains to something that will change your life. For example, a cancer diagnosis, um, or, you know, something that would be considered quite confidential or sensitive. And so, uh, both California and Kentucky are actually the only two states that have passed a bill that would allow, uh, a slight delay for the physician to review and then reach out to the patient, um, if they, if this happened. Uh, interestingly, Massachusetts does not have such a law. So, OK. Hopefully you're not like, oh wow, there's some can of worms here. So, My argument is let's actually use this as an opportunity. How can we use OpenNotes to improve patient-provider understanding and relationships in the peri-op world? And what can we do as individuals, as institutions, and at the professional society level? So because we're releasing these notes, it feels like, oh, the spotlight's on us now. How do we move from reacting to these changes to transforming with them instead? I think early on, we, um, we conducted a survey at both Stanford and UCSF about anesthesiologists' awareness and knowledge of the Cures Act in 20, late 2021, which was after the Cures Act had already gone live, and there were many of my colleagues who were saying, can we undo this? Is this, and we're like, oh no, we can't, the train already left the station. And Because of that, that really spurred us to think, oh, we really need to educate and help people get on board and think more creatively about what is actually happening with the CURES Act. Now, in our world, you know, think of a time when you helped a patient or family member or colleague understand what we do, whether that's anesthesiology, surgery, critical care, pain medicine, right? In pediatric, in the pediatric world, We are really good at delivering care, which is great, but also helping people understand it in a compassionate and kind of patient-centered manner. And so, we're really positioned to do this really well. So how can we do that? As I mentioned earlier, OpenNotes has really studied this quite thoroughly. Um, so, the senior author here, Tom Del Blanco, is an internist at Beth Israel, who was one of the co-founders of OpenNotes. And as I said, in the outpatient world, you know, we found that patients find reading their notes to be important for their health management. They share their notes with loved ones and caregivers, and patients are rarely, uh, freaked out or troubled by what they see. And interestingly, another signal that came out of this is that traditionally underserved patients, uh, particularly benefited because oftentimes, you may be in the clinic and your physicians speaking at you and you don't really register all the things, uh, that they're saying, but then you can look it up afterwards in your, um, in your, in the written clinic note to remind yourself, oh, right, that's what we discussed. This is the plan of care. So their conclusion, this study came out about 5 years ago, um, OpenNotes brings benefits to patients that largely outweigh the perceived risks. There are really minimal downsides for the outpatient world. So what about, as I said, in anesthesiology, surgery, per-op environment? Um, a couple of years back, uh, you may recognize Doctor Marsiglio was a former fellow here. Um, we came together and said, well, we actually haven't thought carefully about this, uh, in anesthesiology. We really think we should have a framework for how we should move forward with sharing anesthesia records with patients. And so What if we seize this opportunity to engage with patients and families, and where might this take us? So I'm not saying that you individually have to take on the burden of figuring this out, but there are things that you can do on an individual basis that can help move this in the right direction. So, the first thing is think about what you are documenting and why you're documenting it. Because as we've learned over time, um, words do matter, right? And so, here, here's a study I'll highlight about how stigmatizing language can change the outcomes of patients. Here, um, they compared discharge summaries for patients, uh, who had opioid use disorder and for patients in which the discharge summary used the term opioid opioid use disorder as opposed to something more pejorative, such as, here's a drug abuser, um, they were far more likely to have an appropriate treatment plan and additional specific follow-up care. Because when we use person-first language that's non-stig like to de-stigmatize the kind of conditions that our patients are coming in with, it does change the way we think and frame um our patients, even though, you know, our, we would like to think that we don't have unconscious biases like this. And this has been well studied in the um internal medicine world, but less so in per-op. So, you know, in our pre-op evaluations, our surgical notes, nursing notes, flow sheets and screening forms, all of these things, I'm not sure that we've carefully thought about all the things that we are documenting and how a patient might perceive reading those things. We know that in OpenNotes, People are not really distressed by abbreviations or jargon. Those things, they can figure out. But I think the principles of not using pejorative language, um, and