This article is timely because policymakers want to see a major shift to Patient-Centred Care with SDM as one of the major techniques driving the change. Historically, the NHS has used targetted incentives, based on simple measures, to motivate clinicians to change. Although there is a strong faction who have adopted new ways of working with patients without incentives, overall, progress is slow. Quite a lot is known about the disincentives to adopt new ways of working but to date, no simple formula for overcoming resistance has been found.
The assumption that aggregating idiographic patient information creates normative performance data is an attractive one to service managers – even more so if the data can be portrayed as a quantitative outcome which creates an instant ‘measure’. Framing an output measure using a logic model that effectively says ‘good’ EBM leads to ‘good’ SDM might lead to the conclusion that a simple measure of the existence of the latter demonstrates the appropriate use of the former. The paper goes as far as to suggest that a measure based on the question ‘Did SDM occur or not?’ might be seen as sufficient evidence to underpin a performance framework, and, indeed, such measures do exist.
Despite the N=1 assertion, the authors seem to slip into the trap they initially describe – that a simple statistic at a population level will tell us whether EBM has led to SDM . In the text, SDM, and to some extent EBM, is transformed from a process to an outcome based on the patients view that SDM took place. While fulfilling the managers need for simplicity it does seem to ignore the complexity of the process and the roles of the many people who may be involved in it. Also, a great many consultations neither lead to, nor ask for, a decision – chasing up referrals and providing reassurance or social support occupy a significant proportion of GP’s time. Patients access a wide range of sources in decision making which are never revealed to their physician, nor is the weight given to them.
A number of questions spring to mind about the idea of SDM being the only outcome that matters. How is this assessed by the single question? If N genuinely does = 1, wouldn’t a judgement have to be made on each case according to its merits? What if a choice did not exist (or existed in the evidence base but not in the service provision). Evidence itself is often contradictory or subject to change. How a decision was arrived at might be viewed differently by either party with hindsight, especially in relation to prioritising responses involving multimorbidity. Given the complexity of SDM as a process, might not a single tick box response lead to some fairly formulaic consultations as opposed to the idea of the conversations that are meant to replace them? How does timing effect how the single question might be answered?
A question that occurs to me about the whole way the paper is couched is why SDM is portrayed as such a singular process when in reality it is a member of a family of techniques designed to assist patient autonomy?
The danger as I see it is that policymakers use the single question because it’s simple, not because its the right question. A better question might focus on whether patients have care plans which take account of patient preferences and how they are being jointly used to manage patients’ wellbeing.
Mainstreamnot 