Professional Documents
Culture Documents
Domain Modelling Some Healthcare Sector Concepts
Domain Modelling Some Healthcare Sector Concepts
Domain Modelling Some Healthcare Sector Concepts
Dines Bjrner
Department of Information Technology
Technical University of Denmark
DK{2800 Lyngby, Denmark
E{Mail: db@it.dtu.dk
September 14, 2000
Abstract
As2 part of a student term and a programming methodology research project we
explore the informal and formal description of a health care sector concept. The student
term project shall show that the project students understand the various concepts of
abstraction & modelling, of domain attribute & facet, and of documentation principles
and techniques. The research project shall test the relevance of these principles and
techniques. At the same time the combined project can hopefully also serve as a basis for
further, application-specic applied research and experimental, prototype development of
a software system for the support of patient and health care unit (eg. hospital) operations.
The3 concept of healthcare sector is, in this document, explored along two, of several
axes:
Which are the players of the healthcare sector, to wit: citizens (healthy and sick),
family doctors, pharmacies, community nurses, clinics, hospitals, health insurance
companies, pharmaceutical industries, government regulatory agencies, etc. ? Who
communicates with whom ? What is communicated: information, medicine, pa-
tients, test samples, etc. ?
Patient/hospital treatment cases: Which are the actions that are performed by
patients and by hospitals (their sta etc.), for each of a number of \standard"
treatment cases (to wit: hip operation) ? Can these be semi-formalisable in the
1 The present document is a torso: Incomplete. It drafts the initial parts of a report that a pair of students
in the fall 2000 course, 493_ 51 Advanced Software Specication, is supposed to complete according to all the
\bells-&-whistles" of what is prescribed to be best, modern practices of software engineering. In another
document, avaliable over the web, we detail, for that course, specics for how the eventual report is to cover
the subject, ie. how it is to fullll the project brief (See Sect. 2). Lecture notes for 493_ 51 Advanced Software
Specication can be found at:
http://www.it.dtu.dk/~db/courses/notes.ps
http://www.it.dtu.dk/~db/courses/domain.ps
http://www.it.dtu.dk/~db/courses/wg23.ps
2 This paragraph basically outlines a `project synopsis'
3 This paragraph basically outlines a `product synopsis'
1
2 September 14, 2000. Dines Bjrner
Contents
1 Introduction 3
2 Project Brief 3
3 Synopses 3
3.1 Project Synopsis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
3.2 Product Synopsis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
4 Rough Sketch Scenarios 4
4.1 Scenario 1: Patient Flow in the Healthcare Sector . . . . . . . . . . . . . . . . 4
4.1.1 Flow Partners (\The Player") . . . . . . . . . . . . . . . . . . . . . . . 5
4.1.2 Flow Components: Informational and Material . . . . . . . . . . . . . 6
4.1.3 Flow Component Actions . . . . . . . . . . . . . . . . . . . . . . . . . 8
4.2 Scenario 2: Hospital Treatment Cases . . . . . . . . . . . . . . . . . . . . . . 9
4.2.1 Treatment Case I: A Generic, Surgical Case . . . . . . . . . . . . . . . 9
4.2.2 Towards a Case Treatment Planning Language . . . . . . . . . . . . . 13
4.2.3 Treatment Case II: Hip Operation . . . . . . . . . . . . . . . . . . . . 13
4.2.4 Treatment Case III: Organ Transplantation . . . . . . . . . . . . . . . 13
4.2.5 Treatment Case IV: Acute/Burst Appendicites . . . . . . . . . . . . . 14
4.2.6 Treatment Case V: Pneumonia . . . . . . . . . . . . . . . . . . . . . . 14
5 Analysis and Concept Formation 14
5.1 Healthcare Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
5.2 Hospital Case Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
6 Concept Terminology 15
6.1 Healthcare Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
6.2 Hospital Case Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
7 Narrative 16
7.1 Healthcare Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
7.2 Hospital Case Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
8 Formalisation 16
8.1 Healthcare Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
8.2 Hospital Case Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
c Dines Bjrner, 2000 September 14, 2000, 20:56
Domain Modelling some Healthcare Sector Concepts. September 14, 2000 3
9 Validations 17
9.1 Healthcare Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
9.2 Hospital Case Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
10 Verications 17
10.1 Healthcare Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
10.2 Hospital Case Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
11 Discussion 18
11.1 Healthcare Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
11.2 Hospital Case Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
12 Conclusion 19
13 Bibliographical Notes 19
1 Introduction
The present document, when completed, shall, when reading just the informal text (with
its easy-to-understand diagrams &c), help decision makers in the healthcare sector to decide
whether to embark on a review of their business, or on a business process [re{]engineering
eort, and/or on new software product acquisiton.
