Labs · Power BI

Lab: Build a 4-KPI Card Row for One Industry (Hospital) from Sample Data

By Ravindra Bagale · Last updated: 8 October 2026

Suppose we are…

Suppose we are building the "8 AM view" for the Pune multi-speciality hospital from Lab 35 (similar to Ruby Hall Clinic or Jehangir Hospital). The medical superintendent agreed to our plan: four KPI cards first. We have hospital_visits.csv with 40 visits from March 2026 (35 OPD, 5 IPD). All data is sample data made up for practice.

Goal of this lab

Build a row of 4 cards:

• Visits = 40
• Avg OPD wait (min) = 34.7
• Avg bill (Rs) = 3.65K
• 30-day readmission = 7.5%

Then add a Department slicer and check the cards change correctly.

What you need (all free)

• Power BI Desktop.
• Download hospital_visits.csv (40 visits, March 2026)
• 40 minutes.

The data: before and after

Before. The raw file. IPD rows have no wait time (blank), because admitted patients do not wait in the OPD queue.

After. The 4-KPI card row.

The formula

Visits = COUNTROWS ( hospital_visits )

Avg OPD Wait = AVERAGE ( hospital_visits[WaitMins] )

Avg Bill = AVERAGE ( hospital_visits[Bill] )

Readmission % =
DIVIDE (
    CALCULATE ( [Visits], hospital_visits[Readmitted30d] = "Yes" ),
    [Visits]
)

• AVERAGE ignores blanks, so the IPD rows (no wait) are left out: 1,213 minutes ÷ 35 OPD visits = 34.66. If the blanks became 0, it would wrongly divide by 40 and show 30.3.
• Readmission % = 3 readmitted ÷ 40 visits = 7.5%. CALCULATE keeps only "Yes" rows for the top; DIVIDE avoids errors if Visits is 0.

Steps

1. Open Power BI Desktop. Get data → Text/CSV → hospital_visits.csv → Transform Data.

2. In Power Query check the types: VisitDate = Date, WaitMins = Whole Number, Bill = Whole Number, the others = Text.

   👁 What you should see: WaitMins shows null on the IPD rows (V-5001, V-5010, …). Leave them as null. Do not use Replace Values to turn them into 0.

3. Click Close & Apply. In Table view check: 40 rows.

4. Create the four measures above (right-click hospital_visits → New measure, one at a time).
5. Format: Avg OPD Wait → Decimal, 1 decimal place. Avg Bill → Whole number (display units Thousands gives 3.65K in the card). Readmission % → Percentage, 1 decimal.
6. Add 4 Card visuals (the new Card visual can hold several, or use 4 classic cards). Put them in one row at the top, same size.

7. Rename each card's label: Visits, Avg OPD wait (min), Avg bill (Rs), 30-day readmission. (Double-click the field name in the Fields well to rename it for this visual.)

   👁 What you should see: 40 | 34.7 | 3.65K | 7.5%, as in the after image.

8. Add a slicer with Department and click Orthopaedics.

   👁 What you should see: Visits 12, Avg OPD wait 41.2, Avg bill 4.05K (4,050), Readmission 8.3% (1 of 12). Orthopaedics has the longest queue.

9. Clear the slicer. Add a title text box: Pune hospital – March 2026 (sample data). Save as Lab-36-hospital-kpis.

Common mistakes

Self-check checklist

Try-at-home challenge

Switch industry. Make a 4-KPI row for a school with a tiny table you type yourself (Home → Enter data): 10 students with columns Student, Class, DaysPresent (out of 20), Marks (out of 100), FeePaid (Yes/No). Choose KPIs: Students, Attendance %, Average marks, Fee collection %. Write the DAX for Attendance % and Fee collection %.

Students = COUNTROWS ( School )
Attendance % = DIVIDE ( SUM ( School[DaysPresent] ), [Students] * 20 )
Avg Marks = AVERAGE ( School[Marks] )
Fee Collection % = DIVIDE ( CALCULATE ( [Students], School[FeePaid] = "Yes" ), [Students] )

Samjla ka? COUNTROWS, AVERAGE that skips blanks, and CALCULATE + DIVIDE for a percentage: four cards for any industry. Aata pudhe jaauya: take this row to your own industry and data.