🩺Root Cause Analysis of Readmissions Using SQL and Power BI

The Follow-Up Gap MapIn healthcare, every number has a story and sometimes, those stories hide critical problems. Recently, a multispecialty hospital began to notice a disturbing trend: Patient readmissions were climbing steadily over the past six months. At first glance, the dashboards didn’t look unusual. Departmental admissions were stable. Mortality rates were low. Discharges were timely. Yet something wasn’t right. Patients were coming back. Quickly. As a healthcare data analyst, I was tasked with a single question: Why are our patients returning so soon after discharge and what can we do about it? 

Connecting the Dots with Data The answer wasn’t obvious. Readmissions don’t stem from one issue they’re a symptom of deeper, systemic cracks. What I needed wasn’t just a dashboard, but a way to connect clinical behavior to data behavior. Armed with SQL and Power BI, I began to dig through thousands of rows of patient records. Admissions, discharges, diagnoses, and timelines they all held . But it wasn’t until we began looking at patterns, not just counts, that things started to come into focus. The Pattern That Changed Everything We began segmenting patients by more than just age or diagnosis. We compared discharge types, follow-up appointments, and gaps between visits. Suddenly, something became clear. Patients discharged against medical advice were 2.3x more likely to be readmitted within 30 days. That one insight changed the conversation. It wasn’t just about what happened in the hospital it was about how patients left. 

📊 Turning Insights Into Action The Power BI dashboard visualized every angle of the problem: Department-level heatmaps that lit up with readmission spikes. Age groups that revealed vulnerability among older adults Diagnosis categories that showed chronic conditions like diabetes and stroke were recurring drivers. Even, the timing patients discharged on weekends were more likely to return within a week What had started as a basic request turned into a full-blown root cause analysis powered by data.

💡 Insight Doesn’t Work Unless It Leads to Change Data isn’t magic it only matters when it prompts action. The hospital didn’t just view the report and move on. 

Based on the findings: Follow-up calls were made mandatory for all patients discharged AMA High-risk departments started assigning a discharge review nurse ,Weekend discharge protocols were reviewed to ensure continuity of care. Within three months, early readmission rates had dropped by over 11%. The Bigger Realization You don’t need machine learning to drive meaningful impact. You don’t need advanced AI to change how a hospital functions. 

All it took was: Clean SQL queries Sharp Power BI visuals And a relentless focus on asking “Why is this happening?” instead of just “What happened?” 

Final Thoughts :We often look at dashboards expecting answers but the real magic happens when we ask better questions. For me, this project was more than charts and reports. It was a reminder that data is only valuable when it creates momentum. When it moves people. When it leads to decisions that change lives. And sometimes, the most powerful stories are hidden in the most ordinary data