• Jane A Barrett, John A Baron, and Michael L Beach.
• Department of Community and Family Medicine, Dartmouth Medical School, Hanover, NH 03755, USA. jane.a.barrett@dartmouth.edu
• Osteoporos Int. 2003 Nov 1; 14 (11): 889-94.

AbstractHip fracture has repeatedly been associated with increases in the risks of death and pulmonary embolism (PE), but few studies have considered whether other fractures are also associated with these adverse outcomes. The authors estimated the 90-day risks and relative risks of mortality and PE after fractures, and the longer-term relative risks of mortality, for each of ten fracture sites. Using the 5% US standard sample of the Medicare population, we identified 81,181 fractures of the pelvis, patella, and long bones occurring between July 1, 1986, and June 30, 1990, among beneficiaries aged 65 years or older. The fractures were randomly matched by sex, race, and year of birth with a beneficiary who did not suffer that particular fracture, to form ten control cohorts. All patients were followed up until December 31, 1990. The 90-day risk of death was greater than 9% after hip, nonhip femur, and pelvic fractures, with risks relative to their matched control cohorts of 4.7 to 6.7. For white patients with these three fracture types these relative risks decreased substantially with age. White patients who had a hospitalization in the 2 years before hip fracture were more likely to die in the first 90 days than those who had no hospitalization before the hip fracture, but this effect, too, decreased with age. Patients who survived 1 year after most fracture types had no clinically significant excess mortality compared with their surviving controls. For patients with hip, nonhip femur, and pelvis fractures, however, there remained an elevated risk of 1.6 to 1.8, and for patients with proximal humerus fractures the risk ratio was 1.4. All lower-limb fractures carried a higher risk of PE than any upper-limb fracture.

23 keywords...

hide…