Demand action, not assumption

A data governance model at Vanderbilt University Medical Center supports patient safety by providing complete and accurate information.

Mary Reeves Rita Bowen

Many of us are familiar with the old adage “garbage in, garbage out,” or “GIGO.” If your organization is one that does not have an active, authorized and responsible data governance program, you run the risk of achieving GIGO. Even with a fully implemented electronic health record (EHR) and achievement of meaningful use, the value of your information is compromised without data governance.

Clinicians must have a high level of trust in the integrity of EHR information. Achieving that level of trust is accomplished through data governance. Data governance is the who, what, where, when, which, how and why of content management in the electronic record. And, like Vanderbilt University Medical Center, many of the nation’s leading healthcare providers are actively implementing programs, creating policies and assigning executive teams to ensure it is in place. Data governance of patient information enables organizations to support patient safety by providing complete and accurate information.

A real-world example
Vanderbilt University Medical Center has had a self-developed EHR for more than 20 years. The organization now has 100 percent of its inpatient population’s information available electronically.

When Vanderbilt created its EHR, it started with computerized physician order entry (CPOE). This initial step successfully introduced providers to an electronic environment and encouraged them to use the EHR as an important part of patient care. As many have learned, EHR implementation must be a top-down project with clinician acceptance and support.

The organization quickly found that stronger data policies were needed. Vanderbilt needed policies to answer these specific questions:

What happens to the data?
Who manages the data?
What data is authorized for inclusion in the EHR?
What defines draft, edited and final versions of the data?

Vanderbilt began addressing data governance through its traditional medical records committee. This silo method was a long and difficult approach. Ultimately, the organization enlisted a top consulting group to develop a data governance committee and structure with a new health record executive committee at the top.

Easier and faster if done upfront
It is easier and faster to establish policies and procedures for data governance when done upfront, versus trying to correct “the way we’ve always done it” mentality. Secondly, there will be many groups and departments within an organization doing their own thing, from both a technology and data governance perspective. It is important to corral these groups and feed them through central governance. This is where policies are established, implemented and enforced.

Vanderbilt’s health record executive committee began in September 2009. The charter states the following purpose: “Sets strategy and guiding principles for creation and use of the health record and is responsible for the continued migration and evolution of the health record.” The primary focus of the health record executive committee and the two subcommittees – policy and migration/deployment – is to develop strategies for enhancing standardization of practice, while reducing risk and enhancing compliance.
Members of the executive committee include representation from medical staff, clinical staff, hospital executive leadership, CMIOs, CIOs, director of patient care informatics, director of VUMC medical information services, risk management, compliance, accreditation and standards, and the office of general counsel.

Three of the functions of the health record executive committee are:

Establishing and implementing standards for the health record regardless of the media on which the information is used, collected or stored.
Authorizing and approving policies and procedures concerning the health record.
Establishing standard documentation practices.

One of the first policies developed and approved by the health record executive committee was “Carry forward of clinical information in the EHR.”

Cut and paste (also known as cloning, re-use, carry forward) is ubiquitous – especially with electronic documentation. Every electronic record has a large population of cut-and-paste information. And while the cut-and-paste feature benefits clinician productivity, it can create a data integrity problem for the organization. The feature needs a structural policy that is managed and reviewed.

Key points of the policy are:

Carry forward with caution;
Information that is carried forward is uniquely identified; and
More documentation doesn’t necessarily mean better documentation.

At Vanderbilt, all cut-and-paste information stays a different color until it is reviewed and authenticated by the clinician. Vanderbilt also audits clinicians on adherence to the cut-and-paste policy.

In addition to the “carry forward of clinical information” policy, these policies have been approved:

Definition of legal medical record;
Electronic signature;
Additions, corrections and deletions in the electronic medical record;
Naming convention standards; and
Scanning and indexing standards.

A new role for HIM professionals
Vanderbilt learned that data governance is the foundation of the EHR, but many other organizations have not. HealthPort, a leading ROI service provider, confirms this fact.

Both Vanderbilt and HealthPort report that IT system administrators usually become the owner of data. Within each application this makes practical sense. However, data governance must oversee consolidated data from multiple silos, not just one system.
Health information management (HIM) professionals are well equipped to complement this enterprise-wide focus, as they know how to ensure data integrity, accuracy, completeness and privacy of information contained within medical records. HIM professionals are already trained and deep into ROI, privacy and HIPAA, which, as stated, are essential components of data governance.

As the siloed data came into their EHR, Vanderbilt learned that information is not the same across departments. HIM is adept at inter-departmental translations and knows how to manage all the various data politics.

A consistent, updated data dictionary is required
Organizations should continually review the core policies and procedures driving governance. One major need is a data dictionary. Many do not know what a data dictionary is, or if they do, have not implemented it.

Teams should be effective at mapping data and building their data dictionary as the organization implements and transitions between systems. Each data element must be defined, verified and edited such that it does not lead to wrong information in a patient’s record. Incorrect record data can lead to dramatic adverse events and erode an organization’s valuable patient safety initiatives.

Data governance helps make electronic information usable and available to caregivers, researchers, quality monitors and future reimbursement modalities. To be of value, the data must be normalized and comparable, as well as consistently defined and understood. Data governance is the foundation of that understanding.

Foundation for understanding
Clinical data can be very unstructured and heterogeneous. While this fact may support the originator of the data, it causes problems for other users and the overall organization. For EHR data to become actionable information and used by clinicians in direct patient care, there has to be a high level of trust in the integrity of that information.

Like many other institutional problems, it is important that data governance be encapsulated within visible policies and procedures, and that they remain living, breathing documents – essential ingredients within the institutional culture.

About the authors

Mary G. Reeves, RHIA, is director, medical information services, Vanderbilt University Hospital. Rita Bowen, M.A., RHIA, CHPS, SSGB, is SVP of HIM and privacy officer, HealthPort. For more on HealthPort, click here.