Article Title

Factors Associated With Longevity Among Oldest-Old Male Patients With Hip Fractures

Publication Date



oldest-old male patients, chronic conditions


Background/Aims: Few studies address the management and mortality of oldest-old male patients with hip fractures, although such events are increasing. This study identified factors associated with survival among male patients with surgically repaired hip fractures applying classification techniques to multiple chronic conditions (MCC), focusing on differences between patients ≥ 85 years old and younger patients age 50–84.

Methods: Data from the Veterans Health Administration identified 4,059 male patients age 50–100 with both hip fracture diagnosis and repair (index event) during fiscal years 2005–2009 (i.e. October 2004–September 2009). MCC was defined by Charlson and Elixhauser comorbidity conditions after eliminating overlapping diagnoses. A preliminary Cox regression model examined the association of MCC with survival. Latent class analysis produced classes using MCC, medications, use of physical therapy and demographic covariates (age, race/ethnicity, marital status, VA Priority denoting highly disabled). Cox proportional hazards modeled association between class and 5-year survival.

Results: The sample included 942 patients age ≥ 85 (23%). Survival censored at 5 years postoperation averaged 1,030 days, or almost 3 years. The oldest patients were more likely to be married (48% vs. 41%) and to die within 6 months (36% vs. 17% younger); minority race/ethnicity was similar, as was mean MCC (3.9 [SD: 2.8; median: 4.0] vs. 3.9 [SD: 2.7; median: 4.0]). Preoperative physical therapy was equivalent (24% vs. 24%), but postoperative physical therapy was less common for those age ≥ 85 (20% vs. 32%). Oldest patients also averaged fewer medication classes (5.1 vs. 5.5 preop; 6.4 vs. 7.4 postop). In a preliminary adjusted model, mortality risk increased by 13% per additional MCC (hazard ratio: 1.13, 95% confidence interval: 1.11–1.15). After latent class analysis with three classes identified, 43% of oldest-old were classified primarily white and healthier, 43% white and sicker with more medication classes and 14% primarily black with alcohol abuse. Similar trends were observed for younger patients (36%, 53% and 11%, respectively).

Conclusion: Although the oldest-old male population represents a relatively small proportion of patients, their longevity may be related to various health care factors. Close investigation of classification of individuals with MCC and hip fractures may help target interventions within health care settings and contribute to healthy aging.