This wiki has undergone a migration to Confluence found Here
<meta name="googlebot" content="noindex">
Requirements-Dynamic Model
From HL7Wiki
This page documents one or more V3 Methodology Requirements
Contents
Description
The HL7 methodology includes two parts needed for communication of information between systems, the static model (information content itself and any wrappers) and the dynamic model (what is the receiver supposed to do with the information that has been communicated). This section seeks to briefly describe the dynamic model and to completely document the dynamic model requirements for MIF.
The core of the HL7 dynamic model is the interaction. A formal definition from HL7:
A unique association between a specific message type (static model or document), a particular trigger event that initiates or "triggers" the transfer, and the Receiver Responsibilities (in terms of response interactions) associated with the receipt of the Interaction.
A single Interaction explicitly answers the questions:
- What the particular message type is (static model which may be a document, a external addressing (transport) wrapper, and an 'action' wrapper.
- What caused the message to be sent (triggering event)
- What is expected of the receiving system (trigger event)
- How a receiving system knows when it has to send a particular type of response message(s) - (receiver responsibilities).
Interactions are categorized by type and each type is typified by a pattern of receiver responsibilities.
- interaction-based - An interaction caused by the receipt of another interaction. For example query response receiver responsibility.
- state-based - An interaction caused by a change in status. For example, putting a repeating order on hold (to suspend action on that order). Receiver responsibility is usually a confirm or other acknowledgement.
- user-based - An interaction caused by a user interacting with a system or some other environmental occurrence. Receiver responsibilities varies from a simple acknowledgement to multiple interactions (for example, lab order fulfillment).
Requirements
Dynamic Model
| Requirement | HL7 requires a methodology and structure for both sending information as well as communicating the 'reason' information has been communicated. The reason for the information communication are the dynamic model components. The primary structure is the interaction (described above). |
| Rationale | A formal structure is needed to produce formalized, harmonized, interoperable standards. |
| MIF |
|
Trigger Events
| Requirement | Each interaction must have a trigger event. HL7 needs to be able to define the 'real world' occurrences that result in a need for information to be exchanged |
| Rationale | The trigger events must be able to be precisely referenced when creating interactions. The trigger event is the reason for communicating information. Part of interoperability is not only what information will get shared but also the circumstances where it will be shared. A system that does not share the appropriate information at the correct time will not achieve interoperability. |
| MIF |
|
| Requirement | Each trigger event must have a unique name |
| Rationale | The trigger events must be able to be precisely referenced when creating interactions. The trigger event is the reason for communicating information. |
| MIF |
|
| Requirement | Each trigger event must include a type. |
| Rationale | The trigger event types communicate the 'pattern' expected as far as receiver responsibilities. For example, and interaction-based trigger event is 'kicked off' when a triggering interaction is communicated. |
| Methodology | There are three types:
|
| MIF |
|
| Requirement | EnvironmentalOccurrence may have associated text. |
| Rationale | Describes the nature of the environmental occurrence (e.g. user action) |
| MIF |
|
| Requirement | EnvironmentalOccurrence may have a related state transition. |
| Rationale | Describes the state transition related to the environmental occurrence. A state transition includes the static model, the class, and the transition. A frequent example is when a user requests info to be posted to a patient's EHR (related state transition is to 'complete' an observation) and record it or to request diagnostic testing (related state transition is to 'activate' an order). Trigger isn't actually the state transition but is associated with it. E.g. a request for the transition or request for fulfilllment of a previous transition. The triggering action is something else, but it's still useful to tie these together because of the semantic linkage. |
| MIF |
|
| Requirement | A state transition includes the static model, the class, and the transition. |
| Rationale | Required to know which information model to update. |
| MIF |
|
Static Model References
| Requirement | Structured documents have a document definition (static model) |
| Rationale | Documents the primary payload model as well as any parameterized stubs within the model. |
| MIF |
|
| Requirement | Structured documents may not have a dynamic model. |
| Rationale | Documents are a point-in-time artifact and, as such, many times have no additional instruction to sending a document to a document repository other than to 'store' the document or a link to the document. In this case, there is no state transition. |
| MIF |
|
| Requirement | Each static model must be able to be bound to one or more interactions, and to be bound to one or more other static models. |
| Rationale | Each interaction is bound to a specific payload model, for which is bound a specific control act wrapper model which is bound to a specific transport model. This includes the binding to an appropriate parameter model for query interactions. The parameterized models are not bound until defining the interaction so that static models can be re-used in differing interactions to provide similar business function (acknowledgements, for example). |
| MIF |
|
| Requirement | Each static model has a content type. |
