Welcome to the Augs Fellows webinar series. I'm Dr. Nani Moss, moderator for today's webinar. Today's webinar is Mind the Quality Chasm, Implementing Quality Improvement Methods in Practice, presented by Dr. Aaron Metzold. Dr. Metzold will present for 45 minutes, the last 15 minutes of the webinar will be dedicated to Q&A. Dr. Metzold is a third-year fellow at the University of Iowa. She completed medical school at University of North Dakota and Obegon Residency at Tri Health in Cincinnati, Ohio. She then completed a one-year Quality Improvement Fellowship at Tri Health following residency. During her fellowship, Dr. Metzold completed the Intermediate Improvement Science Series at the James M. Anderson Center for Health Systems Excellence in Cincinnati, Ohio. She has attended the ACOG Quality and Safety for Leaders in Women's Health Care Conference, as well as the Safety Project Collaborative. Dr. Metzold has worked on the Obegon Quality Assurance Committee at Tri Health, and is a current member of the Gynecology Quality Committee at the University of Iowa. Before we begin, I'd like to review some housekeeping items. This webinar is being recorded and live streamed. Please use the Q&A feature on the Zoom webinar to ask Dr. Metzold questions, and use the chat feature if you have any tech issues. OGG's staff will be monitoring the chat and can assist. Take it away, Dr. Metzold. All right. Thank you, Dr. Maas, and thank you to OGG for this opportunity to talk about quality improvement. I'm really excited to talk about implementing quality improvement methods into practice with everyone today. So I have no disclosures. Here are my objectives for today's talk. My overall goal for this discussion is to help everyone understand more about quality improvement methods, and how to implement these into your future practice, as well as fellows' quality improvement projects. I will be touching on the SQUIRE guidelines as well for publishing QI research, so that you're all able to share these findings. Before we dive into how to do quality improvement, I think it is important to review why quality improvement has become a priority in healthcare. So the Institute of Medicine issued several reports in the late 90s to early 2000s about the quality of medicine in the United States. The first was in 1999 with their report to AIR as Human, which highlighted the issue of medical errors in healthcare. And in their follow-up report, Crossing the Quality Chasm, they provided guidance to health systems and how to improve care. So they discussed that when faced with such rapid changes, the nation's healthcare system has fallen short in its ability to translate knowledge into practice, and to apply new technology safely and appropriately. So the major focus of QI is really how to apply our knowledge to our practice consistently and safely. So part of the issue with healthcare is that there is a significant lag between when we find new information on how to best care for our patients, and when this is implemented into practice. So studies suggest that there is a 17-year lag from discovering new research findings to implementing these changes into practice. Another study reported an issue with consistently applying best practices, as they found only 55% of Americans with common health problems received recommended care. And then also, health processes show an average error rate in these healthcare organization processes of about 1 in 10. So I really see quality improvement methods as a way that we can take evidence-based medicine, best practice guidelines, and process improvement measures, and apply quality improvement to apply these to our day-to-day practice in a way that is not only effective, but also sustainable in our clinical setting. So in the Crossing the Quality Chasm report, the Institute of Medicine highlighted six aims that healthcare systems can focus on their improvement efforts on to better meet patients' needs. So these include healthcare that is safe, so avoiding injuries to patients as a result of their care, effective, which is providing care based off of scientific knowledge, so making sure that we are providing evidence-based care, patient-centered, so focusing our efforts and our outcomes on patient preferences, needs, and values, efficient, so avoiding waste, timely, or reducing wait times and delays, and then finally, equitable, which Dr. Brown gave an excellent talk regarding disparities in neurogynecology on the Fellows Lecture Series earlier this year, and I encourage everyone to check out that talk as well. So when starting to think about a QI project, or even while planning a project, it can be very helpful to consider how you are addressing these aims. So before we get started on how to develop a quality improvement project, I wanted to mention the OGS quality improvement work, which you can learn more about on their website. So they list basic quality measures for reporting, which you can see here, and you can see that they have aims, which are addressed by each quality measure, and so you can see here patient safety, and then also effective clinical care. So now we will be getting into how to start a quality improvement project. So as a disclaimer, I tried to highlight some of the quality improvement methods and tools that I found to be most helpful, but there are many other tools and approaches to doing quality improvement, and I will have some resources at the end of