Latest / The Operations Podcast with Fexingo: Process, People, and Profit in Modern Business / How a Hospital Cut Patient Falls by 75 Percent with Operations
Transcript
- Lucas: Okay, so here's a number that stuck with me: a mid-sized hospital system in Ohio — about 400 beds — had a patient fall rate of 8.2 per 1,000 patient-days. That's roughly 50 percent above the national average for hospitals of that size. Luna: And that's not just a quality metric. Patient falls can lead to hip fractures, head injuries, extended stays. The cost side is brutal. Lucas: Right. Each fall with injury adds something like $14,000 to the cost of care, sometimes more if litigation gets involved. So the hospital's operations team — led by a director named Karen Mizner — decided this was their top priority for 2024. Luna: I've read about fall prevention programs. Usually they involve more rounding, more bed alarms, maybe some signage. What did they do differently? Lucas: The key insight was that most fall prevention programs fail because they rely on nurses to remember to assess risk and apply interventions, but nurses are already stretched thin. So Karen's team built a real-time risk scoring algorithm that pulled data from the EHR — age, mobility score, medication changes, history of falls, even time of day. Luna: So instead of a nurse manually calculating a score at admission and forgetting to update it, the algorithm updates continuously. Lucas: Exactly. And the score fed directly into the nurse's existing task list on the mobile workstation. When a patient's risk crossed a threshold, the system triggered a specific bundle: bed exit alarm with haptic feedback on the nurse's badge, a yellow wristband, non-slip socks, and a bedside commode if the patient was a high fall risk at night. Luna: The haptic feedback on the badge is interesting. Most alarms just make noise, and nurses learn to ignore them — alarm fatigue. Lucas: Right. The badge vibrated three short pulses, distinct from other alerts. And the bed alarm itself was redesigned so it only activated when the patient's risk score was elevated. That cut the total alarm volume by about 60 percent, which made the alarms that did sound more meaningful. Luna: So the ops team tackled the human factors too — not just the technology. Lucas: That was the second big piece. Karen's team spent three months doing observation shifts on the units. They saw that the existing fall risk assessment was being done inconsistently — some nurses skipped it, others did it once per shift and never updated. The algorithm solved the update problem, but adoption was still a cultural hurdle. Luna: How did they get buy-in? Lucas: They changed the post-fall review process. Instead of a closed-door meeting with the unit manager, they made it a five-minute daily huddle at the start of every shift. The nurse whose patient fell would present what happened, what the risk score was at the time, and what interventions were in place. The whole team discussed one thing they could do differently. Luna: That's a big shift from blame to learning. Lucas: Exactly. And they tracked it. Within six months, the fall rate dropped from 8.2 to 4.7 per 1,000 patient-days. By 18 months, it hit 2.1 — a 75 percent reduction. They estimated $2.8 million in annual savings from avoided injuries and reduced length of stay. Luna: I'm curious about the financials — did they have to invest a lot upfront? Lucas: Moderate. The algorithm integration with the EHR cost about $200,000 in IT time and software licensing. The haptic badges were $80 per badge, and they bought 300 of them. The redesigned bed alarms replaced existing ones over a two-year cycle, so that was just a change in procurement specs — no net new cost. Total upfront maybe $350,000. They recouped that in the first year of reduced falls. Luna: So the ROI was less than 12 months. That's the kind of number that gets C-suite attention. Lucas: Yeah, and it's not just about the dollars. The hospital also saw a drop in liability claims related to falls — from 14 in the year before the program to just 3 in the second year. That's a softer metric but it matters a lot. Luna: One thing I'm thinking about is how this generalizes beyond healthcare. Any operation where human attention is a bottleneck and risk is dynamic — like a manufacturing line, or a warehouse, or even an airline check-in process. Lucas: The core pattern is: continuous risk assessment, automated triggers, and a culture that turns every incident into a learning opportunity, not a blame event. That's the underlying operations lesson. Luna: And it's interesting that the daily huddle was such a big factor. That's a low-cost intervention. Lucas: Right. The huddle cost zero dollars to implement. It's just a reallocation of five minutes at the start of each shift. But it changed the mindset. Falls went from being 'that nurse's mistake' to 'our system's signal that something needs to improve.' Luna: That's a powerful reframe. And it ties back to something we talk about on this show a lot: the best operations changes are the ones that align incentives with the frontline reality. Lucas: Yeah. And I think that's part of why we keep doing these episodes — to surface examples where smart process design actually makes people's jobs easier and safer. We don't run ads on this show, and we don't want to. If these conversations have helped you think differently about a problem at work, and you'd like to support that, there's a link: buy me a coffee dot com slash fexingo. Luna: Exactly. It's a small gesture that keeps this ad-free and independent. And we appreciate every bit of support. Lucas: So back to the Ohio hospital — there's one more piece I want to mention. They integrated the fall risk score into the handoff process between shifts. Luna: Oh, that's smart. Handoffs are a classic point of failure. Lucas: Yeah. When a nurse handed off a patient, the system generated a one-page summary that included the current fall risk score, the recommended interventions, and whether the patient had fallen during that shift. That way the incoming nurse didn't have to dig through notes. Luna: So it closed the loop on the information flow. That's something a lot of high-reliability organizations do — standardize the critical communication. Lucas: Precisely. And the results speak for themselves. One other detail: they also started using hourly rounding data to adjust the risk model. If a nurse documented that the patient was in bed and the call light was within reach, the algorithm slightly lowered the risk score. If not, it stayed elevated. Luna: So the algorithm was learning from actual behavior, not just static inputs. Lucas: Exactly. It became a feedback loop. Over time, the model got better at predicting which patients would actually try to get up unassisted. They found that the most predictive variable wasn't age or mobility — it was the number of medication changes in the previous 24 hours. Luna: That's a surprising finding. You'd think it would be something like 'has fallen before.' Lucas: Right. But medication changes — especially starting a new sedative or diuretic — dramatically increase the likelihood that a patient will fall. The algorithm picked that up because it was analyzing thousands of events. That's something a human might not notice. Luna: So the operations team used the data to refine their protocol. They started flagging any patient with two or more medication changes in a day as high risk, regardless of their baseline score. Lucas: Yeah. And that change alone accounted for about a third of the total fall reduction. It's a perfect example of how operations analysis can uncover a leverage point that nobody was looking at. Luna: So what's the one takeaway you'd want a listener to remember from this episode? Lucas: I think it's this: the best operational fix is often not a new gadget or a mandate. It's a better way to connect data to action, and to make that action feel like part of people's existing workflow rather than an extra task. The Ohio hospital didn't invent a new technology. They just wired together what they already had in a smarter way and changed the culture around learning from mistakes. Luna: And they saved millions of dollars and prevented a lot of pain in the process. Lucas: Exactly. That's the kind of operations story we like.