Blog Summary
Hospital medicine access centers running ED-to-inpatient distribution out of a repurposed transfer center module and a manually maintained spreadsheet are working with tools that were never built for this. The manual glue holding it together — a coordinator checking treatment teams by hand, a spreadsheet one deleted formula away from failing, and a paging workflow that routes through a dummy provider before the real attending gets updated — creates fragility that doesn’t show up until you’re down to one coordinator on nights and something goes wrong. medaptus Command is built to replace the spreadsheet with configurable, rule-based automation, surface real-time workload visibility across the care team, and give coordinators a single place to manage intake without cross-checking Epic. This post walks through what a live distribution workflow actually looks like, where the risk sits, and what changes when purpose-built distribution automation takes over.
I spent time recently walking through the distribution workflow at an access center, and what struck me wasn’t how manual the process was — that part didn’t surprise me — it was how deliberately they’d built it. Three screens, a spreadsheet, scheduling software open in another tab, and a coordinator who was very good at her job and knew exactly how fragile the thing she’d built actually was. They had repurposed a transfer center module in Epic, layered a custom spreadsheet on top, and created rules in that spreadsheet to mimic the logic a dedicated distribution tool would give them.
They were hitting their goals. Their 10-minute assignment target was real, and their median from order placement to assignment was running around four to five minutes. And yet the coordinator managing that spreadsheet was on call 24/7, because if someone deleted a formula, the whole thing broke.
That is the hidden cost that doesn’t show up in your assignment-time metrics.
Why Is Your Access Center Running Assignment Out of a Transfer Center Module?
When you ask a distribution coordinator why they’re using the Epic Transfer Center for hospitalist assignments, the answer is almost always some version of: it fits, and it was already there. One coordinator I spoke with put it plainly:
“This module is really built more for transfers, not really for hospitalist distribution — we just kind of hijacked the functionality because it fits in with our system.”
— Access center coordinator, large regional health system
That is an entirely rational decision. Transfer center tools in Epic give you a queue, a way to document intake, and a place to capture the assignment. If your goal is to get a patient from the ED to an inpatient hospitalist, you can make it work. And most teams have. But they’ve done it by layering something on top — usually a spreadsheet — because the transfer module doesn’t give you what you actually need when you’re managing distribution: real-time workload visibility across your hospitalist team, a way to see who’s up next, how many admits each provider already has today, and whether a particular hospitalist is already on a case before you page them.
The spreadsheet closes that gap. A coordinator built it, maintains it, and updates it throughout the shift so the next person coming on knows exactly where things stand. It tracks teams, starting numbers, admission counts, point-of-care totals. It is, functionally, a custom-built distribution tool. The problem is that it’s also one deleted formula away from failing, and somebody has to be reachable around the clock to fix it.
What Does the Manual Workflow Actually Look Like in Motion?
I want to walk through what a coordinator is doing on a shift, because I think the operational picture is more complicated than “they’re using a spreadsheet.”
A distribution coordinator at a busy health system is typically managing several bucket queues at once: ED unprocessed cases, ED cases in progress, hospitalist consults, and consults on hold. Each has different routing logic. An ED-to-inpatient assignment is order-driven — the moment the order hits, it surfaces in the queue, and the coordinator opens it, pulls up the spreadsheet to see who’s up next based on team rotation and current workload, and pages the provider through Amion. That part moves fast; the median is around four to five minutes from order placement, against a 10-minute target.
But there is a step that happens before the page that doesn’t show up in that timer. The coordinator checks the active treatment team in Epic to confirm the hospitalist they’re about to assign isn’t already on the case. That check doesn’t surface automatically anywhere in the queue — it’s a manual step, and it’s a real one, because if you skip it and assign a provider already on the treatment team, you’ve created a duplicate assignment that someone has to unwind.
After the page, if there’s no response within 30 minutes, a background report flags the case for a second page. That report refreshes every minute. The coordinator watches it. On nights, when you might be down to one coordinator covering the whole system, that’s one person managing the queue, the spreadsheet, the treatment-team check, the paging, and the escalation cadence — by themselves.
There is also routing logic that has to be carried in someone’s head or built manually into the spreadsheet: pregnant patients at one campus route to a specific physician group regardless of PCP, readmits within 14 days go back to the admitting team if they’re available, and ED-only campuses have different rules entirely. Some of these are in the spreadsheet. Some of them are not, and a coordinator just knows. That’s why we built medaptus Command: a platform specifically for hospital medicine teams, to streamline the entire process from patient intake to assignments to distribution and more.
