The header for a message exchange that is either requesting or responding to an action. The reference(s) that are the subject of the action as well as other information related to the action are typically transmitted in a bundle in which the MessageHeader resource instance is the first resource in the bundle.

Many implementations are not prepared to use REST and need a messaging based infrastructure.

Code that identifies the event this message represents and connects it with its definition. Events defined as part of the FHIR specification have the system value "http://hl7.org/fhir/message-events".

Drives the behavior associated with this message.

The destination application which the message is intended for.

Indicates where message is to be sent for routing purposes. Allows verification of "am I the intended recipient".

There SHOULD be at least one destination, but in some circumstances, the source system is unaware of any particular destination system.

Human-readable name for the target system.

May be used for routing of response and/or to support audit.

Identifies the target end system in situations where the initial message transmission is to an intermediary system.

Supports multi-hop routing.

Indicates where the message should be routed to.

Identifies where to route the message.

The id may be a non-resolvable URI for systems that do not use standard network-based addresses.

Allows data conveyed by a message to be addressed to a particular person or department when routing to a specific application isn't sufficient.

Allows routing beyond just the application level.

Identifies the sending system to allow the use of a trust relationship.

Allows routing beyond just the application level.

Use case is for where a (trusted) sending system is responsible for multiple organizations, and therefore cannot differentiate based on source endpoint / authentication alone.

The time that the message was sent.

Allows limited detection of out-of-order and delayed transmission. Also supports audit.

The person or device that performed the data entry leading to this message. When there is more than one candidate, pick the most proximal to the message. Can provide other enterers in extensions.

Need to know for audit/traceback requirements and possibly for authorization.

Usually only for the request, but can be used in a response.

The logical author of the message - the person or device that decided the described event should happen. When there is more than one candidate, pick the most proximal to the MessageHeader. Can provide other authors in extensions.

Need to know for audit/traceback requirements and possibly for authorization.

Usually only for the request, but can be used in a response.

The source application from which this message originated.

Allows replies, supports audit.

Human-readable name for the source system.

May be used to support audit.

May include configuration or other information useful in debugging.

Supports audit and possibly interface engine behavior.

Can convey versions of multiple systems in situations where a message passes through multiple hands.

Supports audit and possibly interface engine behavior.

An e-mail, phone, website or other contact point to use to resolve issues with message communications.

Allows escalation of technical issues.

Identifies the routing target to send acknowledgements to.

Identifies where to send responses, may influence security permissions.

The id may be a non-resolvable URI for systems that do not use standard network-based addresses.

The person or organization that accepts overall responsibility for the contents of the message. The implication is that the message event happened under the policies of the responsible party.

Need to know for audit/traceback requirements and possibly for authorization.

Usually only for the request, but can be used in a response.

Coded indication of the cause for the event - indicates a reason for the occurrence of the event that is a focus of this message.

Need to be able to track why resources are being changed and report in the audit log/history of the resource. May affect authorization.

Information about the message that this message is a response to. Only present if this message is a response.

The MessageHeader.id of the message to which this message is a response.

Allows receiver to know what message is being responded to.

Code that identifies the type of response to the message - whether it was successful or not, and whether it should be resent or not.

Allows the sender of the acknowledge message to know if the request was successful or if action is needed.

This is a generic response to the request message. Specific data for the response will be found in MessageHeader.focus.

Full details of any issues found in the message.

Allows the sender of the message to determine what the specific issues are.

This SHALL be contained in the bundle. If any of the issues are errors, the response code SHALL be an error.

The actual data of the message - a reference to the root/focus class of the event.

Every message event is about actual data, a single resource, that is identified in the definition of the event, and perhaps some or all linked resources.

The data is defined where the transaction type is defined. The transaction data is always included in the bundle that is the full message. Only the root resource is specified. The resources it references should be contained in the bundle but are not also listed here. Multiple repetitions are allowed to cater for merges and other situations with multiple focal targets.