Avoiding Duplicate Code |
It makes sense to have multiple tables in each message's table grammar, rather than having one monolithic table for each message. This makes mapping easier since you can re-use mappings from message to message, as well as making it possible to reuse code in your application more easily.
Since each table is an object, you can take advantage of this at the application level and write common functions that will take a table and process it. For example, if you have a header table in every message, you can write your code in a manner that every message event calls a single common function for processing the header fields.
Programming in this manner should result in a very lean implementation.
 | ||
|
Multiple IDs Per Patient One question that comes up quite often is how to deal with patients that have multiple IDs. This situation arises frequently within the US health system because different independent hospitals that might have treated the same patient are amalgamated into a larger group. The CDC also has to deal with this problem since they collect patient data from a large number of different organizations. A good technique to deal with this problem is to create Patient ID aliases table in your application. This maps alternate IDs to the specific patient in your system. In this way your application can support many patient IDs for each patient (or person if you want to get really correct in your data modelling). | ||
 |