Community hospital · 32-bed med-surg unit
Finding the real driver behind med-surg turnover
Exit interviews blamed pay. A unit-level Burnout & Resilience wave pointed at recovery and control instead and gave the manager one condition to change first.
Illustrative composite scenario. Built from the HealthDNA™ instrument design and common deployment patterns to show how a wave runs and how leaders read the output. It is not a report of a named client engagement and the figures are modeled examples rather than measured customer results.
- Setting
- Community hospital, single med-surg unit
- Respondents
- 41 RNs, LPNs and techs
- Instruments
- Burnout & Resilience Index™, Just Culture & Psychological Safety
- Cycle
- Baseline wave, 60-day actions, re-pulse at day 90
The situation
- n/aThe unit had lost seven staff in nine months and every exit interview named compensation.
- n/aLeadership had already run a market pay adjustment and turnover did not move.
- n/aNobody could say which shift or which condition was actually generating the exits, so every fix was a guess.
How the wave was run
Scope the wave
One unit, anonymous, all shifts included so night shift could not be identified by subtraction.
Protect anonymity
Reports stayed locked until at least five people in a group responded; no free-text was attributed.
Read drivers, not the average
The team ignored the single index number and worked the nine drivers in rank order.
Change one condition
The manager picked the lowest driver staff could feel within 30 days, not the easiest one to talk about.
What the data showed
| Dimension | Reading | What it meant |
|---|---|---|
| Recovery & Time Off | Lowest driver on the unit | Breaks were routinely missed and staff were contacted on scheduled days off. |
| Control & Autonomy | Second lowest | Scheduling was done to staff, not with them, so every workload fix felt imposed. |
| Meaning & Values Alignment | Strong | People still wanted the work, this was a conditions problem, not a calling problem. |
| Just Culture reporting | Middling with a low response-to-report score | Staff reported events but did not believe anything happened afterward. |
What leadership changed
- ✓Guaranteed break coverage with a named relief assignment on every shift.
- ✓Stopped all non-emergency contact with staff on scheduled days off, in writing, from the director.
- ✓Opened self-scheduling for two of four weeks in each block.
- ✓Started a monthly "you reported it, here is what changed" note in huddle.
The 90-day cadence
- 1Days 1-30: results shared back to staff within two weeks; break coverage rule live.
- 2Days 31-60: self-scheduling window opened; missed-break tracking reviewed weekly.
- 3Days 61-90: re-pulse the unit and compare driver scores, not the headline index.
An engagement average would have shown a tired unit. The driver breakdown showed which two conditions the manager actually controlled and that pay was not the lever.
Instruments used
Ready to run your own wave?
Run this scenario in your organization
Tell us the setting and we will show the instrument, a sample unit report and what a first pulse would look like.
No PHI is ever collected. We only use this to schedule a walkthrough.