interpretation of a negative risk difference

#1
Hi,
I am reviewing a cohort study and trying to make sense of the results. When considering mortality they presented their results as a risk difference which is a little different to what I normally see. Usually reported is RR or OR. In this case the results give a negative number. -35% (95%CI -50 to -21). (risk 100 of 238 with intervention, 31 of 40 with control). Am I to interpret this that there is a 35% absolute risk reduction in mortality with receiving the intervention? Ignore it's negative as simply meaning falling on the intervention side where 1 is the line of no significance? I know the confidence intervals are broad and probably show a lack of precision in measuring the treatment effect.
 

hlsmith

Not a robit
#2
So instead of dividing risks they subtract them, common though not as common as RR.

CI needs to exclude "0" since it is subtraction instead of division!