Hello, and welcome to today's webcast, Mitigating Diagnostic Delays and Errors, 1C is provided for your attendance. The webcast is funded by the Gordon and Betty Moore Foundation through a grant program administered by the Council of Medical Specialty Societies. Be sure to check out the companion podcast, which offers 0.5 credit hours for continuing education. This content will be available today at 2 o'clock Central Time. Registration information is available on your control panel. Welcome to the webcast. My name is Susan Lacy, and I'm the Associate Director of Research and Quality at the Society for Critical Care Medicine in Mount Prospect, Illinois. I will be moderating today's webcast. A recording of this webcast will be available within five to seven business days. To access the recording, log into mysccm.org, navigate to My Learning tab, and click on the Mitigating Diagnostic Delays and Errors course. You will find a handout, evaluation, and recording in the course section. Thanks for joining us. A few housekeeping items before we get started. There will be a Q&A at the end of this presentation. To submit questions throughout the presentation, type into the question box located on your control panel. Please note that the disclaimer stating that the content to follow is for educational purposes only. And now let me introduce you to your esteemed speakers for today. Dr. Mary Jo Grant is a board-certified advanced practice registered nurse. She's also in the prestigious Academy of Nursing and is a fellow. Of course, she's a member of SCCM and is a PICU nurse, practitioner, and researcher at Primary Children's Hospital in Salt Lake City, Utah. So now I'll turn things over to Dr. Grant. Thank you, Dr. Lacey, for moderating this session and thank you to the Society for Critical Care Medicine for the opportunity to speak on this important and timely topic. We have a number of objectives. The first is to define diagnostic errors. The second is to review the causes of diagnostic errors. I'll describe clinical clinician characteristics found to be associated with diagnostic errors, and I'll review system-based failures leading to diagnostic errors and delays. I will also describe methods to mitigate diagnostic errors in terms of clinician behavior modification and changes to limit system-based failures. First starting with the definition. Diagnostic error is defined by the National Academy of Sciences Engineering and Medicine in 2015 as the failure to either establish an accurate and timely explanation of the patient's health problems or communicate that explanation to the patient. This definition focuses on the outcomes of the diagnostic process, recognizing that diagnosis is an iterative process that solidifies as more information becomes available. The diagnosis needs to be timely and accurate so that appropriate treatment is initiated to optimize the patient's outcome. I'm presenting a second definition from 2015, which is a missed opportunity to make a timely and correct diagnosis or take the next diagnostic action step based on available evidence at the time. This definition is a system-based, recognizing that all data may not be available at the time of the initial diagnosis. Diagnostic errors are important for us to understand because of the large scope of the problem. Major diagnoses affecting survival were missed in 20% of autopsy patients who died in a pediatric ICU. And it's important to understanding the frequency of diagnostic errors beyond autopsy patients. In a survey, 81% of pediatric cardiac ICU clinicians reported they witnessed diagnostic errors cause harm in children more than five times each year. Diagnostic error has been reported to be in 7% of adult ICU patients. And diagnostic errors have been reported to account for 70% of medical errors. And as many as 18% of patients in a rapid response experience a diagnostic error. It's also important to note that tools for identifying adverse events do not screen for diagnostic error. As a result of all the work in patient safety, a critical emphasis on clinical decision-making has materialized. Major organizations such as the Society for Critical Care Medicine, the National Academy of Sciences, and the World Health Organization has called for explicit training in diagnostic reasoning, human factors, critical thinking, cognitive heuristics, and biases, probability concepts, and systems thinking. The National Academy of Medicine has released a report on the topic titled Improving Diagnosis that directly addresses diagnostic failure. The patient's diagnostic journey really starts with critical information gathering by clinicians. There's a synthesis of key information and then decision-making and communication. It is important to use diagnostic timeouts or pauses in patient care to reevaluate the patient's diagnosis. The Children's Hospital Association has created a team diagnostic timeout template that can be used for this process. Diagnostic timeouts can be done when there are clinical triggers. Clinical triggers include a transfer to a higher level of care, a lack of improvement or unanticipated deviation from the plan of care, an abnormal finding that doesn't make sense, a concern