14 Analysts, 48 Hours, One Chance to Get It Right: Epic Go-Live Surge Support at an Academic Medical Center

The situation

The CIO of a 600-bed academic medical center had a go-live date that was not moving. The Epic implementation had been two years in the making. The go-live command center was built. The training schedule was complete. The cutover plan was set.

Three weeks before go-live, the organization’s staffing picture changed. Four at-the-elbow support analysts — specialists who would have walked the floors during go-live helping staff navigate Epic in real time — pulled out. Two took positions elsewhere. One had a family emergency. One had a certification issue surface in a background check. Four analysts down with three weeks to go-live is not a staffing problem. It is a risk event.

The CIO’s team made three calls that week. APSIS was one of them. Two other firms said they could have someone in four to six weeks. APSIS said 48 hours.

Challenge and approach

What the work actually looked like

At-the-elbow support during an Epic go-live sounds straightforward. The reality is more demanding. Physicians and nurses encountering a new EHR during an active patient care shift are under stress. They do not have time for a tutorial. The support role requires someone who can read body language, recognize when a user is about to give up, step in before frustration becomes an error, and fix the problem in under 90 seconds without disrupting patient care.

The analysts did not show up on go-live day. They arrived five days before cutover to participate in the organization’s final training sessions, walk the units, meet the charge nurses, and understand the workflows specific to this organization before they had to support them under live conditions. That five-day pre-positioning period proved decisive: when go-live began, the APSIS analysts were not strangers navigating an unfamiliar hospital. They were familiar faces in units that knew them.

The APSIS team averaged 45+ user interactions per analyst per shift during the first week, and documented every recurring issue — a practice that became critical in the post-live optimization work that followed.

Results

The go-live itself was successful by every standard measure. But the APSIS team’s documentation during the first week surfaced something important: three departments were generating a disproportionate share of help desk tickets, and the issues were not random. They were clustered around specific workflows that had been designed without full input from the clinicians who would use them.

APSIS proposed — and the CIO agreed — to extend the engagement through a structured 16-week optimization program. The work involved workflow redesign sessions with department leads, targeted at-the-elbow retraining for the staff generating the highest ticket volumes, and a series of build adjustments submitted through the organization’s Epic governance process.

By week 16, post-live help desk ticket volume was down 62% from the go-live peak. The three departments that had been struggling were among the strongest performers in the organization’s Epic adoption metrics.

APSIS capabilities demonstrated