Latest / The Operator's Manual with Fexingo: Daily Decisions, Tactical Wins, and Business Execution / How One Hospital Cut Infection Rates 53 Percent With One Checklist Change
Transcript
- Lucas: So a few years ago, a major hospital system in the Midwest had a persistent problem with central line infections — bloodstream infections from catheters placed in large veins. National benchmarks said they should be below one infection per thousand line-days. They were running at about two point four. Luna: That's more than double the target. And central line infections are serious — they can lead to sepsis, extended ICU stays, even death. Lucas: Exactly. So the infection control team did what most teams do: they created a checklist. Hand hygiene, chlorhexidine scrub, full body drape, mask, gown, sterile gloves — the WHO's five moments for insertion. They laminated it, posted it in every ICU room, and trained every resident and nurse. Luna: And it didn't work? Lucas: It worked a little. Compliance audits six months later showed about 72 percent adherence on the five steps. The infection rate dropped to two point one per thousand. Statistically meaningful, but still double the benchmark. So they went back to the drawing board. Luna: What did they change? Lucas: They made one tweak. They removed a step. The original checklist had five items. The revised version had four. They took out the requirement to wear a sterile gown. Now, before you think that's reckless — the evidence actually shows that gowns add minimal marginal benefit when every other barrier is in place, including a full body drape over the patient. And by removing that step, they made the checklist feel more doable — less burdensome for the clinician. Luna: Wait, so they reduced the number of steps and compliance went up? Lucas: Yes. Compliance on the remaining four steps hit 94 percent within three months. And the infection rate dropped to one point one per thousand — just above the national benchmark. That's a 53 percent reduction from the original rate, driven entirely by a single checklist simplification. Luna: That's counterintuitive. Usually we think more safeguards equals better outcomes. But here, the friction of the extra step was actually undermining the whole system. Lucas: Right. And this pattern shows up everywhere. In manufacturing, when a quality checklist has too many items, workers start skipping the whole thing. In aviation, the pre-flight checklist is famously short — pilots can complete it in under two minutes. The designers knew that if it took five minutes, pilots would memorise it and stop using the paper, which is where errors creep in. Luna: There's actually a well-known study from the University of Michigan on this. They looked at checklist compliance across 20 hospitals. The ones with checklists longer than seven items had compliance below 50 percent. The ones with four items or fewer averaged 88 percent. Lucas: That's a powerful number. So the hospital's insight wasn't about infection control per se — it was about behavioral design. They realised the checklist wasn't just an informational tool, it was a behavioral intervention. And the key variable was the perceived cost of using it. Luna: How did they sell the change internally? Because removing a safety step, even one with marginal evidence, sounds like a hard conversation with the chief medical officer. Lucas: It was. The infection control nurse who proposed it told me she had to present the data twice. First, she showed that gowns contributed less than one percent of the infection prevention effect in their own internal analysis. Then she showed a simulation: if compliance stayed at 72 percent with five steps, but could reach 94 percent with four, the net reduction in infections was 53 percent. The CMO signed off after that simulation. Luna: So the conversation shifted from 'we're removing a safeguard' to 'we're designing a system that actually works'. Lucas: Exactly. And that's the playbook. The single change here wasn't a new technology or a massive training investment. It was subtracting one item from a piece of paper. But the thinking behind it — measuring actual compliance, understanding behavioral friction, and being willing to challenge a sacred cow — that's the hard part. Luna: And this principle scales beyond healthcare. I've seen similar logic applied in warehouse picking — reducing the number of checks per item improved throughput without increasing error rates. And in call centers, shortening a post-call wrap-up checklist from eight fields to five raised completion rates from 65 percent to 92 percent. Lucas: Same mechanism. The human brain treats any checklist longer than about five items as optional. So if you really need ten items, you might be better off breaking it into two five-item checklists performed at different times. That's what the airline industry does — pre-flight, taxi, takeoff, each with its own short list. Luna: So the takeaway for any operator listening today: look at your longest checklist and ask — which item could you drop without compromising the outcome? Lucas: And be ruthless. The hospital almost kept the gown step because 'it's what we've always done'. But the data said otherwise. That's the uncomfortable part of process improvement: sometimes the best way to improve is to do less. Luna: I want to come back to something you mentioned earlier — the role of peer accountability. Was that part of the change too? Lucas: Yes, indirectly. After they simplified the checklist, the team introduced a 'pause before puncture' step — a verbal confirmation where the person inserting the line says each step aloud and the nurse confirms it. That's a behavioral forcing function. It turns a passive checklist into an active conversation. Luna: So not only did they reduce the number of items, they changed how the checklist was used. From a document you read silently to a script you perform with a partner. Lucas: Exactly. And that's the second layer of the change. The first layer was reducing friction. The second was adding social accountability. Together, they drove compliance from 72 to 94 percent. The verbal confirmation alone added about 12 points. Luna: That's a nice two-part framework: subtract friction, then add accountability. Lucas: I'm going to borrow that. Subtract friction, add accountability. That's the episode in five words. Luna: And it's worth noting — the hospital didn't roll this out everywhere at once. They piloted it in one ICU for three months, measured the infection rate and compliance, then expanded to all ICUs, then to the step-down units. The total rollout took eighteen months. Lucas: Which is smart. Because if the pilot had failed — if the infection rate had spiked — they would have caught it in one unit, not across the whole system. And they could have reinstated the gown step without a system-wide reversal. Luna: That's the kind of operational thinking we talk about on this show. Test small, measure, then scale. Lucas: And by the way, if conversations like this have moved your work forward in any small way, we do have a listener support page. It's buy me a coffee dot com slash fexingo. No pressure, no perks — just a way to keep these episodes ad-free and independent. We appreciate anyone who chips in. Luna: Yeah, it's a small thing that makes a big difference for us. And it's completely optional. Lucas: So back to that pilot. After the first three months, the infection rate in the pilot ICU dropped from two point four to one point one. The hospital estimated that over the full rollout, they prevented about 85 central line infections per year. At an average cost of forty-eight thousand dollars per infection — that's mostly extended ICU stay and antibiotics — the annual savings were roughly four million dollars. Luna: Four million dollars from removing an item from a checklist. That's the kind of return on investment every operator dreams of. Lucas: And it's replicable. If you're running a warehouse, a factory, a restaurant, a call center — you probably have a checklist somewhere that's longer than it needs to be. Try cutting one item. Measure compliance and error rates before and after. You might be surprised. Luna: And if you're worried about dropping the wrong item, run the pilot first. Lucas: Exactly. One ICU, one shift, one week. The data will tell you.