Statistics Matter Thursday, February 23, 2023
Explicit and implicit knowledge - how we change our practice in medicine
Evidenced Based Medicine is Hierarchal
1. Large number (N) (narrows chance - narrows confidence interval)
2. Prospective (less respect for the retrospective because correlation is not causa-tion)
3. Randomized (decreases selection bias)
4. Double-blinded (decreases confirmation bias)
5. Placebo-controlled (decreases treatment bias- placebo effect)
6. Multi-centered (decreases selection bias)
* peer reviewed
* conflicts of interest (COI)
Internal validity vs External Validity
The large, prospective, randomized, double-blinded, placebo-controlled, multi center trial published in a peer-reviewed scientific journal with divulged conflicts of interest is the current top-tier of medical science (“gold standard”). But even at this rarefied altitude, there are several forms of inherit selection bias. To do this kind of study takes considera-ble time and a huge amount of money which in our US system is usually performed by private pharmaceutical companies and medical device manufacturers (funding bias/COI). To enroll people, only about 10% actually do the study (recruitment bias). Studies are of-ten done at extremely large academic centers with the “best doctors” at the “best places” (specialty bias/organizational bias). It is more likely to be published if it has a positive conclusion (publication bias). To have such a tightly controlled experiment, it must speak most often in binary language. While all of these are necessary to create a contemporary gold standard published paper, the conclusion can tell you only about internal validity— within the parameters of the methods and subjects of the designed study. Physicians from all over the US and the world need to know the external validity—how this particular treatment/device performs in their specific place with their population of staff and pa-tients. Ideally, one needs “real world data” of external validity published (or simply locally examined) after the large, prospective, randomized, double-blinded, placebo-controlled, multi-center trial to see how the ideal performs in your real world.
Quantitative Bias
In clinical medical science, almost all of this is put into quantitative measures. Even things like “How bad is your pain?” which is a qualitative measurement is usually placed into an ordinal quantitative measure (eg. 0-10 scale). This is frequently done in satisfaction sur-veys where you rate the service on how well you liked it (qualitative) on a 1-5 scale (quantitative). When you place qualitative measures into a quantitative scale, you introduce a whole host of inherit problems (eg. qualitative measures are not divided up into equal increments— pain going from 2 to 0 is not the same as going from 10 to 8). Reliable and meaningful published qualitative data is difficult to produce in clinical medical science.
Explicit and Implicit (experiential) knowledge
Published medical articles and texts are forms of quantifiable cognitive-based information that is explicit and codified. This medical scientific information is by its very nature always being challenged and disproving itself (eg. “good medicine begins with disbelieving your-self”). We “keep changing what we are saying” ideally as a process of achieving scientific integrity. You confidently but humbly accept the explicit knowledge made available to you with the understanding that it is incomplete, usually plagued with some form of bias and sometimes manipulated. This is quantified explicit knowledge.
We have another kind of knowledge which is qualitative and difficult to quantify and in-cludes things like experience, emotions, values, intuition, and even unexplainable phe-nomena. While this can be experienced collectively and even described collectively, it is more often than not, a N of 1. This kind of qualitative knowledge should never supersede explicit quantified medical science (“…in my experience…”) —but this implicit knowledge must always join the explicit knowledge in the proper professionalization of medicine. The implicit qualitative knowledge is the experience of the physician-patient relationship. From the patient’s experience, there is the moment when “the relationship becomes the treat-ment”. From the physician’s perspective, the relationship becomes not only the motivation to know the explicit knowledge better, but the personal joy of practice comes from the freedom of allowing the patient to participate with you. Implicit knowledge makes the pa-tient a shared partner and ultimately the lead partner in healing instead of the subject of the doctor’s therapy.
MORE CATEGORIES
Daily Calendar (16)
Events (12)
Everyone Has a Story (2)
General (7)
Jobs & Career (3)
Must Read (261)
Podcast (1)