Watch the Recording
Most process improvement webinars work at one of two altitudes. Either they describe a single project in detail -- one department, one problem, one before-and-after -- or they zoom out so far that the advice becomes generic enough to apply to any organization with the proper nouns swapped out. Neither altitude is particularly useful for a CI leader trying to figure out how to operate at the scale of a multi-hospital health system.
This session sits in the more useful middle altitude. Jason Coons walks through how Kettering Health Network built the Process Excellence function from a team of one in 2010 to a team of eight in 2015, expanded the mandate from two hospitals to seven, and developed the training infrastructure, project governance, and analytics integration to support system-wide improvement work. The numbers, the structures, the timelines, and the lessons learned are all specific. The session is essentially the operating manual for a health-system CI function at the stage most health systems are trying to reach.
Several details are worth flagging up front. The function moved from two hospitals to seven with a hiring plan that took roughly a year to complete, because finding the right talent -- people who fit both the skill profile and the culture -- proved harder than expected. The team's training program runs three tiers (Novice, Advanced, and a Certified Process Excellence Leader program that spans 18 weeks) and is built around the requirement that every advanced participant complete a real project. The project governance model uses a strategic alignment score, organizational need assessment, and change readiness assessment before any project gets approved. And the team's recent push has been to integrate process excellence with analytics and strategy -- not as three separate functions but as a single governance structure -- because all three are required for any of them to produce sustained results.
The Information Systems story that opens the session is a useful framing device for everything that follows, because it shows the full progression: training the leaders, mapping the value stream, running a sequence of rapid improvement events over 18 months, building local Kaizen capability, and reaching the point where the department had its own internal leaders capable of running rapid improvement events without external facilitation. By the time of the webinar, IT had 325 submitted opportunities, 234 implemented, and roughly one idea per person per year flowing through daily problem-solving channels alone.
Jason Coons is Network Director of Process Excellence at Kettering Health Network. His team is responsible for driving Lean, process improvement, strategy deployment, and analytics across the network. He has an extensive background in hospital operations, change management, project management, and analytics. He holds a Master's in Health Administration from Ohio University and a Bachelor's of Science in Industrial Engineering from the University of Cincinnati. His certifications include Six Sigma Green and Black Belts, Lean Healthcare, Project Management Professional, and ITIL Foundations.
Jason opened with a story about Kettering's Information Systems department, which approached Process Excellence in late 2013 with a recognition that adding more headcount was not going to produce the outcomes the department wanted. They had about 250 employees and needed to fundamentally change how they delivered value. They picked Lean as the anchor.
The path Jason and his team walked them through is a useful template for any large department considering a similar move. Step one was awareness -- four hours of training for managers and above, ninety minutes for staff, oriented to basic concepts and to defusing the fear of the unknown that any new initiative generates. Step two was the value stream analysis, mapping out the services IT provided across all their service lines and identifying a future state. That future state translated into an 18-month roadmap with six or seven rapid improvement events.
Two things made the IT engagement notable. First, the events compounded. Each one taught the department something that informed the next one, and the leadership team's capacity to lead the work grew along with the work itself. By the time the 18 months ended, eight to ten leaders across IT could run rapid improvement events on their own. Second, the partnership with KaiNexus during this period is what allowed the team to track work that was no longer just rapid improvement events but also daily Kaizen happening at the front lines. The platform became, in Jason's framing, the way they could engage the 200-plus staff who were not in any given rapid improvement event but had ideas worth capturing anyway.
The end-state for IT at the time of the webinar was a department running its own rapid improvement events, addressing every aspect of its service delivery (from ticket intake through fix-and-break resolution), and averaging roughly one idea per person per year in daily problem solving. The 325 submitted opportunities and 234 implemented improvements are the artifact of that work.
That's one department. Now multiply it across seven hospitals, two freestanding emergency rooms, 1,500 licensed beds, 11,000 employees, and 50-plus inpatient units. That is the scope Process Excellence had to figure out how to support.
The Process Excellence function at Kettering organizes its work into three services: training and development, process improvement, and change implementation. The architecture matters because it reflects an operating philosophy that runs through the entire session.
The philosophy is summarized in Jason's framing of the department's purpose: improve people, process, and technology -- in that order. Most organizations, in his view, get the order wrong. They start with technology and assume technology will fix their problems, which usually just speeds up defects or makes already-bad processes more efficient. Kettering starts with people -- the skills, training, and positioning of the staff who actually do the work -- then moves to process, then layers in technology as an enabler.
