4. Page 4 • HL7 New Zealand
Architectures
• FHIR makes no assumptions about the architectural design
of systems
• You can use it for
– Light or heavy clients
– Central server or peer-to-peer sharing
– Push or pull
– Query or publish/subscribe
– Loosely coupled or tightly coupled environments
With history tracking (versions) or without
4
5. Page 5 • HL7 New Zealand
Some possibilities
5
FHIR
Broker
v3
v2
PHR
FHIR
App
Comm.
Interface
DB
FHIR
6. Page 6 • HL7 New Zealand
Repository model
Vendor Neutral Repository
FHIR
HIS GP PACS SystemX Gateway
FHIR FHIRFHIRFHIR
7. Page 7 • HL7 New Zealand
Beyond exchange
7
v3
v2
Other
X12
Broker
v3
v2
FHIR
Repository
Decision
Support
FHIR
FHIR
Other
8. Page 8 • HL7 New Zealand
Bottom Line
• FHIR is a set of tools
– Defined resources
– Extensibility mechanism
– Versioning
– Multi-paradigm
– Set of standard interfaces
– Support for complex services
• Primary purpose is interoperable data exchange
• However, it can be leveraged in many ways
– Many we haven’t even thought of yet . . .
8
12. Page 12 • HL7 New Zealand
REST
• Simple, out-of-the-box interoperability
• Leverage HTTP: GET, POST, etc.
• Pre-defined operations
– Create, Read, Update, Delete
– Also: History, Read Version, Search, Updates, Validate, Conformance &
Batch (Transaction)
• HTTP Headers for various options
– Like resource format or location
• Response codes to indicate outcome
– OperationOutcome Resource
• Works best in environments where control resides on client
side and trust relationship exists
12
Rest
13. Page 13 • HL7 New Zealand
Creating & Updating
• POST vs PUT
• Headers
– Request
• Content-type
• Location (for version aware)
– Response
• Content-type
• Others...
• Status messages
– Standard HTTP
• OperationOutcome
• Version aware updates
14. Page 14 • HL7 New Zealand
Querying by REST
• Base queries
– Simple GET against a resource type
– GET [base]/Patient?name=eve
• Defined rules for query parameters
• More complex
– _include
– Tags
• Queries across reference
– Eg conditions for a patient
– GET [base]/Condition?subject=84238862
– Chained queries
• Compartments
– Syntactic sugar
– GET [base]/Patient/84238862/Condition
15. Page 15 • HL7 New Zealand
Synchronizing
• History of all resources on server
– GET [base]/fhir/_history
• History of all patient resources on server
– GET [base]/Patient/_history
• History of specific patient on server
– GET [base]/Patient/1/_history
• A history of all changes: inserts, updates and
deletions, ordered by newest first
• Limit with _since and _count
16. Page 16 • HL7 New Zealand
What’s a ‘deleted’ Resource?
• Delete using DELETE verb
• Trying read operations will return in a 410 (Gone) result in
stead of 404 (Not Found)
• The resource will not be returned by the search operation.
• You can “undelete” by doing an update (PUT) with fresh
content
• Just a “marker” in a resource’s history
19. Page 19 • HL7 New Zealand
Batch
• POST a bundle to server root
• Processed as a transaction
– All succeed, or all fail
• All operations supported
• Different to /message endpoint
– /message allows custom server processing
– Batch is a sequence processed as a transaction
21. Page 21 • HL7 New Zealand
Test Servers
• Servers
– At least 4 to choose from
– http://spark.furore.com/fhir/Patient/nz100
– http://fhir.healthintersections.com.au/open/Patient/100
– fhir.orionhealth.com/blaze/fhir/Patient
– https://his-medicomp-gateway.orionhealth.com/blaze/fhir/metadata
• Tools
– XML Editor (Oxygen)
– REST Client (POSTMan)
21
Script
• Search patient on name
• GET single patient from list
• XML & JSON
• Update and PUT back
• repeat REST Search
• Show error
22. Page 22 • HL7 New Zealand
When to use REST?
• Want low coupling between systems
– In theory, very little up-front negotiation required
• Small, light-weight exchanges
• Focus is CRUD operations
– Also for publish/subscribe
• Client-driven client-server orchestration
• Server endpoint has fixed location
• Well-suited for Mobile, PHR, Registries
22
Rest
23. Page 23 • HL7 New Zealand
When to avoid REST?