Using person first language can be very helpful, right? Centering the patient and family is really becoming the norm, right? And so, Our pediatricians who trained here use family-centered rounds. Our notes are being shared with patients and families, and clinical records such as the anesthesia record, which is complex and confusing, are also now being shared with, with patients. And as a result, one of the things, um, that other centers who have kind of gone live with this earlier, um, colleagues at Vanderbilt University, for example, they've kind of included in their pre-op consent process. A kind of disclaimer and discussion with the patients and families, hey, we may be releasing records um to your patient portal before we have a chance to discuss it with you. And so it's kind of changed the individual practice of the anesthesiologist there because they know that different parts of the record will be released in this kind of immediate release fashion. And so, one thing to consider, right, is not only how you chart, but also how you discuss the fact that records are going to be released. How might you provide some of this, um, expectation setting before some of the anxiety that was mentioned earlier, uh, results. Another thing that's been interesting is that scholarship on the Cures Act is currently lacking, specifically in anesthesiology and surgery. Um, Family Feud style, guess who are the top three publishers, uh, specialties publishing things about Cures Act? Shout them out. Internal medicine, Oncology. One more. Pediatrics, yeah, pediatrics, internal medicine, emergency medicine are the top 3 publishers of Cures Act scholarship. Unfortunately, way down here, anesthesiology, way down here, general surgery. So, we have some catching up to do, um, in terms of the scholarship with respect to the Cures Act, information blocking and sharing patient records. So, there are a lot of opportunities to engage in research and quality improvement. So remember how I showed that slide that over half of adolescent patient portal accounts were being accessed by parents? Well, uh, we, we didn't just leave it at that. We underwent this quality improvement initiative to ensure that those records, uh, those portal accounts were corrected. In which we reached out to those families and said, here's why it's important for the adolescent to have their own account that isn't being accessed by their parent. And so we had them either change their password, have the adolescent confirm that this is actually the child and not the parent, or close those accounts. So, we were able to rectify that. And in addition, Um, we were able to create a guardrail in Epic that prevents people from activating a portal account using the contact information of the parents. So, when you're signing up for the patient portal and you're signing up for an adolescent-specific account, if it has the parent's phone number or the parent's email, it says, nope, you need to use the child's email or phone number. And so, interestingly, when we developed that, Epic said, oh, you know, that, that is a good thing to do. So now, all Epics do this, right? So, when you see these things that are happening in front of you, and you're like, oh, that's not good, we are able to bring these up to the EHR vendors and have them not just fix it for our own institution, but actually fix it for the millions of customers that Epic now has, which includes most of Boston now. But this is something that we kind of wrote up and said, hey, There may be other things that we haven't noticed, but that you guys might notice and say, hey, this needs to be fixed. And perhaps whether that's from an academic standpoint or just from an operational standpoint, um, these things do impact patient care and patient outcomes because we, as we said earlier, we don't want to breach adolescent confidentiality and have an inadvertent, um, poor outcome where an adolescent Has had something they did not want to share with their parents get shared. So In our um patient portal, right now, the record sharing of what's in the anesthesia record is actually pretty rudimentary and incomplete. And so, um, Doctor Galvez was kind enough to share what, uh, her child's anesthesia record looks like in the patient portal here at Boston Children's. It's no different. So, you see some of the notes, you see the procedures, you see the medications, but no doses, no timing. It's really not what we see on the clinician side. So, what if we showed patients and families what we see as clinicians, right? So, Uh, one of the nice benefits for our ICU colleagues these days is that you don't have to jump into a separate AIMS system to look at what's happening in the OR. In Epic, you can see what's going on because it's all the same record. So, that's really helpful for clinicians. But what if we showed the anesthesia record to patients and families? So, that's exactly what we did. Um, some colleagues and myself, we interviewed 20 adult patients. Um, who had undergone an uncomplicated anesthetic. did a qualitative semi-structured interview and asked them, well, what's interesting to you in this record? What jumps out? What's confusing? What might you, uh, what other things do you want to know? And we did a