Population burden of injury Associate Professor Belinda Gabbe
Population burden of injury Associate Professor Belinda Gabbe School of Public Health and Preventive Medicine Monash University
Introduction Improving measure of population burden has been on the ICE agenda for recent meetings Limitations to the GBD approach to measuring population injury burden Maximising use of existing knowledge and data seen as a priority
Key issues 1. Disability weights a. b. c. d. 2. 3. 4. 5. Panel vs. empirical data Children Proxy LMIC Combining data from different studies Use of ED and inpatient data Injury grouping/classification ICD-11
Recent developments ICE driven manuscript in 2012 edition of Epidemiological Reviews Polinder et al. Measuring the population burden of fatal and non-fatal injury. Epidemiol Rev 2012; 34: 17 -31 Major competitive grant NHMRC Project Grant 2012 -2013
NHMRC Project Grant 2012 -2013 Improving the measurement of non-fatal injury burden – Validating the GBD methods
Investigative team Chief Investigators Belinda Gabbe (Monash University), Ronan Lyons (Swansea University), James Harrison (Flinders University), Fred Rivara (University of Washington), Shanthi Ameratunga (University of Auckland), Suzanne Polinder (Erasmus MC), Sarah Derrett (University of Otago), Damien Jolley (Monash University) Associate Investigators Kavi Bhalla (Johns Hopkins), Theo Vos (University of Queensland), Clare Bradley (Flinders University), Juanita Haagsma (Erasmus MC), John Langley (University of Otago), Gabrielle Davie (University of Otago) Project Officer Pam Simpson (Monash University)
Project aims i. Evaluate the influence of existing and alternative classifications of injuries on disability burden estimates. ii. Identify functional outcome and recovery patterns of multiple injuries and the impact on disability outcomes. iii. Evaluate methods for integrating the functional outcome associated with multiple injuries into burden of injury measurements. iv. Establish the duration of disability, and the percentage with persisting disability, for injury groupings based on empirical data, in order to validate the injury outcome estimates on which DALY values from the original and current GBD projects depend. v. Establish the comparability of the burden of non-fatal injury (Years Lived with Disability) estimates based on different methodologies (i. e. different definitions of health states, disability weights, and measures of the duration of injuries). vi. Develop alternative metrics for the measurement of population injury burden
Datasets – adult participants Study Setting Inclusion criteria Participants Follow-up time points Outcome measures NSCOT US At least one AIS injury severity>2 N=3958 3 and 12 months 30/90/365 day mortality 18 -84 years VSTR Australia ISS>15, ICU admission or urgent surgery GOS, SF-36, SF-12, SF-6 D (3 and 12 -months) FCI, MFA, SIP cognitive, PCLC and CESD (12 -months) N=8213 6, 12 and 24 months GOSE, SF-12, SF-6 D, EQ-5 D, Peds. QL N=15459 6 and 12 months GOSE, SF-12, SF-6 D, EQ-5 D N=8014 2. 5, 5, 9 and 24 -months EQ-5 D N=1219 1, 4 and 12 months EQ-5 D or Peds. QL, HUI 3 n=2856 3, 12 and 24 months WHODAS II, EQ-5 D All ages VOTOR Australia Orthopaedic injury admission >24 hours 15+ years Dutch Netherlands Presentation to ED Adults UKBOI UK Presentation to ED or hospital admission 5+ years POIS NZ ACC claim 18 -64 years
Brief profile Study Mean (SD) age % Male Key causes NSCOT 47. 6 (20. 4) years 65% Road traffic injury (44%), Falls (36%) VSTR 48. 4 (21. 7) years 73% Road traffic injury (46%), Falls (32%) VOTOR 57. 0 (23. 0) years 50% Falls (62%), Road traffic injury (25%) Dutch 49. 5 (21. 6) years 54% Home/leisure (54%), Road traffic injury (24%) UKBOI 48. 2 (21. 0) years 52% Road traffic injury (10%) POIS 41. 1 (13. 0) years 61% Road traffic injury (7%)
EUROCOST 39 categories Study Top 5 EUROCOST 39 health states NSCOT Skull/brain, Internal organ, Fractured pelvis, Fractured femur shaft, Fractured knee/lower leg VSTR Skull/brain, Internal organ, Fractured rib/sternum, Fractured/dislocation/sprain/strain vertebrae/spine, Spinal cord injury VOTOR Fractured hip, Fractured knee/lower leg, Fractured wrist, Fractured/dislocation/sprain/strain vertebrae/spine, Fractured ankle Dutch Superficial injury, Skull/brain, Fractured hip, Concussion, Fractured rib/sternum UKBOI Fractured knee/lower leg, Superficial injury, Open wounds, Fractured wrist, Fractured foot/toes POIS Fractured/dislocation/sprain/strain vertebrae/spine, Dislocation/strain/sprain knee, Open wound, Dislocation/strain/sprain ankle/foot, Superficial injury
EQ-5 D 10000 9000 Number of observations 8000 7000 UKBOI 6000 VSTR 5000 VOTOR 4000 Dutch 3000 POIS 2000 1000 0 1 -month 3 -months 6 -months 9 -months 12 -months 24 -months
GOS-E 16000 Number of observations 14000 12000 10000 VSTR 8000 VOTOR NSCOT 6000 4000 2000 0 3 -months 6 -months 12 -months 24 -months
SF-12 9000 Number of observations 8000 7000 6000 5000 VSTR 4000 VOTOR NSCOT 3000 2000 1000 0 3 -months 6 -months 12 -months 24 -months
Program of work to be completed Classification Mapping to 1990 GBD, 2010 GBD (GBD-IEG), 2010 GBD (final), other existing Disability measures Prevalence of disability Impact of cut-offs on estimates “Cross-walks” New weights Application of alternative classifications, weight, durations to incidence and prevalence data Compare YLD and DALY estimates
Expected outcomes Improved understanding of the relationship between disability measures Improved disability weights based on empirical data Improved classification of injury health states for burden studies Improved understanding of the duration of disability experienced by survivors Recommended methodology for injury burden studies Inform future GBD studies
Key remaining challenges Paediatric injury outcomes, durations and weights (limited data in the study at present) Need for disability data from LMIC to better inform burden estimates Ongoing need for disability data from non-hospitalised injury cohorts (limited data in the study at present)
Acknowledgements National Health and Medical Research Council (NHMRC) of Australia Australasian Epidemiological Association (AEA) for funding BG attendance at ICE
- Slides: 17