• Complex or server-driven orchestration
– Order of operations matters (e.g. complex decision support)
• Unit of work != resource
– “Transaction” may be an option
• Lack of trust in the client for audit, etc.
23
Rest
25. Page 25 • HL7 New Zealand
Documents
• Similar to CDA
• Collection of resources bound together
– Root is a “Composition” resource
– Just like CDA header
• Sent as an ATOM feed
– Usually stored as binary blob
– Can be separate resource with Composition as ‘entry’ point
• One context
– Not propagated
• Can be signed, authenticated, etc.
• Working on conversion between FHIR Doc & CDA
25
Documents
27. Page 27 • HL7 New Zealand
The Composition resource
27
28. Page 28 • HL7 New Zealand
Document Management
• The DocumentReference resource
– Enables ‘XDS-like’ services
• The ‘registry’ is just a FHIR server
– Same query capabilities
• Point to document
– Any document (FHIR Doc, CDA, PDF etc).
30. Page 31 • HL7 New Zealand
When to use Documents?
• No different to CDA
• Focus is on persistence
• No workflow involved
– other than post/retrieve document
• Need tight rules over authenticated content
• Want to communicate multiple resources with control over
how data is presented
• Data spans multiple resources
31
Documents
31. Page 32 • HL7 New Zealand
When to avoid Documents?
• Need for workflow
– Request/response, decision support
• Data is dynamic
– I.e. want view of data now, not at time of authorship
– Multiple contributors over time
• Resources need to be accessed/manipulated independently
32
Documents
33. Page 34 • HL7 New Zealand
Messages
• Similar to v2 and v3 messaging
• Also a collection of resources as an ATOM feed
• Allows request/response behavior with bundles for both
request and response
– /message endpoint on REST server
• Event-driven
– E.g. Send lab order, get back result
• Can be asynchronous
34
Messages
35. Page 36 • HL7 New Zealand
MessageHeader Resource
36
Resource used in both request & response
36. Page 37 • HL7 New Zealand
FHIR messages
• No storage of the “Message” resource implied. Might be used
in a router, converted to v2, etc.
• The server can process them based on the event code and
return the response as another message (again a bundle).
37
37. Page 38 • HL7 New Zealand
Communicating messages
• Message Exchange Pattern
• Synchronous in standard
• Async supported
• Reliability
• Transport
• REST supported but, you can send FHIR Messages over mail, MLLP,
TCP, IPoAC (or Healthlink!)
• FHIR defines an HTTP end-point for messages
• http://server.org/fhir/Mailbox
• Response comes back in the acknowledgment payload
38
38. Page 39 • HL7 New Zealand
Revisiting the Immunization Forecast
TCH ForecasterFHIR ServerFHIR Client
TCH Forecaster
with FHIR Message InterfaceFHIR Client
39. Page 40 • HL7 New Zealand
When to use Messaging?
• Request/response workflow
• Need to drive behaviors more complex than CRUD on a
single resource
– E.g. merge, complex queries
• Need for asynchronous communications
• Need to communicate information about many resources
but want to minimize exchanges
• No “identity” for many resources
40
Messages
40. Page 41 • HL7 New Zealand
When to avoid Messaging?
• Precise control required on how data gets
persisted/displayed to humans
• Need for lightweight communications
• Want to avoid pre-negotiations on behavior
41
Messages
42. Page 43 • HL7 New Zealand
Service Oriented Architecture (SOA)
• Do whatever you like
– (based on SOA principles)
– Ultra complex workflows
– Ultra simple workflows
– Individual resources or collections (in Atom or other formats)
– Use HTTP or use something else
– Only constraint is that you’re passing around FHIR resources in some
shape or manner
43
Services
43. Page 44 • HL7 New Zealand
When to use Services?
• All previous paradigms are a type of services interface
• Use a custom service when capabilities of other paradigms
don’t fit requirement
– Operations other than CRUD on a resource (e.g. decision support)
– Workflow more complex than simple request/response
– Need to mix document persistence with
behavior
44
Services
44. Page 45 • HL7 New Zealand
When not to use services?
• When something else will do the job as well
– I.e. Don’t define a custom service for something that already naturally
is handled by REST, messaging, etc.
