You're not far off. Not an epidemiologist, but I think these sorts of things generally follow a logistic (S-shaped or sigmoid) curve (1/1+e^-x). Initial growth is approximately exponential, then you hit a point of saturation, growth slows, and eventually levels off.