Current manual distribution steps vs. what medaptus Command automates:
| Distribution Step | Current Manual Process | What Command Changes |
| Identify next provider | Check spreadsheet for team rotation, workload, and admit counts | Rule-based recommendation surfaces automatically in the intake queue |
| Confirm provider is not already on the case | Manual check of Epic active treatment team | Automated check via live ADT feed |
| Route to specialty or site exception | Coordinator memory or manual spreadsheet rule (e.g., PCP routing, pregnancy, bounce-backs) | Configurable routing rules per campus, built into the system |
| 30-minute no-response escalation | Background report monitored manually; coordinator watches and re-pages | Configurable escalation thresholds, automated flag in workflow |
| Update Epic with real attending | Manual update after dummy-provider page | Assignment written back to Epic from Command |
What Is the Failure Mode You Are Actually Protecting Against?
When I think about what Command is actually solving here, I keep coming back to one number: one coordinator on nights.
For days and evenings, teams may staff two to three people to work this process. On nights, it can drop to one. That person is managing everything — the queue, the assignment logic, the paging, the escalation — and doing it across multiple campuses that may each have different rules. If the spreadsheet breaks, if they can’t remember a routing exception, if they’re in the middle of a direct admission when an ED case comes in at the same time, there is no backup.
What Command is designed to do is take the rule-based logic that lives in that spreadsheet — and in that coordinator’s head — and make it part of the system. A coordinator opens Command, sees the intake queue, and instead of going to a spreadsheet to figure out who’s up next, Command surfaces a recommendation based on the rules the team has configured: team rotation, current workload, admit counts, and any routing exceptions specific to that campus or patient type. The treatment-team check happens automatically, because Command is pulling from the same ADT feed that keeps the census current. The 30-minute escalation can be built into the workflow, with configurable thresholds by site.
The coordinator is still making the call. But they’re making it with the information already in front of them, without cross-checking three systems. And when you’re down to one person overnight, that matters.
Conclusion
The teams doing this work are good at it — they’ve built something functional out of tools that weren’t built for the job, and they’re hitting their targets most of the time. But the measure of a workflow isn’t whether it works on a good day. It’s whether it holds up on a night when you’re one coordinator short, the spreadsheet is showing something off, and there are four new admits in the queue.
Command is built for that night. The rules that live in your spreadsheet and in your coordinators’ heads can live in the system instead, available to whoever is on shift, consistent across campuses, and not dependent on one person being reachable to fix a broken formula. That’s what purpose-built distribution automation is supposed to do — not replace the coordinator’s judgment, but give them the visibility and the structure to use it well.
FAQs
Does Command replace the Epic Transfer Center, or work alongside it?
Command works alongside Epic and integrates with it via ADT feed. It doesn’t replace the Transfer Center module, but it takes over the distribution logic you’ve been running outside of it — in spreadsheets and manual workflows — and brings it into a single, purpose-built place.
What happens to the campus-specific routing rules we’ve built into our current process?
Routing rules in Command are configurable by site, so exceptions like pregnancy-based routing, bounce-back logic, and campus-specific assignment rules can all be modeled in the system rather than carried in a spreadsheet or in a coordinator’s memory.
How does Command handle cases where a coordinator needs to override the recommendation?
Command surfaces recommendations, not mandates. The coordinator always makes the final assignment — Command gives them the information to do it faster and with more confidence.
We staff as few as one coordinator overnight. Does Command change the staffing model?
The immediate goal is to reduce the cognitive load on whoever is on shift — less manual cross-checking, fewer systems to manage simultaneously. Whether that changes long-term staffing is something each organization evaluates based on their own volume and goals.
Can Command surface whether a hospitalist is already on the treatment team before an assignment is made?
Yes — because Command is connected to the same ADT feed that keeps the census current, it can check active treatment team status automatically as part of the intake workflow, removing the need for a manual Epic check before each page.
About The Author
Jaclyn Corbett is the Product Manager and Release Manager at medaptus, where she has spent over 13 years working in implementation, customer support, and product management. She works directly with hospitals and health systems to understand how their teams use hospital medicine software day-to-day, and translates those workflows into product decisions.
Get the latest updates and news delivered to your inbox.
Subscribe to our newsletter today.
