The availability of data from electronic medical records, claims, smart phones is transforming health and biomedical research. And, as treatment strategies and health care interventions become increasingly complex, the need to develop new methods to extract meaningful knowledge from the analysis of these data could not be greater.
Treatment strategies and health care interventions are becoming increasing complex and the need to compare these strategies on patient outcomes (e.g. survival and re-hospitalization) cannot be greater. Often it is infeasible to compare these complex treatments in randomized clinical trials (RCT). This is because RCTs cannot include a large enough sample of the population, are too costly and time consuming, and the treatments and health care interventions to be compared are too complex to implement a randomization.
More recently, data from electronic medical records, cancer registries and claims and are more available. These data include large populations (e.g. Part A Medicare include 95% of the entire population>65), they allow us to compare complex treatment and health care strategies. However, extracting new knowledge from the analysis of these data present many challenges. For example, these are observational data and therefore methods for confounding adjustment must be developed to assure that a large set of potential confounders are balanced between treatment groups. Second, often there is a misclassification in the treatment assignment, that is the incorrect diagnostic or procedural code is used.
Specifically she in interested in the following research topics:
- Accounting for model uncertainty in propensity score modeling
- Accounting for misclassification of treatment assignment
- Combining heterogeneous databases for estimating average causal effect
- Validating each newly developed method by a side-to-side comparison with randomized clinical trials