Mid Yorks deploys AI-powered rostering to cut locum spend by 50% and fulfil 99% of clinician preferences
Book a demoreduction in monthly locum spend in Paediatrics
monthly locum saving in a single department
of negative shift preferences satisfied
resident doctors across Anaesthetics and Paediatrics
Executive summary
As the NHS works towards the 10-Point Plan for workforce retention, moving from fixed-pattern rotas to flexible, clinician-led scheduling has become an operational priority. Mid Yorkshire Teaching NHS Trust deployed Patchwork’s AI-powered preference-based rostering (PBR) across its Anaesthetics and Paediatrics departments.
Through an iterative implementation, monthly locum spend in Paediatrics fell from £50,000 to £25,000, and 99.83% of clinicians’ negative shift preferences were met.
“This is the first time I have been treated like an adult. It is easily the best roster experience I’ve had in over eight years of training.”
The challenge
Fixed rota patterns, imposed with little regard for personal preference, have driven burnout and an over-reliance on temporary and agency staff. At Mid Yorks, medical staffing teams spent disproportionate hours manually reconciling leave and swap requests against rigid rota gaps.
Less Than Full Time (LTFT) working was especially hard to fit into full-time rota slots. The fractional gaps that couldn’t be reconciled were escalated to the bank, pushing up agency spend.
The approach
Preference-based rostering treats the rota as an optimisation problem rather than a spreadsheet exercise. Built on service plans and job plans, it balances three things at once:
Implementation
The rollout covered 26 anaesthetic and 23 paediatric resident doctors, with Mid Yorks acting as a co-development partner. Two early challenges shaped the calibration:
Weighting time-off requests
Early iterations were adjusted so negative preferences, such as requested leave, were weighted accurately.
Local night-shift rules
Rules on night-shift intensity and recovery periods were mapped into the core algorithm.
This calibration was the direct catalyst for the results that followed, and the matured model is now being deployed at other NHS trusts.
Results
The calibrated system was evaluated during the December 2025 rotation.
99.83% of negative preferences met
Mid-rotation training transitions were handled without extensive manual recalculation.
50% lower locum spend
Paediatrics locum spend fell from £50,000 to £25,000 a month — a sustained £25,000 monthly saving.
Clinicians reported a clear cultural shift: major life events such as weddings and exams were reliably accommodated, an early indicator of better retention and wellbeing.
Conclusion
To maximise autonomy, the algorithm occasionally clustered shifts — up to 72 hours over a period, compared with a typical 50–54 — always within Trust safety standards. Better use of the substantive workforce is what drove the reduction in bank and agency cover.
Mid Yorks shows flexible working is achievable when trusts take an iterative, partnership-led approach: preference-based rostering delivers both financial sustainability and the cultural change needed to retain clinicians.
Could preference-based rostering cut locum spend in your trust?
Talk to our team about what this could look like for your organisation.