Prevention as long term cost control – part 1
“Rising health care costs to derail economy,”
“obesity tsunami will wreck our NHS.”
Not a minute goes by without another tweet or newspaper headline about the inexorable rise of health care costs and our likely future inability to meet this challenge without “radical changes”. Nobody seems up for truly “radical stuff”
Admittedly “radical” is hard, risky and sometimes politically unpalatable.
What about another way.
It is often said that prevention is better than cure. This seems obvious from a quality of life viewpoint. From a cost control perspective … is it?
Consider these two pieces of information.
1) Failure to prevent
Across health care there are hundreds of examples of where we fail to implement cheap and effective preventive interventions fully. For example – stroke prevention in AF, diabetic blood pressure control, smoking cessation in high risk groups, pulmonary rehabilitation to prevent future COPD exacerbations, falls prevention, fracture prevention. This is before we even begin considering primary prevention of obesity, smoking prevention etc.
2) Comparing the cost per case versus the increased number of cases
Diabetes is a good example. A recent paper highlighted that rising prevalence of treated disease has become the main driver of increased spending on health care in the United States.
Higher treated disease prevalence and higher spending per treated case were associated with 50.8 percent and 39.0 percent, respectively, of the spending increase seen in the population ages eighteen and older, while their joint effect accounts for the remaining 10.2 percent.
The proportion of increased spending attributable to increased treated prevalence alone is particularly high in the Medicare population: 77.7 percent, compared to 33.5 percent among the privately insured.
Moreover, the current findings reveal a substantial contribution to the increase in total spending (10.4 percent) from a doubling of the share of the population considered to be obese and from increases in treatment intensity, a component of spending per treated case (11.9 percent), in 1987-2009.
The implication set out in the paper is that constraining the cost of health care will require policy options focused on reducing the incidence of disease, as well as improved understanding of the extent to which more aggressive treatments for chronic conditions do, or do not, result in lower morbidity and mortality.
Another recent US study (of good provenance) found that despite the increased prevalence of diabetes over the past decade, the financial burden borne by patients for treatment of the disease has actually decreased. In essence the cost per patient has decreased (generic meds) but the total population cost had gone up largely due to the increased prevalence.
Of course there are complexities relating to the interpretation of the results, and there are issues associated with generalisation from the US system to the UK. However I believe the overall policy implication is clear – prevention must form a larger part of our policy response to cost control.
Diabetes costs closer to home
Sticking with diabetes, for now, we can draw some limited parallels with the English system, by combining QOF and Programme Budgeting data. Programme Budget data is a source of data on spend. It is imperfect. Nonetheless, it is an important source of data on spend in a programme area.
This (admittedly crude) data tells us that over a five year period there was a 26% growth in total number of diagnosed patients, a 43% growth in commissioner spend on diabetes care. Within this there was a 33% increase in the spend on glycaemic medicine. When you calculate a crude total cost per patient (the bottom row in the figure), the result is interesting – a 6% growth in the cost of medication per diagnosed patient and a 14% growth in the commissioner cost per patient.
So whilst not as clear cut as the US research, and the above has some important caveats, the overall message is the same – overall cost is driven by increasing number of patients AND by increasing cost per patient.
Thus for both data from the USA and closer to home a good chunk, maybe more than half, of cost increase seems to be attributed to more cases rather than rising cost per case. The message is clear – prevention is an important part of long term cost control.
In part two we will explore other areas and some counter arguments
 Thorpe KE, Treated disease prevalence and spending per treated case drove most of the growth in health care spending in 1987 – 2009. Health Affairs 2013 5 pp851-858