Having difficulty in reducing health care-associated infections? Suffering excessive downtime in operating rooms? These and other quality issues often can be traced to one key factor — process variation. James Benneyan, professor of operations research at Northeastern University and co-director of the National Science Foundation Center for Health Organization Transformation, is leading research efforts to help health care organizations narrow the process-variation gap. He discussed his research with Bob Kehoe, H&HN contributing editor.

How are hospitals impacted by process variation?

Variation affects almost every key performance measure and every one of the Institute of Medicine's six key dimensions of a good health care system: efficiency, effectiveness, safety, satisfaction, access and equity.

Three different types of variation are at play here — variability in processes, practices and people. All are important to understand and manage. Practice variations have been studied thoroughly by the Dartmouth Atlas group and others. These variations can affect clinical outcomes and equity in access to care.

How can health care organizations reduce process variation?

Some of the greatest gains are realized by applying principles used in high-performing industries. For example, from the Toyota Production System and the reemergence of Lean principles, we know that standardization, process simplification and reduction of all types of inefficiency — travel, idle time, rework and others — can lead to fewer errors and give personnel more time to interact with patients.

Also, from reliability engineering, such as that practiced in Six Sigma manufacturing organizations and ultrasafe industries like airlines and nuclear power, we know that redundant systems and error proofing through forcing designs are highly effective.

It is critical to consider the design of workflow processes and facilities together. Facilities can be designed to minimize travel time, facilitate team care, make all work visible and be patient-centered rather than caregiver-centered. A classic example is employing systems-engineering tools such as computer simulation and queuing to right-size capacity, flow, staff and space. We currently are supporting two large-scale facility designs in Boston with this type of work.

What has your group's study on operating room turnaround times shown?

A huge piece of the problem again is variability — on-time starts and stops, turning around an OR to prepare for the next procedure, periodic cancellations, even best-practice and evidence-based compliance. These all cause work to jam up like planes on the tarmac.

When it is not possible to reduce variation, more advanced engineering methods can help achieve significant improvements in cost and throughput by scheduling and managing more intelligently.

Essentially, we mimic the algorithmic approaches the airlines use to manage equally complex situations involving interlinked schedules, personnel requirements, no-shows, weather and other unplanned events.

Our group has spent 20 years developing methods to handle this complexity at minimal cost and maximal performance — in airlines or health care to develop simpler rules of thumb that perform almost as well, and that any front-line personnel can use with simple pen and paper.

Perhaps more than in airlines, where computers in some unknown place manage all the scheduling, routing and customer-flow complexities, a key need in health care is to develop methods that are almost as good, but can be used by what we call the "intelligent nonengineer."

Hospitals are focusing on process variation to improve quality and safety, but are we seeing results?

I think our biggest success so far has been in patient safety. The last 10 to 15 years have seen a great deal of focus on safety, and a significant increased use of "systems thinking" about safety. This, in turn, has created a large community of folks who are skilled in process improvement.

A key need now, though, is to deploy, coordinate and focus this incredible resource in a better way. There are a lot of redundant efforts while not all the need is being addressed.

Why haven't quality-management initiatives been more successful?

Being a significantly new way of thinking, most health care organizations have an insufficient understanding among upper- and mid-level management of how to achieve quality.

We have a good start, but we've just scratched the surface of understanding how other industries improve performance and safety. Going back to W. Edwards Deming, the quality guru credited with much of the post-WWII work in Japan, the concept is to achieve higher quality at lower costs and with fewer resources.

Often we are still "solving" problems simply by throwing more — not fewer — bodies at the problem. This actually is the opposite of what Toyota, Motorola and Deming were doing. In safety, this would translate to designing processes and facilities that require fewer people, simpler processes, fewer handoffs and less opportunity for communication lapses.

Which health care processes have the most room for improvement?

Health care is the single largest industry in the United States, with estimates as high as 25 to 30 percent of all the cost attributed to some form of inefficiency or safety issue. It's tough to go wrong starting anywhere. But, typically, there is a huge amount of low-hanging fruit in a few key areas. Inventory and supply chains are a no-brainer place to start. Common estimates are that 25 percent of a hospital's operating budget is caught up in materials. In a $2 billion network, that's $500 million. If you achieve even a 1 percent savings, that's about $5 million a year.

Other places to focus are the high-cost drivers and DRGs — the ED, OR, cardiac patients and chronic conditions. In scheduling and overbooking, it's easy to save several hundred thousand dollars annually per clinic, which adds up quickly if replicated across a system.

How can hospital executives help foster this improvement? 

I don't think it is anything complex. But nothing is free, at least initially. Allocating sufficient resources, communicating and demonstrating commitment to improvement and safety always are the usual replies.

An interesting thought exercise for executives is to ask: How much total time does staff spend working to improve processes?

When Deming worked with Pontiac in the 1980s, the plant stopped production for one hour every week during which all employees worked in cross-functional teams on process improvement. Think about it — that's 2.5 percent of all employee hours formally committed to getting better. In contrast, a lot of the ad hoc work hospitals do occurs on the fly.

It also is useful to have more advanced experience on board to help focus and manage the work to get the greatest impact. Many hospitals actually had internal systems-engineering and process-improvement departments in the 1970s and 1980s, most of which were phased out due to lack of impact.

So, while investing in skill development at all levels is key, it's even more important to establish a process to manage the workings of process improvement.