How do I implement one?

The current state of the technology does not offer a clear path to a solid solution. Numerous vendors offer products that include most of the principal functions. Where they fall short is generally a combination of the state of technical evolution (certain pieces are still too expensive or not fully understood) or social conditions (legislative matters regarding privacy and so on are still unclear). Several factors do influence the successful implementation of a computer-based patient record based on existing technology.

Planning	To keep a computer-based patient record project from turning into an expensive death march requires a lot of up front planning. Successful plans will usually have many (or most) of the following characteristics: A clearly categorized sequence of events in the form of a structured project plan. Clearly defined timelines and milestones so that you know what you're supposed to have and by when. It's always easier to measure progress by the delivery of "things" rather than by a diet of presentation slides. An identified project team including any resources retained on a temporary basis from outside companies. Specifically identified process owners, project owners, and task owners. (A responsibility assigned to a group or department will probably not get done.) A staffing approach that identifies what types of skills in which quantities will be necessary over the duration. An assessment of operational impact—for example, do you have enough electricity in your computer room? Must you buy new workstations? A detailed training plan that describes who needs to know how much of what and by when. Don't be misled by claims of "turnkey" software. Most of these products require sophisticated database skills and solid LAN expertise. Management support for changes to the current business workflow. (For example, changes in how patients or members are registered will almost always be necessary to satisfy data quality requirements.) A list of required interfaces, the timeframes in which they will be complete, and specific vendor commitments where necessary. (Interfaces are among the most frequent causes of delay.) Make sure you really have enough money to complete the project.
Integration	Every good plan ultimately degenerates into work, and integrating the various systems with which the computer-based patient record must coexist takes time and diligent attention to database hygiene. The difference between integration and interfacing lies in your ability to make sure that none of the data flowing among the systems and sub-systems becomes a source of contamination in the database.
Standards	Few standards that support the implementation of a computer-based patient record are firm. A common standard over the last decade is HL7. The key strengths of HL7 are that it is well understood and very flexible. Its flexibility, however, allows variations in implementation from one vendor to the next. The general clarity of the specification will normally expose inconsistencies very quickly. Open Database Connectivity (ODBC) is a database access standard that is frequently referenced. Any reasonable commercial product should support this standard somehow. (Security concerns can justify exceptions to this. Refer to the section "What issues should I take into account?" for more detail.) Visit the Medical Records Institute and CPRI for good discussions of standards and additional links.

Supporting the implementation of systems for healthcare is an industry in itself. The HIMSS web site and the Healthcare Information Systems Directory site include references to many vendors and consultants who can provide additional implementation support services.

< Back

Next >