This item tests a vocabulary distinction that examiners love. Three related terms describe vaccine performance, and each is tied to a specific setting.
Efficacy is measured inside a randomized controlled trial, where participants are selected, doses are guaranteed, and the cold chain is perfect. That protected, idealized estimate here is the higher figure, $90\%$.
Effectiveness is measured when the same vaccine is rolled out through the routine immunization programme, exposed to logistic losses, incomplete coverage, and a heterogeneous population. Because conditions are imperfect, the field value is lower — the stated $85\%$.
A third related term, protective efficacy, is essentially the relative reduction in disease attributable to the vaccine and is usually quoted from the trial alongside efficacy; it should not be confused with the field effectiveness asked for here. So the question maps cleanly: the real-life community figure is the effectiveness. The drop from 90% to 85% reflects programme realities — incomplete cold chain, dropouts who miss doses, and a more varied population than the trial enrolled. The 5% gap or the 87.5% average are distractors with no epidemiological meaning, and 90% is specifically the trial-based efficacy.
\[\boxed{\text{Effectiveness} = 85\%}\]