# Apps and custom software for diagnostic labs

A diagnostic lab should commission software when the gap between collection and an authorised report is where it loses money and accreditation evidence. Off-the-shelf lab systems handle analysers well and handle home collection, doctor referral share-back and multi-centre reconciliation badly.

> Source: MyFloww, a software studio in Bengaluru, India. Canonical page: https://myfloww.in/for/diagnostic-labs/custom-software/. Last reviewed 2026-09-04 by Neel.

## How diagnostic labs in India actually operate

A diagnostic lab in India sells turnaround time and trust, not tests. The operating shape is hub and spoke — collection centres and home phlebotomists feeding a central processing lab — so the same result survives several handoffs before it reaches a patient, and every handoff is somewhere the sample or its identity can be lost. Two customers exist at once: the patient who paid, and the referring doctor whose repeat referrals are the volume.

## The workflow we build around

1. **Booking, increasingly for home collection** — A patient books a test or a package by phone, on WhatsApp, through an aggregator or at a collection centre. Home collection has moved from a premium add-on to an expectation in metros, which puts a phlebotomist’s route at the front of the operation.
2. **Collection and barcoding** — The sample is drawn and labelled, and this is the moment identity is either locked to the sample or lost. A barcode generated at collection and scanned at every later step is the difference between a traceable result and a repeat draw.
3. **Transport to the processing lab** — Samples travel from centres and home visits in batches, some temperature-sensitive with real stability windows. The promised turnaround starts here, and this is the part of the chain with the least visibility.
4. **Processing and pathologist authorisation** — Analysers produce results into the lab information system, and a qualified pathologist reviews and authorises before anything is released. This step is not optional and not automatable, and it is the single most important constraint on anything built around a lab.
5. **Delivery to the patient and back to the referring doctor** — The authorised report goes to the patient, usually as a link or PDF, and separately to the doctor who referred them. The doctor’s copy generates the next referral, and it is the one most often forgotten.

## Integrations that matter for custom app and web application development here

**The lab information system and its analyser interfaces**
: Analysers speak HL7 or ASTM into the LIS, and that system is the source of truth for every result. Custom work wraps it rather than replacing it, because re-implementing analyser interfacing and validation rules means re-earning accreditation for something that already works. The only real decision is which system owns sample identity, and the answer is the LIS.

**Barcode generation and label printing at the point of collection**
: Identity binds to the sample at the moment of draw, including at a patient’s kitchen table, so barcodes have to print from a phlebotomist’s phone and a portable printer rather than only from a desk. Every later scan becomes a traceability record, which is exactly what accreditation asks for.

**Payment collection at the door and at the centre**
: Home collection means cash changes hands in the field, and cash collected by phlebotomists is the reconciliation problem most labs are quietly carrying. A UPI link raised against the specific booking and settled before the sample is logged replaces an end-of-day count with a record attached to the order.

**A referring-doctor portal**
: Doctors chase labs for reports, and the chasing is what makes a lab annoying to work with. Giving each doctor a view of the patients they referred and the turnaround those reports actually took converts an administrative burden into the reason they refer again.

## Constraints that change the build

- **No result leaves the lab before a pathologist authorises it.** An analyser producing a value is not a result. Release requires review and sign-off, so any delivery automation is triggered by the authorisation event and never by the analyser. A system that can be configured to auto-release is a system that eventually will be.
- **NABL accreditation demands end-to-end traceability.** Accreditation requires every result to be traceable through who collected the sample, when, which analyser ran it, which calibration was current and who authorised release. That makes the audit trail structural rather than a feature, and rules out any step done outside the system.
- **Reports are sensitive health data on consumer channels.** A report sent to a phone number is a diagnosis on a device that may be shared, and a wrong number is a disclosure. Under the DPDP Act this is sensitive personal data, so expiring links that verify the recipient are the defensible pattern rather than attachments that live in a chat forever.

## What it costs

| Scope | Price (INR) | Typical timeline |
| --- | --- | --- |
| Home collection and delivery layer over an existing LIS | ₹3,50,000 – ₹8,00,000 | 10 – 16 weeks |
| Multi-centre operations with reconciliation and a doctor referral portal | ₹8,00,000 – ₹18,00,000 | 16 – 28 weeks |

These are bands MyFloww would honour, not indicative ranges. Every project is quoted
individually after scoping.

## A worked example

**Context.** A lab running several collection centres and a growing home-collection service could tell a patient a report would be ready but could not say where a sample was. Turnaround complaints were investigated by phoning people, and cash collected at the door was reconciled from a notebook at the end of the week.

