Skip to content
Sanjeevani Prime

Diagnostics

How to reduce turnaround time in a pathology lab, one stage at a time

Most of the delay in a lab report happens outside the analyser, and you only find it once every stage carries its own timestamp.

By Sanjeevani Prime Editorial DeskPublished 10 min read
Abstract illustration of sample tubes moving along a timeline of clock faces through a laboratory

A physician on the second floor rings the lab at one in the afternoon. The CBC he ordered at half past nine is still not on his screen. The technician checks and finds the result came off the analyser at 11:40. It then sat in the queue for the pathologist, who was reading slides, and nobody noticed until the phone rang. On the monthly quality report, the same lab shows an average TAT of three hours for haematology and calls it acceptable.

That gap between what the report says and what the clinician lives through is where most labs lose trust. A single average TAT hides the stage that is actually slow, and it hides the bad days behind the good ones. The fix is rarely a faster analyser. It is knowing, sample by sample, where the minutes went.

What counts as turnaround time in a pathology lab?

NABL 112A defines turnaround time as the elapsed time between two specified points across the pre-examination, examination and post-examination processes. The lab chooses those points. Most labs use registration to report release for lab TAT, and collection or order to report delivery for the clinician's view of TAT.

Write both definitions into your quality manual and keep them separate. Lab TAT tells you about your own processes. Order-to-delivery TAT tells you what the ward or the walk-in patient actually experiences, and it includes the phlebotomist's round and the courier from the collection centre, which the lab often does not control directly but must still manage.

Set a target per test, not per department. A fasting glucose, an HbA1c, a thyroid profile and a culture have nothing in common except the building. A sensible target sheet might say glucose within 2 hours, thyroid profile the same day if received before noon, urine culture preliminary at 24 hours and final at 48 hours. The numbers are yours to set based on your workload and what your referring doctors need. What matters is that each one is written down and published to users.

Where do the minutes actually go?

In a typical small or mid-sized lab, the time spent on the analyser is the shortest part of the journey. Waiting is the long part: in the registration queue, in a cool box at the collection centre, in a rack waiting for a batch to fill, and in the validation queue waiting for a signature.

The table below is an illustration of one routine biochemistry sample in a 60-bed hospital lab with two outside collection centres. The minutes are not survey data; they are the kind of pattern you will see when you first pull stage timestamps from your own system.

StageTimestamp that marks the endIllustrative minutesCommon leak
Registration and billingBill saved, barcode printed12Single counter at 8 to 10 am, manual entry of doctor name
Sample collectionCollected-at time on the barcode15Patient waits for phlebotomist, wrong tube means a redraw
TransportReceived-in-lab scan70Courier runs every two hours from the collection centre
Accessioning and centrifugationSample accepted in LIS25Samples logged in bunches, centrifuge loaded only when full
Batching and analysisResult received from analyser40Waiting for a batch of 20 before loading
ValidationResult approved55All results, normal or not, wait for one signatory
Report deliveryReport sent or printed30Report printed and kept at the counter until patient returns

Add it up and you have about four hours, of which the chemistry itself took well under an hour. Transport and validation together are more than half. That is the usual shape, and it is good news, because those two stages respond to process changes that cost very little.

How do you measure TAT per test and per stage?

Capture a system timestamp at each hand-off: registration, collection, receipt in lab, acceptance, result from analyser, approval and delivery. Then report, for each test and each stage, the median time, the 90th percentile and the percentage of samples within target, broken down by shift and by collection centre.

Timestamps must come from scans or system events, not from someone writing the time on a sheet at the end of the shift. If your collection centre does not scan samples out, the transport stage will always be a guess. A barcode scan at dispatch and another at receipt costs a few seconds and gives you the courier's real performance.

Avoid the average. One haemolysed sample that was redrawn the next morning can pull a department's mean up by half an hour and tell you nothing useful. The median shows your normal day. The 90th percentile shows how bad the slow tail is, and that tail is what the clinician remembers.

