Fig. 2

Generalized linear regression model between vaccination coverage and reproduction number in schools. A generalized linear regression model (solid line) was more predictive of reproduction number (AIC = −74.55) than other regression curves. High vaccine coverage was associated with lower reproduction numbers among schools. Required vaccine coverage was determined in three vaccine efficacy levels (ε = 1.0, 0.75 and 0.5). As efficacy became smaller, the cross point between the regression curve and the dashed line moved rightwards, indicating that lower vaccine efficacy requires higher vaccine coverage in elementary schools