So You Want To Use The Three Legged Stool
The three-legged stool is a framework most people in healthcare encounter early in their careers and then quietly stop thinking about until a peer review board forces them to justify a decision. It sounds simple on paper. Three legs: best available evidence, clinical expertise, and patient values. If one leg goes missing, the whole thing wobbles. The problem is that nobody actually teaches you what happens when two legs agree but the third contradicts them. I spent years watching clinicians treat this framework like a checkbox exercise. Fill out the form, cite a guideline, document the patient preference discussion, move on. That misses the point entirely. The model exists because real decisions live in the tension between these three elements, not in any single one of them. Here is how it actually works when you stop treating it like paperwork. Start with the evidence leg, but be honest about what grade of evidence you are looking at. Systematic reviews and randomized controlled trials sit at the top, sure. But most clinical questions don't have RCTs behind them. You will frequently encounter situations where the best evidence is a small observational study from 2014 or a consensus guideline written by a committee that has never treated your patient population. Acknowledging the actual quality of the evidence upfront saves you from building a house of cards later. I had a case where a widely recommended protocol for post-operative nausea was based on a meta-analysis that excluded patients over sixty-five. My patient was seventy-one. Following the protocol blindly would have been negligent, even though the evidence technically supported it. I ended up using a different antiemetic combination and documented the reasoning thoroughly. The stool required the evidence to bend to the expertise leg in that instance.
The clinical expertise leg is where most people get sloppy. It is not just seniority. A clinician who has been practicing for twenty years but has never updated their knowledge base since residency does not qualify as expertise here. Real clinical expertise means pattern recognition built from a large sample of direct patient encounters combined with the humility to know when your intuition conflicts with good data. I once saw a colleague dismiss a clear diagnostic finding because "it didn't look like the textbook presentation." The textbook presentation came from a study of three hundred patients in the 1980s. The actual patient had a perfectly valid atypical presentation that the newer literature had already documented. Dismissing evidence because it felt wrong is the number one way this framework fails in practice. Patient values and preferences is the leg that gets the least actual attention, and it is also the one that causes the most downstream trouble. Getting informed consent is not the same as integrating patient values into clinical decision-making. A patient can consent to a treatment while fundamentally misunderstanding what they are consenting to, or they might agree with you on the surface while privately harboring reservations that would change their outcome if you had asked the right questions. I ran into this with a diabetes management plan. The evidence strongly supported an intensive insulin regimen. The guidelines backed it. My clinical assessment said it was appropriate. The patient was a long-haul truck driver who had been managing Type 2 diabetes for twelve years without daily injections. He agreed to the plan during our appointment but made it clear over the next few weeks that the regimen would make his job impossible. The three-legged stool collapsed under its own weight if I pushed the evidence alone. We ended up agreeing on a modified oral medication protocol with closer monitoring. The evidence was weaker for that approach, but adherence went from near zero to sustained. That is the actual point of the model. When you are applying this in a real setting, the workflow is not linear. Most people try to go evidence first, then expertise, then patient preference. That order assumes the evidence will be clear and decisive, which it rarely is. In my experience, starting with the patient conversation usually reveals constraints you cannot see from a literature search. A treatment that looks optimal on paper might be worthless if the patient cannot administer it, cannot afford it, or simply will not tolerate it. Once you know those boundaries, you can scope the evidence more efficiently and then apply your clinical judgment to bridge whatever gaps remain.
There are legitimate limitations to this framework that people rarely acknowledge. The biggest one is time. Properly weighing all three legs on every decision takes significant effort, and most clinical environments reward speed over thoroughness. You will face institutional pressure to standardize care through protocols that flatten the complexity the stool is supposed to capture. Second, the evidence base itself is often contradictory. Two systematic reviews on the same intervention can reach opposite conclusions depending on which studies the authors included or excluded. When that happens, the framework gives you no guidance on which leg should take priority. It just tells you all three matter, which is simultaneously true and unhelpful. Another practical issue is documentation. If you make a decision that deviates from established guidelines because patient values or clinical judgment overrode the evidence, you need to be able to show that reasoning. Vague notes like "patient preference considered" will not hold up. I learned to keep brief clinical decision logs that track the specific evidence consulted, the rationale for weighting expertise or patient preference higher in a given case, and the agreed-upon monitoring plan. This took maybe five extra minutes per complex case but prevented countless headaches during audits or adverse events. If your organization treats the three-legged stool as a compliance checkbox rather than a decision-making tool, there is not much you can do about the system. But you can use the framework honestly within your own practice. Track your own deviations from standard protocols. Note when and why you prioritized one leg over another. Over time you will build a personal evidence base of your own that is arguably more useful than any generic guideline, because it reflects the actual patients you treat, not the idealized population in a study.
Get the Full Details