45
Services
45. Page 46 • HL7 New Zealand
Paradigm guidance
• No absolutes
– Consider a “when to avoid” as a note of caution
– Capabilities/architecture of legacy will often drive approach,
particularly initially
• E.g. If v2 back end, messaging
– Architectures will be driven by legacy requirements, architectural
preferences, enterprise architecture commitments, etc.
46
46. Page 47 • HL7 New Zealand
Combining paradigms
• No requirement for a system to only support one paradigm
– E.g. hospital may be primarily messaging, but use documents for
discharge summaries and reports and expose registries and
appointments via REST with a few custom services for decision
support or specialized workflow
• Data (generally) shared easily across paradigm boundaries
• Caveats:
– If updates come in via document, message or service, RESTful version
id still needs to increment
– Documents should typically be persisted whole, not reconstituted
from parts
• Ensures signature validity
– Legacy messaging systems may not provide the metadata to easily
expose or manipulate discrete resources via REST
47
50. Page 51 • HL7 New Zealand
Specify format using REST
GET /fhir/Patient/1?_format=json HTTP/1.1
HTTP/1.1 200 OK
Content-Type: application/json+fhir;charset=utf-8
Content-Length: 787
GET /fhir/Patient/1 HTTP/1.1
Accept: application/json+fhir
HTTP/1.1 200 OK
Content-Type: application/json+fhir;charset=utf-8
Content-Length: 787
Same for Create/Update
52. Page 54 • HL7 New Zealand
Identity
• 2 different ‘sorts’ of identity
– ID identifies a resource on a server
• Is metadata
• Will change between servers
– Identifier
• Business identifier
• Is an element in the resource
53. Page 55 • HL7 New Zealand
A Resource’s ID
• In fact: an URL
– http://server.org/fhir/Patient/1
endpoint
resource type
identifier
Note: This URL resolves to the current version of a resource
It’s also specific to a server
54. Page 56 • HL7 New Zealand
“Business” identifiers
56. Page 59 • HL7 New Zealand
Tags
• Contain metadata about resources
– Beyond ID, version & date
• Used for different purposes:
– Compliance to Profile
– Security – e.g. sensitivity of resource
– Indicate Document/Message
– User defined
• Transported in different ways:
– REST – as an HTTP header
– In a bundle (document, message, transaction) - in the feed.category
element
57. Page 60 • HL7 New Zealand
Versioning
• Most recent version
– http://server.org/fhir/Patient/1
– Returns single resource
• All versions
– http://server.org/fhir/Patient/1/_history
– Returns bundle of versions
• Specific version
– http://server.org/fhir/Patient/1/_history/_1
– Returns single resource
• Version support is optional
• Works at instance and type level
58. Page 61 • HL7 New Zealand
Resource Reference
Example: part of DiagnosticReport
59. Page 62 • HL7 New Zealand
Coded values: the ValueSet resource
• Most coded elements have a set of possible values
– Large or small
– Static
• (change slowly)
• Can use ValueSet resource to define set of possibilities
– A “true” FHIR ValueSet resource (may be version specific)
• Can contain:
– References to a terminology (like SNOMED)
– A list of specific codes
• Examples
– International examples
• All the HL7 version 2 ‘tables’
– NZ Examples
• Iwi & Hapu
• DHB’s
62
60. Page 63 • HL7 New Zealand
Example of ValueSet
<?xml version="1.0" encoding="UTF-8"?>
<ValueSet xmlns="http://hl7.org/fhir">
<text>
<status value="additional"/>
<div xmlns="http://www.w3.org/1999/xhtml">The value set for confidentiality codes in the elbonian docume
project. The codes are defined directly by the valueset, rather than references to an external terminology</d
</text>
<name value="ConfidentialityOptions"/>
<description value="ConfidentialityOptions"/>
<status value="draft"/>
<!-- Define the options in the resource rather than an external terminology-->
<define>
<system value="http://fhir.moh.elbonia.com/confidentiality"/>
<concept>
<code value='N'/>
<display value="Normal"/>
<definition value="That the document is of normal confidentiality "/>
</concept>
<concept>
<code value='H'/>
<display value="High"/>
<definition value=”That the document is Highly confidential - eg Mental health "/>
</concept>
</define>
</ValueSet>
61. Page 64 • HL7 New Zealand
Contained resources
• When the contained
resource has no separate
identity
• Has quite a few limitations
• Use sparingly