our talk for if you want to learn more. So there are different models for how to go about planning an improvement project, and I will just mention two of the most common approaches, which are the model for improvement and Six Sigma. So the model for improvement asks three main questions to guide improvement efforts. So the first is, what are we trying to accomplish? So this question focuses on setting an aim for the project, which we'll talk about later in this talk, and then the second step is, how will we know that our change is an improvement? So this focuses on establishing measures that are recorded over time to show that our system is improving as a result of our interventions. And finally, what change can we make that will result in an improvement? So I'm sure some of you may or may not have heard about plan-do-study-act cycles, or PDSAs, and I know when I first learned of these, I really struggled with the purpose and how to actually implement these in practice. So we'll talk more about these, but basically, PDSA cycles are really an adaptation of the scientific method, so testing out possible interventions on a small scale with the purpose of guiding learning about the system and intervention so that the project can be successful. And so the other model for quality improvement is Six Sigma. So I think both models are very similar, and it is just important that improvement projects consider these fundamental ideas as they are building their projects. And so Six Sigma follows the acronym DMAIC, which the D stands for Define the Project's Purpose and Scope. Then the M is for Measuring Information on the Current System. So it's to analyze the deep causes of issues within the system and confirm these with data. So what is causing your system to not have the results that you need, and basically having data to show this. Then improve with developing, testing, and implementing solutions that address the deep causes for issues in the system. And finally, control, which basically means making sure that the project is sustainable. So of these two methods, I have primarily used the Model for Improvement. So I'll be focusing on the Model for Improvement throughout the rest of the talk tonight. So a helpful guide as you're starting a quality improvement project is reviewing the SQUIR Guidelines. So SQUIR stands for Standards for Quality Improvement Reporting Excellence, and this provides a framework for publishing quality improvement work, but it's also a great resource for planning a project. So the website has explanations for each component of the guidelines with examples in this SQUIR 2.0 E&E section, and there is also a list of references. So you can review published quality improvement studies. Also, ABOG just approved quality improvement projects for fellow thesis projects, as long as they do follow these SQUIR Guidelines. So I will be highlighting components of the SQUIR Guidelines as we talk about the different quality improvement tools. So many of you may be thinking, why do we need to use methods for quality improvement rather than traditional research methodology? So quality improvement really takes into account the context and processes that were undertaken to improve a system. So this statement was from a paper detailing a large successful initiative in Michigan to decrease rates of central line bloodstream infections. So they state that expanding successful iterations of programs without understanding the social processes and mechanisms that produce the outcomes means that implementers of the programs will not know what they need to do to make the program work and where to focus their efforts. So when a quality improvement project fails to produce the expected result, we have to wonder, why did this intervention not work? So either the intervention did not work in the first place, or the intervention in the new context does not properly replicate the original intervention. So something may need to change with the intervention or the system in order to get it to work. Or the contextual effects of the system mean that the intervention cannot succeed. So basically, the new system that we're trying to implement an intervention in is set up in such a way that the intervention can't work. Or there could be some combination of these. So this makes me think about when we try to take results from research studies or other quality improvement initiatives and try to implement them in our day-to-day practice. We typically know what population the initial interventions were applied to, but not necessarily what changes had occurred to the system to ensure that the intervention was successful. And we know that research is typically done in a very controlled setting, whereas day-to-day practice can be a little less predictable. So quality improvement methods can help better define the context and changes in the system that made the intervention successful. So getting started with a QI project, what are we trying to accomplish? So we need to consider some basic concepts while initially developing our QI project. So does the project have impact to patients, providers, and the healthcare organization? So if you can make a project be in line with some organizational goals or initiatives, you will likely be able to get support for your project from the organization, or the organization may already have methods in place for tracking some useful outcomes that may be useful to you. Then