“The Numbers That Didn’t Match”
It was 9:15 a.m. on a busy Tuesday at Sree Lakshmi Medical College, a mid-sized teaching hospital in South India. The second case on the list was a 62-year-old man, Mr. Raghavan, coming in for a major abdominal surgery to remove a tumor.
Inside OR 3, the team moved with the familiar mix of routine and pressure. Dr. Anil, the surgeon, stood scrubbed and focused. Dr. Meera, the anesthesiologist, watched the monitors with her usual quiet intensity. Two junior nurses, Deepa and Arun, managed sponges, instruments, irrigation, suction and the constant stream of “Sister, gauze… suction… more saline.”
From the outside, it looked smooth.
From the inside, the numbers were already starting to blur.
As the surgery progressed, the field became increasingly bloody. Deepa opened new packs of gauze, passing them as fast as the scrub nurse asked. Used, blood-soaked sponges piled into a kick bucket. The suction canister slowly filled with a dark red fluid — a mix of blood and the clear irrigation saline used to keep the field visible.
“Can you tell me the blood loss roughly?” Dr. Meera asked over the drapes, eyes still on the patient’s blood pressure and heart rate.
Deepa glanced at the suction canister. “About eight hundred, ma’am?” she said, uncertainty in her voice. “Plus the sponges… I’ll calculate.”
There was no automated system. The “calculation” was a mix of habit, mental math and guesswork.
The rough method in this OR was the same as in many: look at the volume in the suction canister, subtract what you think was irrigation fluid, then add an estimate of how much blood must be on the sponges.
But no one knew exactly how much irrigation had gone in. It was written on a whiteboard, updated when someone remembered. The sponges, all different sizes and degrees of saturation, were waiting to be weighed at the end — if there was time. And everyone knew that by then, decisions about fluids and transfusions would already have been made.
Mr. Raghavan’s blood pressure dipped slightly. “Let’s start a bolus,” said Dr. Meera. “And cross-check blood loss again in ten minutes.”
Two hours later, the operation was still underway. The first unit of blood had been transfused. The suction canister had been changed once, the new one already half full. Irrigation bottles came and went — one litre, another half litre, another litre — scribbled in rushed handwriting on the side of the whiteboard.
At one point, when things became more difficult surgically, everyone’s attention shifted fully to the field.
“More suction… more saline… hold that retractor… give me another pack of gauze.”
Nobody updated the board for twenty minutes.
“Meera, how much are we at now?” Dr. Anil’s voice broke through.
She exhaled slowly. “By my count, maybe fifteen hundred? But I need the exact totals.”
Deepa quickly tried to reconstruct the numbers. “How many litres of saline have we used?” Meera asked.
There was a pause. “Four? Or maybe five, ma’am… One bottle we didn’t write down, I think. I’ll check,” Arun replied, looking worried.
Meera knew this was not their fault. This was simply how things were done everywhere she had ever worked: visual estimation, mental subtraction, retroactive guessing. In textbooks and conferences, everyone agreed this was inaccurate. In real life, this was still the norm.
The surgery finally ended. Mr. Raghavan was shifted to recovery, still intubated but stable enough to move. The team prepared to hand over to the post-op nurses.
“How much blood loss?” the recovery nurse asked, pen hovering over the sheet.
There was a silence that lasted a bit too long.
Deepa opened her notebook. “From the suction: two canisters, around sixteen hundred total. Irrigation… probably about nine hundred ml. For sponges, we’ve written seven hundred based on weight.”
“So what do we write?” the nurse asked again.
“Net blood loss… about fourteen hundred,” Meera answered, doing the math in her head, then immediately doubting it. “Write fourteen hundred.”
Later, while documenting, Meera realized that if she added the numbers another way, she could get a completely different total, still “reasonable,” still technically defensible, and still wrong in ways nobody could precisely prove.
That night, in the ICU, Raghavan’s blood pressure dipped again. The intensivist on duty looked at his file.
“How much did he lose in OT?” he asked.
“Fourteen hundred ml,” the nurse replied, pointing at the sheet.
He frowned slightly. “He looks a bit more dry than that,” he muttered, ordering another fluid bolus and an additional blood test.
By morning, Raghavan was stable, but the team knew it had been close. When the department sat down for a routine morbidity and mortality meeting later that week, his case came up.
“We managed,” the head of surgery said, “but our transfusion and fluid decisions were still built on estimates. If that number was off by even five hundred ml either way, our decisions could have been too aggressive or too conservative.”
Meera added quietly, “If you had asked each of us in the OR what the blood loss was at different times, you would have gotten different answers. Our methods are manual, fragmented and subjective. We track blood on gauze one way, suction fluid another way, urine output somewhere else, irrigation in someone’s head or on a smudged whiteboard. All of that, in real time, in a high-pressure environment.”
The quality officer, who was also in the meeting, listened carefully.
“So you’re saying,” she asked, “that in a modern OR with monitors, ventilators, infusion pumps and electronic records… we still don’t have a reliable, automated way to know how much fluid the patient is actually losing during surgery?”
Meera nodded. “Exactly. We are making high-stakes decisions on top of guesswork. Some companies have apps that estimate blood loss from images of sponges. Some devices measure blood content in fluid lines. Some systems track urine output. But they’re all separate. Nothing gives us one integrated, real-time ‘fluid loss picture’ we can trust.”
There was a long pause.
The hospital’s biomedical engineer, who had recently started a small innovation cell with the residents, broke the silence.
“What if we tried to build something?” he asked slowly. “Even a prototype. Something that could bring all this together — blood on gauze, blood in suction, urine output, irrigation in and out — into one live dashboard for the anesthesiologist?”
Meera looked at him with a mixture of hope and skepticism.
“If you could give me a simple, reliable number,” she said, “even simulated at first… That could change how I manage fluids. It could mean safer surgeries. Fewer close calls like this.”
The room went quiet again, but this time the silence felt different. Less like helplessness, more like possibility.
Your Mission
You are part of the small innovation team invited by this hospital to study what happened in OR 3 and propose a tech-enabled solution. You have access to everything you just read: the messy reality of gauze, suction, urine, irrigation, manual notes, human error, time pressure and patient risk.
Your task is to understand the real, underlying problem in this story, map the points where information is lost or distorted, and design a technology-backed approach that could give clinicians a more accurate, real-time view of fluid loss during surgery — starting with a proof-of-concept that a hackathon team like yours could realistically build.
Last updated
Was this helpful?