Forecasts for acute respiratory infection (ARI) consultation rate time-series.

<p>Observed ARI consultation rate time-series are depicted as black lines while red and blue traces correspond to contemporaneous 2- and 3-month horizon forecasts, respectively; their prediction interval bounds are symbolized by dots of the same colors. Forecasts and prediction interval bounds are calculated with a bootstrap-coupled seasonal multiplicative Holt-Winters method (<a href="http://www.plosone.org/article/info:doi/10.1371/journal.pone.0001181#s4" target="_blank"><i>Methods</i></a>). In each panel, the abscissa spans 102 months (01/1996–06/2004) while the ordinate represents the number of newly diagnosed cases <i>per</i> 1000 age category-specific individuals. Panel A: 0–11 months; Panel B: 1–4 years; Panel C: 5–15 years; and, Panel D: >15 years. Forecasts deteriorate slightly towards older age categories owing to seasonality attenuation. Of note, age category-specific 2- and 3-month horizon ARI consultation rate forecasts roughly overlap because of slowly shifting level and negligible trend time-series components. Consequently, 2- and 3-month horizon 95% prediction interval bounds also overlap.