is there evidence-based medicine or best practice statements to support your QI efforts? So if you are early in the process of planning a project, you may consider looking at OGS, ACOG, or AUA position statements, as well as other journal articles to see if there are practices that your patients may benefit from implementing or standardizing a process regarding these. So you also want to make sure that your project will have clear starting and end points so we know when the project is successful. And last, we want to make sure we find measures that have short cycle times so we can easily do small tests of change and quickly assess our interventions for being successful. So as we go through the different QI methods and tools, I will be using a made-up example of decreasing postoperative infections so that you can see how these methods can be applied. Published QI studies do not necessarily use all of these tools, but I think each can be helpful in their own way. So with our initial question of what are we trying to accomplish, so we need to consider what is our aim? The aim of a project should be specific, so again, what are we trying to accomplish, measurable, so have outcomes that can be measured in our system, and then also be actionable, so able to be acted upon, relevant, so have an impact like we mentioned on the last slide, and finally time bound. So make sure that we have a goal for when this will all get accomplished. So an example of how to phrase our aim statement is that we will increase or decrease our outcome in our specific population from our baseline value to our goal value by a certain time frame. And this falls under the specific aim category of our SQUIRE guidelines. So the aim statement differs from traditional research in that the goal of traditional research typically is hypothesis-driven, whereas the goal of quality improvement is really improving patient care or some other aspect of the system. So in research, you'll have a hypothesis, and in quality improvement, you will have an aim. So the hypothesis for a study regarding post-operative infections may be patients undergoing hysterectomy will have fewer post-operative infections with Cefazolin and Mitronidazole versus Cefazolin alone. But a quality improvement project aim would be we aim to decrease the percent of post-operative infections in patients undergoing hysterectomy from 8% to 2% by December of 2021. So research is testing and intervention, while QI is really obtaining an outcome through multiple interventions. So in our hypothetical QI project, we may have reviewed baseline data and found an infection rate in our population of about 8%, and we want to decrease this down to an infection rate of around 2%. So when you start your QI project, you want to collect both quantitative and qualitative data. So quantitative data, so numbers. So this will include your baseline data, like the infection rate of 8% that we talked about on the last slide, as well as doing a deep dive into other processes that are in place. So how often do the correct processes occur, and how often do failures occur in the system? So in our example, we would look at patients who had infections and see if there were possible causes or failures in the system that we could attribute to their infection. And then qualitative data would include feedback from patients and staff, which can be hugely beneficial in understanding which interventions are going to be helpful, as well as using process maps, which can show the flow of a patient or other components through the system, and then a cause-effect analysis that can show how different factors may play a role in our outcome. So mapping out what the current system looks like can be useful in both the available knowledge and the context portions of the SWIR guidelines. So a study by Enger et al. for evaluating the quality of urinary incontinence and prolapse treatment projects convened an expert panel of nine physicians to identify quality indicators for addressing screening, diagnosis, and management of prolapse and urinary incontinence. So they came up with 14 quality indicators for prolapse and 27 quality indicators for incontinence. They then reviewed their system retrospectively by looking at how often these quality indicators were documented in their institutions. So I won't go into all of the quality indicators, but as an example, the AUA recommends that surgeons considering an invasive therapy for patients with stress urinary incontinence should assess a post-void residual urine volume. So this correlates with the quality indicator in the study, which was that women with stress incontinence who undergo surgery should undergo a pre-op post-void residual. So the expert panel had decided this was their basic level of care that their patients should receive. And based on the facility, so there were two facilities that were included, they found that about 50 to 83% of patients had a documented TBR prior to invasive stress incontinence procedure. So I think this is a nice example of identifying a quality indicator based off a best practice statement and showing what the current system looks like in regards to that indicator and a possible area for improvement. So another important step to understanding your system is identifying key stakeholders in the process that you are trying to impact. So these stakeholders are incredibly helpful in understanding issues in the current