Beyond Dashboards: My Journey of Designing a SIRS Email Alert in Tableau (Part 1)
Updated: Jan 11
When I first started working on my Sepsis analytics project, I believed that building a clean and informative dashboard was the final goal. After all, dashboards show trends, values, and patterns clearly - aren’t that enough?
But as I went deeper into the healthcare context, I realized something important:In critical conditions like sepsis, just showing data is not enough. Timing matters more than visuals.
This realization completely changed how I approached my project and eventually led me to design an Email Trigger – SIRS Alert System instead of just another dashboard view.
Understanding the Clinical Context: What Is SIRS?
SIRS (Systemic Inflammatory Response Syndrome) is an early clinical condition that can lead to sepsis if not identified on time. A patient is considered at risk when two or more vital signs or lab values become abnormal, such as:
Body temperature
Heart rate
Respiratory rate
White blood cell count
Individually, these values may not look alarming. But when multiple abnormalities occur together especially over time — the patient’s risk increases rapidly. Doctors and nurses do not just want to know what is abnormal. They want to know when the risk actually begins.
Why Dashboards Alone Are Not Enough
Traditional dashboards are excellent for analysis, but they have one limitation:They wait for someone to look at them. In a real hospital environment many challenges are there like : Doctors are busy, Dashboards are not monitored 24/7, Important changes can happen between check-ins.
So, what will happen if the patient becomes critical at 3 AM, will someone notice it just by having a dashboard?
This question led me to shift my focus from visual reporting to early warning logic.
Introducing the Idea of Trigger Hour
Instead of alerting doctors every hour or for every abnormal value, I wanted the alert to be:
Timely
Meaningful
Non-repetitive
This is where the concept of Trigger Hour came in.
The Trigger Hour represents the first time a patient satisfies the SIRS condition, meaning two or more abnormalities occur together. This is the exact moment when attention is truly required.
To make the story clearer, I also introduced the Transition Hour the hour just before the trigger. This helps in understanding how the patient gradually moved toward risk instead of suddenly appearing critical.
Turning My Analysis into an Actual Alert
After understanding how the trigger hour works, I had to further decide how I should actually show this in my visualization.
At this point, I decided to focus more on how it will be useful in real life scenario rather than focusing on looks and formatting of chart. I asked myself when a doctor would actually need to know about a patient’s condition. That is when the idea of adding an email alert came in, instead of only marking values inside the table.
I made the alert to be sent only at the trigger hour, the exact moment when the SIRS condition is first met. For all other hours, the table clearly shows that no email should be sent. This helps avoid repeated or unnecessary alerts. When the email trigger is activated, it includes the patient ID along with the abnormal vitals such as heart rate, respiratory rate, temperature, WBC, and PaCO₂, so the doctor can quickly understand why the alert was generated without digging into the dashboard.
From the below image, we can get clear idea of how the Email Trigger Alert table looks like.
When the trigger hour is reached and the alert icon is clicked, Tableau opens a pre-filled email draft. The message contains the patient ID along with key abnormal vitals, allowing the clinician to quickly understand why the alert was generated. This keeps the alert simple, clear, and immediately actionable.
How Close This Is to Real Hospital Alerts
The email alert created in this project represents how early warning systems usually work in hospitals. The goal is to identify risk at the right time and notify clinicians only when it is truly needed, along with the most important patient details.
Tableau itself is not a hospital messaging system, but the alert logic mirrors real-world workflows. In practice, similar logic would be connected to clinical databases and hospital communication platforms. What matters here is the decision-making design — knowing when to alert, what to include, and how to avoid unnecessary notifications — which is exactly how real clinical alert systems are built.
Conclusion
This project helped me realize that data analysis becomes truly powerful only when it leads to timely action. Instead of stopping at insights and charts, I focused on designing an alert that highlights the exact moment a patient becomes high risk. This shift from visualization to decision-support changed the way I think about analytics, especially in healthcare use cases.
In the next blog, I will walk through the complete Tableau implementation of this SIRS email trigger system including calculated fields, trigger hour logic, email actions, and the challenges I faced while building it. If you are interested in seeing how this alert was actually created step by step in Tableau, stay tuned for Part 2.
Thanks for reading and happy learning!
