Table 9

Characteristics of identified opinion leaders (odds ratios with 95% confidence intervals)

Hypothesis

Generic sociometric

Condition-specific sociometric

Generic self-designating

Condition-specific self-designating

Social network related

OLs more likely to belong to professional groups

5.27 (2.38 – 11.65)****

3.90 (1.63 – 9.33)**

1.56 (1.13 – 2.17)**

1.13 (0.79 – 1.58)

OLs more likely to have attended local medical school

1.32 (0.62 – 2.82)

0.41(0.08 – 0.90)***

1.02 (0.65 – 1.54)

0.87 (0.55 – 1.38)

Experience related

OLs more likely to have been qualified longer

1.90 (1.10 – 3.28)**

1.18 (0.64 – 2.20)

0.99 (0.72 – 1.36)

1.20 (0.85 – 1.69)

OLs more likely to be in senior posts

6.69 (2.33 – 19.20) ***

5.72 (1.69 – 19.34)***

2.02 (1.23 – 3.21)***

1.35 (0.85 – 2.15)

Qualifications

OLs more likely to have qualifications

1.05 (0.6 3 – 1.75)

1.27 (0.68 – 2.36)

1.80 (1.33 – 2.44)***

0.96 (0.68 – 1.36)

Other

OLs more likely to spend time teaching

0.88 (0.16 – 4.74)

1.35 (0.31 – 5.98)

.93 (0.79 – 4.67)

0.92 (0.34 – 2.50)

OLs more likely to spend time on research

2.30 (0.49 – 10.92)

1.82 (0.41 – 8.11)

2.14 (0.86 – 5.34)

1.10 (0.40 – 3.04)

Keeping up to date score

Mean Opinion Leader Score

3.57

3.47

3.48

3.40

Mean score of other respondents

3.29

3.30

3.25

3.27

Mean difference in up-to-date score

0.28

0.17

0.23

0.13

95% CI and significance⁺

(0.14 – 0.43)**

(-0.09 – 0.36)

(0.14 – 0.32)***

(0.03 – 0.24)*

Use of clinical effectiveness materials score

Mean Opinion Leader Score

2.58

2.37

2.53

2.58

Mean score of other respondents

2.38

2.42

2.36

2.38

Mean difference in up-to-date score

0.3

-0.05

0.17

0.20

95% CI and significance⁺

(-0.02 – 0.41)

(-0.33 – 0.21)

(0.04 – 0.30)*

(0.04 – 0.30)*

Key – * – p < 0.05, ** – p < 0.01, *** – p < 0.001, **** – p < 0.0001, + Independent samples t-test

Grimshaw et al. Implementation Science 2006 1:3   doi:10.1186/1748-5908-1-3