Watch the webinar here: 

KaiNexus CEO and co-founder Greg Jacobson joins host Mark Graban for the ninth episode of the Ask Us Anything series, the recurring session built around questions from webinar attendees. The longest discussion in this episode takes on a hard question from a reader: is healthcare really a decade or more behind other high-risk industries on basic safety? From there the conversation turns to the one element of culture that most promotes continuous improvement, how to train senior leaders to lead Lean rather than just sponsor it, and how to measure engagement in a way that does not mislead.

Here is what the episode covers and the thinking behind each answer.

Is healthcare behind other high-risk industries on safety?

A reader named Sam wrote in reacting to a blog post of Greg's, which had argued that healthcare is a decade or more behind other high-risk industries in its attention to basic safety. Sam pointed to widely cited estimates putting medical error among the leading causes of death in the United States, on the order of 250,000 people a year, and asked whether healthcare might be even further behind than a decade.

Greg's first move was to set the number aside. Whether the figure is 250,000 or 5,000, he said, he does not much care, and he explained why with a story. As an intern in a lecture on chest pain, he was asked what the acceptable miss rate is for a heart attack in the emergency department. The right answer is zero. Everyone understands that over a career a clinician will miss things, but the number to aim at is still zero. He also cautioned against a direct comparison with aviation or nuclear power. Those industries are not killing a quarter million people a year, but everyone interacts with healthcare, and death is part of life, so the comparison is closer to apples and oranges. What he would commit to is the conclusion: whether healthcare is a decade behind or two, it is behind, and it should admit that and adopt what other industries already worked out. He first encountered aviation-style high-reliability practices during his residency training in the early 2000s, only a few years after the 1999 Institute of Medicine report on medical error. Aviation had its reckoning earlier. In the 1970s, planes were colliding and falling out of the sky often enough that the industry, partly as a business decision, decided it had to stop.

Mark picked up the systems thread. You can set a goal of zero and still accept that people are not perfect. That is exactly where the patient safety movement and the Lean movement overlap. Do not expect people to be superhuman. Account for fatigue, for forgetting, for unclear communication, and build better systems rather than telling people not to make mistakes. Two things about healthcare struck him in particular. The first is a lack of transparency. Fear of lawsuits and reputational damage keeps errors unreported, which means there is no good data, and then the estimates that do get produced are dismissed for being imprecise, a problem more transparency would solve. Aviation, by contrast, runs non-punitive mandatory reporting. The second is the uneven public scrutiny. When automakers had safety defects, their CEOs were called in front of televised Congressional hearings, while healthcare's far larger error toll draws nothing comparable. A plane crash cannot be hidden. A medical error sometimes is not even known to the whole care team, let alone the patient's family.

Greg's answer to what it would take was cultural. People have to be expected to report, and when they do, the response should not be scorn or punishment. He used the word carefully: it should almost be celebrated. Most errors fit the Swiss cheese model, a chain of small failures rather than one person's mistake, and people should be able to surface them and feel good about uncovering and solving them. Mark added the mechanism aviation relies on. There is no hiding a crash, but there is every temptation to hide a near miss, and non-punitive reporting is what pulls those near misses into the open. For every instance of harm there are roughly eight to ten near misses, and catching a problem at the level of an unsafe condition is how you prevent the harm. Greg's understanding of the aviation rule makes the incentive explicit: report through the system within the window and you are protected; step outside the system and you are exposed.

The one element of culture that promotes continuous improvement

Asked to name a single element of company culture that drives continuous improvement, Greg did not hesitate: trust. An organization without it will not experiment and will not let people come to the improvement table, because people who feel insecure spend their energy protecting their own position rather than offering ideas. He tied it to a line he uses often. Innovation, looked at under a microscope, is continuous improvement, and the companies that innovate well tend to be high-trust companies.

Mark's addition was that trust is not a fixed condition. It has to be built, and when the starting point is mutual suspicion, between employees and leaders in both directions, leadership has to break the cycle. One concrete way is a commitment to no layoffs as a result of Lean or continuous improvement. That said, a leader can make the promise and employees may not believe it yet, precisely because the trust does not exist, so the promise is the start of a process, not the end of one. Leaders also build trust through humility, by letting go of the idea that they must have every answer. Working a problem alongside employees, asking what they see and what they would try, builds trust in both directions at once.

