Pathology Lab Software: Samples, Results and Trust

Lab software is a chain-of-custody problem wearing a billing system's clothes. From the moment a sample is drawn to the moment a report is released, every step must be attributable - who collected it, who ran it, who validated the result, who released it. Get that right and the reporting, billing and referral features are straightforward.

This guide covers barcoding and the worklist, result entry and validation, report release, and the referral economics that decide whether a lab is profitable.

Barcode at collection, not at the bench

The single highest-value thing lab software does is put a barcode on the tube at the moment of collection, tied to the patient and the ordered tests. Labelling later, or writing on the tube, is where sample mix-ups come from, and a mix-up is the one error a lab cannot recover from reputationally.

Check that the barcode covers the awkward cases: one patient with samples in three different tubes, a repeat draw when the first sample haemolyses, and a sample collected at a home visit hours before it reaches the lab. Each of these should produce a traceable record without handwriting.

• Barcode generated and printed at registration or collection
• Multiple containers per patient handled distinctly
• Repeat and recollected samples linked to the original order
• Home collection recorded with collector and time
• Sample rejection recorded with a reason, not just discarded

The worklist is how the bench actually works

Technicians do not work patient by patient; they work analyser by analyser and batch by batch. Software that only offers a patient view forces them to jump between records all day, and they will end up writing results on a sheet and entering them later - which reintroduces every transcription error the system was bought to remove.

You want a worklist per department or analyser, showing pending samples, letting results be entered in the order the bench works, and flagging anything outside reference range as it is entered rather than after the fact.

• Worklist per department, analyser or test group
• Pending, in-progress and completed clearly separated
• Out-of-range values flagged at entry
• Critical values escalated rather than silently recorded
• Turnaround time visible per sample and in aggregate

Validation and release must be separate acts

Entering a result and approving a result are different responsibilities and should be different actions by different people, each recorded. A technician enters; a pathologist validates; the report is released. Software that lets a result go straight from entry to a printed report with a signature on it has automated away the check that matters most.

Ask what happens when a released report has to be corrected - because it will happen. The right answer is an amended report with its own version, the original retained, and a clear indication to the recipient that it supersedes. A silent edit of a released report is the worst possible behaviour and more common than you would hope.

• Entry and validation as distinct, attributable steps
• Digital signature applied at release, by the validating pathologist
• Amended reports versioned, with the original retained
• Recipients able to tell an amended report from the first issue
• Full audit trail of who did what to a result and when

Referral economics decide whether you make money

Most diagnostic centres take work from referring doctors, collection centres and corporate tie-ups, each on different rates and commissions. If those rates are applied manually at billing, margins leak quietly and nobody notices until a rate that was agreed two years ago is still being honoured.

You want rate lists per referrer, applied automatically, with a commission or share calculated from the same records that produced the bill. And you want a report that shows revenue and payout per referrer, because that is the conversation you have when renegotiating.

• Rate lists per referrer, doctor or corporate account
• Commission computed from bills, not re-entered
• Outsourced or send-out tests tracked with their cost
• Revenue and payout per referrer available as a report

The shortlist checklist

Take this to any vendor, ours included. If a question cannot be answered with a straight demonstration rather than a promise, treat it as unanswered.

1. Register a patient with three tests in two tubes and print the barcodes.
2. Reject a sample and check the reason is recorded and visible.
3. Enter results from a department worklist rather than patient by patient.
4. Enter an out-of-range value and confirm it is flagged at entry.
5. Validate and release a report as a second user, then amend it.
6. Check the amended report is versioned and the original retained.
7. Apply a referrer rate list and find the commission report.
8. Ask what the audit trail records and how long it is kept.