Workforce data can be a powerful resource. When used properly, the data can assist with oversight and assurance of effective and efficient provision of patient care, through appropriate staff utilisation.

Organisations can gain the ability to reduce agency staffing costs while optimising substantive and bank staff resources. The ability to access comprehensive staffing data quickly can ensure safe staffing levels. These levels or ratios can be set by specific patient needs or local considerations. If done well, use of workforce data can also relieve pressure on staff themselves.

Additionally, trust board visibility of workforce intelligence can support strategic decision-making, workforce planning, and performance management, such as establishing their headroom requirements.

Data can also be used to engage a variety of staff groups in effective workforce deployment, in ways that resonate with them, the service they deliver, and their patients.

However, there are challenges with rostering workforce data and these include:

  • There is always lots of data, not all of which are easily presented or well understood – the huge amounts of data can overwhelm staff
  • Data are not tailored in ways relevant to different types of staff: human resources, clinical staff, boards, ward managers
  • Few, if any, departments claim responsibility for rostering data but as the information impacts all staffing groups in an NHS trust, clinical, human resources, and finance teams must all jointly own the data, with an identified accountable officer at board level
  • The rostering data collected are not always used to monitor performance against key performance indicators (KPIs).


Angela Thompson


“It is essential that provider organisations have ward to board reporting and that board members know, through e-rostering performance and KPI compliance, how well they are utilising their workforce. It’s amazing how many hospitals have an e-rostering system but have no idea how efficient it is or how well it is used, yet staff will be the organisations most expensive resource and the greatest asset.

“Providers need to ensure they triangulate the information they review to ensure they not only have efficient and effective rostering at unit level, but that they are also providing safe care and good outcomes for patients.

“Some of the key questions boards should be asking are:

Are units doing rosters six to eight weeks in advance?

Are units completing roster approvals on time?

Are units staying within headroom if not what is the worked total unavailability?

Are units staying within the planned template and budget?

If not, why not are they using additional duties and are these for valid reasons?

Are units producing safe rosters, do staff on duty have the right skills to care for patients?

Are units utilising temporary staff in a safe and efficient manner?”

Developing a roster policy which has clearly defined roles and responsibilities for setting rosters from board to frontline staff will help ensure that e-rostering systems guarantee a dynamic approach to rostering. The policy should also define the key performance metrics that will be used to measure roster policy compliance, including the definitions of each metric, the numerator and denominator, data source and frequency of reporting, including which meetings will receive the reports.

Regular monitoring of adherence to the policy and process for rostering via a series of key performance indicators (KPIs), can be used to review rostering compliance and these should be reported at least monthly. Some organisations use the data daily with live rostering and patient acuity data available in software such as Safe Care, others use weekly reporting. Each trust will need to understand why they are using the data and choose the frequency of reporting accordingly.

Angela Thompson has recommended KPIs for trusts to consider when reviewing their use of rostering. These metrics fall into two groups, baseline good practice metrics that all trust should be working towards irrespective of e-roster organisational maturity. Some of which are based on NHS Improvement’s Good Practice Guide: Rostering.  For Trusts who have embedded e-rostering there is the opportunity for aspirational metrics, that should seek to tackle organisational ‘wicked’ problems, problems that may be difficult to resolve, some examples of the key KPI’s are below:

Basic ‘Good practice’ KPIs:

  • Roster Template/Budget Variation – Linking roster templates and the available shifts to service plan and budgeted establishment for each service. This will ensure financial controls are built into the roster design which will facilitate conversations about productivity. This should include having structures and hierarchies that align to organisational accountability.
  • Net hours – The sum of all contracted hours over worked minus the sum of all unused contracted hours. This should include identifying those staff with no duties assigned.
  • Roster Approval – Approval should be appropriate to service need, with six weeks being the recognised minimum for the two tier approval process. This approval should also focus on the quality of the approved roster from a clinical, quality and financial perspective. With poorly compliant rosters being rejected against agreed criteria.
  • Headroom – Total unavailability is made up of annual leave, study leave, sickness, parenting leave, working day and other leave. This can be considered prospectively for planned rosters (checked during roster approval) and retrospectively for the worked rota to understand the impact on service delivery; the two may look very different.
  • Missing Skills or Charge Cover – Confirming appropriate leadership and skills for each shift is key when writing good clinical quality rosters.
  • Duties with warnings – Service configured warnings that ensure safety and fairness for both staff and patients.
  • Additional duties over funded establishment – All additional work should be recorded on the roster with justification (e.g. enhanced care).
  • Temporary Staffing – Number of temporary staff requests against the total temporary hours worked, and unfilled hours worked broken down by bank and agency staff. Temporary staffing usage should be tied back to substantive unavailability on the roster (e.g. Total hours of vacancy, sickness and special leave are equal to the total time request for temporary fill).