**What was built.** A layer over the existing LIS rather than a replacement: barcodes generated at collection including on home visits, a scan at every handoff, payment taken at the door against the booking, and report delivery triggered by the pathologist’s authorisation rather than by the analyser producing a value.

**What changed.** Turnaround stopped being an average and became a measurement per stage, so delays could be attributed to transport, processing or authorisation instead of argued about. The weekly reconciliation stopped existing, because money was recorded against the order at the moment it was taken.

## What goes wrong with custom app and web application development for diagnostic labs

- Building a patient app before the collection-to-authorisation pipeline is trustworthy. An app showing a status the lab cannot verify converts a private delay into a public one, and every complaint then arrives with a screenshot.
- Allowing release as soon as the analyser produces a value, because it improves the turnaround number. Make auto-release impossible to configure rather than discouraged: a lab shipping an unreviewed value risks a patient acting on it and its own accreditation in the same moment.
- Replacing the LIS to get a nicer interface. The interfaces, validation rules and evidence trail are the expensive part, and they already work.
- Treating the referring doctor as a report recipient rather than as a customer. Referrals are the volume in this business and they are won on how easy the lab is to work with.

## Questions

### Should a diagnostic lab replace its LIS or build around it?

Build around it, almost always. The lab information system holds analyser interfaces, validation rules and the accreditation evidence trail, and replacing it means re-earning all of that for functionality that already works. The gaps worth commissioning sit outside it: home collection, payment in the field, multi-centre reconciliation and the referring doctor’s view.

### What does home collection actually require in software terms?

Four things a centre-based system usually lacks: a phlebotomist app that works on a patchy mobile connection, barcode printing away from a desk, payment taken at the door against the booking, and a route that accounts for sample stability windows rather than only for distance. The last is what separates a working service from one that repeats draws.

### How does custom software help with NABL accreditation?

By making traceability a by-product of the work rather than a document assembled before an assessment. If every handoff is a scan and every release is an authorisation event with a named person against it, the evidence already exists. The risk runs the other way too: any step performed outside the system is a gap in the trail.

### How long before a build like this is usable?

Ten to sixteen weeks for the collection-to-delivery layer over an existing LIS, longer where several centres and a doctor portal are in scope. Phase it so the chain-of-custody scans go live before anything patient-facing, because those scans are what make every later status claim true.

### Can it work offline in a phlebotomist’s hands?

It has to tolerate a bad connection rather than assume none. Draws happen in stairwells and basements, so the app queues the scan and the payment locally and reconciles when signal returns — with the barcode generated on the device so identity is never waiting on a server round trip.

## About diagnostic labs generally

### Why is turnaround time the product for a diagnostic lab?

Because the test itself is a commodity — the same panel on the same analyser gives the same number anywhere. What a patient and a referring doctor choose between is how fast the authorised report arrives and how reliably. A lab that can state turnaround per stage, and hold to it, is selling something its competitors cannot describe.

### Who is the real customer, the patient or the referring doctor?

Both, and they want the same result delivered completely differently. The patient wants their report on their phone the moment it is signed off. The doctor wants a consolidated view of everyone they referred and the turnaround they experienced. Referrals are the volume, so the doctor’s view is a revenue channel rather than a courtesy.

### What does NABL accreditation actually require from software?

That every result is traceable end to end: who collected the sample and when, which analyser ran it, which calibration was current, and who authorised release. If every handoff is a scan and every release is an authorisation event with a named person against it, the evidence an assessor asks for already exists rather than being assembled beforehand.

### Is home collection worth building for?

In metros it is no longer optional. It changes the operation more than it looks: money moves in the field, barcodes have to be printed at a kitchen table, routes have to respect sample stability windows rather than only distance, and the busiest inbound question becomes when somebody is arriving.

### Should a lab replace its LIS?

Almost never. The lab information system holds analyser interfaces, validation rules and the accreditation evidence trail, and replacing it means re-earning all of that for functionality that already works. The gaps worth commissioning sit outside it — home collection, payment in the field, multi-centre reconciliation and the doctor’s view.

## Related

- Industry hub: https://myfloww.in/for/diagnostic-labs/
- Apps and custom software in general: https://myfloww.in/custom-software/
- Nearest industry: https://myfloww.in/for/dental-clinics/
- Nearest industry: https://myfloww.in/for/physiotherapy-clinics/

---

MyFloww · software studio · Bengaluru, India · connect@myfloww.in · https://myfloww.in