I'm delighted to introduce Dr. Karen Shea to tell us about implementing clinical practice guidelines. Dr. Shea, after college, joined the Peace Corps in the Dominican Republic, where she saw firsthand the effects of areas with poor access to medical care. She then trained at UCSF, Mayo Clinic, and CHOP, and worked at academic medicine before leaving to start her own practice in El Paso, where she is the only pediatric epileptologist for 250 miles. So she's the one who's really walking the walk in terms of disparities. Wow, thanks for that nice introduction. So I got the fun topic, which is how do we incorporate these wonderful guidelines into clinical practice? I don't have any disclosures. So our learning objectives are to learn how to critically appraise clinical practice guidelines, and then how to apply them in decisions regarding patient care. So, of course, I want to do a case. So this is a rural community physician, South Dakota. She gets a call from the local emergency department that one of her patients, a 25-year-old male, had a first-time convulsive seizure, and it came on out of sleep. The onset wasn't witnessed. His partner says maybe he was looking towards the left. Kind of hard to say. So in preparation for her appointment with the patient, she searches for an evidence-based guideline, and she locates the open access AAN guideline on management of an unprovoked first seizure in adults. She critically appraises the guideline to determine whether it's trustworthy and whether it applies to her patient. And that really gets down to kind of where we're going with this, right? The ultimate purpose of clinical practice guidelines is to assist the clinician in optimizing patient care. So a helpful guideline will provide clear and actionable recommendations. It'll explicitly state whether these recommendations are strong or weak, and it'll specify the appropriate patient population and whether or not they depend on individual patient circumstances. So in order to use these effectively, providers have to assess the trustworthiness of the development process and then evaluate the extent to which the recommendations are applicable to their patients and in their practice setting. So there are a lot of guidelines out there on how to evaluate guidelines. I guess I shouldn't necessarily call them guidelines. So guides on how to evaluate guidelines. These are just two of them. One's from JAMA, the other's from neurology, and they kind of walk you through the process of kind of deciding whether this guideline is appropriate for your patient. And I go through the JAMA one. So each of these have a list of kind of steps that you can do to assess whether the guideline is going to be useful in your situation. And I think the one on the right is from AGRI, the one on the left is from neurology, and like I said, I'll be sticking with JAMA in the middle. So the first step to assess the clinical guideline is to determine whether the recommendation is clear and actionable. And this involves or requires whether the patients, the intervention, the alternatives, and the recommended action are clear. And then also to discuss the strength of the recommendations. And so as I've been mentioned earlier, there's a lot of different processes or systems that you can do this with. So there's AGRI, there's GRADE, ACC, AHA, that's from the American College of Cardiology and American Heart Association. There's one from the U.S. Preventative Services Task Force, but that usually pertains more to primary care. And then we've already discussed the NICE as well. So in order to determine what the strength of the recommendation is, each guideline kind of has its own process. So the ACC AHA kind of divides it into class 1, class 2A, class 2B, depending on the strength of their recommendation. GRADE uses the terminology strong for or conditional, conditional weak, and so forth. And then once you've determined the strength of the recommendation, you want to look what was the excuse me, what was the evidence summarized with rigorous systematic review, which is kind of what we've been talking about. And again, there's different systems that are used for the ACC AHA. They describe it as the level or quality of evidence, and they take into consideration the study design, the number of studies, the consistency, and then they give it levels such as A, B, R, depending on whether or not they were randomized control trials. For the GRADE, they use the terminology of certainty or quality of evidence, and they take into consideration things like risk of bias, inconsistency, and the results, imprecision, indirectness, publication bias, and so forth. And then they grade the strength of the evidence as high, moderate, low, and very low. This is what I think is one of the most important processes, which is to consider whether the guideline took into consideration all the outcomes that are important to patients. Because a lot of times we talk about seizure frequency and seizure severity, but those might not be as important as certain other things to our patients, like health-related quality of life or adverse events. Are they going to be able to drive mortality, that sort of thing. So when you do a guideline, or when you're evaluating a guideline, you want all important outcomes to be specified. And if there isn't evidence on that particular outcome, then that should be stated. So getting back to our case, based on the guideline, the physician counsels the patient and his partner that the likelihood of seizure recurrence in an adult with a first-time unprovoked seizure is about 40% in the first year or two. But that his recurrence risk is likely going to be higher as his seizure occurred out of sleep. And I think I dropped the line here. Of course, she orders an EEG, but there's none in town. So, you know, they're gonna have to wait a couple weeks before his partner can get off work and drive him for the EEG. So in the meantime, she discusses starting treatment with an anti-seizure medication. She states that if he starts a medication, this would significantly decrease his recurrence risk within the next two years, but isn't going to improve the long-term prognosis for seizure remission. So the