A worked example

Suppose your target for serum creatinine is 3 hours from registration. Last month you ran 1,200 creatinines. The median was 2 hours 10 minutes, which looks fine. The 90th percentile was 4 hours 40 minutes, and only 78% were within target. Breaking it down by shift shows the evening samples are the problem: one technician runs chemistry and haematology together after 5 pm, and biochemistry gets loaded twice a shift. Now you know where to spend effort, and it is not a new analyser.

Review these numbers weekly for the first two months, then monthly in the management review. When a test falls below its within-target percentage, open a corrective action the same way you would for an EQAS outlier.

What does NABL expect from a lab on turnaround time?

NABL 112A, the specific criteria read with ISO 15189:2022, lists turnaround time among the quality indicators a lab must monitor. It also asks labs to make their accredited scope, with method and TAT, available to users, and to define, display and communicate critical limits that need immediate attention.

A few specific points in the current document are worth knowing. For histopathology, NABL 112A says frozen section and squash smear TAT should not exceed 20 minutes, and asks labs to define biopsy TAT by specimen complexity and to strive for more than 75% within-TAT performance in each defined category. For results within critical limits, the lab must communicate them to the clinician or authorised person and document that communication. Assessors will ask to see the TAT indicator data, how often it is reviewed, and what you did when you missed it.

NABL updates its documents regularly. Check the current issue of NABL 112A and the guidance document NABL 112B on the NABL documents page before your next assessment rather than relying on an old copy in the quality cupboard. If your hospital is also working towards NABH, the same TAT data feeds its quality indicators; our guide to NABH entry-level accreditation for small hospitals covers how the two fit together.

How should STAT and routine samples be handled differently?

A STAT sample should never wait for a batch. Mark it at registration with a distinct barcode or tube label, run it on arrival, validate it immediately and phone or message the result. Routine samples can be batched, but on a fixed clock schedule rather than waiting for the rack to fill.

The trouble in many labs is that everything from casualty is marked urgent, so nothing is. Agree a short STAT menu with your clinicians: typically blood glucose, electrolytes, haemoglobin, platelet count, troponin, blood gases, serum creatinine and a few others. Keep the list on one page and give it its own TAT target, often 60 minutes from receipt.

For routine work, change from fill-the-batch to run-on-the-clock. Loading chemistry every 30 minutes, whether there are 8 samples or 25, costs a little reagent on quiet days and removes the worst waits on busy ones. The same logic applies to the centrifuge: a small second centrifuge for STAT tubes pays for itself quickly.

Do analyser interfacing and auto-validation really reduce TAT?

Yes, and usually more than any other single change. A bidirectional interface sends the test order to the analyser from the barcode and returns results to the LIS without retyping. Auto-verification then releases results that pass defined rules, so only abnormal or flagged results wait for a person.

Manual transcription is slow and error-prone. A technician copying 300 results off a printout every morning will make mistakes, and someone then has to check every line. With a host-query interface, the analyser reads the barcode, asks the LIS what to run and sends the results back against the right sample.

Auto-verification needs more care. NABL 112A permits automated selection and release of results for tests that do not need interpretation, under conditions. The rules must be set and approved by authorised, qualified personnel using available standards. Reports released this way should be clearly marked "Auto verified" and traceable to the person who approved the rules. The system must be reviewed once a year, have a way to suspend auto-release quickly if something goes wrong, and come with a contingency plan.

A first set of rules for a routine chemistry panel usually covers:

  • Result within the reference interval, or within a narrower auto-release range you define.
  • No analyser flags such as haemolysis, lipaemia, icterus, short sample or clot.
  • Internal QC for that analyte passed and in date for the run.
  • Delta check against the patient's previous result within a set window, where one exists.
  • No critical value, which always goes to a person and triggers a call.

Start conservatively. Run the rules silently for a few weeks, compare what the software would have released with what the pathologist actually approved, and only then switch on auto-release. In most labs this moves a large share of normal results out of the validation queue, which is exactly the stage that was slow in the table above.

How quickly can reports reach patients by WhatsApp or SMS?

Within a minute or two of approval, if delivery is triggered by the release event rather than by a staff member. The LIS sends an SMS or WhatsApp message with a secure link or PDF as soon as the last test on the request is approved, and the clinician sees the same report in the HIS at the same moment.

