College campuses are vulnerable to infectious disease outbreaks, and there is an urgent need to develop better strategies to mitigate their size and duration, particularly as educational institutions around the world adapt to in-person instruction during the COVID-19 pandemic. Towards addressing this need, we applied a stochastic compartmental model to quantify the impact of university-level responses to contain a mumps outbreak at Harvard University in 2016. We used our model to determine which containment interventions were most effective and study alternative scenarios without and with earlier interventions. This model allows for stochastic variation in small populations, missing or unobserved case data and changes in disease transmission rates post-intervention. The results suggest that control measures implemented by the University's Health Services, including rapid isolation of suspected cases, were very effective at containing the outbreak. Without those measures, the outbreak could have been four times larger. More generally, we conclude that universities should apply (i) diagnostic protocols that address false negatives from molecular tests and (ii) strict quarantine policies to contain the spread of easily transmissible infectious diseases such as mumps among their students. This modelling approach could be applied to data from other outbreaks in college campuses and similar small population settings.

College campuses are highly vulnerable to infectious disease outbreaks, and there is a pressing need to develop better strategies to mitigate their size and duration, particularly as educational institutions around the world reopen to in-person instruction in the midst of the COVID-19 pandemic. Towards addressing this need, we applied a stochastic compartmental model to quantify the impact of university-level responses to past mumps outbreaks in college campuses and used it to determine which control interventions are most effective. Mumps is a very relevant disease in such settings, given its airborne mode of transmission, high infectivity, and recurrence of outbreaks despite availability of a vaccine. Our model aims to simultaneously overcome three crucial issues: stochastic variation in small populations, missing or unobserved case data, and changes in disease transmission rates post-intervention. We tested the model and assessed various interventions using data from the 2014 and 2016 mumps outbreaks at Ohio State University and Harvard University, respectively. Our results suggest that in order to decrease infectious disease incidence on their campuses, universities should apply diagnostic protocols that address false negatives from molecular tests, stricter quarantine policies, and effective awareness campaigns among their students and staff. Our model can be applied to data from other outbreaks in college campuses and similar small-population settings.