From Paper to Pixels

By now, I am the last person who can claim to be “neutral” about our Hospital Information System. I lived with it for two decades—sometimes like a proud parent, sometimes like an exhausted mechanic. Still, if I step back and look at it the way a visiting outsider might, a simple question comes to mind: Did we actually build something useful, or did we merely replace paper with screens?

In April 2009, the New England Journal of Medicine published a survey on health information technology in American hospitals. Its opening line stayed with me: less than 2% of acute care hospitals had a comprehensive electronic records system, and only 8–12% had even a basic one. When I first read that, I smiled—partly in disbelief, partly in relief. If wealthy hospitals in the United States were struggling, what were we trying to do in Sevagram, a village that often didn’t have stable electricity for half the day? No reliable internet. No smartphones. Not even the habit of using a mouse. And yet, we had decided to digitise a 1,000-bed teaching hospital as if it was the most natural thing to do.

We began in 2004, not with grand promises, but with a practical irritation: paper was failing us. Handwritten notes went missing. Files grew fat and unreadable. Lab reports got tucked into the wrong folder. Discharge cards looked like they had survived a cyclone. With nearly 50,000 admissions a year, the system was cracking under its own weight. We needed something that could hold information without tearing, fading, or walking away in someone’s pocket.

The quiet revolution

The first visible change was not dramatic. It was almost boring—registration became faster, test orders became cleaner, reports became retrievable. But in a hospital, “boring” is a compliment. Automation took over the repetitive work: registering patients, printing slips, generating reports, tracking bills, and sending information across departments without a messenger boy sprinting down corridors. Slowly, the hospital began to breathe a little easier.

The discharge summary, in particular, changed shape. Earlier, it was a daily embarrassment. Residents wrote them in haste, handwriting collapsed into hieroglyphics, and crucial details slipped through the cracks. With HIS, the patient’s demographic details pulled in automatically. Investigations came in neatly—date-wise and lab-wise. Operation notes and in-hospital medications got recorded without anyone having to “remember” them later. The resident still had to write the clinical story, but the skeleton was ready. For the first time, discharge summaries looked like something you could actually read without a magnifying glass and a prayer.

The pharmacy felt the impact even more. Every day, about 1,500 patients queued up for medicines. In the old days, a crowded counter and a hurried handwriting were an invitation to error. HIS didn’t eliminate mistakes—nothing does—but it reduced the chaos. Billing became cleaner. Stock tracking became smarter. Pharmacists spent less time deciphering prescriptions and more time dispensing safely. That, to me, was progress.

Did it improve outcomes? I can’t prove it

This is the part where I must resist the temptation to sound triumphant. Did electronic records improve clinical decisions? Did they reduce complications? Did patients live longer because of our software?

I don’t know. And I cannot pretend I do.

We never ran a proper before-and-after evaluation. We didn’t measure errors systematically. We didn’t compare outcomes across years. Like many hospitals, we were so busy building the system that we forgot to study the system. If I were advising my younger self today, I would say: Do the research while you do the work.

What I can say, with honesty, is simpler. The hospital ran more smoothly. Information became easier to find. The number of “missing” things—files, reports, bills—reduced. And when you reduce friction in a hospital, you reduce suffering, even if you cannot calculate it neatly in percentages.

The iPad moment

One day, much later, I saw residents doing rounds with iPads. Radiology images looked startlingly clear. Blood sugars and creatinine values appeared as graphs instead of scattered numbers. Orders could be placed without running back to a desktop computer like a schoolboy sent on errands. That sight gave me a quiet happiness—not because it looked modern, but because it looked useful. The screen was finally at the bedside, where it belonged.

Resistance, frustration, and the slow art of adoption

If technology was the engine, adoption was the fuel. And fuel doesn’t come easily in a hierarchical hospital.

Many senior doctors trained in the 1960s and 70s disliked the idea of a computer between them and the patient. Some felt data entry was beneath them. Some worried it would slow them down. Others simply didn’t trust a system that had been designed outside Sevagram. A few saw it as an administrative obsession, not a clinical need.

So we took a “bottom-up” route. We leaned on residents, clerks, pharmacists, technicians—the people who actually carried the hospital on their shoulders. We sought the passive support of department heads. We tolerated the year of double work—paper plus digital—because that was the only way to build confidence. There were days when the server crawled, transactions failed, and the hospital staff looked at us as if we had personally invented suffering. We absorbed complaints, fixed what we could, and returned the next morning to try again.

The gap between engineers and doctors is real. At times it felt like two different alphabets trying to form one sentence. But over time, we learned each other’s language—slowly, clumsily, and with plenty of irritation. What kept it alive was not brilliance. It was persistence.

What we still didn’t get right

Even today, I can list what we lack without hesitation.

We never built strong clinical decision support into the system. Drug–drug interactions and allergy alerts didn’t become routine in critical areas. Vital signs were not captured reliably across all wards. Problem lists and active medication lists remained incomplete. Some departments stayed stubbornly paper-based. CPOE excited a few units but failed to capture the whole hospital’s imagination. Patients still didn’t have easy electronic access to their own records. Data privacy and security needed deeper attention. And in many ways, we digitised old workflows instead of redesigning them.

In other words, we improved the container, but we didn’t always improve the contents.

A family project, and a hospital project

When I think of those years, I often remember the old Hindi film Chalti Ka Naam Gaadi—a story held together by family, humour, breakdowns, and stubborn forward movement. HIS also became like that. We didn’t work on it only in office hours. It entered our dinner conversations. It sat beside us like a silent guest. Bhavana, Ashwini, and later Shaily were part of that long, unglamorous labour.

But this was never a “family achievement.” The real credit belongs to the people who kept showing up: the pharmacists who learned computers despite fear, the technicians who entered results patiently, the clerks who corrected errors without drama, the residents who used the system even when it slowed them down, and the engineers who tolerated our endless demands and clinical tantrums. The system worked because it became everyone’s system.

We rarely celebrated milestones. We just moved from one problem to the next, like doctors on night duty. Perhaps that is why, even now, the most satisfying part of this story is not the technology itself. It is the memory of a hospital learning—slowly, awkwardly, and with a lot of resistance—to trust a new way of working.

And every time our car approaches that blind turn near the railway underpass between Sevagram and Pavnar, my granddaughters shout, “Baaju!”—making way, making noise, making us laugh. The hospital, too, learned to make way. From paper to pixels, it kept moving.

---