A Controlled Launch

Once a provider organization decides to upgrade its CIS, what next?

HealthCare Partners Medical Group (HCP) transitioned last spring from a 14-year old electronic medical record system (EMR) into Allscript’s TouchWorks Electronic Health Record, which we planned to use as our primary clinical information system (CIS). It had been an enormous project three years in the making that had led to countless long workdays. The TouchWorks system represented a much-needed update for our organization, which required a better way of coordinating care among the nearly 500,000 patients we treat at 38 different sites, as well as our four urgent care centers

Once a provider organization decides to upgrade its CIS, what next?

Melayne Yocum is COO of Healthcare Medical Partners Group, Torrance, Calif. Contact her at myocum@healthcarepartners.com

The transition and implementation, which cost $4 million, represented one of the single largest go-live days for a medical group. We populated our new CIS with more than 225 gigabytes of electronic medical history from the previous system. The transition was the culmination of a year-long selection process.

The Solution for a Problem
Our medical staff and management had indicated that the current system was not meeting their needs. The manufacturer of our earlier medical record system had determined it was going to focus on practice management systems and moved away from designing and supporting electronic medical records. Our old EMR did not interface with hospitals, pharmacies and other relevant providers in a satisfactory manner. Our clinicians wanted features such as electronic prescribing, notes templates and health maintenance reminders.

Moreover, the vendor had stopped providing ongoing support for the EMR as the product aged. By the beginning of 2003, it became obvious to our clinical staff and the members of the management team that a replacement had to be found.

We internally recruited a multidisciplinary team including physicians, technology staff, management and medical records staff to develop the criteria required for the new CIS. Knowing that each constituency in our organization would have different needs, we decided to rank requirements on the basis of whether HCP “must have” such technology; “would need to have” the technology, or finally, whether the technology would be “nice to have.”

The priority was procuring a system that would assist in automating—and thus streamlining—the most common physician activities. For example, the filling of prescriptions could be delayed by numerous phone calls between the physician and pharmacy to confirm which medications were in the patient’s health plan formulary—a process that frustrated all parties. Other tasks that needed to be simplified were the ordering of lab tests, viewing test results and documenting patient encounters. Enhanced patient safety measures, such as automated medical practice guidelines and alerts warning physicians of potentially dangerous drug interactions, were also required. Finally, we wanted a system that was completely HIPAA-compliant and would remain so for years to come. In the summer of 2004, we issued a request for proposal. By early-2005, we had begun winnowing down interested vendors.

We were particularly concerned about disruptions in the coordination of care for our patients. In addition to our nearly 400 clinicians, we have nearly 1,200 nonclinician employees who would be using the system on a daily basis. That meant any significant glitch in the system, or in its support architecture, had the potential of bringing HCP’s activities to a halt and would be unacceptable.

As a result, we closely examined the financial and managerial viability of the vendors. Therefore, any vendor that lacked operational longevity, or the creativity to handle increased traffic over time, was rejected out of hand. To avoid such a scenario entirely, our finance department was employed to “prequalify” all vendors. The initial 24 firms we had shown interest in were reduced to fewer than 20 after the preliminary financial vetting.

The Selection Process
With the financial vetting complete, we began to drill down to the issues of security, workflows for e-prescribing and ambulatory care, radiology, lab, interfaces, training and system reliability. We asked the leading vendors that responded to the RFP to give live demonstrations of their products. The vendors were also required to respond to five different scenarios that were important to us. They had to demonstrate how their system performed under each scenario.

Among the scenarios: Testing how the system would respond to a patient who calls the on-duty nurse in the middle of the night with an itchy rash on his face. How would the system perform when notifying the patient’s physician about the condition? Another scenario was how the system fared in preserving the integrity of “locked” patient records, such as the sensitive notes made by behavioral health clinicians. How each vendor and their system responded to the scenarios were part of the evaluation process.

We chose Allscripts for its financial stability, the robustness of its Web-based system architecture, its history of seamlessly transitioning older EMR systems into its TouchWorks product and its creativity in expanding the platform. Allscripts’ Web-based architecture, which allows the ability to connect affiliated physicians and other key stakeholders (specialists, hospitals, laboratories and other ancillary providers) with our physicians, also played a major role in our decision making. HCP’s previous system did not have a Web capability. Given our desire to build standard interfaces with hospitals, laboratories and contracted specialists, we considered this a must.

We also found that the Allscripts software seamlessly integrated with HCP’s existing practice management systems, simplifying the process of implementing the CIS and integrating it with the key components of registration, scheduling and eligibility. And we appreciated the usefulness of TouchWorks’ iHealth program, which extends the physician’s reach by providing patients with secure online consultations, standards-based automated disease management information and personal health records.

