Belgian Pregnancy Status Implementation Guide
1.0.0 - STU
BE
This page is part of the eHealth FHIR Pregnancy Status Profiles for Belgium (v1.0.0: Release) based on FHIR (HL7® FHIR® Standard) R4. This is the current published version. For a full list of available versions, see the Directory of published versions
Pregnancy and birth involve a broad set of administrative and clinical data, contributed and consumed by many initiatives and stakeholders over the full course of care — from the first confirmation of a pregnancy, through follow-up and screening, to birth and post-natal follow-up of mother and child (eBirth, RIZIV, Born in Belgium, Zoet Zwanger, PALS, the moeder/kind boekje, …).
The pregnancy status is a minimal, well-bounded data set, sufficient to support business cases such as entitlement to the Groeipakket, the on-line pregnancy certificate, and consultation by the woman and her caregivers. The rest of the broader pregnancy care set is out of scope of this iteration, but the information model is designed so that it can be added later without redesign. See the home page for the full scope and roadmap.
Pregnancy status here is for administrative sharing, not clinical care. (For
clinical care, the pregnancy is more commonly represented as a Condition.)
The model is a single pregnancy status Observation (a BeClinicalObservation,
LOINC 82810-3) that contains its detail observations and groups them via
Observation.hasMember — the
Pregnancy Status profile. The
contained details — estimated date of delivery (11778-8, dateTime), expected
number of children (11878-6, integer) and an optional end-of-pregnancy date
(289251005, dateTime) — are constrained directly on the container by slicing
contained by type and then by code, each with a fixed code and value type.
The container is the conformance target. The contained observations are distinguished by code, not by a profile per observation type.
Demonstrative, temporary profiles.
The per-code Observation profiles in this IG are illustrative only and are expected to be removed in the federal guidance.
A federal layer cannot publish one profile per type of observation: there are far too many
.codevalues, spread across many clinical domains, to maintain a StructureDefinition for each. Instead it standardises the genericBeClinicalObservationand lets the container plus the codes carry the specific meaning.
Within the pregnancy-status scope, the IG distinguishes two related but different concepts, each represented by a different FHIR resource type:
Observation, aligned with the IPS Pregnancy status observation.Condition. The detail observations (due date,
expected number of children, etc.) reference it — it is the observations that
point to the Condition, not the reverse.In both cases the supporting detail elements (expected date of delivery, expected
number of children, actual end date) are carried as separate
BeClinicalObservations. The shared, scope-independent information model is the
logical model
PregnancyStatusDataSet.
The detail observations are the robust core of the data — they carry the actual
values and stay unchanged regardless of how the pregnancy as a whole is represented.
The PregnancyStatus Observation and the Pregnancy Condition are then two
framings over that core:
PregnancyStatus Observation asserts the coded finding (pregnant / not
pregnant) and groups the details — the IPS-aligned summary;Pregnancy Condition represents the clinical episode (onset, status,
end).Whichever framing you do not currently use is an optional add-on that can be added later, derived from the same core, without touching the details:
Condition (the current scope), the PregnancyStatus Observation is
the anchor — it is the only resource asserting the finding and grouping the
details. The Condition is the future add-on.Condition exists, it already asserts “pregnant” and anchors the
details (via focus); the PregnancyStatus Observation then becomes an optional,
IPS-compatible summary projected over the same data.The diagram below shows these two framings and how the design is incremental: a
project can start with the PregnancyStatus Observation and its member
observations, and add the Pregnancy Condition later with no redesign of the
existing observations.
The two links are independent and serve different purposes:
Observation.hasMember groups the member observations under the PregnancyStatus
Observation (parent → children). This is internal to the observations and has
nothing to do with the Condition.Observation.focus is how an observation is tied to the pregnancy Condition.
Both the PregnancyStatus Observation and its member observations point to
the Condition this way, so any observation that is part of a pregnancy is known.
This focus can be a literal reference to a Condition resource or a
logical reference (Reference.identifier) — that is, a unique identifier of
the pregnancy.Note. While there is no
BePregnancyConditionresource to point at, thefocuscan already carry a logical reference — a unique pregnancy identifier. That identifier groups all the observations that belong to the same pregnancy, without anyConditionresource being exchanged. Later, when aConditionresource does exist, the very samefocussimply resolves to it (a literal reference) — the observations themselves do not need to change.
Because the pregnancy status Observation contains its detail observations,
fetching the status returns the details inline — no _include is needed for
them.
Given a patient — fetch the pregnancy status (the EDD, number of children and end date come contained within it):
GET [base]/Observation?subject=Patient/123&code=http://loinc.org|82810-3
To list the pregnancy observations about a patient from the other direction:
GET [base]/Patient/123?_revinclude=Observation:subject
Note. When the pregnancy is also represented as a
Condition(clinical-care scope), the observations link to it viaObservation.focus— a literal reference or a logical reference (a unique pregnancy identifier). See the framing discussion above. The recorder is carried in thebe-ext-recorderextension.