Night-Shift Speak-Up Signal After a Merger

What the data support. The 38% response rate is itself a finding: 62% of staff did not answer, so the results describe respondents, not the hospital. The drop from 61% to 38% makes year-over-year comparison unsafe unless respondents are weighted or matched by unit and shift. The night-shift result (28% favorable among 140 nurses) is a credible signal because it is concentrated and consistent with the two near-misses. Low scores on feedback about error and handoffs point to systems and leadership behaviors, not individual attitudes.

What they do not support. The survey cannot tell you that 'culture is bad' hospital-wide, that posters will fix it, or that the two near-misses share one cause. Perception data do not measure hand-hygiene behavior.

Board message. 'We have a signal of low psychological safety on nights and weak error feedback. We are validating it with focus groups and observations before setting targets. We will report response rate, unit-level results, and leading indicators, not a single composite.' Ask the board for 90 days to return with a triangulated baseline.

Reasoning. Boards need uncertainty stated. Overclaiming from a 38% sample invites false confidence and a poster campaign that will further erode trust.

Use existing resources in this sequence:

1. Validate (weeks 1-3). Run 6-8 focus groups by unit and shift, including night nurses; observe 20 handoffs; review 30 recent reports. Ask 'what happens when you raise a concern?' rather than 'do you feel safe?'
2. Close the feedback loop (weeks 2-8). Publish a one-page 'you said, we did' after each unit's focus group and resolve two issues staff name. This directly targets the 39% feedback score.
3. Change leader behavior (weeks 4-12). CNO and unit managers round on nights with a standard question set; teach managers just-culture framing; make clear that reporting will not be punished.
4. Standardize handoffs (weeks 4-12). Adopt one structured handoff, pilot it on two units, and audit 10 handoffs per week with a checklist.
5. Practice speak-up (weeks 6-12). Run short in-situ simulations in which a confederate skips hand hygiene; debrief what made speaking up hard.

Measures: survey response rate and unit-level scores at six months, not six weeks; reports per 1,000 patient days; percent of reports with documented follow-up within 14 days; handoff audit adherence; night-shift rounding logs. Do not measure poster distribution.

Reasoning. The data point to feedback and leadership accessibility. Focus groups and rounding are cheap, diagnostic, and trust-building. A re-survey in six weeks would measure the campaign, not the culture.

Re-fielding in six weeks is too soon. Staff will be asked the same questions before any visible change. That teaches them surveys have no consequences and will lower future response rates.

Gift cards can raise response but introduce bias: staff may complete for the incentive, and if managers can link completion to names, anonymity is compromised. If used, make them unconditional, small, and not tied to scores.

Posters are a weak action. In the hierarchy of effectiveness, education and posters are among the least reliable. They do not change handoff structure, feedback, or leader behavior. A poster saying 'speak up' in a unit where a resident ignored a concern can feel like mockery.

Better: keep the survey closed; run a short pulse survey after 90 days of visible changes; report unit-level results and actions; use focus groups now. Tell the CNO: 'We will re-measure when we have something to measure - after staff see closed-loop follow-up and a handoff standard.'

Judgement. The CNO's intent is reasonable but the sequence is backwards. The political risk of saying no is lower than the credibility risk of a survey that looks rigged.