Skip to main content
. 2016 Sep 21;6(9):e011673. doi: 10.1136/bmjopen-2016-011673

Table 5.

Sensitivity analysis of primary outcome at the two follow-up periods

1–6-month follow-up
7–24-month follow-up
Analysis population N Difference in mean adherence 95% CI p Value N Difference in mean adherence 95% CI p Value
Main analysis: all participants with ≥4 months' depot data 108 −6.2% −13.1% to 0.7% 0.078 116 −5.7%* −13.1% to 1.7% 0.130
All participants as above, setting adherence to 100% for those discharged to GP 130 −5.0% −14.7% to 4.7% 0.316 119 −6.2%* −13.6% to 1.1% 0.097
All participants as above but setting adherence to 0% for refusers 112 −4.9% −14.3% to 4.2% 0.312 124 −6.4%† −15.0%—2.1% 0.142

*Simple linear regression model including only a fixed effect for intervention versus control. Clustering by team ignorable as the ICC was −0.05.

†Linear mixed effects model including only a fixed effect for intervention versus control. The model-based ICC was p<0.001. GP, General Practitioner.