EMRs AND DATABASE STRUCTURES | Behavioral Healthcare Executive Skip to content Skip to navigation


March 1, 2007
| Reprints
What to consider as you tailor an electronic medical record to meet your organization's needs

Behavioral healthcare providers of all types are implementing electronic medical records (EMRs). Justifiably, much attention is being paid to issues of client/patient privacy, clinician work flow, and the EMR's impact on resources and procedures. A sometimes underemphasized element of EMR implementation is content design—how to capture and manage information. To be successful in their EMR deployments, organizations must devote adequate time and resources to designing an EMR that exploits the underlying database to advance the organization's goals, embody its culture, and execute its policies.

Clinical care, business and efficiency goals, and policy considerations should be reflected in the structure of information (tables and types of fields within tables) and the flow of information through the database structure to the outputs (e.g., reports). Questions and implications associated with EMR data management fall into a number of categories:

Specialized Versus Integrated

Questions. Concerning the overall database structure, should you create separate assessments, treatment plans, and progress notes (e.g., by discipline or level of care), or create multidisciplinary forms used across the spectrum of care? Is the clinical culture best represented by a documentation process focused on specialized roles or one that emphasizes collaboration? Does the assessment and treatment information belong to the individual clinician or to the patient?

Implications. EMRs either can support or challenge staff's professional identity and program loyalties. It is important to consider the above questions in the content design process. The use of several forms for a clinical operation, for example, may result in data duplication with the risk of data discrepancies. The database's structure (specialized versus integrated) also affects the user's ease in navigating the EMR, as well as data flow. The flow of information from data entry to reports and decision support tools is a major benefit of databases and should be a priority during content design.

Standardized Versus Individualized

Questions. At the level of discrete data fields, when is it best to use “dictionary” data-type formats with standardized choices, and when are free-text narrative formats better? Is there essentially identical information in different components that can be formatted in the same way (e.g., scales for assessing risk or formats for capturing family history)? What might we want to learn about what we do?

Implications. Dictionary data types ensure data consistency. In addition, dictionary formats make information available for aggregated reporting, while free-text data are difficult to aggregate. Therefore, data-type decisions impact the capacity to aggregate content across programs and populations for efficient report production and more robust analysis. Reducing the number of formats simplifies software development, the creation of user documentation, and staff training.

Focused Versus Comprehensive

Questions. With the entire dataset's scope in mind, how do you decide which data to include and where to draw the line? Must you include an individual item for corporate compliance with your clinical care mission? Do you want to include a data element because it is useful for outcome reporting or because a specialty team wants to include it? Should you use extensive standardized information, or develop a set of required screening questions with optional narrative details as needed?

Implications. It is easy for an EMR design team to have enthusiasm for capturing data while losing sight of the burdens of data gathering being created for line staff. For good clinical care, a primary consideration is the definition of an adequate dataset, and for corporate compliance how to decide what is necessary to achieve complete documentation.

An organization's goals and policies are implemented through the type and extent of controls used in the database structure and information flow. Database controls maximize the data's completeness, consistency, relevancy, and validity. An effective controls strategy enforces desired content while minimizing the inclusion of inappropriate and unnecessary data. Examples of types of controls include:

  • Accessible or “read only” fields

  • Required or optional fields

  • Permit or do not permit operations, such as editing and deleting

  • Event logic (e.g., “If this response is selected, then this will happen”)

Unique Information Versus Document Building

Questions. A database permits users to automatically copy data between components. How do you decide when to use this feature? What information is a “snapshot” and should be entered as new each time (i.e., unique information)? What information is historical and can be forwarded and built collaboratively (i.e., document building)? Should you permit additions to and/or editing of forwarded content? If different clinicians collaborate on the same information, how and to what extent do you maintain accountability for that content?

Implications. Document building can make data input more efficient and reduce duplicate data entry. It also facilitates the creation of a truly longitudinal chart and consolidates data in a shared environment, enabling clinicians to locate information more easily. On the other hand, some contend that document building in the hands of an unmotivated clinician risks the omission of important new information and the inclusion of unedited inappropriate data.