| Rationale | Identifies the type of content represented by the model when entered from this entry point. The contentType determines whether the model is legitimate content for a classStub from another model. |
| MIF |
|
| Requirement | Naming must be driven by the binding of the static models to ensure separation of the models transparent. |
| Rationale | Naming of each static model as bound still needs demarcation from the other model content. |
| MIF |
|
| Requirement | For each query interaction, an additional binding to the parameter list is required. Note this includes query requests as well as query responses. |
| Rationale | The parameter list is required for both requests and responses. |
| MIF |
|
| Requirement | A parameter model is a list of passing parameters with a query request or those used to select records in a query response. Each parameter has a parameter name and may have a traversal name. |
| Rationale | The parameter list is required for both requests and responses. The parameter is a unique name per parameter in the same list. The traversal name is the of the element when traversing from the association end directly to this specialized class |
| MIF |
|
Interations
| Requirement | Each interaction must have a unique name |
| Rationale | The interactions must be able to be precisely referenced esp. as part of receiver responsibilities. |
| MIF |
|
| Requirement | Each interaction must include one or more static models. There are use cases where only the transport wrapper, and/or transport and control act wrapper are communicated without a 'payload' (meaning a static model or document). |
| Rationale | The interactions must be able to be precisely referenced esp. as part of receiver responsibilities. Also, the receiving system needs to kn ow |
| MIF |
|
| Requirement | Interactions must be able to bind to a message model or a document as the payload. |
| Rationale | There are two 'types' of static model content at HL7. RMIMs which represent workflow-type communications and Documents which represent point-in-time information. |
| MIF |
|
| Requirement | Interaction receiver responsibilities must be supported |
| Rationale | Frequently in response to an interaction, one or more 'response' interactions are triggered. There can be multiple responses. For a simple, two-part example, a 'transaction' is not complete until both a request trigger event is and information is communication and it's confirm response is receiver. The confirm response is the 'receive responsibility'. These responsibilities are communicated by the sender and are intended to 'complete' a business function. |
| MIF |
|
| Requirement | Each interaction may include one or more receiver responsibilities |
| Rationale | Frequently in response to an interaction, one or more 'response' interactions are triggered. There can be multiple responses. |
| MIF |
|
| Requirement | Each interaction may have an interaction type. |
| Rationale | Synonymous with trigger event types. See Description above. |
| MIF |
|
| Requirement | Interactions may include a use type. |
| Rationale | Identifies the type of content represented by the model when entered from the entry point. The contentType determines whether the model is legitimate content for a classStub (binding) from another model. |
| MIF |
|
Receiver Responsibilities
| Requirement | Each receiver responsibility must have a reason. |
| Rationale | Indicates the conditions under which this receiver responsibility should be chosen. Should be mutually exclusive with the conditions of all other receiver responsibilities for this interaction. Also, the combined reasons for all receiver responsibilities should be complete. I.e. There should be no circumstance that doesn't fit into the reason of one and only one receiver responsibility. |
| MIF |
|
| Requirement | A receiver responsibility may have one or more 'response' interactions. |
| Rationale | Supports acknowledgements as well as query responses. |
| MIF |
|
| Requirement | A receiver responsibility may have one or more 'response' trigger events. |
| Rationale | Indicates that the specified trigger event should be fired, along with any interactions associated with that trigger event and supported by any of the receiving system's application roles. |
| MIF |
|
Application Roles
| Requirement | Each application role must have a unique name |
| Rationale | The sending and receiving application role must be able to be precisely referenced when creating interactions. |
| MIF |
|
| Requirement | Each application role may have related application roles. |
| Rationale | Used to build composites. |
| MIF |
|
| Requirement | Each application role may be either or both sending interactions and/or receiving interactions. |
| Rationale | Application role is about the functional capability. Need to document for each role when the application expects to receiving information and act; or sends information. |
| MIF |
|
| Requirement | Each application role may either create or consume a document. |
| Rationale | Application roles need to deal with documents as well as other static models. |
| MIF |
|
| Requirement | Interactions, Trigger Events, Application Roles and Structured Documents may have a number of different types of annotations |
| Rationale | See rationales for individual annotations types |
| Implementation |
Future Requirements (not complete)
Note that HL7 is currently re-developing and replacing the current dynamic model methodolgy. Below are SOME of the requirements. Requirements are still being determined and documented as part of the HL7 Enterprise Architecture Framework Alpha implementation projects.
| Future Requirement | Communicate conformance of receiver responsibilities. It is necessary to know which receiver responsibilities must happen (mandatory), should happen (required), may happen (optional), and must not happen (usually based on a previous elaboration of a receiver responsibility). |
| Rationale | It is necessary to know which receiver responsibilities must happen (mandatory), should happen (required), may happen (optional), and must not happen (usually based on a previous elaboration of a receiver responsibility). |
| MIF |
|
