KPI card 1
Total recorded visits: 1,250.
Previous lesson5.2 Preparing Health Data for Power BINext lesson5.4 Comparing Health Facility PerformancePrevious moduleModule 4. Multi-Source Health Data Analysis and TriangulationNext moduleModule 6. Monitoring, Evaluation and Health Programme Performance
MODULE 5 · LESSON 5.3
Building Your First Health Dashboard
Estimated time: 30 minutes
WRITTEN ON THE BOARD
Lesson 5.3 builds the first dashboard.
Slide 1 of 3
Read complete notesREAD COMPLETE NOTES
All figures in this lesson are invented for teaching.
5.3.1 Our dataset
| Facility | January | February | March |
|---|---|---|---|
| Central Clinic | 120 | 135 | 110 |
| Riverside Clinic | 95 | 100 | 105 |
| Hilltop Clinic | 110 | 115 | 120 |
| Valley Clinic | 75 | 80 | 85 |
The questions for this dashboard are: how many patient visits were recorded, which facility recorded the highest total, how attendance changed over time, and which facilities should be examined further.
5.3.2 Calculate the totals
- January: 120 + 95 + 110 + 75 = 400.
- February: 135 + 100 + 115 + 80 = 430.
- March: 110 + 105 + 120 + 85 = 420.
- Three-month total: 400 + 430 + 420 = 1,250.
These are recorded visits. They should not automatically be interpreted as 1,250 unique patients.
5.3.3 Decide what to display
KPI card 2
Number of facilities: 4.
KPI card 3
Reporting months: 3.
Visual 1
A line chart of the January, February and March totals.
Visual 2
A bar chart comparing total recorded visits across the four facilities.
Visual 3
A table of each month's total attendance.
5.3.4 Build the dashboard in Power BI
- Import the Excel workbook.
- Check the data types.
- Review the dataset for possible problems.
- Create a visual of total attendance.
- Add a card for the total.
- Add a line chart for monthly attendance.
- Add a bar chart comparing facilities.
- Add clear titles.
- Check the values against the source dataset.
- Save the report.
A useful dashboard should be readable and accurate before it becomes visually elaborate.
Safe DAX for a rate
Completeness Rate = DIVIDE(SUM(Reports_Received), SUM(Reports_Expected), 0). DIVIDE avoids a divide-by-zero error. The 0 it returns is a fallback, not proof that completeness was truly zero. If expected reports are blank, investigate the denominator before you publish the card.
5.3.5 Understand what the dashboard tells us
- February had the highest monthly total.
- March's total was slightly lower than February's.
- Central Clinic recorded the highest three-month total.
- Valley Clinic recorded the lowest three-month total.
The dashboard does not say why attendance differed. It does not establish population size, service availability, staffing or reporting completeness. Those questions need more information.
Lesson 5.3 knowledge check
Choose an answer for each question, then check. The suggested answers stay hidden until you do.
1. What is the total attendance across all facilities for the three months?
2. Which month has the highest total?
3. Does the highest attendance necessarily mean the best quality of care?
A dashboard should be designed around clear questions. Do not assume that a visual pattern explains its own cause.