I am looking a variation in Medical test requesting practices by Primary care physicians. The tests I am looking at have clear guidelines, the data is normalised by demographics etc. Other variables like experience, special interests, QOF framework, patient pathways, resource allocation, easy of access etc have been factored in.
After all this one would still see 15 to 30 fold variation.
how does one go about defining " acceptable variation" in such a system where in a human being ( Doctor) is interacting with a patient. 2 sets of human factors influencing each other.
For example, I am a secondary care physician- and I am aware , in the last 10 patients I saw for a suspected condition - the number of tests I did vary from person to person.
Thoughts and comments please
HI,
You have raised a very fundamental query in the practise of care !
How do we define "acceptable" and "variation" in the context of an inherently variable / dynamic duel? Each human is a variable and then he / she gels into a predefined role(model) of disease / doctoring at a commonly defined threshold. The in- homogeneity of the system is an NP problem just like climate interaction / change and it's assessment ...(please see a similar discussion on how / wat metrics to use other than "temperature" to define climate change , in this forum).
You may rigorously define just one variable and pin it and see how the other variables "adjust" around it to attain a stable equilibrium!
I agree with above.......The correlation coefficient of ïntrapractise"" tests on a single case (predefined) or coefficient of variation between practices in a single clinic are good suggestions to undertans the äcceptability in the variation!!!
However, the analysis should inform the system of the benefit of such an excercise...the secular trends in variablity of practise is already known and defined.
Hi
The importance of context was already mentioned.
Your example brings the notion that you for yourself feel uncomfortable when you reflect on your work and experience variability. So the context could be "Is deviating from a guideline (if it exists) acceptable?" "What are the reasons of the deviation?" - This could still be scrutinized with the methods of EBM and shared decision making. More qualitative than quantitative.
Quantitative questions and indicators are more in the context of the health care system and economics - "Is extra cost for tests reasonable?" "Is cost saving harmful for patients?"
Last but most important the context of the patient. Do they experience "personalized shared decision making"? Or do they feel injustice by being treated differently? Do their perceptions vary over time or depend on their outcome?
All this could go into acceptance!
- Badges
- Science topic
- Similar topics
- Geoscience
- Asia
- India