Different Ways to do Action Research, QIPs

different ways to do action research: fishing net sunset DSC_0168 copy Anoop Negi via Compfight


Are there different ways to do action research? What different kinds of studies are there?


No, there’s not just one type. The typical action research project includes assessing the current situation, seeing a problem, researching it, instituting a plan for change (in some kind of physician behaviour usually), actually doing it, and then seeing what results, of course measuring something both before and after the change. In real life, you jigger it as it runs, and you jigger it after it runs, and you do the whole thing again. The computer gurus at OntarioMD call it, “Plan, Do, Study, Act.” I think in my little scenario, perhaps it’s closer to, “Study, Plan, Do, Act”. (Doesn’t that make a wee bit more sense?…but what do I know?) That’s called experimental research.


But there are also descriptive studies, which are just as valid for action research. These are historical descriptions, a looking backward to see what happened. This is an example of nonexperimental quantitative research, one of these different ways to do action research. These studies often help with generating hypotheses.


As our clinic moves towards initiating the standard Quality Improvement Projects in patient satisfaction, access, and hospital discharge follow up, I think I can try to do one of these descriptive efforts to look at  recent changes in medical guidelines, and how they have affected my practice. I am personally trying to implement the new US guidelines for lipids/ statin use (yes, I know I’m Canadian and this hasn’t been accepted formally yet north of the border).


It occurred to me the other day that for the last few months, I’ve been discontinuing statin use in at least a couple of people per week, and on the whole ordering less lab tests. So, this is something that has already happened. The standards came out months ago, and fairly abruptly I started using them. I know I’m behaving differently now than I was 6-8 months ago. My approach to lipids, statins, lab tests, risk assessment… my approach to many patients has changed, radically. Looking back, I think this is largely how I’ve managed my practice, reading when I can and implementing changes on the fly, never measuring anything but my gut feeling. Not too scientific, huh?


I’ve decided to follow the US guidelines for a number of reasons: a) Canadian guidelines eventually (don’t we all eventually follow the US?) follow US guidelines, almost exactly. I’ve just decided to get in on the sharp edge of the blade. b) I in no way pretend to be any kind of scientist /expert/ statistician. You can fool me. I can try, I can frown, but throw enough numbers at me in an authoritative tone, and I just may fall for it. As a practising GP, I have to rely on guidance from expert panels that produce guidelines. I don’t have enough time to do a thorough literature search, and I frankly struggle with the statistics. The US put a lot of resources into this study, an entire panel of experts that were able to do a very broad literature search over a period of 2-3 years. They came up with clear statements based on the literature that currently exists. I think I can trust this expert panel. c) Their suggested treatment guidelines just made utter sense to me. d) The US guidelines are simpler. Period. Treat the patient, not the numbers. That just fits with my psyche. High risk patient, no matter what the lipid levels, gets a high dose statin. Simple. No real risk? No statin.


Imagine that. Actually stopping a drug. And, more than that, they gave me actual mg goals for different drugs. Crestor max dose usually 20, as only one study has been done on doses past that! Did you know that? How many patients have you got on crestor 30 or 40? Lipitor is more fuzzy, high dose being 40-80. This simple GP likes crestor; enough decision making! High risk patient, crestor 20. Easy.


Of course, eventually it will all come out in the wash. Maybe targets are right; we just don’t have the evidence yet. I suppose that’s why they suggest still measuring the lipids…


Has that occurred to you? With the new lipid guidelines… why measure it?


I’ve dropped down my testing to yearly, so if some information comes out on targets, I’ll have some data. I look at LFTs sooner, of course. Is there any way to measure this decrease in MD angst? No framingham risk calculation, and pushing the statin up to see where the risk ends up. It’s all based on a fallacy! It’s all theory! Less calculating, less RN calls, less repeated blood tests, less non smoker skinny as a rail marathoners with a bad FHx on statins.


I think I’ll look at 2 weeks to start, perhaps in the spring of last year, and look at the same 2 weeks this year, 2014. If I just stay with my practice, that should eliminate some other factors. I’ll look at prescription doses, statin starts, statin stops, lab test frequency, and possibly other aspects, and try to measure the change that I’ve perceived. Hopefully two weeks will do it…


If I’m not mistaken, that’s called an ex posto facto study (McMillan, 2012). What’s my goal? Probably several. One, I need to stop flying by the seat of my pants. I need to think of what the problem is, measure it, research, reach a possible solution, institute it, measure it. This is punishment, clear and simple. Learning 101. Secondly, there just may be financial implications here. People will be coming off crestor 30, and 40. People will be coming off meds, on the whole I’m thinking… Less lab tests…And, I need to explore these different ways to do action research, ways to do QIPs!


It’ll be interesting. Interesting punishment.




McMillan, J. H. (2012). Educational research: Fundamentals for the consumer (6th ed.). Boston, MA: Pearson Education


Leave a reply