top of page
No tags yet.

Why you REALLY shouldn't make decisions based on e-rostering data

Well, not unless you have a deep and broad data quality programme in place first...

Maintenance and shift entry in e-rostering is considered by most Ward Managers to be a decidedly onerous task. It’s not a clinical activity, yet requires input from senior clinical staff to create plans, decide when to call in temporary staff and to resolve fairness disputes (for example: contentious leave planning, shift swapping or incompatible staff needed to work the same shift).

The skills of “working with computers” - using spreadsheets, or more complex tools like e-rostering solutions is a necessity - but there is usually a hard-core subset of staff resistant to working this way.

Drop in training sessions are useful, but many times frustrated Ward Managers won’t turn up for these, either not motivated, not really interested or assuming there is no solution to their issues. Training has to be top notch and ideally include the proactive monitoring of how well Wards are working with e-rostering. It’s best not to let those frustrations fester.

Ideally, training will be followed by certification of competence on e-rostering, with this skill expiring and renewable.

Planning rosters often takes place with large usage gaps between sessions and it’s tough for Ward Managers to become truly expert in their rostering system on this basis.

  • Shift data input may be in the day to day hands of a variety of staff, often working on the same rosters. How well do they communicate?

  • Daybooks are a source of verification, but are often also wrong. Which system is correct? Without a time and attendance “clockings” system it is very difficult to say.

It is no surprise therefore that shift entry discipline – the entry and modification of roster plans and actuals – is often very poor. This fact can be part of the motivation to a move to departmental or centralised e-rostering (that’s a topic for another post!), which has met with variable success across some Trusts.


What is more surprising (and concerning) is that holds true even where the e-rostering system drives the payroll system (the inbound ESR interface).

It’s a direct cause of board reporting errors, incorrect safe staffing returns, leaking contract hours and overpayments.


As an example, in our studies, an e-rostering solution without a strong data quality programme would report in error (approximately) 3 times more staff owing time to the Trust than reality. And where managers can’t depend on the data, they won’t have the confidence to issue “catch up” shifts. Paid for clinical time leaks away, often replaced by costlier bank or agency hours.

Data quality is therefore of paramount importance, even if not exactly a sexy topic.

Meanwhile it’s important to note that the tenet of “garbage in, garbage out” is most relevant to the accuracy of the Board Reports and Safe Staffing returns. These will be compromised, to varying degree, by poor data quality.

Should external audit occur, it would be super to have demonstrable controls in place maintaining data quality, to avoid exposing the Trust board to unwelcome censure.

Thoughts on solutions

  1. The Trust must adopt a deep and thorough data quality programme, to catch and target errors automatically to those able to fix them (and probably responsible for creating them in the first place – as kind of virtuous educational cycle).

  2. Consideration should be given to a clockings system (Time and Attendance - these are rare in the NHS, often considered to be anti-nurse) to automate the capture of shift start and end times, reduce Ward Manager administration and improve accuracy.

  3. Trusts should identify poorly constructed rosters (our Sentinel solution will drive this) and either support and educate or replace Roster Managers who are struggling.

bottom of page