parallel study in 20 children who, uh, had undergone anesthesia and interviewed one of their parents, uh, to ask the similar questions. And so, um, I won't go into too much depth, but happy to talk about it, uh, sometime later today if you're interested. But I'll highlight a couple of things that, uh, of the qualitative comments that, that, uh, came out of the study. And so, here, I'll just read this to you. Basically, this is a patient who had undergone anesthesia, who has now seen their own record. Basically, I had no idea, and now I feel like I have a very good idea of what went on, and I feel more relaxed now because it all went fine according to the notes. It does make me appreciate what goes on in the operating room and the level of care that I received. And I guess I see there's a lot more than just count backwards from 10. I'm seeing that there's a lot more complexity in terms of like different drugs being given at different times, cause I mean, on Grey's Anatomy, the dude is just sitting there doing a crossword. I mean, it seems like there's a lot more going on. Right? And so, patients do appreciate seeing like, well, this is what's being documented because that's what we're providing as anesthesiologists. What else are they interested in knowing? Um, they're very curious about the medications they're being given, uh, the case tracking events. So, the timeline of what's going on is very interesting because when you're unconscious, If you've undergone anesthesia before, You lose consciousness and then you wake up several hours later, and you're like, oh, what happened? But the timeline of what happened of in-room, induction, intubation, surgery, emergence, extubation, etc. um, that kind of paints the picture of like, well, what happened while I was out? And that's actually quite reassuring to patients and families. Um, a lot of people were interested in kind of a summary or big picture, uh, view of like, well, did it go OK? I mean, obviously, we specifically selected uncomplicated anesthetics, so we didn't have to explain complications or things that went wrong. So, obviously, the child or the patient themselves didn't have anything. Uh, interesting to kind of report, but they're interested in, in knowing, OK, well, here's all this complex stuff. Did it go OK, right? So, currently, that is not something that you can derive, uh, without expert knowledge. And so, um, we also showed them the Gryphen grid. So, remember going back to this, many people see this for the first time and they're like, what are all these squiggles and dots and symbols? Um, patients were actually not disturbed by that, but they were actually curious, like, what does that mean? Oh, there, and we asked them specifically, are you alarmed if there was a low blood pressure? And like, no, I assume you dealt with it. And Usually, that's true because you can see in the drug administration record, if someone was hypotensive, we treat that. Um, and so, there are different components of the anesthesia record that people pointed out. Um, and then I'll highlight a couple of comments from the pediatric study. Um, interestingly, many of the parents, uh, were more concerned about their own child instead of their own health, which I guess kind of tracks with parents, if anyone's a parent out there. So, this parent said, I think if I was looking at my chart, I'd be like, all right, whatever, I'm alive. But because it's her, I'm just like being a mom. And so this particular parent uh was very interested in what had happened to their child. Um, Another quote I'll highlight in the pediatric induction and behavioral assessment in the last line where it says recommendations for next time, the pre-medication, that kind of stuff is what's important to me. So, we have a standardized pediatric induction form that uh kind of allows us to document in a standardized way how the induction went. So, if the child was kicking and screaming and not having a good time with mice conduction, um, We can provide possible recommendations for next time. And this particular uh parent of the child said, oh, that was really, that's kind of helpful for the next anesthesiologist. Another thing that came up in both the adult and pediatric study, and this is a kind of common issue with all electronic health records, um, is the idea of copy paste and, um, macros that are not actually true. So, this person, uh, said to us, you know, you should not be creating documents that you've cut and pasted from somewhere else because once you put this kind of thing in there, it raises questions about, OK, well, what else in here is made up, right? And I'm sure I'm, I'm guilty of it, and you may relate to this too, where you've either clicked a macro or you copy forwarded or uh Use something in your documentation that you didn't necessarily verify 100%, so it's an issue. The other thing that um I'll say about how we can kind of address CURES Act stuff is at the professional society level. So, as I mentioned, three years ago, 2021 is when C CURES Act went, Cures Act went live. But just this October 2 months ago, um, with the help of many friends here in the audience as well as, uh, nationally, the ASA, the American Society of Anesthesiologists, we published this