Hospitals may buy digital tools to save time. Introducing those tools can also change how people do their jobs. A study in Communications Medicine compared staff survey scores before and after a funding program. The results describe NHS organizations that received awards, so tool adoption by individual workers is a separate question.
- Researchers compared 26 funded trusts and 77 unfunded trusts using survey scores aggregated by organization.
- The interval for the overall estimate on thinking about leaving included zero. That survey measure asks about attitudes. Employment records would be needed to count actual departures.
- The abstract reports less favorable scores in several areas for funding of real-time location systems and no detectable associations for automation and AI funding.
- Paper date
- 21 November 2026
- Organizations
- 103 acute or acute-and-community trusts in England
- Survey period
- 2018–2023
- Source checked
- Publisher abstract, summary and official survey guide
Original publication: 21 Nov 2026 · The date above refers to this brief.
The two groups in the staff survey study
NHS trusts
How awards were compared with survey scores
Identify award status
Group trusts by whether they received a program award.
Check the pre-funding period
Compare how the groups were changing before funding.
Compare later changes
Estimate the difference in changes across funded and unfunded trusts.
The exposure was a funding award
The study matched records from the 2021–22 Digital Productivity Programme to NHS Staff Survey results. About 25.2% of the 103 trusts received funding. That share describes organizations. To estimate how many staff actually used a new tool, researchers would need use records and implementation dates.
Source 1 ↗The baseline comparison matters
The authors treated 2018–2021 as the pre-funding period and 2022–2023 as the post-funding period. They compared how staff scores changed across the two groups, inspected trends before funding and used weighting to make groups more comparable on measured baseline characteristics. Trusts selected into the program can still differ in ways the study could not measure.
Source 1 ↗A score about leaving measures survey responses
The overall funding estimate for thinking about leaving was −0.06, with a 95% confidence interval from −0.14 to 0.01. The interval includes zero. The NHS guide explains that summary scores average responses coded on a 0-to-10 scale, with higher scores indicating more favorable answers. Actual departures would require a study using employment records.
Source 1 ↗Source 2 ↗The funding categories tell different stories
The abstract groups automation and AI together and reports no detectable associations. It reports less favorable stressors, engagement, motivation and involvement scores for real-time location system funding. The abstract gives the direction for those outcomes. The authors omitted virtual-reality estimates because the groups had inconsistent pre-funding trends.
Source 1 ↗A follow-up could connect funding to implementation
An evaluation could track when a tool went live by department, which employees used it, and what changed in their work. It could then follow workplace scores, staff retention and the intended productivity outcomes separately. The design would help distinguish the award, the initial rollout and the period when the tool becomes routine.
Source 1 ↗The main numbers describe trusts and survey scores
Primary publisher abstract, plain-language summary and ethics statement for the accepted article. The source check covers the design and results reported in those sections.
01What does the estimate for thinking about leaving describe?
- What was observed
- The overall estimate was −0.06 on a survey score, with a confidence interval including zero.
- Where the conclusion stops
- An estimate for attitudes and an estimate for departures would require different outcome data.
Source 1 · Accepted paper abstract, Results
02Did the program fund a single technology?
- What was observed
- The program included multiple technologies, and the abstract reports different results across funding categories.
- Where the conclusion stops
- Individual use and workflow changes would require implementation records.
