Home / Projects / Clinical Quality & Safety Analytics
Clinical Quality & Safety Analytics
Clinical Quality & Safety Analytics is a project by Akash Renuka Ashok (Analytics Developer (self-initiated project)). A Power BI decision-support dashboard that turns quality indicators, clinical incidents, SIRS notifications, consumer experience and workforce data across twelve residential aged care homes into governance-ready reporting.
- Role
- Analytics Developer (self-initiated project)
- Category
- Power BI Case Study
- Stack
- Power BI, DAX, Power Query, Python, Data modelling, Statistical process control (SPC)
The problem
A clinical governance team receives data from six or seven directions at once, each on its own cycle and in its own shape. A portfolio-level view averages away exactly the variation a governance team needs to see, and a quarterly snapshot shows position without showing direction.
What Akash owned
Designed and built the full pipeline solo: synthetic source data, Power Query cleaning, a star-schema semantic model, DAX measures and the Power BI reporting layer.
Highlights
- Built a star-schema model (11 tables, 29 DAX measures) across 13 mandatory quality indicators, three financial years and 17,847 clinical incidents, with numerator/denominator-stored indicators so facilities of different sizes weight correctly on rollup
- Added direction-aware benchmarking so indicators that improve in opposite directions (e.g. falls vs. consumer experience) still resolve to correct escalation status
- Implemented statistical process control with 3-sigma limits from a fixed 12-month baseline, surfacing sustained drift that a quarterly snapshot against benchmark misses entirely
- Found and corrected two flawed early designs through testing: a threshold that let a home running 50% above benchmark on major-injury falls pass unflagged, and a control-chart baseline that hid a genuine step change
- Reconciled real-world data defects as named Power Query steps: mixed date formats, 140 duplicate incident records, and 48 facility-name variants resolved to 12 facilities
- Built drill-through from any facility on any page to a filtered scorecard, plus row-level security by state and report-page tooltips
- Documented data provenance and project limitations directly in the report: entirely synthetic demonstration data, modelled on the national aged care quality indicator programme structure
Outcome
Produced a governance-focused reporting layer that surfaces which homes need attention, whether a change is ordinary variation or worth investigating, and what specifically to look at, rather than a portfolio-level dashboard alone.
Links
More work by Akash Renuka Ashok:
Last updated