Research Week 2 Part 5.docx

Here are all of the patients that participated in this study. And as you can see, they are stratified according to age, their military duty status, the gender, the race, and of course, the BMIs. And so, this constitutes our database. Each of these individuals gave an informed consent. Each of these...

Here are all of the patients that participated in this study. And as you can see, they are stratified according to age, their military duty status, the gender, the race, and of course, the BMIs. And so, this constitutes our database. Each of these individuals gave an informed consent. Each of these individuals received documentation from the institutional review board, so they knew what they were getting into, and they decided that they wanted to participate. And so, if you count these up, these are 12 individuals. And again, six went to the Trindelenburg group, and the other six went to the Supine group, all randomly assigned there. The question then is, is this a control group or is this an experimental group? Well, it turns out that, as I said, six of these individuals went to one group, the other six went to the other group. So, the reality is that this table that you see here really consists of both groups together before they are assigned. Is this a prospective study or is this a retrospective study? Well, this is something we really haven't talked about, so let's talk about it now. In a prospective study, you gather data as you go along with the study. As the study proceeds, you gather data. Well, that's what we're doing here. We are collecting these patients together, assigning these patients, having them go through the experiment, and we are collecting the data. Since we're collecting the data as we go in real time with patients or subjects or participants, we can say that this is a prospective study. What is a retrospective study? Well, retro, of course, means back. This is the type of study that happens when we go back and we look at information that's already there, for example, like in a patient's chart. If we had done a retrospective study, that means we would have gone into the chart of these 12 patients, and we would have collected information, maybe on their blood pressure, the kinds of medications that they are using, the types of physiologic responses that they typically have during early morning rounds. That type of data, it wouldn't have been in real time. We would have been going back, and we would have been going into a record in order to extract the information. So if we were going into some sort of existing record and extracting information, by definition, that is a retrospective study. Finally, is this a single-blind or a double-blind study? Well, in this particular case, the experimenters did not know who was being assigned to which group. And in addition to that, the patients did not know who was being assigned to each group. As a matter of fact, while the patients were undergoing the experimentation, they simply watched television while they were in the midst of the experimental procedure. So as far as they were concerned, they were not privy to whether they were participating as supine individuals or as Trendelenburg individuals. As far as the methods are concerned, once again, this is drawn from the information that you read. There were lots of cardiopulmonary variables that were measured. Among these were static compliance, airway resistance, chromatoxic elimination, alveolar metabolic volume, crate output, and several others. Specific inclusion criteria were met. For example, these subjects had to be clinically stable on the day of the experiment. They had to be nutritionally repleted. They had to be male. They had to be tetraplegic. Patients were deemed ready, willing, and able to undertake a two-hour spontaneous breathing trial. Now, this doesn't mean that every single one of them was able to perform a two-hour spontaneous breathing trial. It simply means that they had the potential to make it to the two-hour mark and were deemed ready to do so. The statistical analyses that were used in the study that you read about included the t-test, which of course we've made reference to several times. We've also used a repeated measures analysis of variance, which is a special type of ANOVA statistical analysis. We also used multiple linear regression analyses. We're not going to concern ourselves with all of these. We're simply going to focus on the t-test because, as I said before, this is not a statistics course. We just want to introduce some very, very basic concepts, and so we're going to stick with the t-test as our preferred test just to see how things work at a very basic level. Now, we've talked a lot about different things. Let's turn our attention momentarily to the NM3 monitoring, which you'll be using in the SCI Center. And what I'd like to show you here is just one basic element of the volumetric capnography that we use. In volumetric capnography, we're able to generate a volumetric capnogram as it's shown in the lower right panel. You can see that it's the orange waveform, and that orange waveform is something that we're going to have much more to say about later on. But for right now, what I'd like you to realize is that that orange waveform is generated by taking an N-tidal CO2 signal as shown in the yellow tracing and then combining that mathematically with an airflow signal as shown in the red tracing. When you put those two together mathematically, what you end up with is that volumetric capnogram on the right shown in orange. And that is the fundamental function of the NM3 device that enables you then to measure all of those other variables we've been talking about. This is how those variables are measured. So when we talk about alveolar metabolic and cardiac output and the other variables that we've talked about, it all starts out with taking an N-tidal CO2 signal, combining it with an airflow signal, and coming up with a volumetric capnogram. Just a moment ago, we talked about basic experimental designs. Here's a similar design. However, what we're doing is we're going one step further. We're making this design just a little bit more complex, and in doing so, we're going to increase the power, the predictive power of the study. And here's how we do that. Let's start once again by looking at the experimental group and the control group. And as you can see, we have, once again, six individuals that are going to be randomly assigned to the experimental group to receive tryptelemberg. We