A nurse sitting idle waiting for results can't speed up a 90-minute process
At Rutgers Cancer Institute, where some patients once endured three-hour waits before treatment even began, clinicians turned to an unlikely partner — operations researchers — to see what the human eye inside a busy clinic could not. By building a digital replica of their blood cancer clinic and running thousands of virtual scenarios, they discovered that the suffering was not caused by too few nurses, but by a laboratory miles away and a single queue that made no distinction between the urgent and the routine. The insight, once found, was swiftly acted upon — and the results remind us that the architecture of care is itself a form of medicine.
- Cancer patients were spending up to three hours between check-in and the start of treatment — time that, for people already carrying an exhausting diagnosis, was its own quiet harm.
- The instinct to hire more nurses turned out to be wrong; a computer simulation revealed the true culprit was a 90-minute round trip to an off-site laboratory that held every patient hostage to its pace.
- By moving lab processing onsite and splitting the single queue into fast-track and supportive-care pathways, the clinic unlocked capacity that had been invisible — nearly doubling daily infusion patients from 50 to 80.
- Visit times fell by 75 to 90 minutes even as patient volume grew, and bloodwork turnaround collapsed from 90 minutes to under 30 — the simulation's predictions confirmed in practice.
- The framework is now being offered to cancer centers nationwide, though its authors are careful to note that each hospital must do its own diagnostic work — the lesson is the method, not the answer.
Andrew Evens, deputy director for clinical services at Rutgers Cancer Institute, watched cancer patients sit for three hours between arrival and treatment — some of them facing infusions that would stretch the rest of their day. He knew the wait was more than an inconvenience. So he called his former professors at Rutgers Business School and asked them to help engineer the problem away.
What followed was an unusual collaboration. A team led by operations researcher Xin Ding embedded graduate students in the clinic to observe and time every step patients took, then combined those observations with electronic health records to build a digital twin — a three-dimensional simulation of the clinic that could be tested through thousands of scenarios without affecting a single real patient.
The simulation overturned the team's assumptions. Adding more nurses, their first instinct, shaved less than a minute off average visit times. The real bottleneck was an off-site hospital laboratory returning blood results after roughly 90 minutes — results that patients needed before treatment could begin. Compounding this, a single undifferentiated queue meant quick-visit patients were stuck waiting behind those receiving eight-hour infusions.
The model showed that moving lab processing onsite and creating separate fast-track pathways would cut average visit times by 75 to 90 minutes, even with a 20 percent increase in patient volume. The clinic implemented both changes — the fast-track using software features that had always existed but were never activated. The results matched the simulation precisely.
Evens and his colleagues published their findings in the Annals of Operations Research, arguing the framework could apply to emergency departments and outpatient clinics across the country. But he was careful to temper the enthusiasm: every hospital has its own layout, staffing, and constraints, and each would need to build its own model. Rutgers itself is already facing new workflow puzzles in its newly built Morris Cancer Center, where blood draws, consultations, and infusions are spread across separate floors. They may call the business school back in. The work, Evens acknowledged, is never finished — but now they have a way to see the problem before they have to live it.
Andrew Evens was sitting in a clinic where cancer patients waited three hours between walking through the door and beginning treatment. Some of them needed infusions that would stretch six, seven, eight hours into the day. Others came for a quick blood check or a consultation. All of them moved through the same constrained spaces, the same bottlenecks, the same queues. Evens, deputy director for clinical services at Rutgers Cancer Institute, knew the wait was more than an inconvenience. Cancer is difficult enough without adding hours of sitting in a waiting room.
So he did something unusual. He called his former professors at Rutgers Business School and asked if they could help him engineer the problem away.
What followed was a collaboration between clinicians and operations researchers that produced something called a digital twin—a three-dimensional computer simulation of the blood cancer clinic at Rutgers Cancer Institute. A team led by Xin Ding embedded graduate students in the clinic to watch how patients actually moved through it, timing each step, recording patterns. They combined those observations with electronic health records that tracked every moment from check-in to checkout. Then they built a model that could be tested, adjusted, and run through thousands of scenarios without touching a single real patient.
The simulation revealed something counterintuitive. The researchers had assumed that adding more nurses would speed things up. In the virtual clinic, an extra nurse shaved less than a minute off the average visit. That wasn't the problem. The real bottleneck was something else entirely: blood samples were being sent to an off-site hospital laboratory that took roughly 90 minutes to return results. Patients couldn't start treatment without those results. Meanwhile, a single undifferentiated queue meant that someone waiting for a quick blood check might find themselves stuck behind someone receiving an eight-hour infusion.
The simulation showed what would happen if the clinic moved laboratory processing onsite and created separate pathways—a fast track for cancer treatment and a separate queue for supportive care like transfusions. The numbers were striking: average visit times would drop by 75 to 90 minutes, even if patient volume increased by 20 percent. Bloodwork turnaround would fall from 90 minutes to less than 30. The clinic could nearly double the number of infusion patients it treated each day, from about 50 to about 80.
The clinic implemented both changes. The laboratory moved onsite. The fast track was created using software features that had existed all along but had never been activated. The results matched the simulation's predictions. What had been a three-hour ordeal became manageable. The framework, Evens and his colleagues argued in a paper published in the Annals of Operations Research, could work anywhere patients flow through a sequence of constrained resources—emergency departments, surgical units, outpatient clinics across the country.
But Evens was careful not to oversell the approach. Every hospital is different. Every cancer center has its own layout, its own staffing patterns, its own constraints. Other institutions could learn from what Rutgers discovered, but they would need to do their own analysis, build their own models, identify their own bottlenecks. The Rutgers Cancer Institute itself is already facing new puzzles. The newly built Morris Cancer Center has blood draws, doctor visits, and infusions spread across separate floors in the same building. That creates fresh workflow challenges. Evens said they may bring the business school back in to solve them. The work is never finished. But now they have a way to see the problem before they live it.
Notable Quotes
Cancer is difficult, emotionally and physically. So we wanted to make sure patients were able to go through these different junctures of their visit in a very efficient and patient-centered way.— Andrew Evens, deputy director for clinical services at Rutgers Cancer Institute
You could probably take some learned lessons from this article and apply them, but at the end of the day, you're going to need to do your own analysis.— Andrew Evens
The Hearth Conversation Another angle on the story
Why did adding more nurses barely help? That seems like the obvious fix.
Because the constraint wasn't how many people were drawing blood or checking patients in. It was that the blood had to leave the building, go to another hospital's lab, and come back. A nurse sitting idle waiting for results can't speed up a 90-minute process she doesn't control.
So the simulation showed you where the real problem was hiding.
Exactly. You can't fix what you can't see. The clinic felt crowded and slow, but the crowding and slowness had different causes at different points. The simulation let them test solutions without guessing.
Did the staff resist moving the lab onsite? That's a big operational change.
The source doesn't say there was resistance. What's interesting is that the fast track—separating cancer treatment from supportive care—was already possible with their existing software. They just hadn't thought to use it that way. Sometimes the solution is already in your hands.
Can every cancer center do this?
Evens was explicit: no. Every building is different, every staffing model is different. You can't copy Rutgers' answer. You have to ask Rutgers' question: where is time actually disappearing? Then you build your own model and find your own answer.
What happens now?
The new building has created new puzzles. They're already thinking about bringing the business school back. This isn't a one-time fix. It's a way of thinking about how work actually flows.