Logo Jade
Demo

The new generation of JadeIH

Every blood group, established on two samples

JadeIH, the immunohaematology software of the Jade suite, receives blood group requests from your LIS, drives your analysers and only confirms a blood group after two determinations on two samples. It is designed for hospital laboratories and multi-site networks that perform blood grouping and antibody screening.

Request a demo See the journey of a request
  • 2/2two determinations, two samples, two people
  • § 5.1.5readable SVTM rules, cited to the paragraph
  • ✓ Jadereceipt of results verified in your LIS
Fictitious patients, numbers and analysers. Animated demo on fictitious data: request IH26-004122 for KELLER Hans opens from the list, the wells of the gel card are read (AB, RhD positive), Lucie Morel validates technically, then Dr Hélène Vautier validates biologically. JadeIH announces “This will be the 2nd determination: the group will be confirmed.”, the identity moves to two determinations, the card loses its PROVISIONAL watermark and the results are received in Jade at 08:36.

What JadeIH connects

Apart from the analysers, everything comes from the same vendor.

  1. Jadeyour LIS

    Remains the master of patient identity. Sends one HL7 order per tube and receives the validated results.

  2. The middlewarethe relay

    Carries messages over MLLP, holds them during an outage, translates codes and keeps its own log.

  3. JadeIHBlood Bank module of JadeMP

    Checks every entry and every validation, maintains the blood group identity and prints the blood group card.

  4. IH analysersthrough the middleware

    Receive their orders per test and per site, and return reactions well by well. List on request. The middleware is compatible with Instrument Manager (Data Innovations).

  5. Jade Litethe Jade workstation

    Displays JadeIH blood group requests, patients and merges, and tracks received tubes.

Four blood bank stories, told screen by screen

A night shift, a duplicate patient, a rule that changes, a data migration. Each step is timestamped and shows the JadeIH screen at that moment.

  1. From 02:14 to 07:41.A night shiftAn urgent request, an unknown patient, a blood group confirmed in the morning.
  2. From 1 to 2 November.A duplicate patientTwo records for the same patient, merged without losing a determination.
  3. From 5 to 10 November.A rule that changesAn alert added to the rules, reviewed by three biologists before coming into force.
  4. From thirty days before go-live to D-day.Migrating from the old JadeIHThe entire history migrated and checked, then go-live on D-day.

A night shift

Night of 4 to 5 November 2026, Val-Clair Hospital. Yann Berger, technician, is running the blood bank alone.

Fictitious patients, numbers and analysers. Screens recreated from the application.

02:14 The request arrives from Jade.

In the emergency department, Paul Vuillemin, 47, may need a transfusion. JadeIH receives the urgent order, acknowledges it and puts it at the top of the list; the banner counts the emergency.

Recreated screen: Blood group requests

02:15 Unknown to the blood bank.

JadeIH searches for the patient by Jade number, then by AHV number, never by name alone, and does not find him. Yann confirms: the record opens with no validated group, showing the emergency procedure from the Swiss guidelines.

Recreated screen: patient record, no validated group

02:20 The analyser link goes down. Nothing is lost.

At 02:16, the whole tube was sent to IH-VCC-1, routed per test and per site. The outage shows on every screen; the middleware holds the messages, and everything resumes at 02:24.

Recreated screen: request record, link down

02:31 Results come back, well by well.

IH-VCC-1 returns the group, the reaction of each well and the card image: anti-A and anti-D hold the cells at the top of the gel; on the plasma side only the B cells agglutinate. A RhD positive, antibody screen negative.

Recreated screen: analyser results and gel card

02:33 Urgent technical validation: the result goes out immediately.

Yann types in the confirmation code: the results go to Jade, flagged as not biologically validated. The request stays in the biological queue: a single tube does not make an established group.

Recreated screen: request validated urgently (UTV)

03:05 A second tube, collected at another time.

A new tube is collected in the emergency department; at 03:22, IH-VCC-1 returns the same group. On the same tube, JadeIH would have warned: “Same sample as the 1st determination: does not count as the 2nd.”

