Everyone is heading down this path. In all sorts of policy spheres.
Health care, yes def. Social care, ditto. Troubled families – many similarities.
Probably many other areas of social policy.
Invariably with limited resource got to decide who to offer key worker too, mostly we go (for obvious, resource related and pragmatic reasons) for the “high risk” group.
I remain unconvinced it will be the panacea everyone hopes for.
I hope I’m wrong.
Here are 10 reasons for caution
We overestimate importance of freq flyers, the highest risk, most frequent offenders, however you want to frame this question
impactable, CAN you actually make a difference – in health context, often this group are really quite poorly and there may t be a great deal that can be done apart from increasing efficiency in use of multiple services. See the empirical evidence of “integration” here.
Numbers and scale
If you want to make difference at scale – focus on big numbers.
If you’re focusing on the top 5% you’ve got to make a giant change to make any difference in net terms at population level.
See Roland for a neat illustration of this
Say you’ve set a target to reduce event rate (here -admissions) bu 10% in the population.
By concentrating on the 0.5% at highest risk of admission, more than the total number of admissions in this this group would need to be avoided (107.5%).
If the next group down were the focus of an intervention (the 4.5% %of the
population at high risk), you’d need a 40% reduction in events in this group to acheive your 10% population level target- doable?
We ignore regression to mean.
If it was a big year last year, chances are it will be a small one next. Underscores the importance of long term evaluation.
And we ignore supply induced demand.
The more supply then in the context of unlimited demand you may pay more if you increase supply. Make sure your gate keeping is cast iron.
We forget about variation due to chance. Do proper evaluations!
We ignore the evidence
We implement unevaluated interventions and interventions we know don’t work.
We don’t evaluate properly
We forget to consider what’s happening in the control group in our evaluation
What about the economics, including implementation costs.
Words like “where’s the hidden troubled families national evaluation” spring to mind. My guess is that the data says its marvellous when you consider it in a simple way with no comparator, but not quite so shiny in the context of a proper robust evaluation
EVALUATE STUFF PROPERLY
Evaluate your intervention against changes in overall patterns of admission or using a control group
this paper is awesome
If you’re “screening” for something, most cases do not come from the screening process.
Neat example re diabetes. Applies in many other areas.
models are based on an assumption that interventions focus and greatest return should is in people at greatest risk. This is erroneous
Geoffrey Rose pointed out a pitfall in this argument using hypertension over
20 years ago – most admissions come from low risk patients, and the greatest effect admissions will be made by reducing risk factors in the whole population rather than in a small group of high risk people.
(Taken from the Roland article)
In the ‘Prevention Paradox’, Rose (1985) argued that strategies which aim to reduce CVD risk in the whole population are more effective than those targeting individuals at highest risk. The ‘paradox’ is that such interventions might offer only small advantages to each individual, but can achieve large overall health gains for whole populations.
Most cases of diabetes are not in the group that are classed as ‘high risk’. In that high risk group the individual risk is highest, but there are more cases in net terms in those at lower risk, who are obviously far greater in number.
Same applies in pretty much any sphere.
See point 3 also
Nuffield Trust Developing care for a changing population: Supporting patients with costly, complex needs
The reflections in the discussion in the pic here may well be the same considerations that carry pertinence in other policy areas
This stuff def applicable in health care. I’d bet a months pay also applies in other areas of social policy
I hope I’m wrong, but I am cautious
But I’m not of a view that case management in folk at highest risk of something we want to prevent is the panacea we hope it id
High risk case management will save mega dollars?
Commonwealth Fund Models of Care for High-Need, High-Cost Patients: An Evidence Synthesis
State of the art evidence is this one
“Overall, the evidence of impact is modest and few of these models have been widely”
There’s lots of emerging material on models for high need patient groups on the Commonwealth Fund website. Little empirical evaluation though.
Cost savings in ACO model. McWilliams. Mostly coming from population approach not high risk approach