have six that will be assigned to the control group to receive supine. And so they will receive the chest optimization protocol, as we've talked about. However, those individuals in the first group, in the experimental group, who first go through experimental treatment, will then be allowed to cross over. And a few days after they have been exposed to the tryptelemberg position, they will be now exposed to the supine position. On the other hand, those individuals that began the study by being in the supine position, a few days later, will be crossed over and placed in the tryptelemberg position. In essence, what we're doing is we're taking the six individuals that began in tryptelemberg, and they're being exposed to both types of interventions. Those that began in the supine position also are being exposed to both experimental interventions. So in essence, what this means is that at the end of the run, at the end of the trial, we will have not six individuals that have gone through tryptelemberg, but in fact, 12 individuals that have gone through tryptelemberg. And we will also have 12 individuals that will have gone through the supine position. What this means is that instead of having a group of 12 individuals, in essence, this is like having 24 individuals. Because we now have more patients participating in a study, we have a much stronger study that is much more predictive of the population at large. A word about experimental bias. We've talked a little bit about this, but I'd like to show you a photograph here that I think really shows what the problem is. This is a photograph taken during the actual acquisition of research data. And I want you to realize that this shows that experimental bias can be present even when you think that it can't be. In this case, you can see that there's a little table here in the patient's room, and the patient is on the other side of the translucent screen. There's a respiratory therapist back there with an envelope. She opens the envelope, which tells her what position to put the patient in. And while all of that's happening, there's another respiratory therapist standing in front of the NM3 machine, chatting verbally with the respiratory therapist on the other side. And when she says he's ready, since the patient has already been hooked up previously, this respiratory therapist standing on this side of the screen begins to collect the data. Unfortunately, when this first happened, the patient's position was projected onto this translucent screen, so the respiratory therapist then was able to see whether the individual was in trendomper or in supine. Needless to say, it was important to go from a translucent screen shown on the left to an opaque screen like the one shown on the right. And I'm simply showing that to you because in research, things don't go perfectly. They never go perfectly. And even after you feel like you know what you're going to do, there's still going to be a lot of adjusting and a lot of tweaking. And so you should know that as a researcher going in so that you're not necessarily surprised by that. All right. Well, the study is done. The data is recorded, and now it's time to analyze the data statistically. I'm not going to go through each and every one of the physiologic variables, but we will look at just a couple of them. These are changes in cardiac output before and after the delivery of the chest optimization protocol. And as you can see from this slide, there is a group that received trendomper. There's a group that received supine. And so if you look at the group that received trendomper and supine before chest optimization, they're virtually identical. There's really no difference whatsoever between those two groups, and we wouldn't expect there to be too much of a difference. That's a good thing. If they're widely different, well, then we have to begin to wonder why is it that they're so different? They shouldn't be that different. If we've randomized these patients, if we've carefully stratified these individuals, there should be virtually no difference between them at the beginning of the experiment. At the end of the experiment, after they have completed the chest optimization protocol, what we find is that the group that was placed in trendomper has a cardiac output that is significantly greater than the cardiac output of the group that was placed into supine. And if you think about it for a moment, of course, it makes perfect sense because if we place someone with their head down, we're going to augment the cardiac output. And we do that intentionally in the ICU all the time for those individuals that have decreases in blood pressure, for example. So we're able to see here that the cardiac output is, in fact, a statistically significant amount. Similarly, in this slide, we're looking at the alveolar minute volume. Again, this is alveolar minute volume measured with the NM3 device. And what we can see is that both groups are fairly similar before they receive chest optimization. But there's quite a divergence in terms of the value of alveolar minute volume for the trendomper versus the supine position with the trendomper being much greater. So needless to say, when we put someone in trendomper and they have a spinal cord injury and they receive chest optimization in that position, we can expect a rather sizable increase in their alveolar minute volume. And that's a good thing, particularly when you realize that by having an increase in your minute volume, you're probably going to be able to do very well when it comes time to come off the mechanical ventilator and do a spontaneous breathing trial. Well, the last two variables that I showed you were actually intermediate variables. This is the dependent variable, which of course means that we want to find out what happened with the spontaneous breathing trial duration. And you can see if you look at the purple column, this is the column that pertains to trendomper. The orange column pertains to supine. Those individuals that did their chest optimization in trendomper were able to do a spontaneous breathing trial that averaged about 88 minutes, plus or minus 38 minutes. So that's a good, long, solid time. And then those individuals that were placed in supine for their chest optimization were able to do a spontaneous breathing trial of about 34 minutes, plus or minus 30 minutes. Is this a statistically significant difference? Well, when we run a t-test, we find out that it is, because we have a probability value equal to 0.0001. Since