Transition to fracture and mortality: development of a risk assessment tool from two large population-based prospective cohort studies — ASN Events
  1. Schedule
  2. Welcome Reception & ANZBMS Plenary Posters
  3. Transition to fracture and mortality: development of a risk assessment tool from two large population-based prospective cohort studies

Transition to fracture and mortality: development of a risk assessment tool from two large population-based prospective cohort studies (#120)

Thach Tran 1 , Dana Bliuc 1 , Hanh Pham 1 , Tineke van Geel 2 , Jonathan D Adachi 3 , Claudie Berger 4 , Joop van den Bergh 5 , John Eisman 1 6 7 8 , Piet Geusens 9 , David Goltzman 10 , David A Hanley 11 , Robert G Josse 12 , Stephanie M Kaiser 13 , Christopher Kovacs 14 , Lisa Langsetmo 15 , Jerilynn C Prior 16 , Tuan V Nguyen 1 6 , Jacqueline R Center 1 6 8
  1. Garvan Institute of Medical Research, Darlinghurst, NSW, Australia
  2. Department of Family Medicine, Maastricht University, Research School CAPHRI, Maastricht, The Netherlands
  3. Department of Medicine, McMaster University, Hamilton, Ontario, Cadana
  4. CaMos National Coordinating centre, McGill University, Montreal, Quebec, Canada
  5. Department of Internal Medicine, Maastricht University Medical Centre, Research school Nutrim, Maastricht, Netherlands
  6. Faculty of Medicine, University of New South Wales, Sydney, Australia
  7. School of Medicine Sydney, University of Notre Dame Australia, Sydney, Australia
  8. Clinical School, St Vincent's Hospital, Sydney, Australia
  9. Department of Internal Medicine, Maastricht University, Research School CAPHRI, Maastricht, The Netherlands
  10. Department of Medicine, McGill University, Montreal, Quebec, Canada
  11. Department of Medicine, University of Calgary, Calgary, Alberta, Canada
  12. Department of Medicine, University of Toronto, Toronto, Ontario, Canada
  13. Department of Medicine, Dalhousie University, Halifax, Nova Scotia, Canada
  14. Faculty of Medicine, Memorial University, St. John's, Newfoundland, Canada
  15. School of Public Health, University of Minnesota, Twin cities, Minneapolis, United States
  16. Department of Medicine and Endocrinology, University of British Columbia, Vancouver, British Columbia, Canada

Existing fracture risk assessment models are not designed to predict fracture-associated consequences. We aimed to develop a predictive model for individualisation of progression from no fracture to fracture, re-fracture and mortality according to different comorbidity risk profiles using the multistate Markov model.

There were 11,000 people (70% women) aged 67.5 (±9) years from Dubbo Osteoporosis Epidemiology Study and Canadian Multicentre Osteoporosis Study. Incident fracture was identified from X-ray reports and questionnaires, and death ascertained though contact with a member family or obituary review. During a median follow up of 12.5 years (IQR: 5.8, 15.0), 2,500 individuals fractured (28/1,000 person-years in women, 15/1,000 person-years in men), 700 re-fractured [65/1,000 person-years (women), 44/1,000 person-years (men)] and 2,800 died [2.1/100 person-years (women), 4.4/100 person-years (men)]. The mean age of initial fracture, re-fracture and mortality was 75 (±5), 78.5 (±8) and 81 (±9) years, respectively. Predictive models included age, BMD, prior falls, prior fracture and comorbidities. In general, the more severe the initial fracture, the less the comorbidities added to subsequent risk of re-fracture and mortality. A 70-year old low-risk woman (defined as T-score = -1.5 with no comorbidities) with an initial hip fracture had a probability of 8% of re-fracture and 53% of dying within 5 years compared with 5% re-fracture and 71% mortality probability for a similarly aged high-risk woman (T-score =-2.5 with history of falls, prior  fracture, cardiovascular disease and diabetes). By contrast a low and high-risk woman with an initial distal fracture had re-fracture probabilities of 12% and 20%, and mortality probabilities of 21% and 44%, respectively. Similar trends were found for men.

This study used a novel, robust technique to develop predictive models for individualisation of progression to fracture and its outcomes which will allow informed decision making about risk and thus treatment for individuals with different risk profiles.

58c21e4fe39c9-Predictive+models+for+progression+to+fracture+and+its+outcomes-Table.jpg