What the doctor ordered

MIKE BROSIE, ENGINEERING VICE PRESIDENT, MEDCO HEALTH, FRANKLIN LAKES, N.J., USA.

The OMEGA project closed 11 weeks ahead of schedule on 5 October 2001 and 2 percent under budget. What's more, the volume on 31 December 2001 reached 343,000 prescriptions a week—more than 600 percent of plan.

In 2001, Medco Health filled 548 million prescriptions through its own pharmacies and a network of retailers. The firm built the first fully automated pharmacy—dubbed the new dispensing protocol (NDP)—in Las Vegas, Nev., USA, in the mid-1990s. As competition has intensified, Medco Health's management team planned a second, even more complex, automated dispensing pharmacy, says Kenneth Klepper, Medco Health executive vice president and chief operating officer. “One of Medco Health's differentiating competitive advantages involves leveraging advanced technologies to improve the quality of the prescription health care we provide to our members, while at the same time containing costs for our clients.”

A technological marvel, the new OMEGA facility was completed early and under budget in 2001, largely due to Medco Health's reliance on project management, according to Klepper. “By investing in project management tools and training, we can embark on major projects with the confidence that we will be able to deliver on our commitments,” he says. “These projects and the investments they require ensure we will continue to provide our clients and members the type of services that have redefined the benchmark for quality care and have earned Medco Health accolades for customer service.”

The new facility is the world's largest automated pharmacy, dispensing 780,000 prescriptions a week. The 282,000-square-foot building is located in Willingboro, N.J., USA, and includes a high-tech system of filling equipment linked by conveyor belts. To maintain quality control, each prescription is scanned several hundred times as it makes its way through the pharmacy. (See sidebar, How the Pharmacy Works.)

Leadership in Place

The project began with an integrated plan. More than 50 project managers, engineers and support personnel participated.

“We set up the reporting requirements, identified the dependencies between the subcomponent project plans and established the project management office,” says Engineering Vice President Mike Brosie. “The aggregate project planner managed the documents and followed up with the project managers regarding compliance. The risk mitigation plans were also developed. This was the road map. Without it, we surely could not have attached such a high level of success.”

All team members attended a Myers-Briggs session to learn more about each other. Individual roles and responsibilities were defined and published. “Myers-Briggs gives a good understanding of the characteristics of your team,” says former Program Manager Wayne Rice. “You could tell who was introverted and extraverted, which ones like to get down and dirty into the details and which prefer the big picture. In some areas, we saw where we were lacking in skills. Most of all, it was a good icebreaker.”

In addition, project managers attended a two-day project management class. “Some of us knew project management forward and backward, but a lot of people on the team only had a general understanding,” Rice says. “We had a fairly short time frame and limited budget, and many team members had not been involved in a project this big with tight control, so the project approach was extremely beneficial.”

Gaining Leverage

Based on lessons learned from the NDP project, Rice conducted intensive and exhaustive “improve-the-process” sessions. “It's amazing how smart you become the second time around,” he says. “There were a lot of lessons we learned. To try to manage the integration development project with 22 vendors was a significant project itself.”

Although the same types of machines used in the NDP were planned for OMEGA, virtually all of them had substantial design improvements. As a result, the OMEGA team prepared multiple mitigation strategies and contingency plans. (See Table 1 for high-impact risks and how they were handled.)

To avoid any production delay, the team hired operations staff during the equipment design phase. These early hires later became supervisors.

Most notably, Rice inserted a four-month contingency buffer, 10 percent of the 40-month project, in the schedule, a technique common in the critical chain method.

This contingency buffer absorbed several project delays, including the effort to secure the Willingboro site. “If we had to do it again, I would have used some more risk/reward/ penalty clauses,” Rice says. All in all, the project consumed six weeks of the 17-week buffer and most of the contingency money.

Contract Work

For NDP, Rice managed the facility construction, computer systems, equipment design, warehousing and operations team, and a dozen equipment sub-contractors. The combined oversight was overwhelming. For OMEGA, a consultant helped develop vendor selection criteria, and the team delegated the sub-contracting to ATS, a major equipment manufacturer in Cambridge, Ontario, Canada.