system, sometimes even issues that you don't realize yourself, like challenges that nursing or pharmacy may have with the current process that can even make new interventions more challenging. So they can also help, your stakeholders can also help with the implementation of the plan and problem solving when certain aspects of the intervention need to be adapted to work. So ideally meeting with the stakeholders throughout the project can be helpful to get their feedback as well as keep people invested in the project. So these stakeholders can be faculty or residents, nurses, either in your office, pre-op or PACU, you can include anesthesia, social work, pharmacy, and also your patient can be a great resource in order to get feedback on what the patients need and being patient-centered. So this would be included in the interventions portion of the SQUARE guidelines with kind of the intricacies of the team being involved. So an article in the Green Journal highlighted key stakeholders and possible roles for them in future improvement efforts on improving access to second line therapies for overactive bladder. So they highlight the potential roles for healthcare professionals, professional societies, industry and insurance companies that could impact future improvement efforts. So the last tool I'll be talking about with understanding our current system is developing a process map. So this can be a useful tool for understanding where failures in the system are occurring, where interventions can be implemented and which stakeholders may need to be involved in the intervention. So after talking with your stakeholders, you may find that your process map may have changed based off of learning more about your system. So in our post-op infection example, we may want to map the process prior to the patient having surgery. So if we want to implement a kind of infection prevention strategies, we may want to know that what order these things are done in so that we know where interventions can have an impact. So it may have a process map that looks like this. Surgery orders are placed. The patient sees maybe a nurse practitioner for your pre-op visit. Anesthesia calls the patient two days prior to surgery. The patient presents for surgery. And then finally, antibiotic orders are released to the pharmacy. And so this can provide information for the available knowledge and the context portions of the SQUIRE guidelines as well. This is an example of a more complex process map. In this study, they mapped out the OR instrument reprocessing from the OR to decontamination to assembly. So process maps can range from simple to complex, but they give good information on what the current system is like and how change in one area might impact areas around it. So next, we'll be looking at how do we know our change is an improvement. So we've talked previously about identifying where issues were occurring in a system and how frequently these were occurring in our quantitative data. So with the study of the instrument reprocessing for the OR, the authors evaluated when trade defects were occurring based off of their reprocessing stage. So they found out that most defects were occurring during assembly, and the most common defect was missing instruments. So after looking at your system and where issues are occurring, you can now narrow your initial intervention to the most common issues and hopefully make the greatest impact with the fewest number of changes. So with our example of postoperative infections, we would look at patients who had infection and identified where issues were occurring in their care. This is an example of a Pareto chart, which shows where issues are occurring from the most common to the least common. So focusing QI efforts initially on the categories on the left will likely lead to greater change. Also, I know this is a very simplified example, and there are often other factors playing a role in our example of postoperative infections. So you may be thinking that patients who had infection may have uncontrolled diabetes or morbid obesity that played a role, and that can certainly be the case in things that you may look at in your population. And this would also lead you to possible different interventions. So it's very important to try to think about all of the different things that could be impacting your outcome. So in our hypothetical example, though, we found that 16 patients with infection had issues with the timing of antibiotics, and 12 had incorrect antibiotics given, and 12 had an incorrect skin prep. So those would be the areas we would want to maybe start our interventions on. Once you have identified these areas for improvement in your current system, you can look into starting to develop a key driver diagram. So this includes your AIM statement with your new key drivers. So we can adapt the issues from the previous slide to fit into key drivers, which are then going to be what drives change in our system. So making sure every patient has the appropriate antibiotic selection, making sure there's appropriate timing of the antibiotic, appropriate skin prep, and appropriate re-dose of antibiotics for longer procedures are going to be our key drivers. Then after we identify these drivers, we are going to develop interventions that are going to target these drivers. And this diagram can also be used for the interventions portion of the SQUIRE guidelines. So this is an example of