To prevent death: 333.
In a bitterly complex analysis of Center for Medicare and Medicaid Services data, these authors describe a relationship between admission rate and subsequent cardiac adverse events. Based on a statistical sample of Medicare patients visiting acute care hospitals, these authors calculate an admission rate for chest pain for each, and divide the sample into quintiles. Then, the authors follow index visits for chest pain to those hospitals, and measure 30-day acute myocardial infarction or death. Thus, a relationship between admission rate and poor outcomes.
The mean adjusted admission rate for chest pain ranged from 37.5% in the lowest quintile to 81.0% in the highest quintile. Owing to the large sample size, many of the differences between hospitals in each quintile meet statistical significance. However, the difference that leaps out at me the most, for-profit hospitals represented 24% of the highest quintile for admissions, while for-profit hospitals were only 7.8% of the lowest.
And, what was that massive variation and expenditure associated with, in terms of beneficial outcomes? An inconsistent reduction in subsequent AMI and death which, through multivariate logistic regression, was equal to about 3.6 fewer AMIs and 2.8 fewer deaths per 1,000 patients – with very wide 95% CIs.
And, thus, to oversimplify and overstate the soundness of the analysis, the NNT of 333.
It seems very reasonable to suggest a relationship between intensity of care and 30-day cardiac outcomes. Such intensity of care, however, is quite expensive – on the order of probably $1.5-$2M per life shortened and captured in this 30-day window. As our population ages, we are simply going to have to do better – in order to maximize the value of the limited healthcare dollars.
“Variation in Chest Pain Emergency Department Admission Rates and Acute Myocardial Infarction and Death Within 30 Days in the Medicare Population”
http://www.ncbi.nlm.nih.gov/pubmed/26205260