statement. On the 21st Century Cures Act final rule, information blocking compliance. So basically, prior to this, there were no national society guidelines on, well, OK, well, what do you do about this law with respect to um how we release perioperative records. And what we recognized earlier. is that we should not be releasing the anesthesia record in bits and pieces and give people an incomplete picture while intra-op, right? I've heard stories from across the nation and even here where parents may have seen bits and pieces and then become quite alarmed, called into the operating room and said, wait, what's going on in there, right? And that's actually quite disruptive to the care of the patient who's in the operating room. And also not in the spirit of what actually the 21st Century Cures Act wants, right? We're not against transparency. We want patients and families to have the medical record, but we want it to, we want to give it to them when it's done, right? In the same way when you go to your primary care, uh, provider's office. You will see them and they'll, you know, do their history and physical, discuss the plan of care, but you don't see their note. Until it's signed. Why, why would they release their halfway done note, um, before it's ready? So, in the same way, why would we release the halfway done anesthesia record in real-time? That doesn't make sense and what we've advocated for here in this statement from our national society is, let's think more carefully about that and can we provide guidance to EHR vendors to tweak some of these settings. Cause the EHR vendors are actually very amenable to adjusting it to whatever we want. It's just that they need the expert and professional input of the, uh, the physicians and whoever is kind of looking at this carefully and thinking about it, right? And going back to uh the two state laws in California and Kentucky that do provide some degree of um ability to delay, the American Medical Association actually has a model legislation, basically a template of what a law would look like should a state house want to enact something similar to California and Kentucky. And one of the things that they cite, they had a survey of 1000 patients and They asked them, well, would you want, uh, the choice for your physician to reach out to you before you received life-changing, um, test results. So instead of, you know, the results of your bone marrow biopsy popping up in your patient portal, and then it's the weekend and you can't get a hold of the physician until Monday. We want people to have the choice of being able to say, no, I'd really rather my physician contact me with those results as opposed to the default, which is this rigid idea that anything that comes through electronically must be released immediately once it's signed, right? Cause there are many of our oncology colleagues have had to deal with this uh this idea, whether it's radiology scans or pathology things where These have popped up in the patient portal, but there's no one to guide that patient or family through that situ situation. And so, one of the mitigating factors for that is, yes, you could tell patients if you are anxious or you don't want to see that result, don't open your phone. But as you can imagine, People are not great with self-control. So, It's still this kind of idea of How do we give people the things that they want without putting an undue burden on themselves, right? And so, On an individual basis, be mindful of how you document, knowing patients may read it. Try and set expectations and discuss, provide anticipatory guidance of what are the implications of immediate record sharing. So, if, you know, if I'm A surgeon having multiple operations all day, and I'm signing my op notes and they're being released, and the parent has a question about it. Well, you might not be able to get back to the parent till the end of your operating room day, right? So, letting them know, here's what to expect. There's, as I mentioned, surgery and anesthesiology, we're way behind in the scholarship world, at least in terms of publications. I mean, maybe there are people out there doing quality improvement, um, projects out there but haven't published. But, you know, think about joining or collaborating on research and QI efforts in per-op record sharing. As an institution, ensuring that we have consistent messaging about how we're releasing records and working with our IT teams and legal teams to say, hey, you know, actually, it doesn't make sense to release only pieces of the anesthesia record. Let's Put a delay until the whole thing is signed, right? Many institutions have moved toward that based on the statement that we were able to uh put out from the ASA or individually, even before that statement was published. But that is a thing that we now can stand on and say, you know, we have thought about this carefully as professionals and we are not against transparency, but we want to ensure patient safety during the surgery. And provide parents and caregivers of patients who are undergoing anesthesia, a way to kind of understand what's going on once everything is done. And as I mentioned, at the professional society level, developing, um, and advocating for recommendations