Study design: Cross-sectional questionnaire-based workforce survey together with collection of retrospective data on work absence.
Objectives: To determine if psychosocial ‘blue flags’ are related to back pain and/or sickness absence due to back pain.
Summary of background: The original description of the psychosocial ‘yellow flags’ for back pain chronicity included a mixture of individual psychological parameters and parameters related to perceptions about work and the workplace. It has recently been suggested that these latter parameters should be considered separate and distinct from the individual parameters , and can be termed ‘blue flags’. To date, however, there has been no attempt to explore the specific relationship between the blue and yellow flags or their relative relationship to symptoms and disability.
Methods: The workforce of a large multi-site company was invited to complete a booklet of questionnaires, which included the standard Nordic instrument for obtaining back pain data, and specific instruments to obtain data on ‘yellow’ and ‘blue’ psychosocial flags. The blue flags included psychosocial aspects of work, attribution and elements from the demand/control model, with psychological distress used as a yellow flag comparator. Of the 7,500 workers, 60% responded. Sickness absence records identified workers who had taken absence for back pain. The exploration of the data involved determining statistically significant relationships between psychosocial scores and both back pain history and absence. Appropriate statistical procedures were then used to establish cut-off points for the psychosocial variables. Odds ratios were calculated for two particular outcome variables: self-reported back pain in the previous 12 months and recorded absence over the same period.
Results: Cut-off points were established for each variable, along with the odds ratio (OR) that this score or a score above or below (depending on the scale direction) is associated with reports of back pain or absence. The ORs for psychological distress were 1.9 and 2.4 respectively for LBP and absence in the last 12 months. The ORs for the blue flag variables varied from 1.1 to 1.5 for LBP and from 1.8 to 3.2 for absence.
Conclusions: The psychosocial blue flags reported here are statistically significantly related both to reported back pain and absence. The effect size is less than that for distress in respect of back pain, but variously higher and lower for absence. Whilst prospective studies are needed to determine cause/effect, the results offer tentative support for the suggestion that blue flags should be addressed in clinical interventions.