Recreated screen: patient record at 03:24, two tubes

07:30 In the morning, another person validates.

Dr Hélène Vautier neither entered results nor validated technically: JadeIH lets her validate both requests. The 2nd confirms the group, and the card loses its PROVISIONAL watermark.

Recreated screen: confirmed identity and blood group card

07:41 Received in Jade, not just sent.

The validated results go back to Jade as ORU^R01. JadeIH then queries Jade and writes it to the request log: the loop is closed, and you know it.

Recreated screen: request log, receipt verified

What this story shows

Next story: A duplicate patient

A duplicate patient

Sophie Lambert has been known to the blood bank since 2023, with one determination. On the evening of 1 November, she arrives at the emergency department.

Fictitious patients, numbers and analysers. Screens recreated from the application.

01.1122:40 In the emergency department, a second record.

The record received from Jade has no AHV number: JadeIH cannot match it with the 2023 one. An O RhD positive determination is validated on it that same evening.

Recreated screen: the second record, without AHV number

02.1102:00 At night, JadeIH spots the pair.

The duplicate search runs every night and for every new patient. It proposes the two records: same last name, same first name, same date of birth, score 95.

Recreated screen: Duplicates and merges

08:12 In the morning, “Compare” puts the records side by side.

Dr Marco Rossi compares identity, group, determinations and requests, column by column. The “In Jade” panel shows the latest results; he ticks “I have checked the tests shown”.

Recreated screen: Merge two records, the comparison

08:16 The checks pass; the merge waits for a biologist.

ABO and RhD are identical: a discrepancy would have blocked the merge, with no override. Since it brings the 2nd determination, Dr Rossi's request goes to a biologist for approval.

Recreated screen: checks before the merge

09:05 Another biologist validates, and everything goes through at once.

Dr Hélène Vautier did not request the merge: she can validate it. Record P0109871 becomes read-only and turns into an alias; the identity moves to 2/2 and JadeIH offers to resend the results to Jade.

Recreated screen: the record kept after the merge

09:20 An unmerge remains possible.

A biologist can undo the merge from the snapshot taken at 09:05: each record gets back its requests and its determination.

Recreated screen: the Merges tab

What this story shows

Next story: A rule that changes

A rule that changes

A Swiss laboratory applies version 3 of its rules. The head of department asks for an alert on unclassified RhD variants.

Fictitious patients, numbers and analysers. Screens recreated from the application.

05.1110:20 A revision starts from the rules in force.

Dr Hélène Vautier opens a revision: version 4 starts from the rules in force, nothing is re-entered. Until it is validated, version 3 applies.

Recreated screen: Rules, revision v4

10:34 An in-house rule, taken from the catalogue.

From the catalogue, she adds “Unclassified RhD variant”, a rule written by Lysoft, at Practice level. The draft files it under “What changes”.

Recreated screen: Rules catalogue

10:41 Relax the antibody screen validity? No.

She tries to extend the antibody screen validity from 96 h to 120 h. It is a Requirement in Switzerland: JadeIH warns her and asks for a reason; she gives up.

Recreated screen: a Requirement someone tries to relax

11:02 The test cases pass, the simulator confirms.

“All checks pass: 32/32 test cases.” In the simulator, a weak D of unspecified type triggers the new rule, and the result does not count as a determination.

Recreated screen: Rules simulator

06.1109:00 Three biologists, hidden opinions.

The revision goes to the committee. Each member gives an opinion rule by rule and only sees the others' after giving their own.

Recreated screen: committee review, as seen by Dr Marco Rossi

10.1108:00 A biologist who is not the author puts v4 into force.

Dr Anne Collet keeps the rule as is and puts it into force with an effective date of 10.11.2026.

Recreated screen: decision and entry into force

08:05 A locked report, a cited version.

The report “Blood bank rules, version 4” keeps the opinions, the decision and the source of each rule. Every future trigger will cite v4.