patient doesn't want to have any more seizures, but he's not so sure about whether or not he wants to take a daily medication. He's concerned about side effects and whether or not they would interfere with his ability to work. So again, using the guideline, she counsels that the adverse events associated with the older anti-seizure medications can range from 7% to 31%. That was level B, but they're usually mild and reversible. She isn't as familiar with the newer anti-seizure medications, but she says she'll do some research and they decide on a virtual appointment the next day. So the next step in appraising a clinical guideline is to determine whether the panel made appropriate judgments in the interpretation of the evidence and the decision of the final recommendation. So this includes aspects like, did the panel appropriately consider the magnitude of effect and the relative importance of the outcomes? And this is really tough. Balancing benefits and harms requires inferences on the part of the panel trying to determine the preference that the patients would have when all outcomes were considered. And these inferences can be based on a number of sources, including studies that address the relative importance that patients place on outcome, but there's not a lot of these in the literature. There's also consultation with patients and patient organizations, and then opinion of panel members, and hopefully the panel will include patients. And then the guideline developers have to specify the relative importance of the outcomes before they gather the evidence and then again when the evidence summaries are complete. So the next is whether the panel considered all relevant factors for formulating their recommendations. And this is very important because if you're writing a guideline, you want people to be able to follow it. So you have to take into consideration things like whether or not resource-intense interventions are available to the community in question. I can give you just an example of that. I do some remote EEG reading, and I read a lot of pediatric, I should say, neonatal hypothermia patients. And we had one institution that ordered just a routine EEG for a patient that was undergoing hypothermia. And so of course I call her up and I say, have you considered doing a longer study? And she was like, we only have one machine. So if we hook that baby up, we're not going to be able to monitor anybody else. So you always have to take those limitations into consideration. And then additionally, you want to think about whether the interventions are going to be acceptable to the patient. So is this something that patients are going to be willing to receive or that they're going to be able to adhere to? And then what is the feasibility? So again, this is both on a personal and a systems level. And now conflicts of interest. This has already been spoken about. But you definitely want to take in whether the panel had any conflicts of interest and whether you think they might have influenced their judgments. So anytime you develop a clinical practice guideline, you're going to make judgments. And this includes the magnitude of the effects, the quality and certainty of the evidence, and then again patients' values and preferences. And any of these judgments are potentially vulnerable to conflicts of interest. So as was discussed earlier, the most common conflicts of interest that can affect guidelines are financial and intellectual interests of the authors. And the extent to which these conflicts of interest can influence panel members, it's not like a yes or no sort of thing. So this is really a continuum and it can range from really no important influence to completely biasing. So guidelines have to judge the extent to which conflicts of interest might have influenced panel members' judgments. And trustworthy guidelines are ones that report the conflicts of interest of each guideline panel member and of the organizers who may have accepted industry funds for the guideline. And this should be done explicitly and transparently. And then the guideline should also describe in detail the mitigation strategies that they used to decrease the risk of bias in their guideline. And they should provide easy access to the organization's policies. So the next step is to determine whether the recommendations apply to a specific patient. So obviously recommendations from a guideline can be directed to patients who are different from those that a particular clinician encounters. So the providers have to judge to what extent the characteristics of the population for whom the recommendations were formulated are similar enough to those of the patient that they are treating. So getting back to our case. Overnight the physician consults the AES-AAN 2018 guideline on the efficacy and tolerability of the newer anti-seizure medications for new onset epilepsy. She feels that the guideline is appropriate because the author's disclosures are reasonable and appropriate steps were taken to mitigate conflicts of interest. The evidence was summarized with rigorous systematic review. The guideline panel considered many important outcomes, although she didn't see any patients among the panel members. But the recommendations were clear and actionable. And the panel's judgments in the interpretation of the evidence and the decision of the final recommendation seem appropriate. And she felt that the recommendations appeared to apply to her patient. So the physician calls her patient the next day. The patient reports that he and his partner had discussed the information that had been provided at the first visit, and they felt that the risk of adverse events were significantly less than the risk of another seizure. So they decided that they do want to start treatment, and they're eager to hear the physician's recommendations regarding treatment options. And that's it. The impacts on clinical care are that clinicians can and should evaluate the trustworthiness of the guidelines, and that there are several tools that can be found online that can help walk you through this process. Thanks.

clinical practice guidelines

guideline appraisal

patient care

seizure case study

evaluation tools