and legislation is another way that you can get involved and ensure that based on your own experiences, whether that's as a surgeon, anesthesiologist, critical care physician, pain medicine, there are many ways in which this affects all of us, but My call to action to you is get involved. And even if you're not gonna get super involved, your own clinical notes will be shared with patients. So think about the implications of that. Where do we go with this? Um, As I mentioned, uh, Epic, which is our EHR vendor at Stanford, which is your EHR vendor. Um, they provide a patient portal called MyChart. And so, as I mentioned earlier, that kind of redacted rudimentary view, uh, could be improved, right? And, well, how will they improve it? Well, we have to provide input on what do patients actually want to know and see. And so we have an ongoing relationship with them to help, uh, provide them with recommendations on what we might show them, what we might show patients and families through the patient portal application. Um, another kind of obvious next step is, well, what do patients and families want to know from surgical documentation? Um, what's in operative notes that is interesting or not interesting? Um, I have several, many surgical friends who have told me that in their clinics, some patients do ask after the fact like, oh, I didn't understand this part of your op note, and she was surprised to hear that they even read the op note, but That's the thing, we don't know. We, if, if we don't study it, then how can we say other than anecdotally, that this is what patients and families are thinking and feeling about seeing these things. Is there a role for explainer apps? Basically, uh, could you build a You know, either I Hot topic these days, artificial artificial intelligence, large language models. Can we feed some of the medical text and data into uh such an app and have it provide uh more patient-centered or um Uh, more kind of user-friendly explanations of what happened. Maybe there's a role for that. And so can we combine, and here's the, like, next, next step to that is could we even combine the clinical narrative text that we're putting in with the waveform data that's in our monitors, right? Can we have um these AI models explain, oh yeah, this anesthetic went fine, or uh this patient is at risk for vasoplegia, they're gonna, you know, be hypotensive for the next 18 hours in the ICU. Could we Take some of that, um, data to integrate what we see as clinicians that pattern recognition and expertise, could we model that, um, in an AI model? And perhaps this could serve a dual purpose, not just clinically, but also interpretively for the family. So, remember, medical records belong to the patient's family, but we're still there, surgeon, anesthesiologist, and we can and should do more to optimize communication with our patients regarding the care that we provide. So, thank you. Um, happy to take questions. I really want this to be a discussion. Thank you. All right, great. James, thank you so much, uh, very thought provoking in a lot of ways. I'm willing to bring the mic to anybody who has a question. I would just start maybe with, um, from an anesthesia perspective, we have a lot of records that are started by one person and finished by another person, and accepting the large language models and AI stuff, currently, many of us will take information, make a diagnosis, or. Make a conclusion that may not be the same from one provider to another, so I open a record. I leave, somebody else finishes the case and finishes that record. The record goes out to the patient and they have thoughts, conclusions in there that I actually don't agree with or don't, uh, I have a different opinion about what happened. Have you guys thought about that or dealt with that at all? And now the patient comes back to me and says, why is this in the record? And I say, well, I, I actually don't agree with that. And sort of this this mixed messaging that can occur in records even among one record or from one record to another where you're getting different results or different conclusions from the same data. How do you deal with that from one of the things that came up in our interviews with both patients and parents. is that they did see that there may be multiple people involved in an anesthetic, and that's something that's very routine in our world, handing over a case or having a trainee or CRNA or uh any number of different care team members, which can all have different opinions or different conclusions as you mentioned. And so, they didn't specifically say, oh well, this, this idea of attribution, right? Um, they didn't say anything specific to, oh, why did this person do this at this time point, but what they did say is they recognize that, um, different people are involved in the care, and if we explain to them, Again, this is part of the anticipatory guidance cause sometimes If I'm solo, uh, taking care of patients, I may say to the family, you know, if you see me in the cafeteria, it's because one of my colleagues is covering for me, right? Because otherwise they're like, I just met you in pre-op, why are you here in the cafeteria? So, Thinking about how do we say, OK, there may, at some point, uh, during