identified by any one of the members of the clinical care team, and diagnostic uncertainty. A timeout should occur when what is going on doesn't fit our explanation of the problem. It's a time to update and reprioritize the problem list, review all the diagnostic tests, determine if new tests are needed, if other people are needed. The question is, is the organ system or process focused on the only one that could cause the symptoms that we see? Is this an atypical presentation of a common disease or rare diagnosis? What's the worst-case scenario for this patient? There are a number of causes of diagnostic errors that include cognitive failures, admitting a critical part of the physical examination, misinterpreting a laboratory value, not considering an appropriate differential diagnosis, premature closure, or not expanding the differential diagnosis. There are system-based failures, poor coordination of care, inadequate mechanisms to convey clinical results, such as lab results that are not reported properly. And then there are both cognitive and system-based failures. We work in a very unorganized and chaotic clinical work environment that's gonna affect cognitive processes by way of distractions and interruptions. Two commonly diagnosed conditions, pulmonary or misdiagnosed conditions, pulmonary embolism and stroke, can be the result of both cognitive and system failures. The surviving sepsis recommendations from 2021 applied diagnostic delays and errors to the guidelines. We need to establish a work system and culture that supports the diagnostic process. Healthcare organizations are responsible for developing a culture that provides a safe place for professionals to identify and learn from diagnostic errors. The culture should emphasize quality, safety, professionalism, and the intrinsic motivation to promote the diagnostic performance. A culture that emphasizes discipline and punishment for those who make mistakes presents a barrier to the reporting process. Leaders need to collaborate to achieve quality goals, as well as successful change initiatives. We need to prioritize diagnosis to get support from senior management. We need systems in place to identify a diagnostic error early, so that actions can be taken to avert negative effects resulting from the error. Diagnostic errors may be prevented by having adequate staffing, by addressing fatigue, which includes decision fatigue and sleep deprivation. Cognitive processing is affected by the environment. Impacts of the physical environment include the layout of the ICU, noise, and lighting can all have an impact on the performance of the clinicians. We as clinicians are subject to heuristic thinking as a mental shortcut used to simplify problems and avoid cognitive overload. Part of this is how the human brain has evolved and wired, allowing individuals to reach reasonable solutions to complex problems. However, the very thing that makes heuristic thinking helpful, such as decisions based on logical assumptions gained from experience, can also lead to a systemic bias and incorrect decision-making when assumptions are wrong. Generally, intensivists begin their diagnostic evaluation after other clinicians have already evaluated the patient and assigned diagnostic labels. This creates a strong bias. Cognitive bias are systemic over cognitive dispositions in human thinking and reasoning that often do not comply with the tenets of logic, presumably reasoning and plausibility. Anchoring the tendency to be overly influenced by the first list of diagnosis we hear is the first impression. A recall bias is a systemic error caused by differences in the accuracy of events based on past experiences. Implicit bias occurs automatically, but that nevertheless affects the decisions that we make. Rational failures arise from cognitive biases, logical failures, and general shortcomings. They can be worsened by fatigue, sleep deprivation, cognitive load, stress, and burnout. A lot of work has been done recently on teaching cognitive bias mitigation, particularly in medical schools. By understanding how clinicians make clinical decisions, examining how errors due to cognitive bias occur, cognitive bias training and deep biasing strategies may be developed to decrease diagnostic error and patient harm. Studies of the impact of critical thinking skills and bias training have mixed results that are really very limited by their methodology. There should be cognitive bias awareness training for all staff. Differential diagnoses should be established and reiterated at diagnostic timeouts. An escalation plan should be established and reiterated at diagnostic timeouts. Trigger for an escalation plan include a patient with a high acuity score, such as high PRISM or PEW scores, a lack of response to treatment, and acute clinical change, such as vital signs outside of a normal range. And when there's a knowledge gap, members of the care team need to feel empowered to question decision-making of other team members. Patients use a gap analysis tool to assess for internal communication gaps. All human decision-making involves both intuitive and analytical decision-making. The trick is to weigh both types. Decisions vary