Section 1
The Activity Diagram
Guide to notation:
Activity Diagrams
Activity Diagrams are used to describe a complex system in terms of smaller functional components called Activities. An Activity represents a dynamic process during which objects are manipulated and transformed, giving rise to a Product. The Product of one Activity can be required for another Activity to function. The circles represent Activities; the boxes represent the Products. A Product can be represented together with its sub-products. Such partitioning is discussed more fully in the guide to Concept Diagrams.
Each Activity can be analysed separately. One of the uses of an Activity Diagram is that it can provide an architecture for the analysis of a complex process. Activities can also be analysed in terms of their interaction, through the Products they exchange.
The result of such analysis is represented in the Concept Diagrams of Section 2 (which illustrate the necessary properties of the Activities and Products) and in the Event Diagrams of Section 3 (which illustrate the causal sequence of events both within Activities and during the exchange of Products between Activities).
The arrows shown on Activity Diagrams show only the direction of production and use. They do not represent the sequence in which Activities run. Activities often run concurrently. The representation of several Activities within another Activity illustrates the coordination of the subordinate Activities as part of the process of the overall Activity.
The Activity Diagram
The Activity Diagram assists in defining the scope of the project’s work. It illustrates the clinical activities that we have found it useful to distinguish. During work on the Clinical Process Model the activities have been considered separately as well as in terms of the interactions between activities.
As emphasised in the Introduction, the object of care can be an individual person or a population (including groups of people such as family groups), in other words any person(s) who can become the focus of clinical care. The object of care makes demands on the clinical process (represented in the diagram by the large arrow and shadowed box). It is understood that the demand made on the clinical process by the object of care is for the maintenance and enhancement of good health, as well as for care required as a result of ill-health.
The activity of observing, results in the observations that are the basic materials of clinical assessment. Moreover, such assessment is itself regarded as being an act of observation.
The activity of planning arises from the observations made, and generates clinical plans to meet the object of care’s requirements, to overcome, palliate or prevent clinical problems, or to make further observations. Planning defines the procedures to be used and their inter-relationships.
Plan implementation comprises the performance of procedures, which will be the interventions necessary for the clinical needs identified, and the observations required for further clinical assessment or for monitoring progress. The observations that lead to planning can be of a wish to continue in good health; and the interventions performed when plans are implemented can be intended to enhance health, as well as protect or restore it.
The activities of observing, planning, and plan implementation, constitute the clinical process, and their products (observations, plans, and procedures) are all regarded as clinical actions.
The activities of the clinical process all make demands on clinical knowledge and skills, and on the authority (in the form of accountabilities) and resources necessary for the use of those skills.
Beyond the scope of this project to date is any detailed consideration of the process of clinical learning, but it is recognised that as clinical procedures are performed on the basis of what clinicians understand to be possible, so clinical knowledge and skills are acquired through an accumulation of experience and experimentation.
Section 2
The Concept Diagram
The components of the Concept Diagram and their relations to each other are described in stages. Each stage of the description is supported by an illustration. The illustration does not however necessarily depict all the Concepts to which those in the illustration are related. For a complete view at any stage, reference should be made to the fold-out illustration of the whole Concept Diagram which is included inside the back cover of Volume 1. The Glossary of Concept Names (Volume 2) also illustrates each Concept in turn, together with all the other Concepts to which it is related.
Concept Names
A brief description of each Concept is given in the Glossary of Concept Names (Volume 2), as well as details and an illustration of all the other Concepts to which it is related. The Glossary also contains cross-references to all significant mentions made of each Concept in Section 2.
Words and phrases used as Concept Names are not used in any other context in Section 2. Thus every time a Concept Name is used in Section 2 it is referring to the Concept. When reference is made to a combination of two Concepts the Name used is simply a combination of the two 'parent' Names. (Explained in more detail in the Notation Guide).
Concept Names are distinguished in Section 2 by use of a font that is slightly different to that of the main text, eg observations and interventions.
Clinical Examples
¶All paragraphs that deal with clinical examples to support the Concept Diagram are indented as this one, and lead with the character ¶. Single examples in the main text are in parentheses.
Guide to notation: Concept Diagrams
A concept represents a collection of objects that share one or more defining properties.