That ordering shows up in how the three services are designed.
Training and development is the foundation. The Novice program is four hours, oriented to the basics, with a six-week post-class requirement to conduct a waste walk, identify an opportunity for improvement, log it in KaiNexus, and drive it through implementation using 5S, visual management, standard work, and process mapping. The Advanced program is a full day, focused on A3 problem solving and the PDCA approach, with a twelve-week post-class requirement to identify and work a real problem with a small team, with the A3 attached in KaiNexus for visibility. The Certified Process Excellence Leader (CPEL) program is the most demanding -- 18 weeks, seven courses, starting with the cultural conditions that need to be in place before improvement work can succeed, then moving through data analysis, value stream analysis, and rapid improvement event leadership. CPEL participants lead their own large-scale project during the 18 weeks, mentored by the Process Excellence team throughout.
Process improvement is the bread and butter -- the project work, the operational data analysis, the simulation work, the facility design modeling that lets the team mock up 3D scenarios to test patient flow before construction starts. Most projects use a value stream analysis approach with three to five rapid improvement events spread across three to six months, though the team adapts to the constraints of the area they are working in. Some areas can't free up team members for three- or five-day events; in those cases, the work happens through task forces meeting weekly or every other week for multi-hour sessions.
Change implementation is project management and strategy deployment work, increasingly attached to the network's major strategic initiatives. Opening new facilities, launching new strategic processes, hitting milestones on initiatives that span the full system -- Process Excellence is involved in planning and execution because the work fails without that discipline. This is where the function moved from being a services group that responds to requests into being an embedded part of how strategy gets executed.
The CPEL program produced its first cohort in 2014, and the project examples Jason shared are useful for understanding what "real project" means at this level of training.
A flash sterilization rate reduction project initially targeted the industry standard of 5 percent but reduced it all the way to zero, with flash sterilization becoming an exception rather than a routine practice. Three months of operation at 0 percent followed.
An orthopedic pre-admission testing project produced a 50 percent reduction in pre-admission testing time for total hip and total knee patients.
A pain management project at one facility improved Press Ganey scores by examining the process of getting medications to patients and resetting patient expectations.
An IT and Physician Network hardware deployment project achieved a 50 percent reduction in hardware issues and eliminated build issues during the extraction training rollout.
A care coordination project in an emergency department at one of the hospitals improved follow-up rates with e-visits and increased follow-up charges.
The point of listing these is not the specific numbers but what they signal about the depth of the training. CPEL participants are not learning Lean concepts abstractly; they are running real projects with real impact while they learn. That changes both the engagement and the durability of the training -- and it changes the kind of leader the program produces.
Kettering's project governance is one of the most explicit and practical parts of the session. Every project that wants to enter the Process Excellence pipeline starts with a standardized business case signed off by a VP sponsor. From there, the project moves through three evaluation gates.
The first is a strategic alignment score. This was developed by an executive director in the IT division who wanted a more rigorous way to prioritize capital needs. He spent time meeting with over a hundred leaders, mapped the network's strategic plan against key prioritization factors, and built a weighted scoring model. Projects get scored, ranked, and prioritized against the alignment criteria before any other consideration.
The second is organizational need -- what Jason called the "gut check." A project can score well on strategic alignment but still not fit the immediate organizational moment. The network has a specific cross-system strategy called "One Best Practice" that drives standardization across the seven hospitals, and projects get evaluated against that lens as well.
The third is a change readiness assessment. This is the part most CI functions skip, and Jason's team treats it as critical. The assessment looks at vision and business case clarity, leadership engagement, implementation effectiveness, and sustainability. It runs through a series of yes/no questions covering each of those areas, and the output is not a green-light/red-light decision but a list of areas that need to be addressed before the project starts. If a project is not ready, the team gives feedback to the requesting area and re-engages when the conditions improve.
The change readiness piece is what protects the function from being pulled into projects that are doomed for reasons that have nothing to do with methodology. A project led by a manager who is not bought in, in a department that is exhausted from previous changes, with no clear sponsor, is going to fail regardless of how well it is run. Saying so upfront and helping the area get to readiness is a more honest service than agreeing to lead a project that will burn the area's appetite for future improvement work.
Beyond governance, Kettering uses a standardized project framework that runs from business case through initiation, planning, execution, and closing. The framework includes a standard checklist project managers use to ensure key steps and approvals don't get skipped, and explicit "stop signs" where the project pauses to confirm alignment, sign-offs, and executive understanding before proceeding to the next phase.
