For example, this 2010 Dartmouth Atlas Surgery Report from the Dartmouth Institute made much of the national variation in Medicare hip replacement rates in 2000-01. Noting that the rate ranged from 1.2 per 1,000 in the hospital referral area of Alexandria, Louisiana, to 6.7 per 1,000 in the hospital referral area of Boulder, Colorado, it concluded that
[b]ased on the data presented here, it appears that patients in some regions and among some populations may not be getting adequate access to the procedures, while patients in other regions and among other populations may be undergoing the procedures at higher rates than necessary.
Reaching such a conclusion from local hip replacement rates requires assuming that the U.S. population is composed of identical people identically distributed over a featureless geographic plain that is everywhere the same.
Unfortunately, the Dartmouth report failed to inform readers that the most common reason for hip replacement is osteoarthritis and that it is well known that the prevalence of primary osteoarthritis of the hip occurs at much higher rates in Caucasians than in other population groups. In 2010, the proportion of blacks or African Americans in Alexandria, Louisiana, was 57.3 percent. In Boulder, Colorado, it was less than one percent. Hip replacement rates also vary with age, comorbidities, socio-economic group, and employment history.
I did not know this. It does seem relevant to know this, before concluding that one region is doing "too many" hip replacements.