The intended early contact (within first week) of workers absent with musculoskeletal disorders only occurred at one experimental site; the control sites had no procedure for early contact. Absence rates improved over the four years at the intervention sites compared with the control sites: a decrease of 2.0 v an increase of 0.9 days/1000 working hours. The median return-to-work time for early intervention compared with controls was 4 days v 5 days (P=NS). Considering return-to-work time irrespective of whether the intervention was delivered early or late, the median durations were also 4 days v 5 days (P<
0.05). When looking at work retention over 12 months, the median duration of subsequent absence for early intervention was 5 days compared with 11 days for controls (P=NS). For the larger number of workers receiving a late intervention, the median duration of subsequent absence was median 4 days v 11 days for controls (P<
0.05).
Previously defined cut-off scores were used to categorise hypothesised risk; scores beyond the cut-off point were considered detrimental, and the ‘flag’ was considered to be ‘flying’. Odds ratios (OR) were calculated to explore the association between the flags and taking sick leave; a statistically significant association was found with ORs between 1.5 and 2.9. The cut-off scores were then used to compare the length of absence between workers who had zero flags flying and those who had one or more flags flying. Absence over the ensuing 15 months was significantly longer for those people who had one or more flags flying (mean 10.6 days compared with 6.1 days, P<
0.05). There was a trend for longer absence with more flags flying.