One form of relation between concepts is illustrated by a line directly from one concept to another. The relation is named: in this case it could be ‘married to’. The number of objects that can be related between concepts is determined by the type of line used to illustrate the relation (broken, solid, with or without a ‘crows foot’):
Each A must be related to one BEach A may be related to one B
Each B must be related to one AEach B may be related to one A
Each A may be related to one BEach A may be related to one or many B
Each B must be related to one AEach B must be related to one A
Each A may be related to one BEach A may be related to one B
Each B must be related to one or many AEach B may be related to one or many A
Each A must be related to one BEach A must be related to one or many B
Each B must be related to one or many AEach B must be related to one or many A
Each A may be related to one or many BEach A may be related to one or many A
Each B may be related to one or many A
The property of a relation that determines the number of objects that can or must be related between two concepts is known as the ‘cardinality’ of the relation. Each relation has two cardinalities: one cardinality is from A to B and determines how many B can or must be related to A; the other cardinality is from B to A and determines how many A can or must be related to B.
Each of the directions that a relation runs in (from A to B, and from B to A) has a name as well as a cardinality. In the illustrations in Section 2 only one name is used (that which best conveys the ‘meaning’ of the relation); both names are found, along with the cardinalities, in the Glossary of Concept Names (Volume 2).
Within a collection of objects, sub-collections can be identified by additional distinguishing properties. The concepts representing the sub-collections are called subtypes. The ‘parent’ concept is called the supertype. The subtypes illustrated here are the only possible alternatives of the supertype. The box they are represented in is therefore called a complete partition. The line from a supertype concept to the box that represents the partition of subtype concepts is always solid and ends at the edge of the box.
If the subtypes illustrated are not comprehensive, then a box with a double line for its bottom edge is used. This is called an incomplete partition. Incomplete partitions need not be mutually exclusive. A person can be female and a doctor, but not all doctors are female, nor vice versa. In the document such combinations are referred to by using both titles, eg. 'female doctor'.
A concept can represent a collection of objects that are also members of two supertype concepts. Each male nurse is also a member both of the set ‘man’ and the set ‘nurse’.
Summary
B is a subtype of A.
C is a subtype of B.
Each B must be related to one D.Each C must be related to one or many E.
Each D must be related to one B.Each E may be related to one C.
Each illustration in Section 2 does not necessarily depict all the relations of the concepts represented (they are found in the Global Concept Diagram which folds out from inside the back cover of Volume 1). In certain cases, to limit the number of concepts being described at any one point, a ‘short-cut’ through the model is marked in the illustrations by means of a diagonal line. Thus, adapting the illustration above:
All objects represented by C are also members of the collection of objects represented by B. Therefore:
C is a subtype of A.
Each C must be related to one D.
2.1 Action
2.1.1 Timepoint & Location
We consider the clinical record to be concerned with actions, clinical acts carried out within health care. An action is carried out at a location.
An action can be complex, made up of a number of different component actions. If all the component actions are performed at the same location then the location of the parent action is also that of all its component actions.
If a complex action is performed at a number of different locations,then one separate location is recorded for each component action that is subordinate to the parent action.
Start and conclusion timepoints can be ascribed to an action, recording when the action began and ended. These timepoints mark the timespan of an action. There is no provision within the model as yet for other examples of timepoint, such as those that would allow the recording of the time actually spent on an action, where that differs from the timespan.
2.1.2 Object of Care & Procedure
Actions include observations of an object of care’s state of health, and interventions that attempt and/or risk a change to that state of health (physical, mental, social, etc). The actions of observation and intervention are grouped together as clinical procedures. They are focused on individual objects of care.
Objects of care, as already emphasised in the Introduction, represent populations as well as single patients. Modelling work has been done by ourselves and others on representing means of defining populations (such as family groups defined by consanguinity and other ties, including role within the group). The work is not complete enough however to incorporate into the current version.
2.1.3 Protocol
It is considered that procedures are performed as an expression of what the clinician understands it is possible to do. This may be the clinical knowledge of how to observe or measure a particular characteristic, or of how to intervene in an attempt to achieve a particular end. The representation of this understanding in the Cosmos model hinges around protocols.
The performance of any procedure is regarded as being the implementation of one or more distinct protocols. Individual protocols can be defined and distinguished by a number of properties (circumstances of use, expected outcome, necessary resources, etc) and these properties can be referred to when planning and scheduling procedures. The representation of these properties in the Concept Diagram is considered in Section 2.4.
The term protocol has been chosen rather than ‘technique’ because its popular use carries a sense of etiquette, an understanding of how things should be done. A narrower usage is sometimes found in clinical care to describe how complex, but well-defined and tightly-bounded procedures should be performed (eg chemotherapy regimes). Protocols in the model have this capability, but can also represent less complex tasks (eg taking the blood pressure).
¶ An immunisation programme has protocols that determine both the identification and enrolling of subjects for immunisation, as well as the actual immunisation procedure itself (injection, etc).