Recreated screen: version 4 report

What this story shows

Next story: Migrating from the old JadeIH

Migrating from the old JadeIH

A laboratory is leaving the old JadeIH. Thirty days before go-live, it prepares the migration of its entire history.

Fictitious patients, numbers and analysers. Screens recreated from the application.

D−30 The source: the old JadeIH, read-only.

JadeIH connects directly to the old JadeIH database, with a read-only account. “Test” checks the connection before any trial.

Recreated screen: Data migration, the source

D−28 A mandatory dry run.

Everything is read and checked, nothing is written: 12,480 patients, 31,206 requests, 214 antibodies. Eight anomalies remain to be resolved before the import.

Recreated screen: dry run, fictitious counts

D−27 Corrections, each with its reason.

The six site codes and the unknown status are mapped. A patient born on 31.02 is excluded, with a reason, and the rerun trial passes with no anomaly.

Recreated screen: corrections before the import

D−21 The import, batch by batch.

The import replays exactly the last trial, in batches, and can be rerun. Identities come in exactly as they were validated in the old JadeIH.

Recreated screen: batch import

D−20 Every identity discrepancy goes to a biologist.

JadeIH recalculates each identity from the migrated requests: three serious discrepancies. For CHAPPUIS Henri, a 2017 request says RhD negative; Dr Anne Collet decides “To correct”, with her reason.

Recreated screen: identity check

D-day On D-day, only what has changed goes through.

A second pass only writes what has changed since the import. As long as nobody has touched what it created, the import can still be undone.

Recreated screen: migration history

What this story shows

Next story: A night shift

Guided tour

Try JadeIH like your technicians

The tour follows a request from the fictitious laboratory, from scanning the tube to receipt in Jade. It runs on its own; click a step to take control.

Fictitious patients, numbers and analysers. This morning, IH-VCC-1 is disabled: ABO and RhD go to the backup analyser, phenotype and Kell are entered by hand.

The journey of a request, from Jade to Jade

An order leaves Jade at 07:44. At 08:31, JadeIH has checked that the validated results actually arrived in Jade. In between, every message goes through the middleware and every step is written to the request log.

Fictitious patients, numbers and analysers. Accelerated clock.
  1. Jade sends the request

    One OML^O33 order per tube, relayed by the middleware. JadeIH saves it, then acknowledges it: 54 ms on average, measured in acceptance testing.

  2. Order to the analyser

    ABO, RhD and antibody screen go to IH-VCC-1, the site's preferred analyser for these tests.

  3. Analyser results

    An ORU^R01 carries the value, the reactions of each well and the card image.

  4. Two validations, two people

    Technical by Yann Berger, biological by Dr Anne Collet. The group was already confirmed: the request serves as a check.

  5. Received in Jade

    JadeIH queries Jade and writes it on the request. Otherwise, it shows “Results never arrived in Jade”.

Each test goes to its analyser, with a backup

Routing is set per test and per site, in order of preference. A tube goes out as one order per analyser. If an analyser is disabled, the next one takes over; if there are none left, the test switches to manual entry, with a reason.

Tube 261105042301 (GRP): a single order, to IH-VCC-1, for ABO, RhD, RH phenotype and Kell.

Two samples to confirm a group, two people to validate it

JadeIH recalculates the blood group identity at every biological validation, from validated requests only. The 2nd determination only counts if it comes from another tube and another collection time. A discrepancy with the known identity blocks validation, and the discrepancy is named.

341of 463

In 2024, 341 of the 463 serious or very serious near misses reported in Switzerland were stopped in the laboratory, including 168 “wrong blood in tube”: another patient's blood in the tube.

Source: Swissmedic, 2024 haemovigilance report.

What does not count as a determination

The 2nd determination only counts on another tube, collected at another time: the night shift shows it. Other results are excluded automatically, and the screen says why.

TABO, TRHphenotype and Kell not requested
FAVRE Léon born 03.11.2026

Another validator

Whoever entered a result or validated technically does not validate biologically. The setting is flagged as sensitive and every refusal is logged.

