RT Journal Article SR Electronic T1 Evaluation of the Latrobe Valley Better Health Injury Prevention Program JF Injury Prevention JO Inj Prev FD BMJ Publishing Group Ltd SP 66 OP 69 DO 10.1136/ip.7.1.66 VO 7 IS 1 A1 L M Day A1 J Ozanne-Smith A1 E Cassell A1 L Li YR 2001 UL http://injuryprevention.bmj.com/content/7/1/66.abstract AB Objective—To evaluate the Latrobe Valley Better Health Injury Prevention Program, a regional community based intervention in south east Victoria, Australia. Method—The evaluation design was quasiexperimental including pre-intervention and post-intervention observations in a predominately town dwelling population of approximately 76 000. There was no comparison community. Process measures included key informant interviews. Impact evaluation utilised self reported changes in injury risk and protective factors, gathered by a random household telephone survey. Outcome evaluation was based on five years of emergency department injury surveillance data for the Latrobe Valley. Results—The program built strategic partnerships, increasing the emphasis on local safety. Activities were implemented in the targeted areas of home, sport, and playground injuries. Some 47 000 educational contacts were made with the community and at least 6000 resource items distributed. There were significant increases in home safety knowledge. Some changes in the areas of playground and sport safety were achieved after partnership development with relevant agencies. Poisson regression models showed significant decreases in the presentation rate for all home injury and for the more severe home injuries. Conclusion—This study clearly demonstrates the difficulty of conducting robust evaluation in the absence of readily available and reliable data and adequate budgets. The Latrobe Valley Better Health Program activities contributed to structural, environmental, and organisational changes that have the potential to reduce injury. The extent of this contribution beyond that made by the statewide injury prevention strategy is not able to be determined.