Source 1 · Accepted paper abstract, Methods and Results
Numbers you can inspect
| Measure | Value & unit | Origin & method |
|---|---|---|
| NHS trusts | 103 trusts | Reported Number of organizations in the study. Source 1 · Accepted paper abstract, Methods |
| Funded trusts | 26 trusts | Reported Organizations receiving a program award. Source 1 · Accepted paper abstract, Methods |
| Comparison trusts | 77 trusts | Reported Organizations without a program award. Source 1 · Accepted paper abstract, Methods |
| Share of trusts with funding | 25.2 percent of trusts | Calculated 26 divided by 103, multiplied by 100 and rounded to one decimal place. Source 1 · Accepted paper abstract, Methods |
| Thinking-about-leaving score estimate | -0.06 survey score points | Reported Estimate for the overall funded group compared with unfunded trusts. Source 1 · Accepted paper abstract, Results |
| Lower 95% confidence bound | -0.14 survey score points | Reported Lower confidence bound for the overall estimate on thinking about leaving. Source 1 · Accepted paper abstract, Results |
| Upper 95% confidence bound | 0.01 survey score points | Reported Upper confidence bound for the overall estimate on thinking about leaving. Source 1 · Accepted paper abstract, Results |
Compare the actual experiments
These studies answer different questions. Read the unit and endpoint before comparing results.
| Study | Unit & setting | Readout | Interpretation boundary |
|---|---|---|---|
| Organization averages and individual survey change Source 1 · Accepted paper abstract and ethics statement | NHS trusts and US adults | Trust staff survey scores and repeat survey response profiles in a separate study | An average can change when respondents change or when their experience changes. |
| Funding before and after an award Source 1 · Accepted paper abstract, Methods | NHS trusts and dental care boards | Trust survey changes and monthly dental claims | Both studies use pre-policy observations. The digital study compares funded and unfunded organizations as its main design. |
The export includes claims, available numbers, methods and source locations. It contains our reading notes and published summaries; it is not raw participant data or an independent reanalysis.
Evidence update · 21 Nov 2026
First publication. Any calculations in this brief use reported figures. The reading notes identify possible follow-up questions. AI source check; no human editorial or clinical review.
Read the results at the level reported
| Funding group | Result reported | How to interpret it |
|---|---|---|
| All digital funding | Thinking about leaving −0.06 (95% CI −0.14 to 0.01) | Adjusted score difference for the overall award comparison. |
| Automation and AI | No detectable associations reported | Applies to the measured outcomes and the funding category studied. |
| Real-time location systems | Less favorable scores in several areas | The abstract gives the direction. Magnitudes require the full results. |
| Virtual reality | Estimates omitted | Pre-funding trends were inconsistent. |
Findings come from the accepted paper’s abstract. The official survey guide explains the scores.
Your questions, answered
What does the sample of 103 represent?
The count is for NHS trusts. The researchers compared aggregate survey results published for those organizations.
Were actual resignations measured?
Thinking about leaving was a survey subscore. Employment records would answer a question about actual departures.
Does the AI result cover every kind of tool?
The comparison uses funding for a combined automation-and-AI category within one program. A tool-specific conclusion needs information on that tool and its use.
Why are there no virtual-reality estimates?
The authors report inconsistent trends before funding, which makes the changes after funding hard to interpret as a program association.
Limits of this interpretation
- Awards were allocated without randomization. Weighting can account for measured differences, leaving the possibility of unmeasured selection.
- Survey scores are aggregated by trust. The study used organization-level scores and funding records.
- The post-funding window contains two survey years.
- The source check covered the abstract and summary. Detailed technology subgroup estimates were beyond that scope.
Sources & transparency
- Zhou, Leiva-Granados, Yi et al. (2026) — A differences-in-differences study of staff experience in the English NHS after investment in digital technologies
Primary publisher abstract, plain-language summary and ethics statement for the accepted article. The source check covers the design and results reported in those sections. · Accessed 21 Nov 2026
DOI: 10.1038/s43856-026-01999-4 - NHS Staff Survey — 2023 Basic Guide, section 4
Official survey documentation for the meaning and direction of 0-to-10 theme scores and subscores. · Accessed 21 Nov 2026
Prepared and source-checked with AI for Press-news Team. The source notes identify the material available during preparation. The article received no human editorial or clinical review. The source list supports the reported results and the calculations shown.
Source check: AI source check — primary research and stated source limits
Clinical review: Not applicable to this educational guide
Suggest a correction