Writing on BMJ blogs, Muir Gray makes the case that any distinction between public health and clinical care is artificial – population vs personalised care are two sides of the same coin.
He thinks of this in terms of value. For public health, benefits rise fast as investment increases, then there is a levelling out (the law of diminishing returns). Harm however rises in a straight line despite investment because the more interventions that are done the more scope for harm there is. This is particularly true for an interventionist medical model. Overdiagnosis and overtreatment (think depression, ADHD, chronic pain medication) can result in a reduction of the benefit to harm ratio. See graph to illustrate the point here.
For individualised treatment there is a slightly different issue, illustrated by this graph. Gray explains:
As we put more resources into a particular intervention or operation, the offer for the individual patient changes. When there is only enough resource to make interventions available to a few, they are offered to people who have the most to gain, and who are therefore more willing to accept a risk. However, as investment increases, interventions are offered to people who are less severely affected; therefore the maximum benefit they can expect is less, but both the probability and the magnitude of harm is the same.
Could this value model be applied to medicalised addiction treatment? After all our whole ‘harm reduction’ ethos is premised (obviously) on the reduction of harms. If we work more intensively with more people, kept in treatment longer, is it possible we cause more harm over time? Would safe exit from treatment reduce harm or would it have a paradoxical effect? Premature death, for instance, is a pretty important harm to prevent. Do abstinence based treatments resulting in discharge from treatment reduce harm more effectively over time or not?
There is something a bit different here though. We have the concept of choice in addiction treatment which is often much more limited in medical interventions that are often based on guidelines with stepped approaches – think asthma or high blood pressure treatments. Does this value model translate?