Thus the present document shall illustrate, to IT professionals of the healthcare sector,
that new ways of looking at their business are being studied by academia, and that informatics
candidates will soon emerge from universities, candidates who are able quickly, at a high
level of computing and mathematical modelling (operations analysis, etc.) professionalism |
candidates that they should probably be demanding and hiring.
Meanwhile we invite healthcare sector management to collaborate with us on our better
understanding their domain and on advanced prototype development projects.
2 Project Brief
The healthcare sector project has to deliver on two fronts: (i) Explorations of what is meant by
the healthcare sector concept, and (ii) methodology experiments wrt. how one might develop
healthcare sector [computing] systems | from domain via requirements to software design.
The project brief is to document both [(i){(ii)], intertwined: Present domain, requirements
and software design proposals for healthcare sector systems, thus, at the same time, providing
answers to both (i) and (ii): What is a healthcare sector system, and how to develop software
for it !
3 Synopses
There are two kinds of synopses for the undertaking of this document4 :
4 See the Abstract (Page 1) footnotes 2 and 3, respectively.
pm
ih hg One or more
ck N km instantiations
dp ip
ik ch mh kg
pm nh Communication
pg
P H
ph
hk
K
A healthcare process graph
The material and information communications, for such situations are therefore as for
medical doctors.
Doctor['s Oce] ! Patient: The medical doctor | or his oce secretary or nurse |
leaves a message with, sends a letter to, the patient: Please contact us by phone or come
to our oce, or: here is the result of the clinical tests you recently underwent.
Patient $ Clinic:
$ :
$ :
$:
$ :
$ :
4.1.3 Flow Component Actions
Above we have `sans serif' emphasized certain terms. They designate actions: observer and
generator functions performed on
ow materials (including people) and information. We now
examine these:
schedule: (make an appointment)
interview:
analyse:
diagnose:
prescribe:
treat:
medicate:
refer:
order:
invoice:
pay:
:
:
:
:
This ends our scenario on
ows in the healthcare sector. It is, as scenarios usually are,
incomplete, but, we believe, sucent for subsequent analysis, concept formation, narration
and formalisation.
4.2 Scenario 2: Hospital Treatment Cases
The objective of presenting this scenario is to start the process of identifying the ways and
means of treating | especially | conventional cases of patient hospitalisation: We refer to
such cases as treatments for: (i) cardio arrest, (ii) burst appendicites, (iii) kidney stones, (iv)
hip replacement, etc.
By a treatment case we mean the set of regular, commonly established procedures: inter-
views, analses, diagnoses, tests, medicination, other treatments, sugrical operations, rehabil-
itations, etc., that a patient undergoes during hospitalisation.
The longer range aim of understanding varieties of treatment cases, is to understand the
resource management consequences: Schedulings and allocations, involved in patient treat-
ment so as to provide for optimum patient satisfaction and hospital resource management.
4.2.1 Treatment Case I: A Generic, Surgical Case
We distinguish between (i) a preamble, the actions performed by patient and hospital, before
an unscheduled treatment plan is established, (ii) the establishment of an unscheduled treat-
ment plan, (iii) the scheduling and allocation of resources to fulll the treatment plan | we
refer to the result of this phase as the scheduled treatment plan, and (iv) the treatment plan
execution.
Preamble:
(IN) The patient, somehow (by phone, or letter, or at the family doctor's oce, or by
own initiative) turns up at the hospital for a standard treatment.
