Study Design. Quasi-experimental, nonrandomized, nonequivalent, parallel group-controlled study involving before-after mailed surveys of general practitioners. Objectives. To measure the magnitude of any sustained change in general practitioner beliefs and stated behavior about back pain 4.5 years after cessation of a media campaign designed to alter population back pain beliefs. S ummary of Background Data. A media campaign providing simple advice about back pain carried out in Victoria, Australia between September 1997 and December 1999 resulted in significant improvements in physicians' beliefs and stated management. Methods. Mailed surveys of general practitioners in Victoria and the adjacent control state, New South Wales, were performed before, immediately after, and 4.5 years after the Victorian campaign. We elicited knowledge and attitudes about back pain by asking respondents to indicate their level of agreement with a set of statements. We also elicited their likely management by presenting 2 hypothetical scenarios. Results. A total of 635 Victorian and 511 NSW GPs completed the latest survey. There were sustained improvements over time in Victorian physicians' beliefs about back pain and their stated behavior: e. g., at Survey 3, compared with baseline, Victorian physicians were 2.0 (95% confidence interval [CI], 1.3-3.0) times as likely as their NSW counterparts to know that patients with low back pain need not wait to be almost pain free to return to work; 1.78 (95% CI, 1.27-2.49) times as likely not to order tests for acute low back pain and 0.47 (95% CI, 0.33-0.69) times as likely to order lumbosacral radiographs. They were also 0.49 (95% CI, 0.34-0.73) times as likely to prescribe bed rest and 1.62 (95% CI, 1.19-2.22) times as likely to advise work modification. Conclusions. A population-based strategy to shift societal views about back pain has had a sustained effect on GP beliefs and stated behavior 4.5 years after its cessation.
