Rebounding case notifications of chlamydia : an epidemiological game of 'Clue'?