SLR - January 2016 - Donny Sellenriek

Title: Morbidity and Mortality Associated with Geriatric Ankle Fractures: A Medicare Part A Claims Database Analysis

Reference: Hsu RY, Lee Y, Hayda R, DiGiovanni CW, Mor V, Bariteau JT. Morbidity and Mortality Associated with Geriatric Ankle Fractures: A Medicare Part A Claims Database Analysis J Bone Joint Surg Am. 2015 Nov 4;97(21):1748-55.
 
Scientific Literature Review
 
Reviewed By: Donny Sellenriek, DPM
Residency Program: Kaiser Permanente, North Bay Consortium
 
Podiatric Relevance: Ankle fractures are the third most common extremity fracture in the United States in people 65 or older. According to the US Census bureau, the geriatric population is expected to almost double in size by 2050, reaching an estimated 83.7 million. With a high likelihood that podiatrists will manage an increasing number of geriatric ankle fractures, a better understanding of this patient population is meaningful to our practices. The goal of the study was to better understand the mortality and morbidity of geriatric ankle fracture patients that required inpatient admission and compare that cohort to a hip fracture cohort and a control group.  

Methods: One hundred percent of the 2008 Medicare Part A claims were screened by the authors for new diagnosis of ankle fracture and hip fracture. Anyone under 65 and anyone that sustained the same fractures the preceding year were excluded. The control group consisted of all other Medicare inpatient admissions in 2008 for any reason, other than ankle or hip fracture.  

Primary outcome was the one year post-injury mortality rate of the ankle fracture cohort, the hip fracture cohort, and the control group. Secondary outcome measures were length of inpatient stay, discharge to nursing home, readmissions and medical complications. Medical complications included: surgical site infection, deep vein thrombosis, pulmonary embolism, congestive heart failure, pneumonia, urinary tract infection, pressure ulcer, myocardial infarction, Clostridium difficile infection, and gastrointestinal bleeds.

An age subgroup analysis was performed. Pre-existing health status was determined with the mean Elixhauser and Charlson-Deyo comorbidity indices, although, only the Elixhauser score was used for calculating comorbidities. Cox regression analysis was used to find the relative contribution of the injury to one-year mortality, independent of age or comorbidities.  

Results:  The ankle fracture cohort had 19,648 patients. The hip fracture cohort had 193,980 patients. One-hundred-eighty patients sustained both ankle and hip fractures and were included in both cohorts. The control group was comprised of all other admissions except hip and ankle fractures and had 5,801,831 patients. They found a statistical significant difference (p < 0.001) in mean age, sex and comorbidity burden between these groups. One year mortality was 11.9 percent for the ankle fracture cohort, 28.2 percent for the hip fracture cohort, and 21.5 percent for the control group without standardizing for age and comorbidities.

The mean age for the ankle fracture cohort was 77.5 with an age group breakdown as follows: 42.7 percent between ages 65 and 74, 37.4 percent between ages 75 – 84, 18.2 percent between ages 85 – 94, and 1.7 percent from age 95 and older. The hip fracture cohort mean age was 83.6 with an age group breakdown as follows: 14.6 percent between ages 65 and 74, 38.9 percent between ages 75 – 84, 41.2 percent between ages 85 – 94, and 5.3 percent from age 95 and older. The control group mean age was 78.9 with an age group breakdown as follows: 35.9 percent between ages 65 and 74, 39.3 percent between ages 75 – 84, 22.4 percent between ages 85 – 94, and 2.4 percent from age 95 and older. The ankle fracture and hip fracture cohorts had similar gender distributions but there was a statistically significant difference between them. The gender distribution was 21.6 percent male and 78.4 percent female for the ankle fracture cohort and 26.6 percent male and 73.4 percent female for the hip fracture cohort. The control group was 42.4 percent male and 57.6 percent female. The mean Elixhauser scores were 2.3 for the ankle fracture cohort, 2.5 for the hip fracture cohort and 2.4 for the control group.  

Hazard ratio for one year mortality after admission for the ankle fracture was 0.520 when compared to the control group. The hazard ratio for one year mortality after admission for a hip fracture was 1.361 when compared to the control group. After standardizing for age and comorbidities the hazard ratio for hip fractures was 1.081 and for ankle fractures was 0.557, with the control group being 1.  

As far as secondary outcomes, ankle fractures had statistically significant less morbidity in all measures of the 90 day complications except surgical site infection and deep vein thrombosis. The ankle fracture cohort surgical site infection rate was 3.6 percent verse 1.5 percent for the hip fracture cohort. The rate of DVT complication was equivalent for the ankle and hip fracture cohorts.  

The ankle fracture cohort had: a shorter hospital stay at 4.6 days versus 6.0 days for the hip fracture cohort, a significantly lower discharge to nursing home at 59.2 percent verse 71.4 percent for the hip fracture cohort, lower thirty day readmission rate at 13.3 percent verse 15.0 percent for the hip fracture cohort, and lower 30 day mortality at 1.9 percent verse 6.8 percent for the hip fracture cohort.  

Conclusions:
The study showed the ankle fracture cohort was younger, had a higher prevalence in females, had less comorbidities, and had less complications than the hip fracture cohort. According to the author, hip fractures may represent a sentinel event that marks decline, whereas ankle fracture tend to demonstrate an active lifestyle. The authors suggested that this and the significantly lower one-year mortality rate may warrant more aggressive and definitive management of geriatric ankle fractures.  

The higher surgical site infection rate of the ankle fracture cohort is not surprising given the tenuous soft tissue envelope of the ankle. However, the study was not designed to better understand the cause of the higher surgical site infections though; one of the limitations of this study. This is an area that could benefit from more research to better characterize risk factors unique to the geriatric population compared to a younger population.  

This study identified a need for more research directed at identifying potential areas for more aggressive management that could better serve this patient population. 

Educational Opportunities

Upcoming