This relates to Dr. George Lundbery's video editorial available in yesterday's blog post here. According to his explanation the right clinical pathway (regarding a specific condition in his example) depends upon the stability of the patient. Apparently physicians are continuing to follow evidence-based practice when the patient has not been stabilized but then are not shifting to the less costly equally effective evidence-based practice when the patient has become stable. So what is at issue is the grain of the specification of what is the best evidence-based practice. I imagine that if the less costly practice was to be followed on both stable and unstablized patients, physicians would more likely conform. As it is, their conformance requires an adjustment in practice during the course of the clinical pathway.
What I am thinking now is that if clinical pathways are written at a high grain (relatively few dimensions of criteria) conformance by physicians is likely to be high. But as additional dimensions are added to the criteria (such as stability of the patient) rates of conformance are likely to drop because conformity becomes more difficult. As the specificity of the pathways becomes higher conformity will become more and more labor intensive and cognitively demanding. This will clearly lead to automated systems and physicians will be asked to input a lot of data, surrendering judgment to the expert system software. It will become increasingly time consuming to document why the physician disagrees with the output of the system and physicians will feel a loss of autonomy as their role becomes more nearly that of highly credentialed technicians.
There is a differences between pay-for-performance and pay-for-conformity. Under pay-for-performance judgment remains within the physician's discretion. Financial rewards are based upon outcomes rather than processes. Under pay for comformity, most judgment shifts to the automated system. I am not advocating anything here. The grain of the specification of clinical pathways and the distinction between pay-for-performance and pay-for-compliance have many implications I hope to explore in future posts.