Challenge 1: Identifying drug-condition associations within an entire observational dataset

There is great interest in finding relationships between drugs and adverse events through retrospective exploration of existing observational databases. Methods that can correctly classify drug-condition relationships as true or false associations could greatly contribute to the current pharmacovigilance practice by providing all key stakeholders additional information to inform therapeutic decision-making. This challenge seeks analysis methods that, given an observational database containing multiple different types of drugs and conditions, can identify the relationship between all drug-condition pairs. Grand Prize: 1st place on March 31, 2010 $10,000 Progress Prize: 1st place on November 30, 2009 $2,500 Progress Prize: 2nd place on November 30, 2009 $1,000 Challeng...

Challenge 2: Identifying drug-condition associations as data accumulates over time

Timely detection of drug-related adverse events as part of an active surveillance system would allow patients and health care providers to minimize potential risks and inform their therapeutic decision-making as quickly as possible. This challenge seeks to evaluate a method’s performance in identifying true drug-condition associations and discerning from false association as data accumulates over time. Grand Prize: 1st place on March 30, 2010 $5,000 Progress Prize: 1st place on November 30, 2009 $1,000 Progress Prize: 2nd place on November 30, 2009 $500 Challenge 2 Description (pdf) OMOP Cup Rules (pdf) Common Data Model Data files (zip files contain tab-separated text files, uncompressed size is about 5 times compressed size): Some unzip utilities do not support l...