PMean: A quick summary of my research

I might be giving a very brief (5 minute) overview of my research for students in the Department of Biomedical and Health Informatics. Here are some details of that work, with links if anyone wants to dig deeper.

I am interested in research about the research process, and one very simple thing that would be easy to replicate would be an audit that I made of IRB approvals at Childrens Mercy Hospital over a five year period. I found that more than half of the studies failed to recruit the number of patients that they promised to get. I never got this research published, but it would be pretty easy to do so. I do have a brief summary on my website.

A big reason for the shortfall on sample size is that it takes longer to get patients than you might think. I have built some Bayesian models that can forecast how long it might take to accrue the required number of patients. Details are available on my website and my blog.

I am working on mining the electronic health record at KUMC and while that research is proprietary, it would not be too hard to replicate the work on Health Facts. Here is an early account of what I have been working on, and I hope to have a publication soon based on this work.

There are a couple of other studies that I ran using the NHAMCS database of CDC that could very easily be replicated on Health Facts. One is a look at the care of children with asthma who report to the Emergency Room and are sometimes given bad therapies or not given good therapies.

Another is the 10 most common Dermatology diagnoses that a pediatrician is likely to see. The link in Pubmed is not very helpful because there is no abstract and the article is behind a paywall, but I can send more details if you are interested.

I am also interested in developing case studies in research ethics. Here is an example.