Welcome back. This is section E of the lecture framing the question, and this section is entitled Some Examples of Framing the Question. I'm K Dickerson. So now I want to cover a few examples of how to frame a question, and we've been talking in sort of somewhat theoretical terms about the considerations you have. Now let's talk about how you actually do it. Remember that were going to go with PICO or P-I-C-O or P-E-C-O. And we're going to use that as a framework for asking our question. So here's the question that I'm interested in, is drug therapy associated with long term morbidity and mortality in older persons with moderate hypertension? So our P, our population or patients or people or older persons with moderate hypertension. Our intervention is drug therapy, our comparison isn't stated. And you'll often find, in the answerable clinical question, that the comparison group is not stated. So, I just wanted to tell you that ahead of time, but you will have to think about that and that's what I'm going to show you next. And then O is long term morbidity and mortality, those are the outcomes that we are interested in. Of course we have to say what morbidity and mortality are, as you know from previous sections. So here we have the question. So it's a little different order, it's not PICO but all the elements are there, I-O-P, or setting, and condition of interest. So the comparison goes on through the intervention there. And as I mentioned in this particular question the comparison group is not mentioned. So when we talk about drug therapy, which is our intervention, what drug therapy are we considering? What's eligible for our systematic review? So you have to decide this. And we decided ahead of time that drug therapy for moderate hypertension, we would include ACE inhibitors, angiotensin receptor antagonists or ARA. Beta adrenergic blockers, combined alpha and beta blockers, calcium-channel blockers, diuretics, central sympatholytics, and direct vasodilators. So those are all drugs that we would consider all right in answering our question. And so you can see already that it's sort of a broad question and that we've decided to include all these different types of interventions for moderate hypertension, because they come from very different classes. But we're going to include all of them because it's really a much bigger question which is, morbidity and mortality. No matter what you're taking, does it work, is basically what we're asking. And what's long term? Well, you have to define what long term is and we're saying at least one year, greater than or equal to one year. What's morbidity, what's mortality? So we're including then as outcomes fatal and nonfatal strokes, fatal and nonfatal coronary heart disease, cardiovascular events, and total mortality. And you would have to define each of these very specifically, and I will say, again, that sometimes people choose just to use whatever definitions authors use in the individual studies included in your systematic reviews. And that probably is okay, unless you're in a field where there's a good deal of disagreement about what constitutes an outcome or a diagnosis. What's an older person? Well I'm very sorry to tell you that it's people older than 60 years old who are outpatient. That is, I wouldn't call those older persons necessarily but I'm afraid our healthcare setting does. And outpatient, so they're not interested in people who are inpatients. And then what's moderate hypertension? For this systematic review, it's a systolic blood pressure of 140 to 179 millimeters of mercury, and diastolic is 90 to 109. So you can see that even though the question looks pretty simple, there actually behind the scenes are definitions of what each of these terms means. So let's try an epidemiology example so that you have some practice doing both. Here's the question. Is a history of exercise training associated with falls in community dwelling and institutionalized people? So how does that parse out? So here we have it, the exposure is history of exercise training. The outcome is false, the setting is community dwelling, and the population is institutionalized people. And we're putting, in this case, settings and population together because I tend to do that. So history of exercise training, we're saying has to have been in the past two years. And it could be anything from balance training, mobility training, physical therapy, strength training, Tai Chi. So any of these count as exercise training. And again, here's a place where some experience with the subject matter probably is useful, because people know that Tai Chi has been tested to look at its association with falls. What about outcomes? Well, the outcomes that were determined to be of importance by these systematic reviewers were number of falls, falls that were injurious and that has to be defined, hospitalization, fracture, and death. So just having a fall Is an outcome. Well, what's a fall? Is it losing your balance but catching yourself on the edge of the counter? What if you fall in your house and nobody knows about it? Is that a fall? So you have to consider all of these things and how you're going to record those falls. Do people keep diaries? Is it only in certain situations and so forth. And in this study, institutionalized and community dwelling people were defined as outpatients and people in nursing homes, that's considered community dwelling. And even though it didn't say anything about the elderly, it is just patients greater than or equal to 65 years old. So those are the definitions that are behind the scenes for this shorter sentence. So those are two examples of how you would go about asking your question, and the behind the scenes eligibility criteria associated with that shorter version. So here's a question that someone who took the course, or group that took the course, in 2009 came up with. They looked for their topic at the association between the level of alcohol consumption and the incidence of stroke. Their research question then, with a question mark, was does moderate to heavy alcohol consumption reduce the risk of stroke? Their population was adults without prior stroke. Their exposure was alcohol consumption and it could be presented in the paper as drinks per day. It could be measured over the past month or longer, binge drinking was all right but so was looking at it as short term. So exposure had a pretty broad definition here. The comparison group was non-drinkers, total non-drinkers. The outcome was ischemic or hemorrhagic stroke or both. And enough information for the authors of the systematic review, to be able to estimate for each individual study, a relative risk and odds ratio. And attributable risk in the 95% confidence intervals. And, then, finally, these epidemiologists and clinicians, who were in this group in 2009 said, we are only going to look at cohort studies. So that was a very interesting situation, it wasn't an intervention, it was really an epidemiology study or etiology study you could say. So that ends our examples in section E, and our next section will be section F where I will talk about analytic frameworks.