this is less than or equal to 0.05, we know that this is a statistically significant difference. So at this point, we have reason to believe that by putting an individual in trendomper and doing the chest optimization in that position, it definitely has a statistically significant impact on the outcome that we have chosen, in this case, the spontaneous breathing trial. Well, now that we know this information, what can we do with it? Well, this information is from a randomized control trial. And this information now enables registered respiratory therapists to be able to design evidence-based protocols. We can modify existing protocols that will guide the neurorehabilitation and the weaning of ventilator-dependent patients with cervical spinal cord injury. So by doing this clinical research-- and it's clinical research because it's being done, of course, in the clinic, in the hospital, at the bedside-- by doing this clinical research, we have actually gathered data that we can now reinvest back into the rehabilitation program in order to improve how that program works. And by continually doing that, by continually following that process, we're able to make better and better and better refinements as time goes on. So that as we continue to receive patients, patients that come after this group that we have experimented with should be able to count on progressively better and better care, higher quality care. Well, now that you've read Article 2.6, we're going to go on to Unit 2.7, Educational Research in the Clinical Research Practical. And as a result of having read the article, I think you now realize that in addition to there being clinical research, there's also educational research. In short, there are lots of different kinds of research. But what we do at Hillsborough Community College is something that's very special because we have a commitment to scientific research and to training our students in how to do that research. And we also, as educators, perform research using our students as subjects. So we want to look at the role of a research practicum in helping sophomores to learn not only research knowledge, but also research skills. We would like to explain how an educational research study can actually help improve educational practice and improve learning outcomes for students. And we'd also like to look at the role of SPSS, which is a statistical package. How does this type of package help us to mine for data so that we can understand perspectives in educational practice? We recently published this article. It's the article that you read in Respiratory Therapy. And the importance of this article is that it's really the result of a natural experiment that occurred. We did not plan it this way. And we wanted you to read this article because we want you to see that sometimes, as I mentioned in the last unit, things happen in research that you simply can't foretell. You can't predict. And so you have to be ready, willing, and able to kind of bend with the circumstances. I'd like to discuss very briefly the history behind the Clinical Research Practicum. It was in 2017 that a partnership was established between the Department of Respiratory Therapy and the HCC Foundation. That year, we received a grant that funded our research and also helped us to purchase equipment with which our students would be able to go into the clinical environment and not only acquire research knowledge, but also acquire research skills. So we're thankful to the contributions from the HCC Foundation, also from the Phillips Medical Equipment Company, as well as the Smiths Medical Equipment Company. And of course, we're very grateful to the Spinal Cord Injury Center at the Tampa VA for allowing us to use their facilities to provide this knowledge and these skills for our sophomores. Well, the Clinical Research Practicum really has two major components, one of which is educational, the other one of which is clinical. From an educational standpoint, we are attempting to provide our students with clinical research knowledge and skills acquired in a two-day in-hospital Clinical Research Practicum. And then of course, they take a quiz that enables us to be able to measure how well they've been able to internalize concepts like you're doing right now didactically and concepts that they encounter in the clinic at the bedside. In addition to that, because we know so little about ALS patients when they're on a mechanical ventilator, our students play an exceedingly important role in using the NM3 monitor to be able to measure physiologic variables of patients that are on a mechanical ventilator and have ALS. Clinical educators are researchers also. Your professors are doing research, and very often they're doing research from the data that students provide through tests that they take or through procedures that they perform in a laboratory, for example. That educational data very often is placed into a statistical database as shown in number one. That allows us to be able to graphically demonstrate how there are changes that occur over time. This leads to the production of information that we can then use in class as shown in number two. In number three, we see an educator who is now imparting knowledge to students based on the information that was gleaned from the research. And finally, in number four, we see hopefully an improvement in the skills of the students based upon the research that we've done. So just like a clinical researcher would use data in order to improve the outcomes of patients, clinical educators use research in order to improve the learning outcomes of students. One important statistical package that you should know about is the use of the statistical package for the social sciences, so-called SPSS 27. Now every year it changes. Next year it will be SPSS 28, and then after that 29, and so forth. But what I'm showing you here is just a small part of SPSS. It happens to be the variable part of SPSS where we put in all of the variables that we want to measure. And there are other parts of this as well. This happens to be the data view. We just saw the variable view a minute ago where we put in variables. Well, the variables obviously are going to have numbers associated with them, and so you can see the numbers here. This relates to data that we have captured regarding our students and their performance over the long term, over the last few years, and it pertains, of course, to the article that you read before beginning this particular unit. And what is so helpful about