a key driver diagram in the Green Journal that one of my mentors, Dr. Murcott, was involved in for decreasing the rate of premature birth in Ohio. So it nicely demonstrates what drivers contributed to the change, and then what interventions were used to accomplish them. So when planning a quality improvement project, it is important to identify what will be measured. So, your outcome of your outcome measure is typically what you described in your AIM statement. So, in our example, it would be cases with postoperative infection. Process measures look at your interventions and whether they are being done for each patient. So, in our example, this may be the percent of cases that had antibiotics administered in an appropriate timeframe. And for balancing measures, these are going to be consequences that may happen due to your interventions. For our example, we may have cases that were delayed for infection prevention measures. So, as another example, if you're doing a project on implementing changes that increase patients going home same day after surgery, you may track readmissions, or you may track phone calls or office visits as your balancing measures. And so, these would, of course, be under the Measures section of the SQUIRE Guidelines. Now we'll look into what change can we make that will result in improvement. So, we will first talk about developing an intervention, then talk about measuring these interventions with run-and-control charts. So, when thinking about interventions, you want to consider interventions that will be highly reliable. So, this means that we have a high number of actions achieving our intended results out of the total number of actions. So, different interventions are going to have different levels of reliability. So, things like training or memory aids are going to have a 10 to the minus 1 level of performance or 90% reliability. However, interventions that change the structure of the process so that you have to actually try to do the wrong thing are going to be more reliable. So, you can increase your reliability 99.9% of the time if you're really changing the structure of how things are done. So, as an example, this is the door leading to the back stairwell by our offices. So, if the door doesn't shut, an alarm will sound after about a minute of the door being open, at which time the person who came through the door is already long gone. So, you can see that there are numerous signs or memory aids indicating that you need to shut the door. But this door, unfortunately, still alarms several times a week. So, both the alarm and the signs are not highly reliable interventions to make sure that the door is shut. A highly reliable intervention would be replacing the door with one that automatically shuts. So, with our key driver diagram, an intervention like reviewing antibiotics at a pre-op conference would likely be less reliable because you still have to actively do something to make sure the correct antibiotics are placed. However, an intervention with an order set that automatically selects the appropriate antibiotic for the case would be more reliable. However, you do have to use caution when changing the system so that it is very selective to your process and won't have unintended consequences on other systems. So, once you have identified your interventions, then you can start making small tests of change using the Plan, Do, Study, Act cycles or PDSAs. So, these help to develop and refine your interventions before implementing them on a larger scale. So, this is where we apply the scientific method by planning an intervention and forming a hypothesis on what will happen and then comparing those results. Once you have planned the who, what, where, and when of an intervention, you carry out the plan and document your observations. Then you analyze the results. You may find that there might be, you know, an issue with pharmacy for one of the antibiotics and that that is why it has been missing or not delivered on time. So, then after you do your PDSA, you either adopt, adapt, or abandon. And so, typically, you will adapt. So, you'll learn something and based off of that observation, you'll adapt the PDSA for the next time. Once you find that it is working smoothly, you'll adopt it. But if it's not working at all, you may abandon it. So, again, most PDSAs will end with adapting the plan, adapting the change, and implementing it again. This is an example of a PDSA worksheet that I found on AHRQ.gov. So, it's easy to kind of forget all of the different iterations of things that we did. Sometimes we may even be doing small PDSAs without even realizing it. So, these PDSA worksheets can help you to keep track of different small tests of change that you did so that you can also describe these with sufficient detail in the intervention portion for the SQUIRE guidelines. How this looks in practice is that you may try the system, your new intervention, for one patient. You would then talk with your stakeholders involved, look at what the result was, get feedback on how your intervention worked. And based off of what you learned, you'll choose to adapt, adopt, or abandon. And then you'll try with another patient. So, this works best if you are actively involved in your patient's care. So, you're on your event service. You can try this out with different patients and see how your new intervention or process works. So, after trying it with another patient, you