(ADM) At the admissions oce the type of case is [re{]conrmed by a medical sta (a
medical doctor, a nurse, or other).
(TPL) The case is registered and a treatment plan established.
Unscheduled Treatment Plan:
An unscheduled treatment plan essentially prescribes:
(I&T) The plan may st precribe a series of interviews and tests by hospital sta:
{ (RIS) For example: X{rays are to be taken,
{ (ECG) an ECG (electro-cardiogram) likewise,
{ (BLD) blood tests for blood sugar, cholesterol, etc.
{ (TST) blood pressure, urine sample tests, faeces tests,
{ (EYE) eye tests,
September 14, 2000, 20:56
c Dines Bjrner, 2000
10 September 14, 2000. Dines Bjrner
{ (ETC) &c.
(BED) The plan also prescribes; registration and `bedding' in an appriate ward.
(SOP) and the
lowing sequence of procedures:
{ (FST) fast the night before the surgical operation,
{ (WAK) wakening up the morning of the day of the surgical operations,
{ (PRE) bathing, medication and dressing for the operations,
{ (MOV) movement to the operating theatre,
{ (SUR) the surgical operation itself,
{ (IPO) immediate post-operation observation and treatment
{ (CWD) possibly ina special, critical observation ward, and
{ (MOV) movement back to the ward.
(POR) The plan then goes on to prescribe the daily post-operation routine:
{ (MED) medication,
{ (MEL) normal or special dietary meals,
{ (REG: RIS, ECG, BLD, TST, EYE, and/or ETC) regular tests of various kinds:
blood pressure, temperature, blood tests, urine tests, ECGs, eye test, throat in-
spection, etc.
{ (INT) daily interviews by medical doctors,
{ (RTR) retraining,
{ &c.
(?) Finally the plan species conditions for
{ (RPL) alteration of plans, ie. replanning | in case of unforeseen contingencies:
Failure of prescribed treatment, lack of healing, etc. | possibly leading to another
surgical opration.
{ (TRT) termination of individual treatments | in case of foreseen healing,
{ (OUT) and release from hospital | when cured.
SOP
I&T
POR
FST
RIS BED
WAK MED
ECG
PRE MEL
BLD
MOV REG
IN ADM TST
SUR INT
TRT
EYE
IPO RTR ? REL
ETC
VWD RPL
MOV
continuously
until test (?)
TPL
\subjective measures" into account, for example patient, patient family and hospital sta
preferences.
Treatment Plan Execution:
The treatment can now be set in action and according to the scheduled treatment plan.
We can, colloquially, speak of an \execution" of the scheduled treatment plan.
Such a scheduled treatment plan \execution" can be characterised in a number of ways:
{ We think of someone, one or more hospital sta, monitoring & controlling the
\execution" of the scheduled treatment plan.
{ At any point in time the locus of monitoring & control resides at one sequentially
\executed" action, or at any number of such actions in a cluster of concurrently
executable actions.
{ \Execution" of an action usually takes time.
{ \Execution" of a next action takes place either \after" the end of execution of a
previous action, or, when of a cluster of concurrently \executable" actions, only
\after" execution of all of these have terminated.
{ By \after" we sometimes mean `immediately after', and sometimes, `sometimes
after' !
{ Execution of an action may fail for a number of reasons:
One or more of the prerequisite resources, incl. the patient, hospital sta, fail
to be present at the time and place for which the action was scheduled,
or some equipment does not function properly,
or seemingly performed action leaves an inconclusive result: The patient dies,
or test measurements are contradictory, or other.
{ A (scheduled) treatment plan can prescribe what should be done in case of failures.
To keep matters simple in this explorative document we have concentrated such
\exception{like" conditions to the ? (TRT, RPL, REL) test two of whose outcomes
which prescribes \repetitions" of actions or actions clusters.
The actual behaviour of a patient/hospital system seen in the context of a scheduled
treatment plan for that patient and in that hospital, can be recorded | by attaching
information to the nodes, clusters and arrows of the scheduled treatment plan.
