Process Re-engineering Can Produce Results, Lumeon Finds

A rigorous look at organizational processes, perhaps bolstered by new technology, can produce big savings in almost any industry. In health care, Lumeon finds that this kind of process re-engineering can improve outcomes and the patient experience too–the very Triple Aim cited as goals by health care reformers.

A bad process, according to Robbie Hughes, Founder and CEO for Lumeon, can be described as, “The wrong people have the wrong information at the wrong time.” One example is a surgery unit that Lumeon worked with on scheduling surgeries. The administrative staff scheduled the surgeries based on minimal contact with the clinicians–a common practice throughout the industry that might seem efficient. But unfortunately, people who are uninformed about the clinical aspects of the surgery make sub-optimal plans, directly leading to poorer outcomes. The administrative staff don’t use rooms and other resources effectively, and stumble over risks that the clinicians could have warned them about. Lumeon uncovered the problem during a single morning meeting with this particular hospital. By enabling the clinicians to better coordinate with the scheduling staff, the surgery unit more than doubled its presurgical screening capacity without asking for increased funding.

I recently wrote about a controversy over patient loads that erupted into a major political controversy (rarely a formula for rational process engineering). Thus, when talking to Hughes, I was sensitized toward the importance of good processes. The health care field is stuck in the kind of blindness toward process seen in the fictional medieval setting of Monty Python’s Jabberwocky, but some of the more forward-thinking institutions are doing the hard work of streamlining their processes. These include:

  • Cleveland Clinic, which reorganized their recommendations for patient behavior before and after surgery, called Enhanced Recovery After Surgery (ERAS)
  • BUPA, a major British insurer that has a formal process model
  • U.S. giant Kaiser Permanente, which uncovered enormous waste when clinicians search for supplies

The higher you rise above the scene, and the more you can think about the system rather than one silo, the more efficient you can become. The Kaiser inquiry covered the entire supply chain for each hospital. BUPA is fortunate to possess actuarial information that help it assign a predicted cost and likely outcomes to cancer cases, where the company can assign caretakers to patients as needed throughout the whole recovery process.

Another useful scope is the sequence leading from a patient’s initial contact to a successful outcome, a process or “pathway” that goes far outside the hospital’s walls and beyond the time in the doctor’s office or surgical unit.

Typically, Hughes says, one day is enough to find process improvements. Through interviews and through observation–because staff misunderstand and misrepresent their own processes–Lumeon can develop a process map, expressed visually like the post-operative pathway in the following figure.

Typical pathway, describing post-operative process
*Click to see Full Size – Typical pathway, describing post-operative process

The best motivation for taking a longitudinal view, of course, is risk-sharing. A doctor who will be rewarded or penalized for outcomes will be willing to invest in producing better outcomes. Similarly, an insurer such as BUPA will be motivated to reduce readmissions if it has a long-term responsibility for patients. Bundled payments are a round-about, highly diluted approach to risk-sharing.

Fee-for-service models mean having to define a deliverable that everybody can understand and achieve. A bundled payments model is far from this. UK outcome measures truly place risk on the provider. In the US, bundled payments dilute risk.

But Lumeon can find ways to improve processes even within a fee-for-service model by enabling health organizations to guide patients more successfully through their entire health journey. For instance, with the company’s Care Pathway Management solution, doctors can remind patients to come back in five years for a colonoscopy, thus potentially saving lives while ensuring the institution’s own revenue stream under fee-for-service. Other simple goals can be to make sure the patient has a complete list of tasks prior to surgery (such as not to drink water in the morning) in order to eliminate late starts or last-minute cancellations, which are very expensive as well as frustrating. Predictably, Lumeon finds a certain set of common problems over and over, regardless of medical disciplines or institutions. Hospitals sometimes optimize within each department, but not across multiple departments. Usually this change comes down to maximizing compliance with a known protocol, rather than trying to use sophisticated artificial intelligence techniques to look for new approaches that theoretically offer benefits.

Lumeon also works to minimize disruptions to existing workflows. Large institutions such as Kaiser can tell everybody to adopt a whole new way of doing things, but staff within most institutions might be more resistant. The staff can still be trained to do things like create quality standards and follow them, or call patients at certain intervals or after a procedure, but these processes need training before they become reliable and predictable. Culture and habit, not technology, turn out to be the biggest barriers to process improvement.

Software, too, must be molded to current ways of working. We all experience little tolerance in our work or everyday lives for non-intuitive computer interfaces that appear to be putting barriers in our way. For instance, I have never forgiven my phone vendor for changing the most common activity I do on the device (turning airplane mode on and off) from a three-step process to an eight-step process.

The most effective persuasion is evidence-based. If an institution can get one department or doctor to adopt a new process, and can then collect data showing that it improves outcomes and cut costs, other departments are likely to follow along. In contrast, staff are likely to be oblivious to a study from a journal with statistics from clinical trials, no matter how scientifically valid the study may be. Hughes says that resistance to change is often attributed to doctors, but he thinks that this resistance is primarily caused by change being forced on them without evidence. With proper, objective data supporting a change, doctors are often the first to lead new initiatives in the spirit of delivering better patient care.

New kinds of records are needed to keep track of outcomes and make use of the valuable data they provide. Ideally, Lumeon would integrate with electronic medical records, but the EMRs are rarely set up to hold and provide such information. Instead, Lumeon installs software on top of the EMR, calling their addition an “agility layer.”

Hughes identified two common practices that can interfere with process improvement. The first is the growing focus around “patient engagement,” which can be as superficial as sending reminders for online check-ins or as fundamental as giving patients access to data.

However, patient engagement by itself is not sufficient to deliver meaningful process improvement. Patient engagement measures can make a difference as an integral part of an effective operational process. For instance, there is no point in getting patients to fill in data online if it’s not going to be used by the clinicians.

Second, the focus on documenting compliance with standards, such as meaningful use, often becomes a documentation exercise rather than a way of improving care. Unfortunately, this is a problem that is seen all over the world by well-intentioned governments and funders who want to offer incentives for good behavior by paying for better processes. But this all too often ends in additional costs and effort to administer the care, rather than actually focusing on the basics.

About the author

Andy Oram

Andy Oram

Andy Oram writes and edits documents about many aspects of computing, ranging in size from blog postings to full-length books. Topics cover a wide range of computer technologies: data science and machine learning, programming languages, Web performance, Internet of Things, databases, free and open source software, and more. My editorial output at O'Reilly Media included the first books ever published commercially in the United States on Linux, the 2001 title Peer-to-Peer (frequently cited in connection with those technologies), and the 2007 title Beautiful Code. He is a regular correspondent on health IT and health policy for He also contributes to other publications about policy issues related to the Internet and about trends affecting technical innovation and its effects on society. Print publications where his work has appeared include The Economist, Communications of the ACM, Copyright World, the Journal of Information Technology & Politics, Vanguardia Dossier, and Internet Law and Business.