Chassin, Loeb cite high reliability as 'next stop' of QI journey

TJC taking leadership role, developing tools to help hospitals

To paraphrase an old TV ad, "When The Joint Commission speaks, people listen." So when Mark R. Chassin, MD, FACP, MPP, MPH, president, and Jerod M. Loeb, PhD, executive vice president, Health Care Quality Evaluation, co-authored a recent article in Health Affairs entitled, "The Ongoing Quality Improvement Journey: Next Stop, High Reliability," you can bet healthcare quality professionals stood up and took notice.

For while the idea of employing high-reliability principles to healthcare quality improvement is not new, an endorsement of this significance is.

Chassin and Loeb assert that it is possible for the industry to both achieve and sustain high and consistent levels of excellence by focusing on three key areas:

  • Leadership: Leaders of healthcare facilities and systems have to make high reliability a priority.
  • A culture of safety: This entails an emphasis on trust, reporting unsafe conditions, and improvement.
  • Robust process improvement: This entails adoption of a range of QI tools and methods, including Six Sigma, Lean, and change management.

What is high reliability?

The term "high reliability" certainly sounds like a laudable goal, but what, exactly, is it — and can it be measured? "You judge the existence of high reliability based on no major quality failure, as in airline safety, or nuclear power," says Chassin. "Those industries have close calls but are pretty close to zero in major quality failures. What we're really talking about is the kind of quality problem that has a substantial effect on patients."

In other words, he continues, high reliability entails having reached a "very, very" high level of safety and maintained it over time across all services. "We know there are no highly reliable healthcare organizations anywhere in the world, and it's pretty far out there in terms of what's possible," Chassin says. "Our intention in describing high reliability is to be able to look at other industries that have reached that state and maintained it, and set the bar much higher than it has been in the past."

Small increases in safety from year to year are nice, Chassin says, "but we have to get to no wrong-site surgeries. We have to get to none."

"My definition would be minimal variation — seeing process variation getting less and less and less," adds Patrice L. Spath of Brown-Spath & Associates, based in Forest Grove, OR, and lead author of a chapter on 'High Reliability and Patient Safety' in Error Reduction in Health Care (Jossey-Bass; Second edition, 2011). "To get to perfect is impossible without spending gobs and gobs of money on high tech and eliminating all your people, because whenever you have people involved, you have variation. The question becomes, which processes need to be highly reliable and in which is it OK to have 90% reliability? You need to sort those things out in setting priorities."

It's important, she continues, to define what you're trying to achieve before working toward your goals. "You can say zero errors is doing it right every single time; the problem is that what might be an error for one patient might be the right thing for another patient, so I don't know if zero errors is my definition," says Spath.

In setting priorities, she continues, "You look at those areas where if our processes are not highly reliable something really bad is going to happen. You can say, is it really important that the admissions person collects information 100% accurately, or is 90% okay? On the other hand, you'd better make sure the right patient gets the right surgery, and that a timeout that occurs better be 100% reliable. In other words, you focus your energy on areas where your 'return on investment' is better."

Is a reliable hospital possible?

Yosef D. Dlugacz, PhD, senior vice president and chief of clinical quality, education and research at Krasnoff Quality Management Institute, a divison of the North Shore - LIJ Health System, and Spath's co-author for the high reliability and patient safety chapter, goes even farther. "In my opinion it's hard to talk of a hospital as a highly reliable organization," he says. "You need to look at small portions of delivery of care, like focusing on the OR or the ICU."

Dlugacz explains: "Those two environments are closed environments with homogenous processes of care. You have a beginning, an end, and a unique culture, which is very important for a highly reliable organization." He points to the Navy as an example. "They identified the process in which they can land a plane safely on a ship in any circumstances — rough seas, and so on. If a plane lands they catch it, and people know exactly what to do when they land."

Dlugacz believes that safe environments can be developed, for example, in ICUs all over the country. "In other words, if you have patients on ventilators you know what to do on self-extubation, or with UTIs you know you can develop measures to prevent sepsis," he says. "If you can implement and measure processes, you can create a potential safe environment with zero problems — no pressure ulcers, infections, or falls."

On the other hand, he notes, units such as EDs do not lend themselves to such an approach. "Because the interests are so varied by unit and each one is episodic care, there is not a defined beginning and end. In the ED you have one diagnosis, and on another floor you'll have another. Also, the patients are all different; a 90-year-old is different than a 30-year-old, and the care processes are very different."

"That's pretty pessimistic," Chassin responds. "I prefer to be optimistic and think we can get there. Just think: A few years ago people thought it was impossible to get rid of central-line infections, but a number of facilities have now done it for years."

