EHR System Replacement: Making Sure the Next Time Is the Last Time

iStock 000053971326 Large

The Tipping Point

Spurred by the rapid changes in the healthcare IT environment and government pressures to “meaningfully use” their EHRs, a growing number of hospitals and medical practices are deciding that their legacy systems are inadequate and maintaining them is too expensive and too risky. According to a study by Black Book Rankings released in February 2013, almost 20% of physician practices are considering switching EHR systems within the next year. If your organization has completed a thoughtful analysis and determined that your incumbent system cannot be optimized, the next challenge is to plan and execute upon a successful replacement EHR implementation strategy. This article addresses four critical components that administrators and their teams must thoroughly vet in preparation for an EHR system replacement.

The Game Plan

Replacing an EHR system is often more complex than the implementation of the original EHR. It is critical to invest in up-front planning to define the business and end-user needs, implementation approach, data migration method, and technology integration strategy. To vet these areas, organizations should consider four components:


As with any complex project, the involvement and support of the individuals who are affected can ultimately dictate whether or not the outcome is successful. In the case of a replacement EHR implementation, the end users must buy in to the new initiative, have strong leadership, and receive the proper training to help them adopt the changes.

  • Organizational Buy-In – Many EHR system implementations fail due to a lack of buy-in. Staff may be apprehensive about the new system, given the shortcomings of the old one or memories of a difficult implementation process. All users must understand how the new system will enable them to be more successful in their jobs and how the organization will avoid the pain points of the previous EHR implementation.
  • Executive Leadership – Clinical system implementations cannot achieve true success if they are managed as IT projects and/or lack clinical and operational leaders who are willing to hold people accountable for the success of the system. IT-led projects result in a system that does not meet the needs of the end users, a user base that does not feel included in key decisions and thus not responsible for making the system successful, and a leadership team that is unwilling or unable to support the system once discontent begins to grow.
  • Change Management – Unlike the first implementation, staff now have experience using an EHR and may want to recreate that experience with the replacement EHR, rather than adapt their processes to work within the new system. The key will be to acknowledge and leverage the valuable input experienced users may have while setting realistic expectations about what users will and will not have an opportunity to influence. The quality of your communication will need to be much more sophisticated than during the original implementation, with more detail and transparency. Whereas users might have felt overwhelmed or unconfident participating the last time around, they will have a rich foundation from which to base questions – and criticism, if they are not effectively involved.


One of the most common mistakes made when implementing a replacement EHR is trying to recreate the work flows optimized for the original system. Organizations need to perform a thorough current-state work flow analysis of clinical processes, then involve clinical staff and system application analysts in designing future-state work flows that both promote efficiency and take advantage of functionality in the new system. Once these future-state work flows are created, they should be thoroughly tested by a variety of users to ensure that the nuances of the work flows that impact day-to-day patient care are vetted. They should also be created with quality initiatives and population health targets in mind. Lastly, mock go-lives should occur in which patient visits and other common processes are simulated from beginning to end to validate the handoffs between clinical staff within the system. By having staff and physicians engaged from the start, you can create sound work flows while eliciting user buy-in.

In addition to work flows, a strategy must be devised to transition staff from one system to the other. A discussion should occur as to whether to move to the new system on an application by application basis (e.g., first the practice management system, then the EHR) or all at once (“big bang”). A big bang would be the right decision if you have a short timeline and the infrastructure to support everyone going live simultaneously. Bringing on the practice management system followed by the EHR is more appropriate if some applications in the old system are in need of replacement now, while other applications can meet your needs in the short term.


IT staff will most likely focus on the obvious technology impacts of the replacement system, such as servers, security, mobile devices, and networking. The area most often overlooked is interfaces. Many organizations incorrectly assume that because they had a working interface in the old system, they can simply repoint this to the new system. A new system may involve new work flows and data mappings. The interface data that will be passed between systems should be reviewed in detail. This should include an in-depth testing process that accesses each data element being passed between systems, any data translation rules that each system may be applying to the data, and triggers inserted into the work flows to pass off data at the appropriate times.

Next, a data migration strategy needs to be developed to determine when data will flow to the old system versus when it will flow to the new system. For example, if you perform your system cutover on January 1, does a lab result from a December 20 visit go into the new system or the old system? Typically, a cutover date of service is selected in which every patient service that happened before that date flows to the old system and everything that happens on or after that date flows to the new system. Furthermore, should the interfaces feed data to both the original and replacement EHRs for a period of time to ease the transition and ensure that data is not lost? Organizations often send data to both systems for 1 to 3 months until they feel confident that the interfaces are working as expected. How long should the old EHR be maintained as a read-only archive to ensure that users can access historical data not present in the new system? This is usually dependent on the organization’s internal data retention policies. Lastly, how will the EHR replacement impact the practice management system?


The final area to consider is your data conversion strategy – a discussion many organizations had when moving from paper charts to their original EHR. In the first implementation, they tried to decide what to capture in the new system as discrete data versus as scanned documents. The replacement poses a new problem in that there is already a wealth of discrete data in the old EHR, tempting the organization to move it all over. Many organizations that tried to move everything from paper charts into the EHR found that the electronic charts became cluttered with data they never use. Treat the existing EHR as you did the paper chart and ask the same question: What do you really need? While there is value in a thorough health history that spans years, organizations must evaluate the cost of the data conversion versus the value that historical data provides in day-to-day operations. Data that is unlikely to be needed in routine patient care may be best left in the archived system if an electronic conversion proves prohibitively difficult and/or expensive. Active problem lists, allergies, active and discontinued medications, immunizations, and possibly, past medical, surgical, family, and social history, last visit note, as well as lab results, are the recommended data elements to convert. Many data elements that are more than 3 years old are best left in the old system, including acute conditions and older scanned documents. Lastly, you might consider bringing forward any long-term meaningful use, Physician Quality Reporting System (PQRS), or other population reporting data required by your organization.


The transition to a replacement EHR poses unique challenges not present in the original implementation. The first three things your organization should focus on are change management to achieve organizational buy-in, identification of strong leaders to champion the system, and optimization of work flows for the new EHR.