· Daramola Joseph Omoyele
Counting fairly: why staff quality metrics need a denominator
Raw incident counts can make your most conscientious staff look like your biggest risk. A short note on the design choices behind Care Quality Intelligence.
When a care provider starts reviewing quality data at staff level, the first chart is usually a count: incidents per person, complaints per person. It is easy to produce and easy to misread.
Counts reward the people who do less
A support worker on four shifts a week will appear in more records than a colleague on one. Someone who reports concerns promptly will appear more often than someone who does not. A raw count mixes exposure, reporting culture and genuine risk into a single number.
Three design rules
- Always show a rate next to a count. In Care Quality Intelligence every staff-level figure is shown as a count and as a rate per 100 delivered support sessions.
- Hold back judgement on small samples. Below a minimum number of sessions the system shows the data but does not rate the person.
- Treat good reporting as a strength. Timely, complete reports are flagged positively rather than added to a risk score.
What this means for your own reporting
You do not need special software to apply these rules. A spreadsheet or Power BI report can carry the denominator alongside every count, set a minimum sample before colouring anything red, and separate reporting behaviour from incident severity. The aim is a supervision conversation based on evidence that both sides recognise as fair.
The demo of Care Quality Intelligence uses synthetic data, so you can explore these rules without any real people's information.