Table 2 Days and spells of sickness absence for participants (n = 2236) and matched controls (n = 8944) during follow-up
Measure of sickness absenceMean rate/100 person years(95% CI)Years in relation to the intervention
2 years before1 year beforeThe year of the start of intervention1 year after2 years after3 years after4 years after5 years after
Days
    Controls1682 (1641–1724)1.001.001.001.001.001.001.001.00
    Participants1782 (1696–1872)1.161.171.000.931.031.021.121.18
(1.07 to 1.27)(1.08 to 1.28)(0.93 to 1.08)(0.85 to 1.01)(0.94 to 1.13)(0.93 to 1.12)(1.02 to 1.23)(1.07 to 1.30)
Short spells (1–3 days)
    Controls129 (128–130)1.001.001.001.001.001.001.001.00
    Participants144 (142–145)1.151.131.101.101.101.111.121.12
(1.08 to 1.22)(1.06 to 1.19)(1.04 to 1.17)(1.04 to 1.17)(1.04 to 1.17)(1.05 to 1.17)(1.05 to 1.19)(1.05 to 1.19)
All long spells (>3 days)
    Controls76.3 (75.6–77.0)1.001.001.001.001.001.001.001.00
    Participants94.7 (93.2–96.2)1.211.261.251.161.201.221.281.26
(1.12 to 1.30)(1.18 to 1.35)(1.17 to 1.34)(1.09 to 1.24)(1.13 to 1.28)(1.15 to 1.30)(1.20 to 1.36)(1.18 to 1.34)
Very long spells (>21 days)
    Controls14.7 (14.4–15.0)1.001.001.001.001.001.001.001.00
    Participants17.2 (16.6–17.9)1.181.231.031.061.121.111.271.28
(1.01 to 1.37)(1.08 to 1.42)(0.90 to 1.19)(0.93 to 1.21)(0.98 to 1.27)(0.98 to 1.26)(1.12 to 1.44)(1.13 to 1.44)
  • Rate ratios (95% CI) derived from repeated measures Poisson regression analysis. 95% CI, 95% confidence interval.