in clinical characteristics, and it's unlikely one set of rules will apply to all patients. Though we're frequently in an intuitive mode, all decisions need to be monitored for rationality. If it's not rational, then analytical reasoning should be a force of habit. The care team may contribute 30% of diagnostic errors and delays. This is the result of multiple factors, including burdens of documentation and competing demands from other elements of high-quality ICU, such as adherence to protocols that may distract the clinician. Healthcare organizations can implement mechanisms that improve systemic feedback at all levels. Feedback can help clinicians assess how well they're performing in the diagnostic process. Correct ego and overconfidence, identify when remediation efforts are needed, and reduce the likelihood of repeated mistakes. As a first step, diagnostic errors may be reported in morbidity and mortality conferences, root cause analysis, department meetings, or walk-rounds. Patients with suspected sepsis but unconferred infection, we recommend continuously re-evaluating and searching for an alternative diagnosis. A system problem is that when missed or delayed diagnosis becomes obvious to a practitioner, it often doesn't lead to any learning or feedback for others who were involved earlier in the diagnostic decision-making. I'm an advocate for an ICU-centric approach to diagnostic error that emphasizes learning from missed opportunities in the diagnostic process that stem from either cognitive or systemic failures, really focus on errors that have caused or could potentially result in preventable harm. There's a growing body of research using information technology as differential diagnosis generators based on clinical systems and patient history. In these simulation models, as soon as the clinician enters the reason for the encounter, the system generates a list of diagnostic suggestions and groups them according to published incident rates. This may help with the many steps in the diagnostic process, including cognitive overload. This is a new, novel way for information technology to help mitigate diagnostic errors. In terms of assessment, hypothesis generation or failure to delay in considering the correct diagnosis may lead to diagnostic error. Notable weighting or prioritization, such as too much weight given to lower priority or lower probability diagnoses, too little weight given to higher probability and lower priority diagnoses, too much weight given to a completing diagnosis that may be novel or more interesting but less likely, a failure to appreciate the urgency of the illness, delay in recognizing complications, and there's also a risk in overtesting and ordering unnecessary tests that can cause harm. Risks and benefit must be carefully considered when ordering diagnostic tests. This is a role for diagnostic timeout. The risk of diagnostic error increases with increased acuity of illness. Differential diagnoses should be documented in the electronic medical record. Patients are often acutely deteriorating on arrival to an ICU. A comprehensive diagnostic evaluation, such as chart review, collateral history taking, or advanced imaging may be deferred until the patient stabilizes. This is where heuristic thinking dominates. Also, critically ill patients are unable to be interviewed or participate in the physical exam. Rapidly changing physiology generated a constant stream of distractions and a new diagnostic data point may be created. A recent study used triggers to identify patients with unplanned hospitalization within 14 days after a primary care visit and found a misdiagnosis rate of 20%. A manuscript published in the Journal of Patient Safety in 2021 reported that patients evaluated in a rapid response, 18% experienced a diagnostic error and 16% may have benefited from earlier referral to a specialist. The use of consultants may lighten the cognitive load and lead to fewer diagnostic errors or delays. The electronic record may also lessen the cognitive load. For example, clinical decision support tools that are refined enough to improve specificity without sacrificing sensitivity. The electronic record can be used to create novel triggers that automate detection of potential errors. What about an electronic record that triggers for substantial discrepancies between the emitting ICU diagnosis and the discharge ICU diagnosis? Referrals are enhanced by the use of telemedicine. Telemedicine expands the ICU intensivist reach into smaller community-based hospitals and emergency departments. However, specific data are lacking on how telemedicine influences diagnostic accuracy. However, we know that timely consultant with an expert is a recommended strategy for reducing diagnostic errors. There are a number of patient factors that may increase the likelihood of a diagnostic error. A failure or delay in eliciting a critical physical examination finding may be related to complexity and atypical presentation of the disease. Inadequate or misrepresented clinical exam findings, suboptimal weighting of critical examination findings, and a failure or delay in follow-up on critical examination findings. This is an opportunity for the