The execution phase varies in cadence based on what the area can support. Some teams can free up staff for three- to five-day rapid improvement events. Others -- especially 24/7 clinical environments where pulling staff means the work stops -- need a different rhythm. The team adapts the cadence rather than forcing a standard format that doesn't fit, but the underlying framework remains consistent.
Reports out at the end of every rapid improvement event are mandatory, with VP sponsors expected to attend. The team's framing of the report out is worth noting: they are looking for a green light from leadership, but the conversation is structured as "here is where we are headed, here is where we are going, do you see any major issues or concerns." Leaders can voice concerns and the team can redirect. In Jason's experience, they have never had a leader walk in and shut down an event at the report out -- because the expectations are set early and the sponsors have been engaged throughout.
The most forward-looking part of the session was Jason's description of how the team integrated three functions that most health systems run separately: process improvement, strategy, and business intelligence/analytics.
The integration starts with the network's BI governance team, which includes a core team and an executive team that evaluates data analytics needs against the strategic plan and prioritizes them. From there, Process Excellence sits at the center of the analytics initiatives -- because the team's view is that creating reports and dashboards without driving meaningful action through process improvement is not worth doing. Every new analytics initiative launched across the organization links back to strategy and has an associated Process Excellence initiative attached to it.
The length-of-stay work is the worked example. Length of stay was a number-one operational priority across the system in 2015 -- seven hospitals, fifty-plus units, with the goal of reducing actual length of stay to the geometric length of stay benchmark from CMS. The Process Excellence team was pulled in late 2014 to organize the work at the network level. The analytics platform launched in early 2015 surfaced length-of-stay data daily, tracking actual versus geometric and the gap between them.
KaiNexus became the organizing platform for the project work itself -- the overall project plus the associated rapid improvement events, all loaded into one place, with all leaders and project teams added so the work was visible to everyone involved. Five months in, the team had achieved a 34 percent reduction in the gap between actual and geometric length of stay, with about $1 million in associated savings. Jason flagged the platform's role explicitly: it drove collaboration, kept individuals engaged, generated ideas that didn't only surface at scheduled events, and created the accountability and transparency that holding fifty units' worth of work together would have been impossible without.
By the end of 2014, the Process Excellence function had generated over 700 opportunities for improvement system-wide, with roughly 500 completed and a 76 percent change rate. Cost savings exceeded $1.2 million. Revenue gains were just under $400,000. Hours saved totaled just over $1 million -- time freed up so staff could provide more value to the organization rather than absorbing waste in their daily work.
About a third of the opportunities tracked were daily Kaizen items led by front-line staff, not project work facilitated by the Process Excellence team. That ratio matters. It is the signal that improvement is happening in the daily fabric of the work rather than only at the project level.
Year over year, total savings since inception came to just under $5 million. The ROI trajectory was a 2-to-1 return in 2011, nearly 5-to-1 in 2012, then below 1-to-1 in 2013 -- the only year the function operated below break-even, driven by onboarding new resources and the time it took to bring them up to speed -- and back to 2-to-1 from 2014 forward.
The training numbers are equally instructive. By the time of the webinar, the team had trained nearly 2,500 leaders since 2011. About a thousand had gone through formal training (Novice, Advanced, or CPEL); the rest had been trained inside projects. Follow-through on the Novice waste walk and OI requirement runs at 55-65 percent. Advanced and CPEL participants are required to complete a project, and CPEL has an application process with a VP sponsor and project pre-selection, though some participants still drop out when they realize what the program requires.
Several of the lessons Jason called out at the end are worth pulling forward because they are the kind that don't show up in generic CI advice.
The hiring problem turned out to be larger than the team expected. Finding people who met the skill profile and were also a culture fit took a year for the five positions added in 2013. The team built a six-step interview process to assess that fit -- phone interview, project presentation, skill-based panel interview, case study under time pressure, behavior-based panel interview, and culture-fit assessment. The case study is worth noting: it puts candidates on the spot with too much information and too little time, not to test whether they get the right answer but to surface how they think under pressure.
Runway length is something most CI leaders don't ask about until they are too far in. Jason flagged it as a critical lesson from starting at Kettering: understand how long leadership sees the journey taking before you accept the role or take on the mandate. If they expect transformation in a year, that's a different job than if they understand it's a multi-year build.
