Orus Studio
Healthcare

Replacing four disconnected systems with one patient record

A 120-bed multi-speciality hospital in Maharashtra

A hospital running separate software for OPD, pharmacy, lab and billing, where a discharge bill took two hours to assemble by hand.

Client
A 120-bed multi-speciality hospital in Maharashtra
Industry
Healthcare
Duration
14 weeks
Team
4 engineers, 1 designer
Year
2025

The situation

OPD registration, pharmacy, pathology and billing each ran on separate software from different vendors, none of which shared a patient identity.

A discharge bill was assembled manually by walking between departments collecting printouts, typically taking about two hours per patient.

The same patient existed as four different records, so clinical history was effectively unavailable at the point of consultation.

Insurance and TPA claims were rejected frequently because supporting documents were collected after the fact rather than during treatment.

What we did

01

One patient identity before anything else

We did not start with features. We started by building a single patient master and a deduplication process, then ran it against three years of historical records from all four systems. About 18% of records turned out to be duplicates of an existing patient — that number alone reframed the project for the hospital's management.

02

Billing as a consequence, not a data-entry task

Every chargeable event — a consultation, a dispensed drug, a test, a bed-day — posts to the patient account at the moment it happens, from the department that caused it. The discharge bill is then a read operation rather than an assembly exercise. This is the single change that removed the two-hour delay.

03

On-premise deployment with encrypted offsite backup

The hospital was unwilling to place patient data in a public cloud, which is a common and reasonable position. We deployed on a server in their own server room, with encrypted incremental backups pushed to India-hosted object storage and a documented restore procedure tested during handover.

04

Claim documentation captured during treatment

TPA and Ayushman Bharat claims each require a specific document set. We encoded those requirements into the treatment workflow, so the system prompts for what is missing while the patient is still admitted rather than at submission.

The parts that were actually hard

Problem

The existing pathology system had no export function and no API. Its vendor was unresponsive.

How we handled it

The data sat in a local SQL Server instance the hospital had credentials for. We read directly from it, mapped the schema by inspection, and validated the migration against 500 manually checked records before cutover.

Problem

Power cuts of 20 to 90 minutes were routine, and the billing counter could not stop.

How we handled it

Billing runs against a local cache and queues writes, so the counter continues through an outage on UPS power. The reconciliation logic on reconnect was the hardest part to get right, particularly around invoice number sequencing, which GST rules require to be unbroken.

Problem

Consultants refused to type clinical notes, which had sunk a previous software attempt.

How we handled it

We built templated notes per speciality with the common diagnoses pre-filled, reducing a typical note to a few taps. We also accepted that some consultants would keep writing on paper, and made the system work when the clinical note is a scanned attachment rather than structured text.

What shipped

  • Single patient master with deduplication across four legacy systems
  • OPD, IPD, pharmacy, pathology and billing modules on one record
  • TPA and Ayushman Bharat claim workflow with document checklists
  • On-premise deployment with tested encrypted backup and restore
  • Migration of three years of historical patient and visit data
  • Role-based access control with per-record audit logging

Outcomes

  • Discharge billing moved from roughly two hours to a few minutes, since the bill is now assembled continuously rather than at discharge
  • Duplicate patient records eliminated at migration, making prior history visible during consultation for the first time
  • Hospital staff reported a marked drop in claim rejections attributable to missing documents, though we did not have a clean baseline to measure against
  • The system has run through routine power outages without billing downtime since go-live

Outcomes are described qualitatively where no clean measured baseline existed before the work started.

Built with

Next.jsPostgreSQLPrismaRedisDocker

Have a problem shaped like this one?

Tell us what you are dealing with. We will come back with a scoped estimate and an honest view on whether we are the right fit.

Explore