your child's care, another anesthesiologist may be involved and they may, uh, document things that are different than, uh, what we originally discussed, but either way, uh, I still have confidence in my colleagues' judgment and safety. So even though I may not agree with their management, it is still hopefully acceptable and safe. Are there other questions? Anybody else have issues? James, uh, really amazing talk and something that I think we're needing to think about more and more going forward, and I'm sure it'll just amplify as the years go on. With regards to the anesthetic notes that get generated in the procedure notes, it's amazing. You go, um, have a big surgery, and the record the next day reflects like 6 anesthesia notes and then 1 little tiny op note, and, um, it sort of dominates the medical record, but those notes are very Um, clinical, right? It might just say airway attempts, this, that or the other. It's not really a patient-centric note. Do you think that the opportunity for us going forward is to try to include additional language or explanations, sort of thinking about that the patients are looking at these records, or do you think there's an opportunity for vendors like Epic to. You know, just like on a Kindle, right, if you're reading and you click on a word that you don't understand, it'll pop up with an explanation, like something like airway attempts, for instance, like, if you hover over the word attempts, it might give you a nice little explanation of attempts can be multiple because of endotracheal tube size, this, that or the other, and it does not reflect a negative, um. Yeah, so I would say it's definitely the latter, right? And I have this slide. This is a, uh, app out of Duke where they took radiology reports and did exactly that. So you're able to mouse over different jargony terms in your CT scan. What's an appendix, and they scan like highlight over it, and it pops up a little image that explains this is, you know, part of, you know, it's close, it's part of the intestinal tract. And this is what's being cut out, right? And so, yes, there are opportunities for more patient-centered kind of third-party or like overlays that don't distract from the efficient clinical documentation that we have to use. Like, a lot of people fear like, oh, well, now I have to like add additional text to explain what I'm doing in plain language. I don't think that's what we're, that's not what I'm advocating for, and I don't think that's our goal either because Patients are smart enough to figure stuff out, but what people have repeatedly said in our interviews, they've said, it would be nice if there was an explainer, and that's exactly what you kind of proposed. So, if you, you could build the next anesthesia explainer app. James, um, great, great talk and summary, um, couple of, um, one comment, um, and question. You know, as far as like the, the Work that you so well, um, highlighted, you know, it's what's interesting to me too is that there's like m this has been going on now for multiple administrations um of the United States government and the philosophy is that yes, patients may not necessarily understand whatever it is that they're reading, but the responsibility like it's their right to read it and also if they need help figuring it out to seek that and that's been like the overwhelming. Message. Um, so, it, it, I'd love to focus on like how to do this kind of stuff. Um, as far as explainer apps, I, well, the other thing I'm curious about is like language. So we do have like consent forms that are in different languages and things like that, you know, but how, uh, I mean, Epic, the MyChart, I've tried to like change the language on mine to see if I can get things to show up in Spanish. Um, any work being done on this? Um, right now, there is work, so Epic does have MyChart in Spanish, Mandarin, and Vietnamese, um, but it does not translate the clinical notes, and, you know, I think that's gonna be a real challenge because You would have to validate that the medical translation is accurate, and that's not something that we're, I don't think. AI is ready for that, and even if we were to have medical translators translate stuff, that would be an enormous lift and so much resources to make that happen. Um, what you were saying at the beginning of your comment is basically this, regardless of how you feel about Joe Biden, you know, he went toe to toe with the CEO of Epic and said, Uh, so, you know, Judy Faulkner, uh, still the CEO of Epic, said, why do you want your medical records? There are 1000 pages and you only understand 10. Joe Biden said, none of your business. If I need to, if I need to, I'll find someone to explain it to me. By the way, I'll understand a lot more than you think I do, and, and so that really highlights that, um, idea that, you know, patients should own their records, they do own their records and they should have their records. Well, I'm putting that out there, could you just address, so with what you just said, how far are we from a standpoint or a time frame where you do own your record, that it's a blockchain cloud-based record that is available wherever you go, whenever you go, and you own it, it doesn't belong to the