Table 3

Incidence per 1,000 person-days of different types of musculoskeletal injuries and hazard ratios for changes in incidence between the intervention and control companies during prestudy and study periodsa

Variable

Company

Prestudy period

(n = 508/436)b

Study period

(n = 501/467)b

Age-adjusted HR

(95% CI)

HR adjusted modelc

(95% CI)

Number

Incidence

Number

Incidence


Acute injuries, all

Int

246

3.16

150

2.14

0.74 (0.52 to 1.06)

0.75 (0.51 to 1.09)

Ctrl

149

2.73

155

2.44

Lower extremity

Int

136

1.75

90

1.28

0.84 (0.55 to 1.30)

0.82 (0.52 to 1.31)

Ctrl

91

1.67

96

1.51

Knee

Int

50

0.64

48

0.68

1.05 (0.55 to 2.00)

1.32 (0.65 to 2.67)

Ctrl

35

0.64

38

0.60

Ankle

Int

37

0.48

17

0.24

0.38 (0.17 to 0.86)

0.34 (0.15 to 0.78)

Ctrl

21

0.38

37

0.58

Upper extremity

Int

53

0.68

31

0.44

0.57 (0.28 to 1.16)

0.52 (0.24 to 1.12)

Ctrl

26

0.48

31

0.49

Total number of off-duty daysd

Int

917

11.8

546

7.8

0.46 (0.26 to 0.83)

0.55 (0.29 to 1.04)

Ctrl

419

7.7

677

10.7

Discharged from military servicee

Int

34

0.44

42

0.60

0.78 (0.41 to 1.51)

0.81 (0.42 to 1.57)f

Ctrl

26

0.48

52

0.82

Follow-up days

Int

77,871

70,222

Ctrl

54,620

63,494


aHR, hazard ratio; 95% CI, 95% confidence interval; Int, intervention company; Ctrl, control company. HRs were calculated by using the Cox proportional hazard model if not otherwise mentioned. Statistical significance level was set at P < 0.05. HRs are based on the interaction term of each study group (intervention or control), and study period was entered into the model to analyse the difference in the change in incidence between the groups. bNumber of conscripts in the intervention and control companies per study period; cadjusted for age, urbanisation level of the home residence, smoking, alcohol intake, earlier musculoskeletal symptoms, orthopaedic surgeries, chronic disabilities due to earlier musculoskeletal injuries, school success, previous physical activity, waist circumference and conscript's physical fitness index (n = 11 adjusting variables); dbecause of acute injuries, rate ratio was obtained using a negative binomial model; eafter the 2-week run-in period; fnot adjusted for waist circumference or physical fitness level, since 36 discharged individuals had missing information.

Parkkari et al. BMC Medicine 2011 9:35   doi:10.1186/1741-7015-9-35

Open Data