– Bundle always better if
possible
<Composition
xmlns="http://hl7.org/fhir">
<extension>...</extension>
<text>...</text>
<contained>
<Organization id="org1">
<!– other data -->
</Organization>
</contained>
<information>
<!-- other attributes -->
<custodian>
<reference value="#org1" />
</custodian>
<!-- other attributes -->
<information>
</Document>
63. Page 66 • HL7 New Zealand
Review
• Atom as the bundling mechanism
– Query results
– Document
– Message
• Request & Response
– Batch/Transactions
• XML or JSON
– Conversion routines available
– (Can’t use ordinary ones)
• Use tags to identify Documents & Messages
– First resource significant
65. Page 68 • HL7 New Zealand
Resource Entry
Version-specific id
Resource id
Last modified
Resource content
Human-readable form,
just like Resource.text
Tags
66. Page 69 • HL7 New Zealand
Bundle as a serialized Object Graph
68. Page 71 • HL7 New Zealand
The need for Profiles
• Many different contexts in healthcare, but a single set of
Resources
– A desire to keep the resources manageable
– Most elements in base resources optional
• Need to be able to:
– Describe restrictions based on use and context
– Support elements not in the core resource
• Allow for these usage statements to:
– Authored in a structured manner
– Published in a repository
– Discoverable at Design and Run-time
– Used as the basis for validation, code, report and UI generation.
71
69. Page 72 • HL7 New Zealand
Profiling a resource. For example...
72
Require that the identifier uses the
NHI – and is required
Limit names to just 1 (instead of 0..*)
Limit maritalStatus to another set of
codes that extends the one from
HL7 international
Add an extension to support “Iwi”
Note: hardly any
mandatory
elements in the
70. Page 73 • HL7 New Zealand
Capabilities of Profiles
• Descriptive
– Information about the Profile (what, why, who)
• Extending a resource
– Adding new elements
– Same capabilities/datatypes as existing ones
• Constraining a resource
– Specifying datatypes for an element
– Modifying datatypes (eg translations)
– Defining terminology bindings
– Specifying multiplicity (eg make an element required)
– Further refining repeating values (‘Slicing’)
• Example of Composition.section (Document)
– Cannot specify default values for resource instances
• Everything explicit in an instance
71. Page 74 • HL7 New Zealand
Structured & published
• A Profile is just a “normal” Resource
• Any FHIR server could serve Profiles (just like Patients, Observations,
etc…)
• So, any FHIR server is a profile repository!
• A profile is simply referred to by its URI:
– Like any other resource
• e.g. http://hl7.org/fhir/Profile/iso-21090
– Extensions in an ‘instance’ of a resource refer to the profile that defined
them using the URI
– Need to think about re-usability when designing profiles
74
72. Page 75 • HL7 New Zealand
Who publishes a Profile?
http://www.hl7.org/fhir/Profile/iso
-21090
http://www.hl7org.nz/fhir/Profile/
patient
http://www.health4all.org/fhir/Pro
file/
73. Page 76 • HL7 New Zealand
Using profiles
• If we are a server then a client can read our conformance
resource, which will list the profiles we support, and provide a link
to them.
– Conformance also has other functions like specifying supported resources,
operations on those resources & interoperability paradigms
• If we are a client, we can tell a server that we conform to a profile
using a tag (the server will probably still check).
– We could also ask a server just to validate
– Could set up ‘distributed’ validation with (eg) National Validation servers
used by local/regional servers
• If we receive a resource with an extension we don’t recognize, it
will point to the profile in which it is described so we can decide
what to do with it.
– Note that a single resource can reference different profiles
76
74. Page 77 • HL7 New Zealand
What’s in a profile, and its relationships?
77
Descriptive
Identifier
Name, Version
Publisher
Description, Code
Status
Date (of publication)
Resource and
Datatype
Constraints
Extension
Definition
Conformance
(server)
Resource
(via Tag)
ValueSet
ValueSet
Extension in
Resource
Tag
75. Page 78 • HL7 New Zealand
Scale / design
• Make a profile with just extensions
• Make a profile with just 1 resource + extensions on that
resource
• Make a profile with all (say national) resources + extensions.
