Adjunct Professor of Epidemiology
Dr. Walker’s research involves studies of drug utilization and the intended and unintended effects of pharmaceuticals and medical procedures. Essentially all his work is done in collaboration with doctoral students in the pharmacoepidemiology program.
Recently completed studies include an evaluation of the determinants of discontinuation of lipid lowering therapy in two HMOs. Patients’ persistence on therapy was much poorer than had been suggested by clinical trial results. Earlier work on medical devices included an examination of the risks, benefits, and costs of an implanted occlusive device as an alternative to surgical closure of patent ductus arteriosus. The occluder proved to be more costly, less effective, and more burdened with unwanted outcomes than surgery.
A study of the incidence and determinants of osteoarthritis is the only recent work quantifying the incidence of this hard-to-define but common condition. Current studies include exploration of health events in large insurance and HMO data bases using neural networks and other pattern-recognition schemes. Taking hepatitis as an example, Dr. Walker is attempting to select out true cases from among the population of persons with hepatitis-related claims and services.
A common theme of much of the current work of Dr. Walker and his colleagues is the use of very large data bases derived from HMOs, insurance schemes, and governmental health authorities. Ongoing methodologic research is aimed at the efficient and accurate use of these resources.
Dr.P.H., 1980, Harvard School of Public Health