Jeffrey Klann, M.Eng., a Ph.D. candidate in health informatics at the IU School of Informatics and Computing at IUPUI, has been named a 2010 best student paper finalist by the American Medical Informatics Association (AMIA).
Klann, also in his third year as a medical informatics fellow at the Regenstrief Institute, was recognized for his paper titled “A method to compute treatment suggestions from local order-entry data.” The paper details his work to produce condition-specific treatment suggestions based entirely on physician treatment data. Reminder systems have been proven to help physicians take better and more efficient care of their patients, but are few in number because content must be meticulously maintained by local experts.
In his 2010 paper, Klann builds upon his previous research that applied an e-commerce metaphor to physicians’ treatment decisions and recommendations. This research earned Klann best student paper finalist recognition by AMIA in 2009.
Though insightful, e-commerce algorithms do not capture the complexities and nuances of patient care. So instead, Klann’s newest research uses Bayesian networks algorithms to produce condition-specific treatment suggestions. Outcomes were evaluated during one year of physician visits involving pregnancy and were found to produce meaningful, non-trivial and generally correct suggestions that mimicked treatments suggested by an obstetric nurse in 71% of instances.
Klann hopes to continue expanding this basic methodology to support larger and more complex models that can be used to develop general tools for decision and content support, compliance monitoring and even decision support systems driven entirely by local data.