WIZARDS FOR DOCUMENTATION Wizards give RNs an OR record that matches their practice and workflow.

My EMR project has me feeling a bit deflated, I'll admit. Deflated, but not defeated. I was hoping and expecting that our electronic medical records would be up and running by the time this article, the last in a year-long, 6-part series, ran. Instead, our go-live date continues to elude me. I'm left leaping the same high hurdles that were standing in the way of a successful EMR implementation when last we met in September (see "The Stalled Install" at tinyurl.com/9444vdk).

As I write this in the middle of October, our EMR project is not where I expected it to be. Not even close. We've accomplished so much over the past several months, yet we're still not ready to go live, we're still not ready to be paperless and we're still not ready to congratulate ourselves on a job well done (and, well, done!). We're not stuck in neutral, but it sure feels as if we're spinning our wheels. We're still troubleshooting our interfaces, those computer bridges that let data move smoothly across our old and new software platforms. The sorry truth is that we don't have a single interface that works right every time.

Garbage in, garbage out
We've all heard the computer axiom, "garbage in, garbage out." We've found out the hard way that human error in data entry impacts interface testing to the point that your tests can be worthless. Three of our hospital's departments need to interface with our health information system (HIS) during scheduling: the main OR, the outpatient surgery center and the endoscopy center.

Here's how things should work: After we enter the patient's data in the EMR, and register the patient in the HIS, the interface should link them both for supply management, charging and billing. But if a single space, comma or number is out of place, the systems won't merge the patient. During testing, we've had schedulers from each department completing data. Each scheduler was entering the data a little differently, however, so we've had many failures in our testing. This has eaten up about a month's worth of work.

We're now focused on only 1 department and have only 2 people entering data to send to the interface for IT, the interface builder and the EMR staff to review for functionality. Waiting for each of them to complete their part of the review is a big time-consumer. Sometimes we wait for days until we know if the test patient passed the interface successfully. All these reviewers are in different time zones and have different priorities, so the testing isn't complete until they get to it. The solution would be for all areas to commit focused time, for as long as it takes to complete this testing.

Don't do as I do, do as I say
I can't share the joy of a completed implementation with you, but I can share what we did wrong and what we did right throughout this process.

Our failures:
• Not hiring dedicated EMR staff. We didn't understand the need to hire or assign a dedicated project manager and 4 other dedicated full-time staff from the beginning. A dedicated project manager would have streamlined many steps of this process and prevented many hours of correcting others' data entry. One of the dedicated positions should be an IT person as a partner to the project manager. If I hadn't spent so much time driving the interface testing and trying to learn the IT processes, the clinical side would have been completed 2 months ago.

• Not holding everyone accountable. We've had weekly status update meetings since we began in March. Everyone was invited, but only the vendor and the clinical staff participated. From the very beginning, we needed to hold everyone accountable for their participation and ownership of their part of the implementation.

• Not communicating well with the team. This was one of the greatest challenges. There are so many facets to an EMR implementation. It's difficult to know who needs to be involved in what. I didn't want to bog down the clinical staff with the interface issues, yet they needed to know how these interfaces were affecting the whole project. I finally made a spreadsheet to track all the issues and people I needed to update daily, weekly or occasionally.

Our successes:
• EMR selection. This was one of the longest and hardest pieces of an EMR implementation. It took more than 2 years of concentrated effort to get to the point where we could take our request to the board of trustees. Our best plan was to include all of the stakeholders and add their needs to the selection matrix we used to document each EMR system we evaluated. We included risk management, nursing quality, process improvement, physician relations, the health record information department, senior leaders, information technology and many more departments. Everyone throughout the hospital had some input into the selection of our new EMR, so when it came time for implementation, they all knew what we were working on. An important note when selecting your software: Don't let IT drive this process. Otherwise, the end result will be what meets their needs, not your clinical needs.

• Team champions. Many people knew that we needed a new EMR, but didn't understand the processes that had to change to get us there. By including staff in frequent updates and planning meetings, we had about half of the staff involved from Day 1. This gave the staff a stake in our project's success. The core team members brought their peers on board as we progressed, which led to easy end-user training and hopefully a smooth go-live.

• System design. The greatest part of our EMR is the use of wizards for documentation. The wizards are selected according to the type of procedure being documented and focus the RN's documentation on the items needed for that specific procedure. The RNs aren't trying to fit patient care into a generic OR record, but now have an OR record that matches our practice and workflow. Although we haven't tested this live yet, I'm seeing nurses fly through the documentation in our testing phase because it complements the actual care they're giving the patient.

Accurate, actionable intelligence
Once we're live and have a month's worth of data to evaluate, we'll complete our analytics training. I'm looking forward to turning my perioperative statistics into accurate, actionable intelligence. I'll finally have some tools I need to complete my job efficiently and effectively. The ability to drill down to the data I want to collect without having to contact IT to have a report written is priceless.

We'll update you with a final chapter of this EMR diary. Thank you for letting me share this journey with you and thank you for all the great comments and contacts that have stemmed from this series.