may say that you're doing pretty well. And then you adapt, adopt, or abandon. So, you're adopting it and using it with three patients. And then you may increase using it on all urogyne patients. And then you may even transition to using it to all gyne patients. And you may learn different things at different stages of implementing the intervention. And so, the intended output of the PDSA is really learning how the system will react to the intervention, the intervention to the system, and having informed action for further changes. So, as an example, you may decide to change your antibiotic order set to make sure that correct antibiotics are being selected. But then you find that the person putting in the orders is actually using an old order set. Then you have to go back and make sure that either the old order set is removed or making it easier to use the correct order set. So, we will finish up by talking about the model for improvement with how we measure our intervention, which in quality improvement typically means a run chart or a control chart. So, if you're wanting to publish a quality improvement project, typically they do recommend having some sort of run chart to demonstrate your data. So, again, the main measure of quality improvement initiatives is the run chart. So, the run chart displays your quality indicator on the Y-axis plotted in some chronological order on the X-axis. So, you may see different units of measure. So, it might be weeks, months. You may even have, like, days in between for rare events. The center line is your median value. And then you also have a goal line, which will show what you're trying to get to. And this can be for your outcome measure. It can also be for your process measures, typically for a balancing measure. You don't want to have any runs or shifts or trends, which I'll talk about on the upcoming slides. You kind of want that to stay level as you're doing your quality improvement work or if it changes, you know, trying to figure out the cause for that change. And so, you also want to be making sure that on these run charts, you are annotating them with important steps in your quality improvement process. So, this would be included in your results portion of the SQUIRE guidelines. So, with traditional research, we typically see our results demonstrated like this with a bar chart. So, you might have your average before your change and your average after the change. And you can see that the intervention was statistically significant. However, we can also show this data in a run chart. So, in this example, you can see that you have a baseline of around 70 here before the change and around 30 afterwards. So, if we add some median lines to this, it would look like this here. And so, basically, it looks like the change caused the intervention that you implemented causes a change in the system. However, in this example, it has the same data, the same averages, but if you see the data chronologically, it appears that something was already going on to decrease the outcome in our system and the change is unlikely due to the intervention. The last example has the same average before and after. However, you can see that there is a big decrease after the change was implemented, but there may be an issue with sustainability as the numbers at the end are now increasing closer back to the baseline before the change. When you're analyzing a run chart, they do have certain probability-based rules to objectively analyze the run chart data for non-random patterns, and these are based off an alpha error of P less than 0.05. So, the first non-random pattern is a shift. So, this is going to be six or more consecutive points, either all above or all below the median line. So, you can see in this example two shifts occurring in the data. The second rule is a trend. So, this is where you have five or more consecutive points, all going up or all going down. And again, here you can see two shifts occurring in this run chart. And here's an example of making a run chart in Excel, which is actually pretty easy. Here you have the date and followed by the numerator and the denominator. So, the numerator is those meeting your outcome, and the denominator is the total patients in your population being studied. So, from this, you can calculate the percent and then determine a median value. You can then create a line graph with your outcome and your median value. So, this would be an example of our baseline data for postoperative infections with a median of about 8%. As we prospectively collect data, we would continue adding the data with keeping the median from our baseline data. So, you don't want to be changing. So, initially, as you're prospectively collecting data, you don't want to change your median line until you start noticing a non-random pattern. So, with this example, we're starting our intervention at week 14 and obviously having great results and noticing that a shift occurred. So, we have more than 8 points below the median line. So, this is a non-random pattern that we're noticing with this example data. Once you notice that a shift occurred and you can actually change the median to reflect that new data. So, in this example, our new data has a median value of about 2.6%. This is an example of run charts from an article in the Green Journal looking at process measures for enhanced recovery. They studied early ambulation, morphine equivalents, early