Cheryl Kraemer Module 2 Week 2

Kahn, M., Baston, D., & Schilling, L. (2012).  Data Model Considerations for Clinical Effectiveness Researchers.  Medical Care (50)7, S60-S67.

Why was this article, blog, post, or multimedia chosen?

     I chose this article because the term data model was in the title and it is a topic for Module 2.  It describes considerations needed when deciding on a data model for healthcare research.   I also chose it because while I am familiar with the term, until reading this article, I didn’t understand what the term meant.

What makes it interesting, appropriate, or reputable?

     I found it interesting because it explains what data modeling is and considerations for model selection at a level that someone new to data models can understand and use.  It provides examples of data models and examines how they would be used in a clinical research setting but also explains their limitations.  It also differentiates which example is better under certain research circumstances.

Is it an opinion? Case study? Research study? Product review?

     It is a case study about the Scalable Architecture for Federated Translational Inquiries Network (SAFTINet) to create a scalable, distributed network to support Comparative Effectiveness Research.  Comparative Effectiveness Research (CER) studies focus on health outcomes, clinical effectiveness, risk and benefits of medical care in real world clinical settings.  CER uses data from clinical, administrative and billing areas.

What was the need, problem, issue or trend addressed in the article, blog, post, or multimedia?

     The need was to determine a data model that would meet the needs for storage, processing and analysis of a vast array of clinical and financial data.  There need to be a decision whether to develop a model from scratch or adopt and modify an existing data model.  Factored into the decision was the resources need to create a new model and the ability to modify an existing model.

What was the solution for which technology had an answer?

     The decision was to modify an already existing data model as it was felt that it was a better use of resources.  It was also determined to provide a return benefit to the original data model’s community with the new modifications.

What implications might this have in healthcare delivery?

     The implications for healthcare delivery are based upon whether the data model chosen can support the research.  Multidisciplinary teams comprised of informatics professionals, clinical investigators and biostatisticians need to evaluate entity relationship diagrams to determine the data model’s ability to meet the research needs for data storage and querying.  An additional consideration is the quality of the data which is outside the scope of the data model. 

What did you learn from it that might have an application for your practice?

     I learned that are better ways to use data than the way it is currently being done with the data from the systems I support.   I also learned that there are challenges to both creating a data model from scratch and modifying an existing data model and the decision must be thoughtfully considered by both the people creating/modifying and using the data model.  Additionally, futures uses must be considered.  The article provided Moody and Shank’s 8 dimensions to be considered when evaluating a data model which will be useful when reviewing data models.