Data management and governance is where most CI functions discover they don't have the foundation they assumed they had. Kettering had moved to a system-wide EHR in 2011, which gave them data -- but as Jason put it, the question of whether all seven hospitals defined, tracked, and measured the data the same way is its own project. The phrase "if you can't measure it, you can't improve it" sounds like a cliche until you discover that two of your hospitals are counting the same metric differently.
Hierarchical management as a roadblock came up in the Q&A. Jason's response was that the cultural readiness assessment is partly designed to surface this upfront. How does the leader treat their employees? Are they pushing improvement down on staff or creating space for staff to drive it? The defined roles for projects (VP sponsor, physician champion, project leader, facilitator, project team) help establish norms during the project that gradually shift the culture afterward. Setting expectations early and reinforcing them often is the mechanism.
President-driven culture, raised in another Q&A question, has a specific operational meaning in Jason's framing. It is not about the president giving speeches. It is about whether the president visits the gemba, whether VPs spend time inside rapid improvement events at least annually, whether senior leaders use Lean concepts in their own management work, and whether they connect the methodology to the strategic problems they are trying to solve. Without that modeling at the top, the function never breaks out of being a middle-of-the-organization initiative pushed down toward the front lines while resistance builds above.
A question came in about whether organizations could achieve 30-50 ideas per person per year, the way some manufacturing companies do. Jason was direct: he doesn't recommend aiming for those numbers in a health system. The risk is that you generate volume for the sake of volume, the organization can't respond fast enough, and people disengage because they feel they are submitting ideas into a void. Quality over volume. Two to three ideas per person per year, taken seriously and acted on, produces more durable change than fifty ideas per person ignored.
Mark added the framing that the manufacturing 30-50 number usually reflects small improvements at very short cycles -- the two-second-Lean style of constant micro-improvement -- which works in some manufacturing settings but maps awkwardly onto healthcare workflows. The right number is whatever the response infrastructure can actually support without breaking the trust that makes participation work.
The Process Excellence function at Kettering was building toward something specific: improvement work that operates as a coordinated system rather than as a collection of well-intentioned local efforts. That goal puts specific demands on the infrastructure underneath it.
When IT was running its 18-month roadmap of six or seven rapid improvement events, the team needed visibility into who owned what, what was being worked, what was complete, and where attention was needed -- across a department of 250 people. A spreadsheet would have worked for a department of ten. At 250, it stops scaling. The platform's role at that stage was to make the in-flight work visible to people who were not in the room for any given event.
When the network-level length-of-stay work spread across fifty units and ten facilities in 2015, the same dynamic operated at a different magnitude. Loading the overall project, the associated rapid improvement events, the leaders, and the project teams into a single shared platform was what made it possible to keep fifty units' worth of work coordinated without requiring weekly all-hands meetings that no one had time for. The platform's role was the accountability and transparency that distributed work requires.
When the team trained nearly 2,500 leaders -- with every Advanced and CPEL participant required to complete a real project, and Novice participants expected to do a waste walk and a real OI -- the platform became the place where the training requirements got tracked, where A3s could be attached and reviewed in something other than email, and where the Process Excellence team could give feedback during the 12-week project window rather than waiting for the next scheduled meeting. The platform's role at that stage was extending the team's coaching capacity beyond what eight people could do face-to-face.
None of this changes the fundamentals Jason called out. People come before process. Process comes before technology. The leadership behaviors -- President-driven culture, hierarchical management addressed honestly, sponsors actively engaged -- have to be real, and no platform substitutes for them. What the platform does is remove the friction that otherwise prevents the leadership work from compounding. The improvements get captured. The work stays visible. The accountability holds. The local wins become spreadable across the network rather than trapped in the unit that produced them.
How did Kettering Health structure its Process Excellence function? The function reports up through the Information Technology and Analytics division and provides three services: training and development, process improvement, and change implementation. The team grew from one person in 2010 to eight people by 2015, expanding from two hospitals to all seven hospitals plus the freestanding emergency rooms and the employed physician network. The structure deliberately integrates process improvement with analytics and strategy rather than treating them as separate functions.
What is the three-tier training model? Novice training is a four-hour foundational class with a six-week post-class requirement to conduct a waste walk, log an opportunity for improvement, and drive it through implementation. Advanced training is a one-day class on A3 problem-solving with a twelve-week post-class requirement to work a real problem with a small team. The Certified Process Excellence Leader (CPEL) program is 18 weeks across seven courses, with participants leading a large-scale project mentored by the Process Excellence team. Participation in the next tier requires successful completion of the previous tier's project work.
