Dartmouth Atlas Project finds variations in joint replacement surgery

The number of hip, knee, and shoulder replacement surgeries is rising significantly for Medicare patients, but the rate of increase varies widely among regions of the country, according to a new report from the Dartmouth Atlas Project.

The report indicates that rates among regions vary by 10 times for shoulder replacements and by four times for hip and knee replacement surgeries.

Researchers analyzed Medicare data from 2000 to 2001 and compared it to data from 2005 to 2006. They found a 15 percent increase in the overall rate of hip replacement, a 48 percent increase in the overall rate of knee replacement, and a 67 percent increase in the overall rate of shoulder replacement between the two periods.

Researchers said the patterns suggest both overuse and underuse of these procedures. Some patients who could benefit from joint replacement may not be offered the procedure, while others may be having a procedure that they might choose to forego if they had received balanced information on risks and benefits.

“A likely interpretation is that the decision to undergo surgery is being influenced more by physician judgments than by the preferences and values of individual patients,” said lead author Elliott S. Fisher, MD, MPH. “Another explanation might be that patients in some geographical areas do not have adequate access to joint replacement.”

“These findings highlight the need for improved physician and patient education and the use of shared decision-making to determine whether a patient should undergo joint replacement,” said coauthor John-Erik Bell, MD.

Regional variations
Marked variations were found in the rate of hip replacement (per 1,000 beneficiaries) during the five years of the study period. In 2000–2001, Alexandria, La., had the lowest rate of hip replacement, at 1.2; the 6.7 rate in Boulder, Colo., was more than five times higher. In 2005–2006, Bryan, Texas, had the nation’s lowest rate at 1.8, and Ogden, Utah, had the highest rate at 7.2.

Rates of knee replacement (per 1,000 Medicare beneficiaries) also varied widely. In 2000–2001, Honolulu had the lowest rate (2.5) and Elyria, Ohio, had the highest rate (10.5). In 2005–2006, Manhattan, N.Y., had the lowest rate (4.0), while Lincoln, Neb., had the highest rate (15.7).

Although rates of shoulder replacement were lower than rates of hip and knee replacement, use of the procedure still varied widely by region. In 2000–2001, the lowest rate of shoulder replacement (per 1,000 Medicare beneficiaries) was found in Lexington, Ky., (0.2) and the highest was found in Fort Collins, Colo., (1.8). In 2005–2006, Syracuse, N.Y., had the lowest rate (0.3); the rate in Provo, Utah, was 10 times higher (3.0).

Differences by race
It is well documented that black patients are less likely than white patients to undergo hip or knee replacement, and Dartmouth Atlas data from Medicare enrollees in 2000–2001 and 2005–2006 show that differences persist. For example, the 2000–2001 rate of knee replacement per 1,000 black Medicare beneficiaries was 4.0; for all other enrollees, it was 6.1. Both rates increased by a similar degree over the next 5 years, but the disparity remained, with a rate of 5.6 for black enrollees in 2005–2006, compared to a rate of 9.1 for all other beneficiaries.

The Dartmouth Atlas Project is run by the Dartmouth Institute for Health Policy and Clinical Practice. A link to the full study can be found at www.dartmouthatlas.org