“ATS was held to rigorous contract specifications and was held responsible for sub-contractor's work,” says George Smith, who managed planning and scheduling for the project. “The OMEGA team performed the project management responsibility over ATS and their sub-contractors as well as over all the many other parts of the project.”

Risk Management

The automated project plans were consolidated into one master plan of more than 5,000 tasks. The project management team used an engineering change request process for tracking changes through the design, fabrication, assembly and installation process. The team described design changes, evaluated the impact and sent recommendations to a decision committee.

For example, in the first quarter of 2001, top management proposed increasing capacity from the original 500,000 prescriptions per week to 780,000.

“We quickly recognized that, with a relatively small additional investment in the support infrastructure, we could increase capacity by 50 percent—more than double the payback—and push out the need for additional capital investment,” Brosie says. “The decision to up the deliverables in midstream was made possible because of a superior process development methodology and excellent risk management.”

The original budgeted cost for the project was $85 million when the scope called for 500,000 prescriptions per week. When the scope increased to 780,000 prescriptions per week, the capital budget increased accordingly.

“As expected, we were over budget in some areas and under in others,” Smith says. “A windfall of labor saving occurred from a much shorter final validation phase, requiring far fewer new hires in operations. Our validation phase was shorter because the rigorous installation and operational qualification and validation processes for the equipment and systems flushed out most of the bugs before the final validation phase.”

The Grapevine

To monitor the schedule, Rice held periodic meetings to review the minutes, the issues log and the project plan control reports. The project plan control reports included a filtered report showing tasks that should have finished or started by the day before the meeting, plus tasks due in the next two weeks. That report also included how far the end date had slipped from the revised baseline.

Another report showed tasks on the critical path and tasks less than 11 days off the critical path (tasks with total slack less than 11 days). Quarterly, Rice conducted half-day status presentations to top management to manage expectations and get timely feedback.

Meeting frequency depended on the stage of the project as well as the complexity of the work. “Meetings were conducted to be productive and move the project along,” Rice says. “We would go through plans, actuals and issues, old business and new business. We didn't dwell a lot on problem-solving unless it was critical. These were working meetings, not just for the sake of communication. They were about accountability.”

The minutes and project plans were distributed to a wider audience, which included major stakeholders such as the entire OMEGA team, supporting personnel throughout Medco Health, top management, vendors, local government, the New Jersey Board of Pharmacy and state development groups.

Quality Control

In the design phase, consultants conducted a technology review survey and a competitor analysis to ensure Medco Health was using state-of-theart technology. From the pill-dispensing machines to the conveyors, the consultants prepared integration testing and production qualification protocols.

The validation steering committee (VSC), which included representatives across the team's functional areas, had the final say. The test results were submitted to the VSC for approval and sign off. The result was a “short, speedy ramp-up” during the final validation testing, Smith says.

At the end of integration testing (and the start of the final validation testing), the team initiated the Six Sigma quality improvement process throughout the entire operation. “A data collection system was embedded into the controls architecture, key personnel in Willingboro were trained in Six Sigma and a half dozen Six Sigma teams have been in progress at any given time since,” Smith says.

Because the equipment had to work cooperatively under the control of the central computer, Medco Health engineers and computer system analysts prepared elaborate protocols to be executed by ATS’ installation staff.

In addition, the building construction contractor had to prepare the production areas in the correct order. “The room for the automated drug-dispensing machines with the longest installation and testing duration had to be ready first,” Smith says. “I provided the construction contractor each week with task activity sheets for each sub-contractor. The carpenters, electricians, plumbers, fire protection installers each had their list of weekly tasks. The building construction was modified to integrate and accommodate the installing of the equipment.”

The Project Legacy

The OMEGA project closed 11 weeks ahead of schedule on 5 October 2001 and two percent under budget. What's more, the volume on 31 December 2001 reached 343,000 prescriptions a week—more than 600 percent of plan.

Today, Rice is Medco Health's vice president of operations at the Willingboro PDC overseeing the day-to-day processes in the world's largest fully automated pharmacy. Between Medco Health's two automated pharmacies in Willingboro and Las Vegas, 1.4 million prescriptions are filled per week. Should demand exceed that capacity in the future, the pharmacy does have the space and capability of expanding operations—and project managers will lead the way. PM

This article is copyrighted material and has been reproduced with the permission of PMI. Unauthorized reproduction of this material is strictly prohibited.