In the world of healthcare analytics, data driven is only as good as the integrity of your timestamps. For health systems running on Epic, there is a common and high stakes trap that clinical analysts and C suite leaders alike often fall into: confusing the Order Time with the Actual Event Time.
If your dashboard shows that your Door to Meds time is improving, but your nurses on the floor are still feeling underwater, your metrics might be lying to you. Here is why that distinction matters and how to fix your reporting.
The Mirage: Why Your Metrics Look Better (or Worse) Than Reality
Epic is a massive relational database. For every single action, such as a medication ordered, a lab drawn, or a patient discharged, there are multiple timestamps captured in Clarity and Caboodle.
The most common mistake is pulling the ORDER_TIME (when a physician clicked Sign) and using it as a proxy for the EVENT_TIME (when the patient actually received the intervention).
The Reality Gap
The Stat Order Illusion: A physician places a Stat order at 10:00 AM. The nurse, busy with another patient, administers it at 10:45 AM and documents it at 11:00 AM.
The Data Error: If you report based on the Order Time, your data suggests the patient was treated at 10:00 AM. In reality, the clinical window of treatment did not begin for another 45 minutes.
The Three Critical Timestamps You Need to Know
To get an honest look at your hospital's performance, you must differentiate between these three pillars of Epic data:
System Time (The Wall Clock)
When the user hit Save in Hyperspace. This is great for auditing productivity but terrible for measuring clinical outcomes.
Order Time
When the intent was created. This measures provider efficiency and decision making speed, not patient care.
Event or Administration Time
When the physical act occurred (for example, the LINE_ADMIN_TIME for meds or the SPECIMEN_TAKEN_TIME for labs). This is the Truth of the patient experience.
Why Health Systems Miss This
Most reporting errors happen during the transition from Chronicles (real time) to Clarity (relational) to Caboodle (dimensional).
In Caboodle, many data models simplify things for the sake of ease of use. For example, in the MedicationOrder fact table, it is much easier to pull the order date than it is to join across to the MedicationAdministration table to find the actual time the drug entered the patient's system. Many analysts take the path of least resistance, leading to clean reports that are clinically inaccurate.
The Fix: How to Audit Your Reporting
If you suspect your metrics are skewed, take these three steps:
1. Audit Your Joins
Ensure your SQL developers are not just looking at the ORDER_PROC table. For clinical efficacy metrics, they must bridge the gap to the flowsheets (IP_FLWSHT_MEAS) or administration records (MAR_ADMIN_INFO).
2. Account for Back Charting
Epic allows clinicians to document an event after it happened (for instance, a nurse documents at 2:00 PM that a med was given at 1:30 PM). Your reports should always prioritize the Clinician Entered Event Time over the System Entry Time. If you use System Time, you are measuring how fast your staff types, not how fast they treat.
3. Use SlicerDicer for Validation
Before finalizing a permanent dashboard, use SlicerDicer to compare Order to Event lag times. If you see a massive spike in orders with no corresponding event time, you likely have a documentation workflow issue that is poisoning your data.
The Bottom Line
When we use Order Timestamps to measure clinical success, we are not just reporting bad data; we are making strategic decisions based on a fantasy.
By shifting your focus to Actual Event Times, you gain a transparent view of your bottlenecks, your staffing needs, and most importantly, the actual care your patients are receiving.
Need Help Auditing Your Epic Reporting?
We've helped health systems uncover the truth hidden in their data. Let's talk about what your metrics might be missing.
Schedule a Conversation