clinical team to help with the decision-making. As noted in a model created by Dr. Barwais, and it was presented in a manuscript published by her. It outlines patient factors that influence the generation of diagnostic errors. These patient factors include patient complexity and atypical presentation of disease, the patient or parent's knowledge about their condition and their health literacy, their narrative skills, language barriers, cultural barriers, and behavioral issues. There are a number of interactional failures between all the domains, patient factors, clinical factors, clinician factors, and organizational factors. We need to enhance professional education to include information on diagnostic error prevention, including communication tools, critical thinking, and an awareness of biases. For example, clinicians learn from case studies that reflect a prototypical case, but actually they're faced with complexities of real patient cases in a clinical practice. We need to focus education on instruction and practice and generation and refining of a differential diagnosis, documentation of a differential diagnosis, and developing an appreciation of how diagnostic errors and strategies to mitigate them, such as coordination of care, structured handoff that includes articulation of the differential diagnosis, and supporting evidence for the working diagnosis. There is a concern about the use of copy and paste functions in the electronic record. It may increase efficiency. It comes with a number of risks, which include redundancy. It contributes to lengthy notes and cognitive overload, as well as the propagation of inaccurate or outdated information. Electronic medical records, such as sepsis detection tools, as well as in the presence of sepsis tools that can be used to dose antimicrobials based on pharmacokinetic properties may be helpful. Clinical surveillance tools may be used to identify comorbidities that may affect a patient's clinical course. But the electronic medical record is not optimized for the communication of diagnostic reasoning and certainty. There are reasons why hospital sepsis programs don't achieve the hoped-for results. This can be related to the inability to identify sepsis accurately without causing alert fatigue. Alert fatigue leads to late identification and delayed treatment. Late or error-prone delivery of the bundle. Disjointed programs that don't sync people, processes, and technology changes. I'm concerned that clinical documentation in the electronic record is not promoting high-quality diagnosis, but instead aimed at meeting billing and legal requirements, forcing physicians to focus on ticking boxes rather than fully documenting their critical thoughts. Thank you. In terms of infrastructure, an infrastructure may lack the ability to report a diagnostic error. The process of reporting diagnostic error in near-miss remains undeveloped and lacks standard measurement tools. Diagnostic error may result in death and are the leading causes of legal claims associated with death and disability. In fact, until a decade ago, much of what was known about diagnostic error was learned through the evaluation of data from malpractice claims. The diagnostic process is a team-based endeavor. Patients and family are partners in the team. Many of these issues are related to the electronic record. Because the diagnostic process occurs over time and can involve multiple healthcare professionals across different care settings, the free flow of information is critical. In regards to laboratory tests, close loose reporting systems that ensure that test results are reported back to the treating team in a timely manner. Some tests can take days or even weeks before they're resulted. Systems need to be in place to ensure that the data is passed on to the correct team. Unrelated to lab reporting are systems that report errors. We need to develop a reporting environment that facilitates the inaccurate diagnosis. Characteristics of a successful reporting system include that reporting is safe, that there is a need for safe environments without the threat of legal discovery and disciplinary action where diagnostic errors can be analyzed and learned from in order to improve the quality of diagnosis and prevent future diagnostic errors. Reporting leads to constructive response. Reporting of diagnostic error depends on identification of the errors and triggers help with the identification. Some ideas of triggers include an unplanned transfer to a higher level of care, cardiopulmonary arrest or emergent procedures. Expertise and financial resources need to be available to allow for meaningful analysis of these diagnostic errors. Healthcare organizations should have in place a system that's capable of disseminating information on areas that identify to be at risk within that organization and give recommendations for change. This framework is from work performed by Dr. Amelia Barwise at the Mayo Clinic in Rochester who's published extensively on this type of work. Generally, in terms of clinician factors, there are two types of healthcare professionals. Diagnosticians such as physicians and advanced practice providers and secondly, all healthcare