78
76. Page 79 • HL7 New Zealand
Profile Descriptive information
• To describe the purpose of the Profile
• To help with location
• Details
– Identifier: Universally unique identifier, assigned by author value=“urn:hl7.org:extensions:iso-21090“
or “2.16.840.1.113883.10.20.2.1”
– Version: Version of the profile, manually maintained by author. Suggested format: a timestamp (e.g.
2013-01-01T12:34:45)
– Name: Free natural text name (e.g. “Extensions to support Orion Medications”)
– Publisher: organization or individual responsible for publishing. Should be populated (e.g. “FHIR
Project Team“)
– Telecom: one or more contact points of the publisher (telephone, email, website etc)
– Description: longer description of the contents of the profile
– Code: one or more coded descriptions to help with finding the profile
– Status (fixed choice of draft, active, retired) + experimental Y/N
– Requirements: scope & usage (the “need” or “why” of the profile)
– Date (of publication), e.g. “2013-07-07"
79
77. Page 80 • HL7 New Zealand
Constraining Structures (elements)
• Defining terminology bindings
– ValueSet is useful
• Specify element datatypes
• Further refining repeating values (‘Slicing’)
• Specifying multiplicity
From the spec
Possibilities
78. Page 81 • HL7 New Zealand
Profile example
Medication.code with a defined valueset
79. Page 82 • HL7 New Zealand
Profile Example
• MedicationPrescription.asNeeded required, and must be
boolean
80. Page 83 • HL7 New Zealand
Slicing
• Constraining an element with 0..* , 1..*
• FHIR Document as an example
• Composition resource is the CDA Header equivalent
– Composition.section is 0..*
– A CCDA equivalent profile might specify the sections allowed
• Different document types would have a profile
82. Page 85 • HL7 New Zealand
Extensions
• Resources design for the 80% - Allow easy extension for the
remaining 20% of elements
• Note - You’re not extending a resource per se, but you
specify its characteristics and where an extension may
occur:
– A specific element within a resource
– A specific element within a datatype
– All elements referencing a datatype
– All elements mapping to a specific mapping target
– An extension
83. Page 86 • HL7 New Zealand
Extensions
• In FHIR, extensions are “normal”
– Consequence of the 80% rule – keep the simple stuff simple
– Extensions can exist anywhere
• Yes, even inside boolean or date
– Conformant systems can’t reject instances just because they contain
unrecognized extensions
– They could:
• Display them
– Should be in resource narrative
• Store as a ‘Blob’
• Make a conscious decision to ignore (unless ModifierExtension)
– (Could lookup profile)
86
85. Page 88 • HL7 New Zealand
An Extension in a resource
88
86. Page 89 • HL7 New Zealand
Extending a multiple birth
Key = location of formal definit
Value = value according to definition
87. Page 90 • HL7 New Zealand
Modifier Extensions
• Also a core part of FHIR
– Needed because some extensions can’t be safely ignored
– Can’t compute on an element containing an unrecognized modifier
extension. However, can:
• Reject instance
• Just display narrative
• Retrieve definition & seek human review
90
88. Page 91 • HL7 New Zealand
Modifier Extension decisions
• When should you introduce them?
– modifierExtension breaks interoperability so:
• If you can accomplish your objective without one, do that
– Consider a new resource or Other
– Could requirement be met by an element that doesn’t change other
element interpretations?
– Best used when already part of existing practice, but in too narrow an
area to justify being part of core
91
89. Page 92 • HL7 New Zealand
‘Other’ resource
• When there’s no way to extend an existing one
• Has:
– Identifier
– Code
– Subject
– Author
– Created
• Rest is extensions
• Should discuss with community first
90. Page 93 • HL7 New Zealand
Designing a Profile
• Define Use Case
• Choose resources
– ‘Other’ if you have to
• Constrain base elements
– Remove
– Make required
– Define ValueSets and bindings
• Define extensions
– Try to re-use existing if possible
– Note ‘hierarchy’ of extensions:
• HL7
• National
• Local
• Build profile
• Publish
91. Page 94 • HL7 New Zealand
The need for Governance
• Profiles & extensions allow a user to customize their use of
FHIR
• HL7 is establishing publically available registries
– For reference purposes
• With great power...