Installation and Implementation
Installation was overseen by several teams, which included both HCP and Allscripts personnel. We gave particular scrutiny to security and continuity issues. We wanted to ensure that preparing for the switchover would not interfere with the functioning of the existing system until the moment the change would be made. We also wanted to assess impacts on process flow for various operations. These included ambulatory care, pharmacy refills and patient telephone lines for our nurse triage program, which provides after-hours advice and care.

The conversion proved complex because our old system was still being used in daily clinical practices. It was constantly being updated with new information and could not be shut down while data was moved to the new application. We implemented a “gap” plan, “freezing” the EMR over some weekends when there was less usage, and using that opportunity to move information to the new system. The old system would then be reopened to allow work to be done. The gap methodology was utilized several times to ensure all information entered into the new system continued to function operationally within the old system.

Our transition was not trouble-free. For example, medications that existed in the early 1990s didn’t have a “match” in the drug list of today in our new system. Each and every anomaly and error had to be researched and resolved. The good news was that we enjoyed a 99.5 percent successful filing into the new system on the first run. The bad news was that .5 percent error involved 250,000 medications, each of which had to be resolved and filed appropriately.

Another problem was bringing old EMR information into the correct location in the new CIS. Our physicians had to have significant input to this process as decisions had to be made in the conversion that impacted how medical data now appeared in the new system. Interfaces had to be built for each of eight sections of the old EMR corresponding to the correct section of the new CIS. Each interface was then tested and errors rectified.

The sheer volume of information meant that our hardware was running nonstop for days or even weeks at a time, as each section of each patient’s individual medical chart was converted. Fixing errors and discrepancies took months.

Training
Training our clinicians and staff required a combination of e-learning modules for self-learning via the web, classroom training and individual training. In addition, HCP implemented the concept of “CIS specialists.” These personnel volunteered to receive additional training in order to become “super-users” of the new system. They then fielded questions from their colleagues before and after the “go live” date. We attempted to have a minimum of one clinical and one non-clinical CIS specialist in each site.

After this major conversion from the old system to the new CIS, we had all of our clinician and nonclinician users doing the same functionality in TouchWorks that they had previously done in the old EMR. The next step was to systematically roll out new functionality such as e-prescribing, online verification of lab results, adding and updating problem lists and allergies, as well as tasking and notes. In addition, we now needed to put computers into each exam room. We chose to roll out this new functionality on a site by site basis.

We employed the same process of training each site on the new functionality before they were scheduled to go live, but in addition, we conducted dry runs about two weeks prior to the go-live. The dry runs presented many opportunities to determine how the new system would work with real patients in the normal workflow of an office and several central departments. This simulation included having the customer support desk in the information systems department up and running as our clinicians and staff worked through a variety of scenarios, simulating treatments or examinations on actors posing as patients. Observers also were on hand to note any issues they spotted during this process.

One example of what observers noticed was that the configuration or placement of the computer arm containing the keyboard and mouse was not optimally positioned for the staff and clinician to use from a standing position. This observation resulted in changes being made to other sites before the next site go-live. Debriefs after each dry run served to document anything that needed additional training or work. Each site was then accountable to work through the areas that needed improvement from their dry run. This resulted in much smoother go-lives that otherwise would not have been possible.

We experienced an approximate 50 percent reduction in clinician productivity for the first couple of days after the go-live. However, most clinicians were back up to their original productivity within two weeks of the go-live date.

Benefits and Results
The new CIS enables clinicians in all 38 HCP facilities—plus four urgent care clinics—to instantly access, read, edit and share up-to-date clinical information on all patients, providing a seamless connection to the entire continuum of care. The real-time connectivity has fostered operational efficiencies that improve patient safety, reduce practice variation and incorporate evidence-based medical care processes.

All contents of the medical records, with the exception of a small fraction, are now available to our providers during patient visits. This compares to about 70 percent availability prior to the rollout.

Another significant benefit to clinicians and patients has been the availability of the CIS for after-hours nurse triage services. When patients contact our nurse triage service after hours, not only does the nurse have the most current and complete medical information on that patient, but the nurse also sends an electronic note to the patient’s clinician with a record of the discussion and patient instructions. The clinician can follow-up the next morning or insert an electronic note into the patient’s chart.

We plan to eventually extend our new system to nearly 2,000 of our affiliated physicians throughout Southern California. We also want to work with other medical groups and IPAs who are implementing a CIS to ensure that we can take advantage of this opportunity to improve patient care throughout our communities by giving patients and their clinicians’ access to their health records whenever and wherever they need them.

HCP’s goal is to deliver dignified and compassionate care to our patients. Implementing a CIS is one significant step forward in continuing our vision.

For more information on Allscripts Touchworks, www.rsleads.com/608ht-204

August 2006