professionals who support the diagnostic process. Working together depends on the quality in a professional education and training. In this model, it brings forth the relationship between organizational factors, clinician factors, patient factors such as I've discussed and then an interaction of all those factors. The prevention of diagnostic error begins with the critical care team by collaborating across disciplines such as nurses and respiratory therapists and engaging consultants, intensivists and emergency room physicians gain valuable diagnostic insight and therefore disperse their own cognitive load. Ideally, the intensivist develops a leadership style that empowers bedside nurses and other members of the interdisciplinary team to take part in the diagnostic process. For example, ICU nurses frequently gather collateral history from family members and experienced nurses often recognize when the patient's course or response to therapy does not fit the presumptive diagnosis. Here are some references that I use when preparing this talk. The diagnostic error research is underrepresentative. This may be because of a lack of awareness or perceived inevitability of the problem or poorly understood characteristics of the diagnostic and clinical reasoning process. Currently, there's a disease-focused approach to medical research funding. The topic of diagnosis cuts across various diseases and body parts and it's not centralized within the NIH research portfolio. As a step in increasing awareness, the Society of Critical Care Medicine has asked me to give this webinar and created a toolkit to begin to address this important topic. The goals for improving diagnosis and reducing diagnostic error include effective teamwork, enhancing education and training in the diagnostic process, capitalizing and maximizing efficiency of health information technology to identify, learn from and reduce diagnostic error in clinical practice and to establish a work system and culture that supports the diagnostic process. In summary, the road ahead begins with a system-wide recognition of the presence of diagnostic misses and errors, training and recognition of limitations in human behavior, the use of toolkits, such as those created by the Society of Critical Care Medicine in conjunction with this webcast to limit diagnostic errors in our critical care systems. Thank you. Thank you, Dr. Graham. That was excellent. So I'd like to see if we have questions from the audience. This is a larger problem in pediatrics as the frequency of certain serious medical problems in pediatrics is much lower and clinical experience may be limited. For example, pulmonary embolism is very frequent in the adult population and the adult clinician treat these kinds of patients at least weekly. In the pediatric population, the pediatric intensivist may only see that type of patient only once or twice each year. How do we address limited clinical experience with infrequent diagnoses? That's from Dr. Rubin. Would you like to answer that, Dr. Graham? Yes, I think that kind of goes down to the awareness of cognitive bias and human behavior. If we know that there is going to be some infrequent diagnoses that we will not see in our current clinical work environment, then you need to be aware of your own biases and your own factors that will limit your ability to make a broad differential diagnosis. I think that presenting a differential diagnosis and reevaluating at the time out to see whether this differential fits in with the clinical characteristics of the patients will help us to move a little bit faster towards making a diagnosis on something that's of lower frequency in a certain clinical context. Thank you. So I had a question for you, Dr. Grant. How do you role model if you're pretty sure that the diagnosis is wrong because of your experience and say you're working with a less experienced physician and you're pretty sure you're right about your diagnosis and you're in rounds with a group of people, how do you handle that situation where you feel like you need to speak up and talk to the physician about why you believe this diagnosis is potentially incorrect? I think that that comes to having a flat playing field where anybody feels empowered to speak up and question those regardless of their level of expertise and training. So first is the ability to speak up and working in a culture where speaking up is advocated by the administration and by the culture of the ICU. And secondly is to just, you know, question the rationale for a certain diagnosis, ask for supporting evidence for that diagnosis and potentially throw out ideas that may mitigate that area of thinking and help to broaden the scope of the differential. I really do like the idea of including the differential diagnosis in medical charting. That's something that I don't see very often and it's a very important way to convey what we're thinking because we all know that handoffs are done in a rapidly changing ICU environment where there often isn't time to perform an appropriate handoff that goes through, you know, everything that you're thinking on the differential diagnosis and supporting factors for or against that diagnosis. And