94. Page 97 • HL7 New Zealand
Overview of a server
97
HTTP / REST interface Encoding/decoding,
param validation,
syntax validation
Fhir Service
Indexer / Search Storage
Implement service
operations as
described in spec
95. Page 98 • HL7 New Zealand
From wire to store
Storage
Fhir Service
REST
interface
JSON/XML
POCO/POJ
O
DBMS
O-R Map
FHIR Parser
98
JSON/XML
POCO/POJ
O
Serialize
FHIR Parser
NoSql
(Xml/Json)
JSON/XML
POCO/POJ
O
DBMS
Serialize
FHIR Parser
96. Page 99 • HL7 New Zealand
Conformance resource
• Allows a server to indicate:
– Metadata about the server
– The resources they support
• The queries against those resources
– The paradigms they support
• REST, messaging, services
– Specific (custom) queries
– The profiles it supports
• Is a normal resource
– Supports discoverability (think UDDI for SOAP)
97. Page 100 • HL7 New Zealand
Conformance
• Which FHIR version?
• Which resources?
• Which paradigms ?
• What search operations?
• What formats?
• What profiles?
• Is this a test server?
• Who can I contact?
• Do you support History?
• Do you support JSON/XML?
• Is a normal resource
• Supports discoverability (think
UDDI for SOAP)
98. Page 101 • HL7 New Zealand
Operation Outcome
• When something goes wrong….return the
OperationOutcome Resource!
99. Page 102 • HL7 New Zealand
Special endpoints
• Binary
– Store/retrieve a blob.
– No query
• Document
– Works at ‘document’ level
– Query on composition
• Mailbox
– Processing endpoint
– Exact actions depending on Trading Partner Agreement
• Note MessageHeader properties – eg event, reason
101. Page 104 • HL7 New Zealand
Workflow
• Resource describe ‘things’
• Workflow says what to do with those things
• Eg
– ePresciption
– Lab order
– Referral
• Order resources
– Order
– OrderResponse
• Combination of Order(for workflow) and specific resource
– Eg Order + MedicationPrescription
102. Page 105 • HL7 New Zealand
Referrals workflow
105
104. Page 107 • HL7 New Zealand
Overview
• Signatures
– On bundles
• Tags used to ‘annotate’
resources
– Eg private resource, ‘Special
Patient’
– Application applies policy
• Resources are atomic level
• Some ‘security-related’
resources
– SecurityEvent
– Provenance
• Philosophically FHIR delegates security to infrastructure
– Others do that better
– But does have comments
105. Page 108 • HL7 New Zealand
Wrapping Up: Tools for Developers
• Specification
– The full specification (included build tools) can be downloaded from the CVS at:
http://gforge.hl7.org/gf/project/fhir
• (Free to download, need account to update)
• Examples
• Validation files
– XSD, Schematron
• Test Servers
– Open source
• Reference Implementations
– Java, .net, Objective-C, javascript
– Open source
– Reference Implementations from www.hl7.org/fhir
• Tooling
– XML/JSON editor
– REST client
108
106. Page 109 • HL7 New Zealand
Further Information & Help
• Information:
– Specification: www.hl7.org/fhir
• Fully hyper-linked
• Includes many sample files and links to codesets & terminologies
– Wiki pages: http://wiki.hl7.org/index.php?title=FHIR
– HL7 Lists server: from HL7.org
– Blogs (listed on spec)
• Assistance
– ‘Stack Overflow’ – tag hl7-FHIR
– Skpe conversations (email me – david.hay@orionhealth.com to be
included)
Published as HTML
Published using validation process that performs consistency checks – like a software build
Really shouldn’t require much guidance to read, but a few things to call out
Objective of spec is developer can skim and decide in < day
Show the search (search for tag)
You’re a message broker routing and translating between v2, v3 and FHIR
Your app interfaces with a PHR using FHIR natively
Your application can communicate using FHIR, but your software uses a proprietary RDBMS
Use FHIR as the common language for a Vendor Neutral Archive
…Or a combination….
You might see data as a nested structure of XML, a series of tables with keys, class-diagrams….