hospital. Well, uh, based on law, you do own the record, but it's just the challenge of like how do you export it. So right now, patient portals are a window into The epic instance that, you know, belongs within Boston Children's, and if you said to uh hymns, kind of the medical record people, give me the whole record, they'll print it out for you or they'll put it on a flash drive or CD for people who know what CDs are. Um, but they'll do that and you'll own it, but The ability to export is like not super robust depending on where you're at. Uh, there's a question, yeah. Uh, James, thank you. Uh, very intriguing talk. Um, I'm a surgeon, uh, full disclosure, um, in a, uh, a typical, um, scenario, a patient may have an operation, they may come back to see us and follow up. Um, what if they ask me, uh, a bunch of questions about the anesthetic record at that time? What should I do? Yeah. Um, so, when this has happened at our institution, they get redirected to the anesthesiologist on call, and so there have been some, uh, When I was on call a couple of months ago, my resident received a phone call about an anesthetic record that was released to the family, and we just walked them through it and said, what's, what's the question? And um, You can redirect to your anesthesiology colleagues. Again, this is one of those things where we haven't worked out workflows, like institutions like this in Stanford, we, during the weekdays have the benefit of a pre-op clinic that is well-staffed with nurse practitioners and a physician, and so, we have the like flexibility to have someone respond during the weekdays, but when it's overnight on call, parent calls in and says, I, I really need to ask you about something anesthesia-related. And we're like, OK, and then we just have to give the caveat, I was not the person who provided the care. What's the question? Can I explain it to you? And if, if the person who's on the receiving end of that phone call doesn't feel comfortable or isn't able to explain it, then we can redirect it to the original anesthetist. But if they're on vacation, well, then we just have to say, well, they're not available right now. Can we get back to you? Oh The Mica. Isn't any better? OK, good. Sorry, um, just following up on the record sharing across institutions and patient ownership, um, you know, another legislation that went into effect with meaningful use as hospitals, uh, rolled out all this stuff, um, is this, um. Mandate that any. Vendor electronic health record vendor has to be able to share with other vendors, and there's health information exchanges set up around the country that essentially enable this. So like Epic isn't able to say I'm not gonna share my records with Cerner. Are you an OR 12? OK. Are you OK in there? And they, they, they can face similar penalties to the million dollar thing that you saw, um, at from a federal level standpoint and. There's two little nuances to this. One is that unfortunately, we don't have like a national identifier, like medical record identifier for patients. So they, the, how a patient gets matched from one hospital to another can be really tricky, and there can be mistakes which are insidiously hard to find. It, as a provider, you may open a patient's chart, look at care everywhere, think you're seeing their cardiac cath report from another hospital, and it actually belongs to someone else with a similar name and a close enough match. So, if it, if the data doesn't really fit the picture, pay attention to those issues. Um, personally, my own story, um, you know, I legally changed my name, transition genders, etc. have had care in multiple hospitals and like, literally, I, I mean, I just one phone call to an institution, like all the records get connected and people can see them everywhere. Um, it's pretty uncanny. Thank you, Julia for that comment. Um, thanks, I, I think we're about at time. Um, one more question, right. Especially in responding to patients' notes are starting to have AI generate responses. And I know that different institutions are dealing with different ways of how they're letting patients know they are or aren't generated by AI. And now, then physicians are sometimes expected to then review the AI generated stuff and make sure it's accurate, which I don't know if it's really saving them work, but what Kind of your thoughts about that, um, right, so we do have a, we are one of the pilot sites for, um, AI generated responses to clinical messages at Stanford, and the anecdotal response from some of my colleagues who are inundated with, you know, in, inviskit messages is that you're right, it doesn't save them that much time because unless it's a very straightforward request. Um, the more complicated responses still have to take a lot of input. So, AI is not going away, it will get better, but right now the current pilot is like, uh, it's OK, it's not saving a ton of time, but that doesn't mean that it won't in 1 to 2 years. This does seem like there's a lot of nuance to trying to help people understand the subtleties of what they might see in that record and there probably is a lot of work to be done. Anyways, James, I wanna thank you so much. I think this was fantastic, gives us a lot to think about, a lot to improve.