Latest / The Operator's Manual with Fexingo: Daily Decisions, Tactical Wins, and Business Execution / How One Snowstorm Exposed a Hospital Supply Chain
Transcript
- Lucas: February 2021. A winter storm hits Texas — not the worst on record, but bad enough. For one hospital in Dallas, the power doesn't go out. The backup generators kick in. Everything's fine on the surface. Luna: So the lights stay on — but something else fails? Lucas: Right. The supply chain. The hospital relied on a single supplier for a critical component: the valve that regulates oxygen flow in ventilators. That supplier was in Houston, which got shut down by ice. Luna: One valve, one supplier, and the entire ICU backup plan collapses. Lucas: Exactly. The hospital had a backup ventilator pump — but they'd never tested it with the actual valve. When they tried to swap it in, the threads didn't match. So for three days, they had to manually bag patients while procurement scrambled. Luna: That's terrifying. And it's a classic single point of failure — something operations people warn about but rarely see in action. Lucas: What's interesting is that the hospital thought they were resilient. They had a backup pump. They had safety stock of valves — about two weeks' worth. But they never asked: what if the supplier itself disappears? The backup pump came from a different manufacturer, and the valve interface was proprietary. Luna: So the inventory was there — but the compatibility wasn't. That's a failure of integration, not just stock levels. Lucas: Right. They had a 'two-bin system' for most supplies — when one bin empties, you order more. But for the critical valve, they treated it like any other part. No one had done a single point of failure analysis. Luna: And I'm guessing after the storm, they did. Lucas: They did. And this is where it gets tactical. The operations team mapped every critical component — about 200 line items — and asked two questions: can we source this from at least two suppliers? And if not, can we standardise the interface so any backup works? Luna: That second question is huge. Because sometimes you can't dual-source — the market is just too small. Lucas: Exactly. For the ventilator valve, there was only one manufacturer that met FDA specs. So they couldn't dual-source. Instead, they did two things: they bought an adapter kit that made the backup pump compatible, and they changed their ordering rule from 'order when stock hits two weeks' to 'order when stock hits four weeks.' That doubled their safety stock. Luna: But doubling inventory costs money. Was there pushback? Lucas: There was. The CFO pushed back initially — said it tied up working capital. But then they ran the numbers: the cost of carrying extra inventory for a year was about forty thousand dollars. The cost of a single day of manual bagging — overtime, risk, patient safety — was estimated at over two hundred thousand. So the trade-off was clear. Luna: Yeah, that's a no-brainer when you frame it that way. Lucas: And that's the core lesson here: operational resilience isn't about eliminating risk — it's about making the cost of failure visible. The hospital's system looked fine on paper. The backups were there. But no one had stress-tested the actual scenario. Luna: It reminds me of something I heard about NASA. They famously test every component in the worst-case conditions — and then add a margin. But hospitals usually don't have that luxury. Lucas: They don't. But they can borrow the principle. One thing the Dallas hospital did post-storm was run a 'failure mode and effects analysis' — FMEA — on the top twenty risks. For each one, they asked: what's the likelihood, what's the severity, and what's our current detection capability? Then they prioritised based on a risk priority number. Luna: So it's a systematic way to find the single points of failure before they break. Lucas: Exactly. And the results were eye-opening. They found that the oxygen valve was actually number three on the list — but it had the highest severity because it affected life support. So it got bumped to priority one. Luna: I wonder how many businesses do that kind of analysis. I suspect most don't until something breaks. Lucas: That's the pattern. And it's not just small companies — large ones too. There's a famous case from 2011 where a flood in Thailand shut down hard drive production for months. Companies like Apple and Dell had been running lean, just-in-time inventory. They hadn't mapped that single source. Luna: Right, the Thailand floods. That cost the industry billions. Lucas: And afterwards, a lot of them built in more redundancy. But the key point from the hospital example is that redundancy alone isn't enough. The backup pump was there — but it was incompatible. Redundancy without integration is just expensive inventory. Luna: So the fix isn't just 'buy more stuff.' It's 'buy the right stuff and make sure it works together.' Lucas: Exactly. And that's what the hospital did. They created a 'critical spares' list with photos, part numbers, and compatibility notes. They put it on a laminated card in every supply closet. And they trained the biomed team to do a quarterly compatibility test — literally plugging the backup pump into the ventilator and running it for ten minutes. Luna: Something so simple, but it would have caught the problem immediately. Lucas: Right. So the operational lesson here is: map your single points of failure, but don't stop at the inventory level. Trace the interfaces. Test the backups under the conditions they'll actually face. And when you find a gap, the fix might not be more inventory — it might be a fifty-dollar adapter. Luna: That's a great takeaway. And honestly, the kind of thing that makes this show valuable — thinking through these real-world breakdowns. Lucas: Yeah, and it's one of the reasons we're able to do this episode ad-free. A handful of listeners chip in monthly through buy me a coffee dot com slash fexingo, and that literally covers the time it takes to research and produce these deep dives. Luna: It's true. And it means we don't have to chase sponsors or clickbait. We can just focus on the operations story that actually matters. Lucas: So back to the hospital. After the storm, they also changed their procurement contracts. They added a clause requiring suppliers to notify them at least ninety days before any manufacturing change — like a valve redesign. That alone would have flagged the incompatibility before the snow ever fell. Luna: That's smart — moving from reactive to proactive. Instead of waiting for failure, you create a signal upstream. Lucas: Right. And the results speak for themselves. Eighteen months later, they had another supply disruption — a different supplier had a fire. But because they'd done the mapping and had the adapters and the higher safety stock, they didn't miss a single ventilator hour. Luna: So the investment paid for itself in the first real test. Lucas: Exactly. And that's the thing about operational resilience: you never know when the test will come. But when it does, the cost of failing it is almost always higher than the cost of preparing. Luna: So for anyone listening — maybe start by picking one critical component in your own operation and asking: is this a single point of failure? And if so, what's the cheapest way to make it not? Lucas: Great challenge. That's it for today. We'll be back tomorrow with another episode.