The Second KDD Workshop on Mining Multiple Information Sources MMIS ′ 08

This work is motivated by the real-world challenge of detecting Adverse Drug Reactions (ADRs) from multiple administrative health databases. ADRs are a leading cause of hospitalisation and death worldwide. Almost all current post-market ADR signalling techniques are based on spontaneous ADR case reports, which significantly underestimate the true incidence. On the other hand, various administrative health data are routinely collected. They, when linked together, would contain evidence of all ADRs. To signal unexpected and infrequent patterns characteristic of ADRs, we proposed the Unexpected Temporal Association Rule and its interestingness measure, unexlev. Its associated mining algorithm, MUTARA, could short-list ADRs from real-world administrative health databases. In this work, we establish a new algorithm, HUNT, for highlighting infrequent and unexpected patterns by comparing their ranks based on unexlev with those based on traditional leverage. Experimental results on real-world databases substantiate that HUNT short-lists more ADRs than MUTARA. HUNT, e.g., not only short-lists the drug alendronate associated with the condition esophagitis as MUTARA does, but also short-lists alendronate with diarrhoea and vomiting for older (age≥60) females. Similar improved performance is found for older males as well as on other drugs. The techniques are promising for post-market drug monitoring based on linked administrative health databases and are included in a health data ∗Huidong Jin is currently with NICTA Canberra Laboratory, Locked Bag 8001, Canberra, ACT 2601, Australia. The work was done when he was with CSIRO. †Jie Chen is currently with SigNav Pty Ltd, Australia. Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee. MMIS ’08, August 24, 2008, Las Vegas, Nevada, USA. Copyright 2008 ACM 978-1-60558-273-3 ...$5.00. mining system delivered to a government agency.