Education to prevent low back injuries changes knowledge of safe behaviour but not injury incidence or cost

Objective. To determine the efficacy of an educational program designed to prevent low back injury in postal workers. Design. Randomised controlled trial. Setting. Two mail processing facilities in the United States of America. Subjects. Approximately 4000 postal workers took part in the study over a 5.5 year period. Intervention. The 34 work teams at the mail processing facilities were randomised so that all mail handlers and clerks in a team either received training or no training. The training sessions were mandatory and occurred on paid company time. The educational intervention was conducted by physical therapists who taught the workers in groups of 10-12. The educational package consisted of classes on back safety, correct lifting and handling, posture, exercise and pain management. In addition the physical therapists visited each work station with the workers and supervisors and suggested physical and procedural modifications to the work environment. The therapists provided reinforcement training six months after the initial visit, and then annually. Main outcome measures. Low back and other musculoskeletal injury rates were expressed as injuries per 1000 worker years of exposure. Work loss due to injury was expressed by the proportion of injuries that resulted in lost work days and time elapsed until return to work. Costs of care were expressed in dollars per back injury. Data were extracted from worker's compensation claims or postal service accident reports. Main results. There were 183 low back injuries in the intervention group and 177 in the control group. Intervention group units had a higher rate of injury than untrained units, however the difference was not statistically significant for back injuries (rate ratio, 1.11; 95 per cent confidence interval, 0.90 to 1.37) or other musculoskeletal injuries (rate ratio, 1.15; 95 per cent confidence interval, 0.93 to 1.41). There was little difference in the proportion of total injuries that resulted in lost workdays in the intervention group (61 per cent) as compared with the control group (56 per cent). The survival analysis model of time elapsed until return to work found no significant effect associated with training. The median total cost per back injury was $309 in the training group and $103 in the control group, however this difference was not statistically significant. Conclusion. The educational program to prevent low back injuries had no effect on injury rates, work loss due to injury or injury costs.