so having this documented in the medical record, I think it's very important to convey this information. Yeah, thank you very much. That's excellent. There's a thank you. Is there a format readily usable for either physician or nurse handoff to enhance communication while reducing diagnostic errors in the ICU since diagnosis can change at any time in the ICU? So are there instruments or tools that you would recommend, Dr. Grant, for the webcast? You know, there are a number of handoff tools. And in my ICU, we use a homegrown patient handoff tool that was based on, it's evidence-based, but it doesn't necessarily reflect the diagnostic process. So that I know of, there is not a handoff tool that really focuses on the diagnostic process and creating an accurate diagnosis. So I'm sorry, I don't know. Okay. But you've developed one for your unit. Yes. Something like a handoff tool for your unit. So do you think that's the most critical issue, is the handoff from, I mean, obviously some patients come straight to the ICU from home or from their primary provider. But a lot of times patients move from within the hospital into the ICU. Which scenario is more fraught with the potential for diagnostic delays and errors? I think the patient that comes within the system, they usually have a diagnostic label already placed on them. We know, as I stated earlier, that a number of patients that experience a rapid response have diagnostic errors. And a lot of that is because they have a diagnostic label assigned to them. You know, it's the patient with respiratory failure from bronchiolitis in bed 3067. And they come down to the ICU, and potentially their respiratory failure is just a harbinger, and they have other diagnoses that we need to consider. But the respiratory failure is what needs to be addressed and what triggers the rapid response. So, you know, not to automatically throw patients into a box that represents what you have, they have been labeled as for their initial diagnosis. Right, so does that go back to your, and I'm not sure I'm clear on this, does that go back to your anchoring? Anchoring bias, yes. Yes, that's your anchoring bias. Okay. Yes, that's very helpful. Anything else that you would like to share with us about this topic that people can maybe take home and work on in their unit? Because we have quite a few people attending this session. Yes, this is, you know, it's really very broad. And like I said, there's a limit in research for this in this topic. You know, I like the idea of using information technology to broaden the differential diagnosis. I thought when I read about that research, I thought that seemed very exciting and cutting edge. I think that, you know, it's a while before you can apply things like that to the clinical context. But, you know, we do have a very robust medical information systems and the opportunity to use things such as sepsis alert tools, which, you know, given their limitations of alert fatigue, they provide the opportunity for us to at least recognize when there's a change or an evolving diagnosis with the patient. I think that, you know, things that you can do tomorrow, I think, is to, you know, talk about differential diagnosis. You know, many, most people think about them in their head, and that's part of our, you know, the heuristic thinking is that we come to immediate and logical conclusions quickly when patients fit into a certain box, but to think a little bit outside the box and see whether the patient truly fits into that and, you know, order diagnostic tests thoughtfully that will either support or refute your working secondary diagnoses that you have. I think that, you know, the thinking part and the communication of what you're thinking part is something that anyone listening to this broadcast can do tomorrow. Great. Thank you very much. Well, I think we don't have any more questions. I have a few more notes to share with you. Of course, we thank you for attending, and again, this webcast is being recorded, and the recording will be available to registered attendees within five to seven business days, and then I'll repeat that you log on to mysccm.org, navigate to My Learning tab, and click on Mitigating Diagnostic Delays and Errors. You will find the handouts, the toolkit, the evaluation, and recording of the course section in the course section. Remember, Dr. Grant did a companion podcast for us, and that is going to be available right now at 2 o'clock Central Time, and that is .5 hours of credit for continuing education. Registration information is available in your control panel. Also, keep an eye out for resources which will be available on the CCM Diagnostic Excellence Program page. Lastly, we will have three more funded webcasts that will all be dealing with different issues associated with diagnoses and diagnostic excellence, and that concludes our presentation for today. Thank you very much.

accurate diagnosis

timely diagnosis

diagnostic errors

cognitive failures

system-based failures

strategies to mitigate diagnostic errors

teamwork

communication

Webcast

Quality and Patient Safety

Sepsis

Sepsis

Evidence Based Medicine

2023

Professional

Select

Sepsis Resources