A few practical rules make this work. Send one message per request when it is complete, with an optional partial report for long tests like cultures, so patients are not flooded. Use a link that needs the patient's mobile number or an OTP, rather than a plain PDF that can be forwarded to anyone. Record the consent for digital delivery at registration, and keep a log of what was sent, to which number and when. Patient reports are personal data, and the Digital Personal Data Protection Act, 2023 applies to how you collect and share them. Labs that are linking reports to patients' ABHA health records should read our ABDM guide for small hospitals.

Most current lab software can do this. If you are evaluating one, check that it captures every stage timestamp listed above, supports bidirectional interfacing with your analysers, has rule-based auto-verification with an audit trail, and sends reports on release. Healthixio's pathology LIMS, made by Kentron Technologies, the publisher of this site, is one example of a lab system built around those features; compare it against others on the same checklist.

A 30-day plan to cut TAT in your lab

None of this needs a large budget. It needs one person who owns the TAT numbers and a month of steady work.

  1. Week 1: Write down your TAT definitions and a target for your 30 highest-volume tests. Publish the list to clinicians and collection centres.
  2. Week 1: Make sure every stage has a system timestamp. Add a dispatch scan at each collection centre if it is missing.
  3. Week 2: Pull one month of data. Report median, 90th percentile and percentage within target for each test, by stage, shift and centre.
  4. Week 2: Agree a one-page STAT list with casualty and ICU, with its own target.
  5. Week 3: Move routine loading to a fixed clock schedule. Fix the courier timetable for the slowest centre.
  6. Week 3: Check which analysers are still manually transcribed and get the interface done for the highest-volume one first.
  7. Week 4: Draft auto-verification rules for one routine panel and run them silently against pathologist approvals.
  8. Week 4: Switch report delivery to trigger on release, with consent recorded at registration.
  9. Every month after: Review TAT in the management meeting and open a corrective action for any test below target.

Run the same report again at the end of the month. If the 90th percentile has come down and the phone calls from the wards have dropped, you are measuring the right thing.

Questions readers ask

What is a good turnaround time for routine blood tests?

There is no single national figure, and NABL leaves it to each lab to set and publish its own targets. Many hospital labs aim to release routine biochemistry and haematology the same working day and STAT tests within about an hour of receipt. Base your targets on what your clinicians need and what your data shows you can sustain.

Should TAT be measured from sample collection or from registration?

Track both. Registration or receipt to release measures the lab's own performance, while collection or order to delivery shows what the clinician and patient experience. Define each clearly so that the numbers in your quality report are not mixed up.

Is auto-verification allowed for NABL-accredited labs?

Yes, for results that need no interpretation, provided the rules are approved by authorised personnel, validated, and reviewed every year. NABL 112A asks for auto-verified reports to be clearly marked, for a quick way to suspend auto-release and for a contingency plan. Critical values should always go to a person.

Can we send lab reports to patients on WhatsApp?

Many labs do, and patients generally prefer it. Take consent at registration, send a secure link or protected file rather than an open PDF where possible, and keep a log of each delivery. Reports are personal health data, so follow the Digital Personal Data Protection Act, 2023 and your hospital's own privacy policy.

Which TAT figure should go in the quality indicator report?

Report the percentage of samples released within target for each key test, along with the median and 90th percentile. A within-target percentage is easy for management to follow, and the 90th percentile shows whether the slow tail is shrinking. An average on its own tends to hide problems.

Sources

  1. 1.NABL documents (NABL 112A specific criteria and NABL 112B guidance for medical laboratories)
  2. 2.NABL 112A: Specific Criteria for Accreditation of Medical Laboratories, Issue 01
  3. 3.NABL 112B: Guidance Document, Medical Laboratories
  4. 4.Journal of Laboratory Physicians: comparing ISO 15189:2012 and 2022 with NABL 112 insights

Sanjeevani Prime is published by Kentron Technologies, which also makes Healthixio and Tatva ERP. Where an article mentions those products, we say so. This article is general information, not medical, legal or tax advice; rules change, so check the sources above before acting. Read our editorial policy.

Keep reading