The use of social media, including Twitter, expanded rapidly over the past decade providing a platform for quick dissemination of information that is easily and freely accessible. At scientific conferences, researchers often use Twitter to make comments on the material presented attaching snapshots of posters or slide presentations, particularly from ongoing clinical trials not yet reporting results in journals. We provide a proof of concept on the use of Twitter as a means to facilitate the retrieval of this information from medical conferences to complement grey literature searches. We harvested data from two scientific conferences: European Lung Cancer Congress (ELCC) 2019 (Geneva April 10-13) and the ASCO annual congress (Chicago May 31-June 4) using the official conferences hashtags. Data were collected prospectively during the conferences using the R package rtweet. A searchable dataset of tweets was created and tweets containing links to pictures were identified, downloaded and classified to allow later access to the presented material. 1,035 tweets containing 748 pictures were obtained from ELCC 2019, which allowed to retrieve podium or poster presentation material from 15 ongoing clinical trials. A manual review of the contents posted from the 15 presentations showed that previously unpublished data were tweeted from 13 of these. Of the 15 podium presentations (2-14 slides retrieved per presentation), 93% included Kaplan-Meier survival data, 66% response data and 60% baseline characteristics. Analyses of the ASCO 2019 material is ongoing with 59,950 tweets containing 19,584 pictures from more than 120 clinical trials. We were able to gather tweets and identify those which could supplement literature reviews. Information obtained from Twitter must be considered carefully (e.g. tweets may not cover all conference material, rigorous review processes need to be established). Limitations are discussed along with ethical concerns and suggestions for optimising information retrieval are presented.