An example operational report prepared in Power BI for a small medical clinic, with possibilities for further expansion, of course. I tried to show how a simple dashboard can illustrate the most important areas of a facility's operations. Here we have basic operational and financial KPIs along with a comparison to the previous month, revenue, visits, and profitability over time, as well as cross-sectional visualizations of visits and medical services. The bar chart includes additional tooltips that appear when hovering over a selected bar, enriching the visualization with extra context. 

Finally, we can delve into an extended analysis of patient flows broken down by gender, age group, and diagnosed condition. Everything can also be filtered by a fictional doctor employed at the clinic and by the selected month. The second tab presents a matrix of the most frequently correlated conditions by the number of patients. The goal was to show how a report based on simple source data can be further enriched with data exploration elements.    

The aim of the project was to show how a report can be individualized for a small clinic and display a much richer context compared to standard management systems for a small medical facility.  

I used a data model developed by AI Copilot based on similar models for small medical facilities reporting according to Electronic Medical Records standards. I later implemented the database schema in an on-prem SQL Server database.  

Power BI Desktop (main environment), Power BI Service (management), SQL Server Management Studio 21 (relational database), Power BI to SQL Server connector, Power Query (data modeling), AI Copilot (testing DAX measures) 

The biggest challenge was designing and implementing the fictional database in such a way as to then simulate the activity of a small clinic with 10 doctors, 500 patients, and 5000 visits over the course of a year in the report. During data preparation, some transformations were specific to certain visualizations, e.g., the double Sankey chart with patient flow required several joins and intermediate table transformations. The correlation map, however, required an additional intermediate table and several transformations to identify the most common pairs of conditions. I also tried to implement several DAX measures specific to the medical industry, such as average revenue per patient or margin on private visits.