nutrition, and whether they met their multimodal analgesia goals. The center line in each of these shows when the intervention occurred. So, I added some median lines so you can see that shift in data. We can also see that prior to the intervention, most of the data started going in the correct direction. In these examples, I'm guessing that they're probably with some sort of education or started trying some tests of change that maybe started moving that data in the correct direction. And so, this would be, again, part of your result section of the SQUIRE guidelines. And then next is the control chart. This is very similar to our run chart, but uses statistical process control to give us additional details regarding the data. The control chart has the primary objective of distinguishing between common cause variation, which is change that is just due to chance, and special cause variation, which can be assigned to an intervention or change in the system using statistical process control. Like we use statistics, our traditional statistic methods in traditional research, this also shows that, you know, our change is not due to chance. So, if we find something that applies to the special cause variation rules, which I'll talk about on an upcoming slide, then we can say that our, that change is not due to chance. This chart shows, so this chart shows a run chart, which appears to have no trends or shifts in the data. So, we don't see any non-random changes in the system. Here is the same data in a control chart. So, a control chart has a mean line rather than a medium line, and you have control limits that are set at plus or minus three standard deviations. So, now you can see that there are actually several points that are outside of our control limits. And so, it is likely that these points were not due to chance, and it's important to note that there are different types of control charts based off the type of data, and they can vary based on whether the data is continuous or attributable or categorical, as well as other factors within the data. So, choosing, you have to choose the correct control chart based off the type of data that you have. These are the rules for identifying special cause variations by making causes that are not due to chance. So, the main rules are having one point outside the upper or lower control limit, also having eight successive points on the same side of the center line or six successive points increasing or decreasing. There are also some other rules based off of how many points or two or one standard deviation from the mean, but these are some good guidelines for figuring out if there is any changes that are not due to chance. So, this is our run chart from our previous example showing our non-random change in the data. This is the same data as a control chart. Again, you can keep the mean the same as you are getting your baseline data until you start to see special cause variation. Here is a point outside of our control limit, so when we first see that point, we'll think that the change is likely not due to chance, and then here are at least eight points on the same side of the center line indicating special cause variation again. Once we identify this change, then we can change our mean line as well as our control limit. So, we'll be able to narrow our control limits, and that also shows that there is less variation within our system. With control charts, you typically have to have some sort of template or software that will help make those, as those are a little bit more challenging to make on your own. Lastly, what about IRB? So, do you need an IRB approval? The answer to that is really it depends, and this really depends on whether your IRB considers quality improvement, human subjects research or not. It's not necessarily on whether or not you're planning to publish or not, so the important thing to note is that not all IRBs are the same in regards to whether or not they require approval. Some have checklists or written guidelines on which projects will need approval, so the best option is to talk with your IRB representative and see what they recommend early on in your process. And here are some resources that I have kind of gone back to again and again. The James M. Anderson Center for Health Systems Excellence does have several webinars, especially there's one on publishing QI that's really excellent. There's also the Institute for Healthcare Improvement that has a lot of modules that are online, as well as different webinars, and the Open School for Quality Improvement and Patient Safety, which can be very helpful for learning about different ways to approach quality improvement. And I also really like the BMJ Quality and Safety Journal, so they publish a lot of quality improvement studies, so if you're looking for a place to submit quality improvement work. But they also have a lot of articles on quality improvement techniques, as well as examples on people who have published in quality improvement. So in summary, quality improvement methods are important for implementing evidence-based medicine into practice. You can use the Model for Improvement and Six Sigma to help with designing and implementing QI interventions. You want to make sure that you design and aim for your QI efforts with a clear goal and timeframe, and utilize your PDSA cycle to really learn about the intervention and adapt it so that it best fits into your