{ These notes record when the respective action (or action cluster) was (were) ac-
tually executed: Time (start and end), by whom, with what results (outcomes),
etc.
{ These notes also record contingencies: Exceptions, remedial actions taken, etc.
{ We call the resulting, so annotated scheduled treatment plan for the case trace.
{ The case trace thus records how many times loops backward to previous clusters
were taken, when, etc.
c Dines Bjrner, 2000 September 14, 2000, 20:56
Domain Modelling some Healthcare Sector Concepts. September 14, 2000 13
{ Thus annotations wrt. any given action box notes to which iteration of treatment
(1st, ith) the pertinent information (the notes, the annotations) relate.
So, surgical procedures, which were prescribed once, may have to be performed repeatedly, or
replanning may presctibe new surgical procedures, etc. Independent of such exception-based
repetitions or emergencies, one can, of course, have treatment plans that explicitly, from the
very planning stage, foresees several surgical events.
6 Concept Terminology
The following concepts have (so far) been identied:
7 Narrative
7.1 Healthcare Flow
8 Formalisation
8.1 Healthcare Flow
9 Validations
9.1 Healthcare Flow
10 Verications
10.1 Healthcare Flow
11 Discussion
11.1 Healthcare Flow
12 Conclusion
13 Bibliographical Notes
The present document builds upon and continues a long research endeavour: [1, 2, 3, 4, 5, 6,
7, 8, 9, 10, 11]
References
[1] Dines Bjrner, Souleimane Koussobe, Roger Noussi, and Georgui Satchok. Michael Jack-
son's Problem Frames: . In Li ShaoQi and Michael Hinchley, editors, ICFEM'97: Intl.
Conf. on Formal Engineering Methods, Los Alamitos, CA, USA, 12{14 November 1997.
IEEE Computer Society Press.
[2] Dines Bjrner and Jorge R. Cuellar. Software Engineering Education: R^oles of for-
mal specication and design calculi. Annals of Software Engineering, 6:365{410, 1998.
Published April 1999.
[3] Dines Bjrner. Domains as Prerequisites for Requirements and Software &c. In M. Broy
and B. Rumpe, editors, RTSE'97: Requirements Targeted Software and Systems Engi-
neering, volume 1526 of Lecture Notes in Computer Science, pages 1{41. Springer-Verlag,
Berlin Heidelberg, 1998.
[4] Dines Bjrner. Where do Software Architectures come from ? Systematic Development
from Domains and Requirements. A Re{assessment of Software Engneering ? South
African Journal of Computer Science, 1999. Editor: Chris Brink.
[5] Dines Bjrner. Software Engineeering: A New Approach. From domains via requirements
to software. Formal specication and design calculi. 2000. Presently this document is
a rather extensive (approx. 900 page) set of lecture notes. It is accessible over the web:
http://www.it.dtu.dk/~db/s2000.
[6] Dines Bjrner. Domain Modelling: Resource Management Strategics, Tactics & Op-
erations, Decision Support and Algorithmic Software. In J.C.P. Woodcock, editor,
Festschrift to Tony Hoare. Oxford University and Microsoft, September 13{14 1999.
[7] Dines Bjrner. A Triptych Software Development Paradigm: Domain, Requirements
and Software. Towards a Model Development of A Decision Support System for Sus-
tainable Development. In ErnstRudiger Olderog, editor, Festschrift to Hans Langmaack.
University of Kiel, Germany, October 1999.
[8] Dines Bjrner. Pinnacles of Software Engineering: 25 Years of Formal Methods. Annals
of Software Engineering, 2000. Eds. Dilip Patel and Wang Yi.
[9] Dines Bjrner. Domain Engineering, Elements of a Software Engineering Methodology |
Towards Principles, Techniques and Tools | A Study in Methodology. Research report,
Dept. of Computer Science & Technology, Technical University of Denmark, Bldg. 343,
September 14, 2000, 20:56
c Dines Bjrner, 2000
20 September 14, 2000. Dines Bjrner