MRDr Marco Rossi validated this request technically

You validated this request technically or entered one of its results: biological validation falls to another biologist.

HVDr Hélène Vautier can validate

A discrepancy with the known identity blocks, and it is named

VUILLE Arnaud is known as B RhD positive, on two determinations. Today's tube comes back A, E+ and K+. JadeIH does not allow validation: it shows each discrepancy, and the message says what to do.

CH.R11 Discrepancy with the blood group identity, Requirement, T-CH 1145 v14, § 5.1.5

VUILLE Arnaud IH26-004133tube 261105044401
3 discrepancies with the known blood group identityValidation impossible: possible identification error, on this tube or on the known determination.
TestKnown identityThis request
ABOBA
EE-E+
KK-K+

Does not match the blood group identity: ABO, E, K. Collect a new sample and look for an identification error.

Fictitious patients, numbers and analysers.

For the biologist

Your country's rules, written in sentences

Each laboratory starts from its country's guidelines. Each rule states its level and its source, down to the paragraph. Your deviations are justified, reviewed by your biologists and tried in the simulator before coming into force.

jademp.your-lab.ch/bloodbank/rules

BLOOD BANK · RULE GUIDELINES

Laboratory immunohaematology rules

Switzerland: SVTM / Blood Transfusion SRC, T-CH no. 1145, version 14, in force from 01.02.2026

Guidelines
CH@2026.1
Content
9 definitions, 6 rules
Status
basis of laboratory layer v3
  1. D07

    Established group § Requirement

    The group is established after 2 complete determinations, on two samples with independent identification.

    T-CH 1145 v14, § 4.2.1 and § 5.1.1

  2. D03

    Antibody screen validity § Requirement

    The antibody screen is collected at most 96 h before the start of the transfusion; extension to 21 d under conditions, on medical certification.

    T-CH 1145 v14, § 4.3 and § 5.6.2

  3. D02

    Interpretation of RhD variants § Requirement

    Weak D types 1, 2, 3 and 4.1 positive, other variants negative; an untyped weak or partial D remains to be clarified.

    T-CH 1145 v14, § 5.1.3 and § 7.1.5

  4. R11

    Discrepancy with the blood group identity § Requirement

    A result that does not match the blood group identity blocks validation: collect a new sample and look for an identification error.

    T-CH 1145 v14, § 5.1.5

  5. R09

    Separation of validators · Practice

    Biological validation falls to someone other than the person who validated technically or entered a result.

    Laboratory choice

jademp.your-lab.ch/bloodbank/rules/simulator

Rules simulator in force

Switzerland, laboratory layer v3

REQUEST PATH

  1. Receptionpasses
  2. Entrypasses
  3. Tech. validationpasses
  4. Bio. validationpasses
  5. Blood group cardpasses, with a note

ON SCREEN

Request record, biological validation

This will be the 1st determination: the patient's blood group identity will be created.

CH.D07 Established group · Requirement

WHAT JADEIH INFERS

  • Girl or young womanyes: woman aged 32, under 50
  • Established groupno: 1 determination after this request, 2 are needed

Trace: CH.D07, v3, facts read: 0 validated determinations, 1st determination

Your opinion, before seeing the others'

Each biologist on the committee reviews the revision rule by rule. Opinions stay hidden until they have given their own, and the author does not put her own version into force. The story of this revision is told above.

Here, you are Dr Marco Rossi: vote, or use the keyboard with 1, 2 or 3.

jademp.your-lab.ch/bloodbank/rules/review
Revision v4 proposed by Dr Hélène VautierIn validationIn force

What changes: 1 rule added

LYSOFT.M01 Unclassified RhD variant · Practice

RhD variant not classified by the rules: state the interpretation retained.

Lysoft, JadeIH rule

All checks pass: 32/32 test cases.

THE COMMITTEE HVAgree MRAgree ACAgree Your colleagues' opinions are revealed once you have given yours.
DECISIONKeep as is by Dr Anne Collet, who is not the author

Fictitious biologists, laboratory and cases. The rule texts, their levels and sources are those of the JadeIH guidelines.

The blood group card, signed by the biologist

In Switzerland, the card follows the patient. JadeIH prints it after biological validation: provisional while only one determination is validated, never for a newborn before the configured age. Its layout is designed in an editor, passes the test sets and only goes live once validated by a biologist.

BLOOD GROUP CARDVal-Clair Hospital · Blood bank
KELLERHansBorn 09.05.1947Patient no. P0102281
AB+RhD POSITIVE
Phenotype CcDEe · K+
1st determination14.10.20262nd determinationto be done05.11.2026
BIOLOGISTH. Vautier
No known irregular antibodies
PROVISIONAL
TRANSFUSION INFORMATION
TRANSFUSION NOTES
No transfusion notes.
LAST ANTIBODY SCREEN
Negative · 05.11.2026
BLOOD BANK
Ext. 4410 · 24/7
P0102281

Show this card at every hospital admission and before any transfusion. Do not throw it away.

Two determinations on two samples: the card is established. Fictitious patient, default JadeIH template, 85.6 × 54 mm.

A versioned template

  1. v4Draftlayout in progressDraft
  2. v3Live since 02.11.2026validated by Dr Hélène VautierIn force
  3. v2Replaced on 02.11.2026kept for the recordArchived

Ten test sets, before every go-live

Everything fits on the test sets 10/10

  • Adult, full card
  • Single determination
  • Several antibodies
  • Multiply transfused, very busy
  • Very long name
  • Weak D
  • After transplant
  • AB negative with autoantibodies
  • Minimal data
  • Child, compound first name

Truncated text, an overlap or a block outside the 1.5 mm print margin blocks going live.

A migration discrepancy, decided by a biologist

The identity validated in the old software is migrated as is, then JadeIH recalculates it from the migrated requests. When the two differ, nothing is silently overwritten.

DUCRET Martine is one of the three serious discrepancies in the migration story. It's your call, with a reason.

Fictitious patient and laboratory.

jademp.your-lab.ch/bloodbank/import

Identity checkMigration compared with JadeIH's calculation

DUCRET Martine 22.06.1952 · FSerious discrepancy
Migrated identity
O+ no antibodies
JadeIH calculation
an anti-K identified in 2019, in a migrated request

The blood bank, live

Every screen's header shows the current emergencies, the age of the oldest one and the status of the links. When a message stops, you see where, without calling IT.

Demo on fictitious data: simulate an outage, then restore the link.

Tracking a tube

All the messages for a tube, step by step, with their time and the delay since the first. A message left unprocessed is replayed after a code is corrected.

2611050398015 messages, no problems, 46 min
  1. ↓Order received from the LISOML^O33 · 3 tests+0 s
  2. ↑Order sent to the analyser, acknowledgedIH-VCC-1 · ABO RHD RAI+8 s
  3. ↓Results received from the analyserORU^R01 · reactions and image+33 min
  4. ↑Results sent back to the LIS, acknowledgedORU^R01 · ACK AA+43 min
  5. ✓Results received in Jadereceipt check+46 min

Downtime mode

If the LIS is unavailable, the request is created by hand, with the paper form number. A 2nd person signs it off before any validation.

UUnknown patient, male, about 40MANUAL

IH26-004132 paper form B-0457, tube 990000131204

Sign-off by a 2nd person: pending

One hand on a record at a time

Two people cannot validate the same request at the same time. Whatever arrives in the meantime is set aside, then picked up automatically.

Opened by Lucie MorelIH26-004128

Analyser result: received, waiting for the lock

Fictitious patients, numbers and analysers.

For your IT team

A JadeMP web module, connected over HL7 v2 through the middleware

Nothing to install on workstations: JadeIH opens in the browser. It takes accounts, permissions and real-time updates from JadeMP, and talks to your LIS and analysers through the middleware.

Application
Web, in the browser. No thick client to deploy.
Platform
JadeMP “Blood Bank” module: accounts, role-based permissions, record locks, real-time updates, saved views, notifications.
Interfaces
HL7 v2 over MLLP through the middleware: OML^O33 for requests, ORU^R01 for results. Each message is logged before acknowledgement, and anything left to process is replayed.
Middleware
The middleware is compatible with Instrument Manager (Data Innovations).
Measurement
Order acknowledgement after saving: 54 ms on average, measured in acceptance testing.
With Jade
jade-api to compare the patient and the request, and to verify receipt of results.
Server and data
Spring Boot on Java, PostgreSQL database.
Jade Lite
Embeds JadeIH blood group requests, patients and merges, and tracks received tubes.
Languages
Interface in French and English.
MSH|^~\&|LYSGW|LYSOFT|||20261105074455+0100||OML^O33^OML_O33|GWVC000121|P|2.5
PID|1||P0104502||ROCHAT^Julien||19581102|M
SPM|1|261105039801
OBR|1|||TABO^ABO group^Jade|R||20261105074200+0100

SPM-2 The tube barcode. JadeIH attaches request IH26-004121 and its tests to it, then acknowledges.

Fictitious messages, simplified real format.

What comes next

The new generation of JadeIH covers recipient immunohaematology. More will be added to this scope, with no date announced.

  1. What JadeIH covers
    • Blood group requests
    • Ab screen
    • Direct Coombs
    • Blood group identity
    • Blood group card
    • Results to the LIS
  2. Next addition

    Transfusion requests

  3. Also planned

    Blood units, issuing, haemovigilance.

    No promised date or order.

Frequently asked questions

What laboratories ask us

Can we migrate our old JadeIH?

Yes. JadeIH reads the old JadeIH database directly, with a read-only account. Other software goes through CSV, Excel, JSON or HL7 files. The dry run is mandatory, the import can be replayed and undone, and every identity discrepancy goes to a biologist. The duration depends on your volumes: let's talk.

How do we validate JadeIH in our laboratory?

Validation remains your laboratory's. JadeIH is designed for ISO 15189 accredited laboratories and gives you the means: replayable test cases, rules simulator, HL7 message log, a trace of every rule trigger, settings history with their author. The documents you will need: let's talk.

What about transfusion requests, blood units and issuing?

Next addition: transfusion requests. Blood units, issuing and haemovigilance are also planned, with no date announced.

Which analysers are supported?

IH analysers connect through the middleware, which translates the codes; each model has its profile. The list is available on request.

Our analysers already go through Instrument Manager.

The middleware is compatible with Instrument Manager (Data Innovations). Let's talk about your configuration.

What happens if the LIS or an analyser goes down?

The middleware holds the messages during the outage and delivers them when it is back. Without the LIS, a request is created by hand and a 2nd person signs it off before any validation. A disabled analyser hands over to the next one; otherwise, entry becomes manual again, with a reason.

Do we need Jade as our LIS?

No. The interface goes through the middleware's HL7 profiles. The “In Jade” features (patient comparison, receipt verification, merges) only exist with Jade.

How do our teams get up to speed with JadeIH?

The gestures stay those of the old JadeIH: same entry syntax, same shortcuts, barcode scanner from any screen. Screens carry their own help. Training, on site or remote: let's talk.

For which countries?

Switzerland, based on the SVTM / Blood Transfusion SRC recommendations (version 14, 2026), and France, based on the decree of 15 May 2018, guidelines under review. The Belgian guidelines are in preparation. The interface exists in French and English.

See JadeIH on your own cases

A demo on the fictitious laboratory, then on your cases: your rules, your analysers, your data migration.

Lysoft SA, based in Plan-les-Ouates (Geneva), has been developing laboratory information systems since 1989.

Tell us about your laboratory and your LIS in the form. Your data is used only to answer your request.

The patients, numbers and analysers on this page are fictitious.

Request a Demo

Fill out the form below and our team will contact you within 24 hours to schedule your personalized demonstration.