the SPSS statistical package is that we can take that data and we can graph that data. The statistical package is a very sophisticated statistical package, and it can provide us with not only a graphical representation as you see here, where you can see the sophomore classes of 2018, 2019, and 2020. And this shows what we call the pre-prep quizzes for those classes that shows that there is a very tiny improvement, still virtually staying the same over time, and that in and of itself is important. Over to the side, you can see some descriptive statistics for those various classes. You can see the mean that they generated, and you can see the standard deviation that they generated as well. And remember, as we've said in the past, if we're going to do a t-test, we're going to need the mean of a group, and we're also going to need its standard deviation. The last thing we're also going to need is how many individuals were in that group. If we have those three pieces of information, then we're ready to do a t-test. Now, in this particular case, we did not do a t-test. We did an analysis of variance. And you'll recall from a previous conversation that we said that as long as you have two groups, you can do a t-test. The minute you have three groups or more, then you have to switch to an analysis of variance. We're not going to go into it much more than that, just to tell you that it exists, because when you do go into your statistics class, you will know a little bit more about the t-test, and you will be definitely doing more work with analysis of variance. And in this graph, what we see is what we call the post-CRP research skill scores for our last three sophomore classes for 2018, for 2019, and of course, you see two subgroups for 2020. We see an on-site group, and we see an online group. And when you look at this, you can see that there is a rather stark disparity between the on-site and the online group. We have analyzed this, and we're still in the process of analyzing this, to try to determine why there was such a disparity. This is a statistically significant difference. You can tell that by the p-value that's down at the right. It's a p-value of less than 0.001. And so, for some reason, the acquisition of skills by the online cohort, the cohort that was unable to go to the VA center, appeared to be statistically lower. So, we are analyzing this. We have made some changes in the educational program, and specifically in the clinical research practicum part of things, in an effort to try to address what we interpret to be some shortcomings. But this is what the article brings out, and so we are anxious to see if some of the changes that we have made will make a difference. In conclusion, this research showed us that our sophomore class that went through a clinical research practicum online did not acquire the skill level that we had hoped for. They certainly did not acquire the skill level that those individuals that went through the clinical research practicum on-site did. So, when we published our findings, we wanted to not only explain what we had found so that other programs might benefit, but what has also happened is that we've established a dialogue with other programs that have had similar issues. And they're learning from us, we're learning from them, and together we're trying to institute changes so that if we do have to go to a format where we have a clinical research practicum online again, this time we will have specific components in place that will help our students to attain the skill levels that they're going to need. All right, well, here we are at the end of Module 2, and just as before, we are going to now be taking the questions regarding the research that was done by ARI et al. This is in an effort to give you additional practice applying research to improving patient care. As you know, we've gone into the discussion section, we've had students answering the questions and providing give and take with their classmates, and that's an exceptional way to be able to learn. The best way to move forward through this article, once again, is to think of yourself as being a registered respiratory therapist in the future. Look at the scenario that's presented to you and try to see yourself using research in order to improve patient care. The objective is the same as before, to discuss the role that research articles play in improving patient care, to explain how understanding the rationale of a published study impacts its use in patient care, to explain how understanding the experimental design and how we do data analysis and how we use statistics in a published study impacts its use in patient care. The message that hopefully is coming through loud and clear here is it's not just about gathering the findings and that's it, we're done. We want to understand how those findings were actually collected. We want to understand the reason for collecting those findings, the environment in which those findings were collected, because that allows us then to determine if those findings may or may not be applicable to patients that we would like to help, patients that we are attempting to treat. You may read through this once again just to reiterate what we talked about last week, just to remind you where we are in the grand scheme of things, and then go on to this week's questions. As you know, week one is now closed. The questions for week two are as follows. Number one, to explain if this study was pre-approved by an IRB, an Institutional Review Board. What are some differences between the two types of nebulizers that are shown in this diagram to the left? You can see there a diagram that shows a test lung, it shows a series of filters, it shows a mechanical circuit, it shows what appears to be a Fischer-Picot heater, and of course over to the right, a mechanical ventilator. And situated strategically along the mechanical circuit are different types of nebulizers. We'd like you to tell us what those types are. Number three, explain the main difference between position one and position two in the ventilator circuit diagram at left for these two types of nebulizers. And as before, we ask you to go to the discussion section in Canvas where you will see these questions listed again. And beginning with number one, write your answers. And then please come back and respond to the answers of one of your classmates. Be sure that you make observations and/or suggestions that are intended to help your classmates learn about how to apply research to patient care.