Measuring success

How do you measure the reliability of processes? In their writings, Spath and Dlugacz suggest a different method from those traditionally used. "For example, now we measure patient falls by the number of falls per 1,000 patient days; but that's a measure of outcomes, not reliability," Spath says. To calculate the reliability of a system at preventing falls, she offers the example of a facility where there are 4.6 falls per thousand. If you subtract 4.6 from 1,000 and then divide that by 1,000, "You end up with 99.5% reliability — which is a whole lot different than 4.6 falls per 1,000," says Spath.

You must also set reliability goals, she continues, such as goals for timeouts or patient falls. "You should say, 'We want our systems for preventing patient falls to be 99.5% reliable — not to say we're going to get down to two patient falls day," says Spath. "Then, you basically turn to the literature about reliability science to see what works best for preventing mistakes."

"The ultimate measure is if we got rid of major quality failures," adds Chassin. "Part of the challenge is that we are not going to be able to rely solely on measurement of quality processes and patient outcomes. Measurement-driven improvement is an important component, but you have the challenge of developing the safety culture that every high-reliability organization has." (For more on safety culture, see the sidebar, below.) "There are metrics that help with that, but with a very big organization, there is a change challenge that supports safe operations — but it occurs in parallel with improvement."

One of the most important measurements, as stated by Chassin and Loeb, involves the initial assessment of where your organization stands. "We've done an inventory of a lot of tools out there; we do not think any of them really address the key changes that healthcare organizations need to make today," Chassin says.

Ironically, he says, part of the problem resides in some of the existing tools that focus on the characteristics of a high-reliability organization. "The problem is that while they do a good job of showing how such organizations work, they do not do a very good job of giving us a roadmap of getting from low reliability to high reliability," says Chassin. "You can't just turn a switch in an organization that does not get reports of close calls, or where there is low trust, had very poor measurement and bad leadership, and have them adopt all the ways in which highly reliable organizations work. It's a very incremental process — and that's what we try to lay out."

Ultimately, he continues, "We are looking at all the leverage points we think we can be helpful with in terms of what can be done, and how we can help with how organizations take those next steps."

Part of that process involves the development of a self-assessment tool. "We're just at the point of putting the first prototype together and 'crash-testing' it," says Chassin. The tool is outlined in the paper, he adds, but says that more details will make it a much more useful and more focused tool to use. "We do not have a fully developed roll-out plan, but we hope within a couple of months to be able to talk more about that," he says.

[For more information, contact:

  • Yosef D. Dlugacz, PhD, Senior Vice President, Chief of Clinical Quality, Education and Research, Krasnoff Quality Management Institute, North Shore - LIJ Health System. Phone: (516) 465-8440. E-mail:
  • Patrice L. Spath, Brown Spath Associates, P.O. Box 721, Forest Grove, OR 97116. Phone: (503) 357-9185. E-mail:

Changing the culture is greatest challenge

Of all the keys to achieving high reliability, perhaps the greatest challenge is the creation of a culture of safety. The Joint Commission's Mark R. Chassin, MD, FACP, MPP, MPH, president, and Jerod M. Loeb, PhD, executive vice president, Health Care Quality Evaluation, outlined a pathway to such a culture in their paper (See the article above.) "We borrowed from someone who is probably, in my opinion, the world's leading expert in safety, Jim Reason, who invented the 'Swiss cheese' model, as outlined in the Health Affairs article," says Chassin. That framework, he says, is "trust, report, improve."

"First, you must establish all levels of trust that are essential if the environment to high reliability is to be established," Chassin adds. "In high-reliability organizations the culture they establish allows, engenders and demands everyone who works there to define and report problems when they are very small — way upstream from causing a catastrophe."

Such an approach, he explains, makes problems much easier to fix before they fester. "In healthcare, much too often we're behind the 8-ball and looking at patients who've already been harmed and searching for the cause; that's no way to develop a highly reliable organization," says Chassin. "In order to achieve this, workers need to trust one another that they will not be blamed for errors, and trust that management won't either, nor will they sweep problems under the rug — but rather that problems will be dealt with appropriately and will be fixed." Although The Joint Commission has for a couple of years required that organization leaders be responsible for establishing a culture of safety, "we're very far away" from achieving that goal on a widespread basis.

"You can't deny the importance of the culture of the organization; otherwise, reliability will go the way of re-engineering, TQM and everything else that's been thrown at the system and not worked," says Patrice L. Spath of Brown Spath Associates, Forest Grove, OR. "You've got to talk about the concept of collective mindfulness, which comes out of the work of Weick and Sutcliffe."

As for quality managers, says Spath, "they should understand the relationship between what happens on the frontlines of care and in an organization so that when expected improvements do not occur or are not sustained at the frontlines, they can make the case to leaders of the importance of looking at the culture and how that is preventing sustainable improvement. That's why they need to become familiar with the concept of collective mindfulness."