The first thing to recognize is that health care itself needs rethinking. The lack of a patient focus is so pervasive that it even finds its way into the database structures as illustrated in the following real example.
We are building a system to support patient entry of health history. After collecting the data, we have to pass it to the Electronic Medical Records (EMR) system. The EMR specifies the interface structure required to pass the data.
For family medical history, the parent-child relationship in the database tables is structured so that medical condition is the master and person is the detail. This means that your data would look something like this:
Coronary Risk
- Mother – Heart attack at 50
- Father – Heart murmur diagnose at 62
This makes all sorts of sense from the doctor’s perspective. The doctor wants to know what family medical history might contribute to your risk of heart disease.
However, this is not at all the way that a patient thinks. When gathering the patient medical history, the question flow should be driven by questions like: “What medical issues did your mother have?” The family member should be the parent record and the medical conditions are the child records. Using this structure, the data would look like the following:
Mother’s Medical Issues
- Heart Attack at 50
- Type 2 diabetes diagnosed at 42
Of course, it is not difficult to refactor the data to accommodate the EMR data structure, but I find it interesting that the natural structure of medical data from a patient’s perspective is different from how it is structured in the EMR.
Ironically, doctors tell me that they are just as unhappy with the structure of the data from a clinical perspective. They assert that everything in the system is driven by the insurance companies, Medicare, and hospital billing systems. Therefore, the EMRs gather exactly the data that is needed to bill the patient, but the information needed to support treating the patient is frequently missing or hard to find.
Patients, doctors, insurance companies/payers, etc. all want the data (and ultimately the entire EMR system) to be support their perspectives. Unfortunately, the payers seem to have won round one.
Recently I handed to my stateside G.P. the 7 pages of test results I got from my cardiologist in Thailand.
He ignored the broad status reports (“heart size: normal”, “abdominal sonogram: normal” etc.) but photocopied only the past 5 years of bloodwork. There are doubtless reasons why chose as he did, but the data were easily accesible in the hardcopy paper report format, and grouped for easy comparison and trending. As you observe, refactoring the data is relatively easy; user needs look to be easily met.