Workforce & Rostering — Roster Design
Tullamore used scenario-based roster modelling to identify the registrar staffing needed to provide appropriate rest, avoid consecutive weekends and maintain clinical coverage.
The existing call model included 24-hour on-site duties. With eight registrars, the hospital needed to understand whether it could remove those shifts while maintaining five operating rooms, ICU cover and other perioperative requirements.
Any new model also had to provide three weeks of leave for every doctor within each 13-week roster cycle.
Mapped coverage: modelled weekday and weekend day, night and off-site call alongside the six to seven registrars required for normal daytime service.
Defined the rules: removed 24-hour shifts, prevented consecutive weekends, built in appropriate post-shift rest and tested against a 48-hour average working week.
Tested staffing scenarios: generated sample 13-week rosters for eight, 12 and 13 registrars, including three weeks of leave for each doctor.
Refined the weekend model: once the 24-hour duties were split, tested whether the off-site weekend day shift was still required.
The eight-registrar baseline could not meet the core requirements for rest between shifts and avoiding consecutive weekends.
A 12-registrar model met those requirements with average weekly hours of 44.7, although each doctor would work five to six weekends in 13 weeks. With 13 registrars, average weekly hours reduced to 44 and weekend frequency improved to five weekends.
The analysis quantified the staffing implications of safer working patterns and gave the hospital a practical basis for workforce planning.
Splitting the 24-hour duties also made it possible to remove the off-site weekend day shift in the 13-registrar model, reducing weekend frequency to four weekends in 13 weeks, creating longer gaps between worked weekends and lowering the overall duty burden.
The modelled 44-hour average supported EWTD planning, although off-site hours were excluded from that calculation.
The model compared staffing levels and roster structure rather than assuming that additional headcount alone would resolve the problem.
Modelled 13-week scenarios
| Scenario | Core rules | Average weekly hours | Weekend frequency |
|---|---|---|---|
| 8 registrars | Not achieved | Not feasible | Consecutive weekends / insufficient rest |
| 12 registrars | Achieved | 44.7* | 5–6 weekends in 13 weeks |
| 13 registrars | Achieved | 44.0* | 5 weekends in 13 weeks |
| 13; off-site day removed | Achieved | 44.0* | 4 weekends in 13 weeks |
* Average weekly hours exclude off-site hours.
Key finding
Moving from eight to 12 registrars was the modelled minimum for removing 24-hour shifts while meeting existing service levels, core rest and consecutive-weekend rules.
What changed for the organisation once the platform was live.
24-hour shifts removed
in every feasible modelled scenario
12 registrars
modelled minimum to meet core rules
44.7 → 44.0 hours
average week at 12 and 13 registrars
No consecutive weekends
designed into the feasible models
4 weekends in 13
with the off-site weekend day removed
3 weeks leave
for each doctor in every 13-week cycle
5 theatres + ICU
clinical coverage maintained
Evidence for planning
clear basis for workforce decisions
We will walk you through how this engagement was delivered and what the equivalent first phase would look like for your organisation.
More case studies