Publication alert: The geography of Medicare’s hospital value-based purchasing

Medicare presently withholds 2% of participating hospitals’ base operating Medicare Severity Diagnosis-Related Group (MS-DRG) payment amounts and redistributes it based on relative performance and demonstrated improvement. It is supposed to function as a reward for hospitals doing well and a motivator for those not. In 2019, the amount redistributed was $1.9 billion, with 44% ofContinue reading “Publication alert: The geography of Medicare’s hospital value-based purchasing”

Improving upon a bar graph

In my preceding post, I showed how to replicate a typical Quickstats graph from the CDC’s Morbidity and Mortality Weekly Report. But such a bar graph is rather information-poor. There are only 24 pieces of data: the mean and the 95% confidence limit for each of the sex-age strata. (You could argue that the upperContinue reading “Improving upon a bar graph”

Reproducing published graphics in R

A wonderful blog post from a few years ago picked five statistical graphics at random from regular media sources like the New York Times and reproduced them with a handful of lines of R code. I thought it was written by Rafael Irrizary for the Simply Statistics blog, but now I am not seeing itContinue reading “Reproducing published graphics in R”

Finding the p-value for a binomial hypothesis test

In addition to the consulting projects I work on, I have been teaching an introductory statistics course online for a number of years. My teaching assistant this semester pointed out a seeming inconsistency that no one had ever raised before. Two widely-used methods in R for calculating the p-value for a two-sided binomial hypothesis testContinue reading “Finding the p-value for a binomial hypothesis test”