Examples of ‘aspirational’ KPI’s:

  • Supervisory Time Utilisation – Ensuring senior clinical leaders have time to manage and lead in their clinical environment. A % of released time against planned is a good example of this.
  • Temporary Staffing Lead Time – Maximising lead time for temporary staffing bookings will help to effectively fill shifts available for temporary staffing usage, minimising the need to use agency when not needed.
  • Intelligent Temporary Staff Modelling – Identifying expected temporary staff reliance based on known roster factors will support appropriate allocation of temporary workforce. Especially at key (high cost/high impact) times. This can also help inform organisational strategy towards agency management.
  • Turnover – Identification of substantive starters and leavers and modelling this into expected operational shortfall as well as helping to highlight cultural problems.
  • Cost per patient day (CPPD) – which can be worked out at trust and individual ward level which is available on the Model Hospital Dashboard
  • Shift/Leave Requests – The proportion of shift or leave requested on the electronic rostering system is a good indicator of system embedding and is an enabler for cultural change when used effectively.
  • Fixed Working Patterns – the degree of fixed (set days) or flexible (available/unavailable) working patterns should support and enable staff to work but not limit a particular service.
  • Auto-roster – the degree of roster automation in the writing process will minimise the administrative burden creating fair and safe rosters.
  • Clinical Safety Rating – Staffing related clinical safety rating and a staff RAG (Red, Amber, Green)

For Trusts that use a real time rostering and patient acuity software  there is the further opportunity to monitor key metrics, relating to the provision of safe staffing, examples of these are:

  • Care hours per patient day for substantive, bank and agency staff
  • Number of rosters approved with the potential for failure in provision of basic care, known as red flag events
  • Red flag patient safety events such as delay in administration of pain relief, falls, pressure ulcers
  • Matrons undertaking regular patient acuity inter-rater reliability audits for each ward

Organisations with e-rostering capability should measure basic good practice metrics, and have plans in place to identify and address outlying areas. Once the system is embedded with established processes to use KPI’s as part of performance management, the forward thinking trust should use data from all workforce systems to highlight and address specific workforce problems. For all of these KPIs to be reviewed effectively and used to make sustainable improvements, Trust leads need to understand all elements of the data at ward level. To do this properly, they need to triangulate relevant data points, for example, triangulating net hours, no assigned duties and temporary staffing usage should help trusts understand what is really happening with unused contracted hours.

Triangulating the data with other metrics such as the Friends and Family Test and outcomes data such as inpatient falls and pressure ulcer incidence by ward will allow staff to understand not only how good the rostering is and how safe the ward is, but also whether patients would recommend the ward to friends and family, giving a good all-round view of how each ward is performing.

The rostering data and aligned KPIs will need to be reported and presented in different ways and with different frequencies depending on the audience and the purpose of the report, whether it is to improve staffing on a shift by shift basis, inform a patient safety initiative or performance manage roster compliance. For trust boards, it should be presented at a corporate level to engage them appropriately and support them in making trust-wide strategic workforce, finance and patient care decisions.


East Cheshire harnesses its workforce data


East Cheshire NHS Trust has made use of workforce intelligence through an e-rostering platform to create a new three-year workforce technology delivery plan.

The strategy was developed in terms of the key elements of an effective workforce deployment methodology, including:

  • Excellence in rostering practice
  • Sound governance assurance for staff inductions
  • Quality assurance process for temporary workers
  • E-job planning and e-leave management
  • Effective workforce analytics.

The trust’s deputy medical director, Dr Darren Kilroy, explains its characteristics:

“The delivery plan describes how, by enacting it, you will improve patient care. The electronic staff record we had previously wasn’t ‘sold’ to the workforce so it was a missed engagement opportunity.

“What is unique about the strategy is that it is being approached and written as a document for the clinical workforce, as opposed to being about the workforce. Hence it makes regular reference to the centrality of safe and high-quality patient care throughout the text and justifies all elements of the strategy as being contributors to the assurance of such care.

“We want something to give to staff that interests them, that will have resonance to different staff groups in different ways.”

The community and acute services provider, which serves a population of over 470,000 published its delivery plan in November 2017.