Training senior leaders to lead Lean

An operational excellence leader wrote in with a good situation to have. After more than a decade of bottom-up improvement, a mix of competitive pressure and some recent success with rapid improvement events had convinced senior leaders that they needed Lean education to become active Lean leaders. What should the team do, and avoid?

Mark's answer was that leaders learn Lean the same way anyone does, by doing it. They should be involved in problem solving, learn the PDSA cycle, and ideally start by improving their own work before asking their employees to improve theirs. Working a problem firsthand shows a leader that problem solving is rarely quick or linear, and that makes them a better coach. Greg offered the medical version of the same idea: see one, do one, teach one. Watching teaches something, doing teaches more, and teaching produces the deepest understanding. The practical rule is just-in-time delivery. Do not pull leaders into a year of training. Give them a small piece, have them apply it, have them teach it. Sometimes a full day out of the workflow is right, sometimes only a five or ten minute lesson is realistic.

Mark named the real obstacle: the expertise trap. People reach executive level by being experts, and Lean drops them into the discomfort of not knowing the answer. Getting past that, being willing to say "I don't know, let's go figure it out," takes coaching. And it takes the leader's explicit permission to be coached, often by a Lean facilitator several levels down. That means getting past a formal power dynamic, the same dynamic aviation studies in the cockpit and patient safety studies in the operating room. Greg offered an image for the coaching relationship. The improvement team is like a control tower in contact with many planes, each plane a department or a leader. The tower checks altitude, suggests a flight path, lets the plane fly, and checks in again. Framing the coach as a guide into a new way of working, rather than as someone telling a leader how to do their job better, is what makes it possible for a boss to take direction from a few levels down.

Measuring engagement without misleading yourself

A live question asked how to measure employee engagement and whether there is an assessment tool. Mark pointed first to signals an organization can read in near real time, more current than a survey: absenteeism and turnover tell you whether people are even showing up. The more telling measure for improvement is how many ideas people bring forward and what share of them get implemented. It is easy to track and it matters.

Greg described what KaiNexus reports. The platform captures engagement data as a byproduct of use, and improvement coaches rely on it to see which departments and teams are active. The headline number is improvements per person, annualized: below one is a warning sign, one to two is solid, and the strongest organizations run above ten to fifteen. But, echoing the earlier episode, the average can deceive. A plant showing ten improvements per person looks engaged until the data reveals that a pocket of twenty people is generating the activity credited to five hundred. So the platform also shows the share of employees with no role in any improvement, a number you want close to zero. Mark closed the episode where it started, on trust. Engagement, like trust, does not simply exist. When leaders complain that their doctors are not engaged, the question to ask is what they are doing to engage them. It is an active responsibility, and breaking an old dysfunctional cycle is leadership's job.

Key takeaways

• The exact medical error count matters less than the target, which is zero. Healthcare is behind other high-risk industries on safety, and admitting it is the first step.
• Safety improves through non-punitive reporting and systems thinking, not by telling people to be careful. Most errors are a chain of small failures, not one person's fault.
• Catch problems at the near-miss stage. For every instance of harm there are several near misses that a reporting culture can surface first.
• Trust is the cultural element that makes continuous improvement possible, and it has to be built deliberately, starting with leadership.
• Leaders learn Lean by doing it, including on their own work, and only with their explicit permission to be coached.
• A per-person improvement count is a useful engagement metric, but always check participation so a few active people do not mask a disengaged majority.

About this series

Ask Us Anything is a recurring series of short sessions answering questions from KaiNexus webinar attendees. It is hosted by Mark Graban, VP of Improvement and Innovation Services at KaiNexus, with Greg Jacobson, the company's CEO and co-founder.

See every episode in the series on the Ask Us Anything main page. Earlier episodes are also available on the KaiNexus YouTube channel and in the KaiNexus podcast archive.

See how KaiNexus helps organizations capture ideas, coach improvement, and connect daily work to strategy. See KaiNexus in action -> 

Bonus eBook: