NHS North West Genomics
2.2.0 - ci-build
NHS North West Genomics
2.2.0 - ci-build
NHS North West Genomics - Local Development build (v2.2.0) built by the FHIR (HL7® FHIR® Standard) Build Tools. See the Directory of published versions
A supplement to iGene Orders and Reports (Alder Hey, MFT, Liverpool), exploring what it would take for Alder Hey, MFT or Liverpool Women's to place a Whole Genome Sequencing order directly with NW Genomics - using the national GMS WGS Rare Disease/GMS WGS Cancer order-entry shape, rather than the generic Genomic Test Order common core iGene Orders and Reports otherwise describes.
LAB-35 sub-order an external
Requesting Genomic Laboratory sends onward to NW Genomics once a national GMS
referral has already been accepted; this page is about Alder Hey/MFT/Liverpool
ordering directly with NW Genomics, the same as every other test iGene Orders and Reports covers, just using
the national WGS forms' shape rather than the generic common core. Nothing here
touches the LAB-35 manifest.
Three Questionnaires already model the Proband/family-member (RelatedPerson) side of
this proposed pathway:
NOS/RelatedIndividual group (Name/Relationship/Sex/
DOB/NHS Number/Hospital Number, plus a Role of Consultand or Proband) every
other Questionnaire below reuses. Used alongside Ask At Order Entry
Questions Common (which
carries the Consent and High Infection Risk items) and the common core -
together these represent a singular order (one Patient, one
ServiceRequest), the same as WGS Local Test
Order and GMS WGS
Cancer below.NOS/Proband (Role fixed to Proband) on its Family Member
ordering pathway - the closest existing precedent for a WGS order naming a second
individual this way.FamilyMembers group carries the same ServiceRequest.supportingInfo ->
RelatedPerson shape, inline rather than via the shared group.GMS WGS Cancer does not currently
use RelatedPerson at all - a cancer WGS order is germline+tumour on one patient, not
a family test - so it isn't part of this particular data model, despite being the
other national WGS form.
Neither Ask At Order Entry Questions
Common nor WGS Test
Additional Ask At Order Entry
Questions is tied
to any one paper form - their Consent/High Infection Risk/RelatedPerson
items were originally part of the Genomic Test
Order common core itself, extracted
into these two Questionnaires for order/test types that don't have their own
dedicated Ask At Order Entry Questionnaire. All three WGS-specific
Questionnaires below do have their own dedicated Questionnaire, so
neither is used alongside them in practice - they're included here only to
show where these fields originally came from.
For the three real paper forms, checked directly against the source PDFs rather than assumed from what each Questionnaire's FSH happens to declare:
| Field | WGS Local Test Order | GMS WGS Rare Disease | GMS WGS Cancer |
|---|---|---|---|
| Order Placer Number | Not on the paper form | Not on the paper form | Not on the paper form |
| Order Group Number | Not on the paper form | Not on the paper form - see the G Number correction below It is believed an Order Filler Group Number is created on submission of the form, but no field on any of the three Questionnaires actually captures/returns it |
Not on the paper form |
| Medical Record Number | "Hospital number" - present | "Hospital number" - present | "Hospital number" - present |
| Account Number / Hospital Spell | Not on the paper form | Not on the paper form | Not on the paper form |
| RelatedPerson | "Family Member (please provide below the Name & DoB of the Proband)" - present, modelled as NOS/Proband |
"Family members to be tested" table - present, modelled as FamilyMembers |
Not on the paper form |
| Consent | "Consent Statement" note ("A complete Patient Choice form must be received by the laboratory before WGS can be initiated") - present, modelled as SNM/74996004-patient-choice-form |
Only "Record of Discussion" attached/to-follow tick - present, modelled as its own item; no consent-for-testing/DNA-storage question like Genomic General's | Same Record of Discussion tick as Rare Disease |
| High Risk Sample | "High Infection Risk? Yes/No" - present, modelled as SNM/281269004 + NOS/InfectionRiskDetails |
Not on the paper form | Not on the paper form |
Order Placer Number and Account Number are both still structurally present on all three Questionnaires (inherited, unchanged, from the Genomic Test Order common core - none of the three overrides or re-declares them) - the table above is about what the paper form itself actually asks for, which is narrower than what the FHIR Questionnaire structurally allows.
G Number (Pedigree Number) was mislabelled "Order Group Number" and has since been
corrected - but a genuine Order Group Number still doesn't exist on any of these
Questionnaires. The field used to live on Genomic Test
Order itself as pedigreeNumber ("G Number
(Pedigree Number) - Order Group Number", mapped to
Patient.identifier:PedigreeNumber) - despite that label, it was never an
order/requisition-linking field; it's a family/pedigree-level concept. NHS England's
own genomics-pedigree-number NamingSystem (https://fhir.nhs.uk/Id/genomics-pedigree-number)
describes it as "a patient's genetic/pedigree number which links their family," and
their own FHIR Genomics Implementation Guide has since moved its equivalent mapping to
a Group resource entirely. It has now been moved to Genomic General Ask At Order
Entry, remapped to
Observation.valueString, coded $loinc#74027-4 "Family pedigree identifier", and
relabelled without the misleading "Order Group Number" suffix. This exact confusion
was also identified once before in dWGS: dWGSAskAtOrderEntry's own
referral_id item (ServiceRequest.requisition) carries a design note distinguishing
the two. ServiceRequest.requisition remains the genuine order-group-linking
mechanism (the HL7v2 ORC-4 Placer Group Number equivalent) - but it still only exists
today as a bespoke field on the dWGS manifest, not on Genomic Test Order, Genomic
General, or any of the three WGS-specific Questionnaires compared here.
The comparison above starts from each WGS-specific Questionnaire and checks the paper form. Going the other way - starting from each paper form's own distinctive fields and checking whether Genomic Test Order plus Ask At Order Entry Questions Common and WGS Test Additional Ask At Order Entry Questions (the generic combination iGene Orders and Reports actually uses today) already has an equivalent - is what would decide whether Alder Hey/MFT/Liverpool could order WGS through the existing generic path at all, rather than needing this proposed WGS-specific one. Of the fields on all three paper forms, only a handful already have a genuine or partial match:
| Paper form field | Which form(s) | Generic combo equivalent |
|---|---|---|
| Hospital Number (MRN) | All three | LN/76435-7 - exact match |
| Proband / Family Member(s) named | WGS Local, GMS Rare Disease | NOS/RelatedIndividual (repeats = true, Consultand/Proband role) - matches the shape, but has no nested Specimen sub-group per repetition the way GMS Rare Disease's FamilyMembers does |
| High Infection Risk? | WGS Local | SNM/281269004 - exact match, and now used by WGSLocalTestOrderAskAtOrderEntry itself |
| Record of Discussion attached/to follow | GMS Rare Disease, GMS Cancer | NOS/RODToFollow (inside the Consent group) - exact match |
| Consent Statement (references a separate Patient Choice form) | WGS Local | Consent group's "Has consent been obtained for tests (Y/N)" - related concept, not the same mechanism |
| Reason for urgency | GMS Rare Disease | Priority (LN/82768-3) - a coded urgency level, not free-text reason |
| Requesting organisation / GMS-GLH laboratory (two org fields) | GMS Rare Disease, GMS Cancer | HL7/ORC-21 "Referring Organisation ODS Code / Ordering Facility" - one field, not the same two-organisation split |
| Test Directory Clinical Indication & code | GMS Rare Disease, GMS Cancer | HL7/OBR-4-r/-h/-c Test Code branches - present for Rare and Inherited Disease/Haemoglobinopathy/Cancer, but none of the three covers WGS specifically |
| Additional clinical information | GMS Cancer | HL7/NTE-1 "Relevant clinical information and family history" - close match |
| Histopathology/SIHMDS Lab ID | GMS Cancer | LN/80398-1-ODS "Pathology Laboratory Hospital/Trust ID" - adjacent concept (identifies the lab), not the same specific field |
| Life status (Alive/Deceased) | GMS Rare Disease | LN/81954-0 "Date of death" - implies deceased status, doesn't capture "Alive" explicitly |
Everything else - WGS test type itself, Family test type (Singleton/Trio/Other), Reason NHS Number not available, Reason for diagnostic test (patient management/reproductive/predictive tick boxes), Additional panel(s), Proband's age at onset, specific rare disease suspected/confirmed, HPO Terms, Main contact, Presentation status, Tumour presentation type/topography/morphology, Haemato-oncology liquid tumour type, % malignant nuclei/blasts, Nucleated cell count, and Neoplastic cell content - has no equivalent at all in Genomic Test Order or Genomic General Ask At Order Entry. HPO Terms in particular is a mandatory field on GMS WGS Rare Disease with nothing resembling it anywhere in the generic combo. This is a fairly direct answer to what motivates this whole proposed use case: the generic order path iGene Orders and Reports uses today could not capture a WGS order's own clinically-necessary detail without the WGS-specific Questionnaires' shape.
Histopathology/SIHMDS Lab ID hints that a cancer WGS order may originate as a reflex
from an existing pathology order, which should itself have its own Order Placer
Number this order doesn't carry forward. GMS WGS Cancer's Histopathology Lab ID and
SIHMDS Lab ID fields (see the Reverse Direction table above) only make sense if a
pathology sample/report already exists before the genomic order is raised - the same
pathology-to-genomics reflex pattern already modelled as its own use case in
Cheshire and Merseyside (Pathology to Genomics
Reflex), where a pathology LAB-1/LAB-3 can
reflex on to a genomic order (LAB-35/LAB-36, or a separate LAB-1). If that's what
GMS WGS Cancer's Lab ID fields are really referencing, the original pathology order
should itself have had its own Order Placer Number (ServiceRequest.identifier:OrderIdentifier,
per Genomic Test Order - Diagnostic
Workflow) - but neither Lab
ID field on the GMS WGS Cancer paper form is modelled as that Order Placer Number, or
as any other structured reference back to the originating pathology order; they're
both free text (Specimen.accessionIdentifier.assigner.identifier.value), which
identifies the pathology lab, not the pathology order. Haemoglobinopathy Genetic
Ask At Order Entry is the
closest existing precedent for a genetic order form carrying content from an original
report - its own LaboratoryResults group carries actual FBC/haemoglobinopathy screen
values (Hb, RBC, HbA2%, HbF%, etc.) forward from a prior report - but even that
precedent carries the prior report's values, not a link back to the prior report's
own Order Placer Number either. So this remains a genuinely open question, not one
this IG has already answered elsewhere.
GMS WGS Rare Disease's one-form-names-several-people shape looks like a data-entry
convenience (from many EPR and LIMS perspective), not a genuine single order - the individual orders it implies must be
linked by Order Group Number for electronic exchange. The paper form itself has no
Order Group Number field at all (see Field Comparison above), which is consistent with
it being designed as a single physical document a clinician fills in once per family,
not as something that maps directly onto one electronic order. In practice, each named
person - proband plus every family member - needs their own Patient/Specimen/
ServiceRequest, so electronic exchange of this form's answers requires decomposing
it into separate singular orders, one per person, each shaped like the other two
national/local WGS Questionnaires in the comparison above. Those separate orders
should/must then be tied back together by a genuine Order Group Number
(ServiceRequest.requisition, the same mechanism dWGS's own referral_id
already uses) - not the common core's own pedigreeNumber field, which despite
its "Order Group Number" label is actually a Patient-level pedigree identifier, not
a requisition-linking one (see Field Comparison above). Since neither Genomic Test
Order nor any of the three Questionnaires
compared here has a genuine ServiceRequest.requisition-backed field today, this
decomposition can't currently be built without either adding one, or reusing dWGS's
own referral_id pattern even outside a distributed sub-order context - rather than
the family relationship being reconstructable only via the composite submission. WGS Local Test
Order's Family Member pathway
already models one instance of exactly this decomposition.
The HPO Terms guide list and the Test Directory Clinical Indication guide list
overlap semantically, but aren't cross-checked. GMS WGS Rare
Disease asks for both HPO Terms (bound to
GMS WGS Guide HPO Terms, 38 phenotypes
transcribed from the form's own guide list) and a Test Directory Clinical Indication
(bound to GMS WGS Guide Test Codes, the 37
$GTD R* codes whose display text names WGS) as two separate items, answering two
different questions - HPO Terms is the observed phenotype on this patient, Test
Directory Clinical Indication is which coded test is being ordered. Comparing the two
guide lists directly shows 10 of the 38 guide HPO phenotypes correspond to a named
clinical indication in the Test Codes guide list:
| HPO guide term | Matching R* WGS test code |
|---|---|
| Cardiomyopathy / Hypertrophic cardiomyopathy / Dilated cardiomyopathy | R135.2 Paediatric or syndromic cardiomyopathy |
| Cataract | R31.3 Bilateral congenital or childhood onset cataracts |
| Ataxia / Cerebellar atrophy / Cerebellar hypoplasia | R54.3/R55.4 Hereditary ataxia (adult/childhood onset) |
| Dystonia / Chorea | R56.3/R57.5 Adult/childhood onset dystonia, chorea or related movement disorder |
| Spasticity | R60.3/R61.4 Adult/childhood onset hereditary spastic paraplegia |
| Microcephaly | R88.3 Severe microcephaly |
| Generalized hypotonia | R69.5 Hypotonic infant |
| Peripheral neuropathy | R78.4 Hereditary neuropathy or pain disorder |
| Abnormality of metabolism/homeostasis | R98.2 Likely inborn error of metabolism |
| Skeletal dysplasia | R104.3 Skeletal dysplasia |
| Multiple renal cysts / Hepatic cysts | R193.4 Cystic renal disease |
No example exists yet in this IG for the proposed direct-order pathway itself, but ten
examples from NHS England's own GOMS FHIR Implementation
Guide - built
from the same underlying Test Order Form Bundle shape our national GMS WGS forms
follow - are already vendored into this IG's own input/resources/ and published
here, each cross-referenced back to its NHS England source in sushi-config.yaml.
Confusingly, nine of the ten are titled Non-WGS, even though they share the same
Bundle structure as the one WGS-titled example - "Non-WGS" here means "this order
form, used for a test that isn't WGS", not "unrelated to WGS":
The HL7 v2 OML^O21 messages above (built by the same
nw-gmsa/Testing notebook series as this IG's own
dWGS examples) are the HL7 v2 counterpart of the FHIR Bundle in
the same row - useful for comparing the same order in both formats. Two further HL7 v2
fixtures in that same
Output/V2/O21 folder -
Bundle-NonWGSScenario3-FetusAsProband-Example-Mother.txt
and
Bundle-NonWGSScenario4-ProbandWithMultipleFetus-Example-Mother.txt -
don't have a corresponding FHIR Bundle example vendored in this IG; only the fetus
participants were carried over.