Skip to main content

In the sprint to abandon paper, EHR adopters often focus on getting data into a system. But, to achieve true success within a customer’s behavioral healthcare EHR, the focus must be on what comes out – the minimum, actionable business intelligence that we need. Too often, reports, visualizations, and data for interoperability are not available at all, or require significant manual manipulation and preparation.

It’s a bit like closet organization. Without an organization system, the closet is a database that clothes are piled into, but when the door opens in the morning, the output isn’t very actionable. Shelves may have been hastily put up a few years ago because at least at that time the clothes wouldn’t be on the floor, but there wasn’t much thought put into how they would need to be retrieved each day. Outfits are a blind grab, and starting completely over might be your only shot at governing the mess.

Unfortunately, an enterprise healthcare software do-over is more expensive than a closet system. As such, every enterprise software purchasing process should start with a basic plan for the data. It’s more common than it should be for organizations to invest significant time and money into behavioral healthcare EHR and complementary systems, only to realize a few years later that all that data entry is not paying off with actionable business intelligence.

The following three steps will help to ensure your organization maintains EHR data governance and avoids the “reset” scenario:

1. Plan around actionable data

Establish the minimal, actionable data needed to operate intelligently and efficiently. What are the deal-breaker requirements for supporting billing and compliance, for example? Get your organization to achieve a meaningful baseline of EHR use as quickly as possible, and then expand from there. It’s so important that it’s worth repeating again: start with the data you need to get out, and then work backward to the data entry process. Going about it the other way around will lead to a pile of clothes on the floor of the closet.

2. Strive for structured data

Free-text data is unstructured and problematic. Unless you’re interested in exploring artificial intelligence technology that’s evolving to structure and index narrative data, think in terms of reporting data for which the end-user entry options are standardized. Narratives are certainly important in behavioral healthcare, but structured options should be ruled out first.

Structure isn’t just about pick lists; it’s also about validation. Proper validation is making sure that all the data required to yield the business intelligence you’re seeking is entered consistently by all end-users. Validation is best when built right into the software. Validation rules are configured such that if user selects “A”, then “B” and “C” must be addressed before the user is allowed to proceed. Software should handle most of the validation because that limits the risk of bad data.

3. Conduct ongoing data performance evaluations

Success with data is as much a mentality as it is about systems. Data “rigor” needs to find its way into the workflow and culture of your organization. When considering data governance, think about how the data you’re looking for will be incorporated into your day-to-day management. This starts with visualizations of key performance metrics that are readily available to everyone, all the time. Constant, real-time visibility of the data drives home the importance of complying with data entry, as well as – of course – keeping everyone apprised of performance to goals.

Once you have the visibility, you can leverage it in one-on-one and team meetings. Include data performance responsibilities in position descriptions. Build a quick dashboard review of performance to goal (and performance to data quality) into all your routine management meetings.

As you evaluate performance and quality of data, continue to question and evolve the data governance plan. Just as you will identify new data opportunities for expansion, you’ll surely find data you thought was important, which actually just isn’t. If you’re not compelled to look at or discuss data regularly, then it’s a good bet the data isn’t valuable. It’s like those clothes in your closet that you think you’ll wear someday, but which you haven’t touched in years. Get rid of it – take it out of the workflow. It’s just clutter that’s diluting the effectiveness of your business intelligence, your data entry, and distracting your team. Maintaining data governance and the data itself as a routine is far easier than a massive project to clean up and try to muscle years of neglected data into usefulness.

A data governance plan shouldn’t scare an EHR vendor, it should be welcomed. Modern behavioral health EHR platforms should be configurable and readily extensible to handle emerging needs. That said, prepare to negotiate and make some compromises – no vendor will address everything exactly as you’ve included it. Regardless of how well your vendor accommodates you, having the right plan for data governance is the key to long term EHR success.