OBJECTIVES:  Therapies for oncology often effect time-to-event statistics like overall survival and progression free survival. In the literature, these time-to-event statistics are summarized by median time-to-event, percentage of people having had the event at a specific point in time and hazard rates, among others. Although pooling based on individual patient data would be preferred, we still encounter situations where only aggregated data is available. For these situations, the hazard ratios can be pooled. However, many high quality papers would be ignored, if publications were only regarded once the hazard ratio is presented. We therefore searched for methods to transform the different outcomes to the same scale and to pool based on as much information as possible. METHODS: A review was performed with respect to the pooling of different time-to-event outcomes. Mixed Treatment Comparisons were performed using the methods to assess their usability. RESULTS: For cost-effectiveness models, a distribution (exponential, weibull, among others) of the time-to-event statistic is often used to obtain the average of the time-to-event statistic for the comparator arm. This distribution can also be used to transform the median and the 'percentage having had an event at a specific time point' into hazard rates. Comparing the transformed medians with the hazard rates for publications in which both are presented implied a way to check the validity of assumptions. CONCLUSIONS: Although the way information about time-to-event statistics is presented may differ across publications, it is often possible to pool the different types of information. This implied the inclusion of papers which otherwise couldn’t be used and a reduction of the uncertainty in the cost-effectiveness outcomes.