FHIR can be your shared persistence layer – nice granularity for storage, extensions for handling “extra” stuff easily
FHIR can be the common model for your mapping layer
FHIR can be the logical model used for decision support, both to expose your data to the decision support engine as well as to define rules (work in progress with vMR-CDS and CQI)
Review
Chained – query across referenced resources - GET [base-url]/DiagnosticReport?subject.name=peter
Tags: security, profile, general
Examples if have time
http://fhir.healthintersections.com.au/open/Patient?name=eve
http://fhir.healthintersections.com.au/open/Condition?subject=84238862
http://fhir.healthintersections.com.au/open/patient/84238862/Condition
Standard search parameters
Example of sync DocumentRefrences – a DHB could maintain a local cache
* the deletion operation should be understood as deleting the record of the resource, with nothing about the state of the real-world corresponding resource implied.
* Now, if you query for /server.org/fhir/patient/@33, you get a 410
* The resource returns back to life!
Not talking about REST in general, but rather FHIR’s implementation of REST
Todo: more specifics from Grahame.
Removed: No natural “server” or no fixed network location
Note that easier in NZ as we locked CDA down.
A Document, no matter how nested, is flattened to a list of entries, the Document’s header being the first.
The document header (and any other the other resources) refer to each other using normal references to reflect the document’s nesting.
Of course, there may be a digital signature (on the whole Bundle) to attest to the content of the document.
Talk about how functions as the CDA header
Furore server – id NZ100
Talk about conversion
You can retrieve any patient using a GET on the patient’s id, which is just an url on the server: /fhir/Patient/<id>
We have our own MIME-type: “application/xml+fhir” and “application/json+fhir”
Mobile, social web, green field are most likely adopters, though significant interest from other quarters
This is not only the URL you use to retrieve the resource, it’s also its id.
All URL’s in FHIR are case-sensitive (and so is the id)
It is *metadata*, you won’t find this in the Resource’s definition
Resource Id’s (=URLs) are infrastructural id’s, they differ from “business” identifier.
Many Resources also have business identifiers, they are explicitly modeled, like Patient.identifier (even more than one identifier possible!)
Business identifiers are completely separate from technical resource id’s
Only the Resources are user-definable, other types are “built-in”
Derived primitives are patterns -> validation consists of regexp matching
Constrained types are defined using invariants (OCL, Xpath, prose) -> validation using schematron, code
Narrative and Extension are both ONLY used in Resources
Resources can use both (derived) primitives and composite datatypes in its definitions
Infrastructural types need special handling, not general-purpose types
Resources SHOULD always contain narrative to support human-consumption as a fallback. However, in a strictly managed trading systems where all systems share a common data model and additional text is unnecessary or even a clinical safety risk, the narrative may be omitted.
generated The contents of the narrative are entirely generated from the structured data in the resource.
extensions The contents of the narrative are entirely generated from the structured data in the resource and some of the content is generated from extensions.
additional The contents of the narrative contain additional information not found in the structured data.
empty the contents of the narrative are some equivalent of "No human-readable text provided for this resource".
References can be in the same bundle, on the same server, on a different server
Can expand value set
See how the resource meta-data is mapped to equivalent Atom members
Author is required by atom, so you’ll have to keep track who authored the resource, which might well be the user that POSTed it to your RESTful endpoint
Summary is optional, but it is easy to fill it with the Resource’s <text> (human readable narrative), so Feed readers have a way to display the contents of a resource. Yes, this means the summary is present twice in the entry.
All elements you see here are Atom spec, not FHIR
See how the resource meta-data is mapped to equivalent Atom members
Author is required by atom, so you’ll have to keep track who authored the resource, which might well be the user that POSTed it to your RESTful endpoint
Summary is optional, but it is easy to fill it with the Resource’s <text> (human readable narrative), so Feed readers have a way to display the contents of a resource. Yes, this means the summary is present twice in the entry.
All elements you see here are Atom spec, not FHIR
Imagine that all resources in a single bundle
Links maintain references
Comment on new recources (cid: id’s)
Specify data types – eg if deceased then must be datetime not boolean
A ‘hierarchy’ of profiles
The higher up the more likely to be reused
Really any FHIR element (Resource, Datatype, Primitive) can be extended. Just nest an <extension> element under the thing you want to extend
You should be able to go to the formal definition endpoint and get the definition of the extension.
Note: birth order is already provided for in FHIR through the multipleBirthInteger
Especially in HTTP space lots of work done – why re-invent?
Application ‘knows’ user and rights, can deliver content as appropriate
No concept of an ‘incomplete’ response to a query (could define tag in bundle if required)