Why does Kettering use a three-gate project governance model? Because there are more project requests than the team can handle, and projects that aren't ready -- regardless of methodology -- tend to fail and burn the requesting area's appetite for future improvement work. The first gate is a strategic alignment score against weighted criteria mapped to the network's strategic plan. The second is an organizational need assessment that includes a check against the network's "One Best Practice" standardization strategy. The third is a change readiness assessment covering vision and business case, leadership engagement, implementation effectiveness, and sustainability. Projects that aren't ready get feedback and re-engagement, not approval.
What's the realistic timeline for building a system-wide CI function? Kettering's journey took five years to reach the state described in the webinar, and Jason was clear that they are still in the middle of it. The function started in 2010, expanded to network mandate in 2013, completed its initial five-person hiring buildout in 2014, and was integrating analytics and strategy throughout 2015. Anyone expecting transformation in twelve months is misreading the work. Asking leadership how long they see the runway before accepting the mandate is one of the most important early questions.
How does Kettering handle the tension between project rigor and front-line daily improvement? By running both in parallel and tracking them in the same place. Projects use value stream analysis with three to five rapid improvement events; daily Kaizen happens on the front lines through opportunities for improvement logged by staff. Roughly a third of the 700 OIs in 2014 were daily Kaizen items rather than project work. The platform is what allows the team to see both flows of activity in one view rather than tracking projects in one system and Kaizen in another.
What should an organization look for when hiring CI talent? Kettering's six-step interview process screens for both skill fit and culture fit: phone interview, project presentation, skill-based panel interview, case study under time pressure, behavior-based panel interview, and culture-fit assessment. The case study is designed to surface how candidates think under pressure with too much information and not enough time -- not to test whether they get the right answer. Finding five hires who passed all six gates took roughly a year, which Jason flagged as something he wishes he had budgeted for from the start.
How do you handle leaders who resist the strategy you've developed? Start by understanding why they resist. Past experience? Fear that the function exists to cut costs? Disagreement with the methodology? Some resistance dissolves when the leader sees results. Some leaders never change, in which case the question becomes whether their resistance is passive (manageable) or actively disruptive (a different problem). The cultural readiness assessment in the project governance model is partly designed to surface this upfront so the team can adjust how it engages with the area.
Why does Kettering track training participation as a metric? Because the function will never grow large enough to lead all the improvement work itself, which means scaling depends on building improvement capability into the leaders across the network. Tracking how many leaders have been formally trained, how many have completed real project work, and how many can lead rapid improvement events on their own gives the team visibility into whether organizational capability is actually growing. The 2,500-leader number across five years is the artifact of treating capability building as a measurable output rather than a hope.
What's the role of the report-out at the end of a rapid improvement event? It's a structured moment to align leadership with the team's direction and surface any concerns before implementation begins. VP sponsors are expected to attend. The team presents what they've done and where they're going, and the conversation is framed around getting a green light from leadership while making space for concerns to be voiced and the team to redirect if needed. In Jason's experience at Kettering, no leader has shut down an event at the report-out -- because expectations are set early and sponsors stay engaged throughout the work.
How should a CI team think about deploying a platform like KaiNexus across the organization? Carefully, and not by flipping the switch for everyone at once. Kettering had roughly a thousand of its 11,000 employees active in the platform at the time of the webinar, with intentional decisions about which departments to activate and when. The risk of turning it on for staff without their leadership being ready is that ideas come in faster than the leadership can respond to them, which trains employees that the system isn't real. Activate departments as the leaders are ready to handle the response load.
Why integrate process excellence, analytics, and strategy under one structure? Because each of the three fails when run independently. Analytics without process improvement creates reports nobody acts on. Process improvement without strategy improves random things rather than the things that matter. Strategy without analytics and improvement infrastructure is words on a planning document. Kettering's integration model has all three functions sitting under shared governance, with every analytics initiative linked to a strategy priority and an associated Process Excellence initiative. The integration is what makes the work coordinate at the network level rather than fragment.
What's the right way to think about ROI on a process improvement function? Kettering's function ran at 2-to-1 ROI in 2011, nearly 5-to-1 in 2012, dropped below 1-to-1 in 2013 during the hiring buildout, and returned to roughly 2-to-1 from 2014 forward. The dip during scaling is worth noting because most CI functions go through it and most don't talk about it openly. Beyond the financial ROI, Kettering also tracks the customer experience net promoter score, project quality, employee engagement, and training participation -- because optimizing only for financial return drives behavior that erodes the function's longer-term value.
.svg)
Copyright © 2026
Privacy Policy