As an action, a procedure can be made up of a number of component procedures. The composing of a complex procedure for a particular object of care, using a number of possibly unrelated protocols, is the subject of clinical planning, described in Section 2.6. Because a complex procedure can be customised for an object of care, there will not necessarily be a protocol for the overall procedure that has been customised. The relation between procedure and protocol is not therefore mandatory.
2.1.4 Accountability
The authority under which an action is to be performed is represented by an accountability. An accountability is an agreement that authorises the carrying out of an action. It is set up between two parties. One party commissions the other; that second party is accountable to or responsible to the first for actions undertaken with the authority of that accountability. An involved party can be an establishment (such as a hospital, health authority, or laboratory) or an individual person, including a patient.
(Each patient is both a person and an object of care).
¶Accountabilities are intended to cover a wide range of agreements, eg:
• the appointment of an employee at a particular establishment (hospital consultant for instance).
• a patient’s registration with a GP.
• the formal consent requested of patients for particular procedures.
An accountability can be an example of an accountability type. The accountability type defines what form of agreement has been reached (eg a particular agreement between patient and doctor is of the type ‘registration with a GP’).
A single accountability can also describe the circumstance of a clinician or an establishment being responsible for the care of a patient over a period of time (defined for instance by a particular illness, however prolonged).
That single overall accountability governs the entire period during which the clinician or establishment has a responsibility towards the patient, however slight, and however prolonged the intervals between consultations.
During this period of time, there can be a number of separate consultations, both outpatient and inpatient. Each consultation is not only governed by that single overall accountability, but can itself involve a number of other accountabilities to cover admissions and/or particular procedures. Each of those accountabilities is confined to the limits of the individual consultation.
Thus, governed by that single overall accountability can be multiple, concurrent accountabilities, distinct with respect to the procedures authorised.
When the patient is recorded as having left the clinician’s care without a provision for further consultation having been defined (whether open, or for a particular appointment) then the single overall accountability ceases. (In some clinical practices a patient’s file is left ‘open’ for a defined period of time, in case a further consultation is required by the patient or clinician.)
¶For example, an accountability is opened for a patient with diabetes mellitus to attend a local hospital. This accountability covers all consultations made at the hospital with respect to the patient’s diabetes, regardless of changes to hospital personnel (consultants, clinic nurses, etc). Each consultation can involve other accountabilities to govern, for instance, inpatient admissions, day-case diagnostic procedures, etc. The overall accountability usually remains open until the patient’s death, or transfer to another diabetic department.
¶A patient’s death can lead to a number of accountabilities being established:
• the clinician can be required by the coroner to perform an autopsy.
• the clinician can organise bereavement counselling for the patient’s relatives.
• authorisation can be sought from the relatives for organ donation.
¶A patient’s registration with a GP is represented as an accountability that is opened on registration, remains open for any consultation the patient is able to make, and closes only when the patient is removed or removes her/himself from the GP’s list.
¶In terms of the model the various NHS Acts are seen as being accountabilities with respect to the populations for which health care is available. The provision of care to any group or individual that belongs to one of the defined populations is governed by that accountability.
Specific acts of care directed at defined populations can require further authorisation, eg immunisation programmes, screening programmes, etc.
More than one accountability can govern the performance of an action: the patient’s formal consent, the hospital appointment of the clinician, the contract which will determine payment, the necessary qualifications and accreditations of the clinician.
A record of any, all, or none of these can be regarded as necessary. As stated in the introduction, the model is not intended to be prescriptive, but rather to cater for a record that is as detailed or as limited as the circumstances require it to be.
2.1.5 States of Action
A record can be required of actions at any stage: when they are being planned, and whether they were carried out or not. These states are described by actions having been:
•proposed.
•or implemented.
Before a proposed action is implemented, it is possible that an alternative action is selected, in which case it replaces the previously proposed action. The replaced proposed action has in effect been withdrawn. The action that replaces it can be a positive intention to do nothing.
The details of a proposed action can vary significantly from those of its complementary implemented action. It may have turned out that the action was performed at a different time, or by a different clinician, etc. The differences can be very important, so the proposed action is kept intact, establishing a relationship between it and the implementedaction that resulted, if any.
Of course a record can be made of an implemented action without there being any need to record its having been proposed. It is not obligatory that for every implemented action there should be a record of its complementary proposed action.
At any one time there can be a record of proposed actions that have yet to be implemented. They are distinct from proposed actions that have been replaced, never to be implemented.
We regard the two types of action (proposed and implemented) as being sufficient to start a description of all the possible different stages in the evolution of an action. Each type will itself have different possible states in which it can be recorded:
•the records of proposed actions will accumulate details of resources to be used, expected timepoints, performers, etc. There is no distinction made in this model between different states of scheduling, other than by providing for a record of the arrangements as they are made.