Validation of a 4-Item Score Predicting Hip Fracture and Mortality Risk Among Elderly Women

Abstract

PURPOSE One in 4 Swedish women experiences a hip fracture, an event that has high concomitant morbidity and mortality. We developed and validated a clinical predictor of fracture and mortality risk, the Fracture and Mortality (FRAMO) Index.

METHODS This was a population-based prospective cohort study with a baseline questionnaire and 2-year outcomes of hip fracture, fragility fracture, and death. The questionnaire was sent to 1,498 women aged 70 years or older in 3 rural populations, asking them about their age, weight, height, mobility, previous fractures, smoking, medication use, and housing. Some women were also asked about previous vertebral radiographs. We defined 2 risk models before outcome data collection and subsequently renamed 1 model (age =80 years, weight <60 kg, previous fragility fracture, and the need to use arms to rise from the sitting position) the FRAMO Index. We used logistic regression analysis to study the association between the FRAMO Index and outcomes in all participants.

RESULTS The participation rate was 83% in this elderly female population (N = 1,248). The 63% of women with 0 to 1 risk factor had a 2-year hip fracture risk of 0.8% and mortality risk of 3.2%. In contrast, women with 2 to 4 risk factors had a 2-year hip fracture risk of 5.4% (odds ratio = 7.5; 95% confidence interval, 3.0–18.4) and mortality risk of 23.7% (odds ratio = 9.5; 95% confidence interval, 6.0–14.9). These differences remained significant after adjustment for age as a continuous variable. Mortality increased with the number of risk factors. The proportion of women reporting previous vertebral fractures was higher among the group specifically questioned about vertebral radiographs (P <.001).

CONCLUSIONS The FRAMO Index identified the majority of women who experienced hip fractures during a 2-year follow-up, who might have been candidates for intensified preventive measures. The FRAMO Index, based on 4 binary risk factors, would be practical for routine use in primary care.