system. And then run and control charts can be used to show the data before and after intervention, and have that added benefit of showing how the system is over time. And so here are my references. Thank you, and I welcome any of your questions. That was a wonderful presentation, Dr. Metzl. Thank you so much for demystifying QI projects. Like you said, we have 15 minutes for questions, so you can submit your questions to the Q&A section. So far, okay, none are up. I have a question. So how do you determine appropriate, attainable, and clinically significant measures for your interventions? That is an excellent question. So with our example of post-op infections, you may see different misclipped databases or studies that have looked at post-op infections around the country. And so some studies have shown an infection rate of about 1.8%, so that's why I picked that 2% number for our example, and it depends also on the type of study that you're doing. So if you have a quality improvement project of like saying you want to increase the number of patients who are going home same day after surgery, you may look at all of your patients that have been having surgery, and you may say like 20% are going home now, but looking at all the patients who stayed overnight, about an additional 50% or so also fit our criteria for going home same day. And so then you may say that we want to increase our proportion of patients who are going home same day from 20% to 50% based off of that information. And with other quality improvement projects, it may be that you are wanting to improve patient satisfaction, and so you may say that I want 90% of my patients satisfied with this intervention, and so you may kind of arbitrarily pick 90%. So I think it ultimately depends on the type of project that you're doing, but you can use either national standards, you can use your existing data, or you can use kind of if you're implementing a strategy that is kind of what you feel is the basic level of care that you should be providing, you may say 90 to even 90 or greater than 90%. So it all depends. It all depends. It really does. We do have a question from Dr. Crisp. What do you typically find, or sorry, what do you typically find to be the most difficult step in starting a QI project for people who are new to following QI guidelines? Oh, great question, Dr. Crisp. So I think it is oftentimes getting your stakeholders involved. So when you're starting a QI project, you want to get people who are involved in each stage of the intervention involved in the project, and even sometimes you may try to get kind of the naysayers to be involved. So people who you know may be a little bit reticent to implementing a change because then you are going to be better able to apply that intervention if you can get people on board who may have very valid reasons for having kind of some reticence to implementing a new change. So I think that initial step of getting others involved in the change, it can be the most challenging part, but once you are able to get people more involved, I think they ultimately appreciate that and also having a voice in that, and so then I think it ultimately makes your quality improvement project smoother if you have people involved and feel like their voices are heard and that you are able to implement your project in a way that you're taking into account all of those different issues that they bring up as well. So most of the QI projects that I've participated in have been pretty small and very local to the institution. Do you think it's possible or even necessary to broaden, you know, QI interventions and projects to other institutions or other, you know, practices? Yeah, I definitely think that that is very important because I think we all kind of operate in our own little system and we may find that something works very well with improving the quality of care that our patients receive, and I think that's why it's also very important to make sure that we are good about publishing our quality improvement projects. So that we can share with people not only what interventions works, but how they've worked in our system. So if people have a similar system, they kind of have a framework for getting that implemented, and then also if they have a different system, being able to identify where those challenges might be as they're trying to maybe implement similar interventions. So that's a great question. I think, you know, it's very important to be able to share our quality improvement work and so that we can kind of broaden that scope. The initial, one of the first examples that I was talking about with like decreasing bloodstream infections with central venous catheters, you know, they really tried to broaden that and make it, they did a great job at implementing it in multiple systems. And so being able to have successful QI projects like that as examples is also really great. Fantastic. Another moment for any additional questions. How about with that? On behalf of AUGS, I'd love to thank Dr. Metzl and everybody for joining us today. Our next fellows webinar will be held on August 24th at 8pm Eastern time. You can visit the website for additional information and to sign up. All right. Thanks, everybody. Have a great evening. Great job.

quality improvement methods

healthcare

model for improvement

Six Sigma

key driver diagrams

run charts

control charts

IRB approval

Augs Fellows webinar series