WHAT CAN ORGANIZATIONAL ANALYS GIVE TO

REQUIREMENT ANALYS ?

Developing an Information System in Hospital Emergency Departments

Anne De Vos, Claire Lobet-Maris,

Cellule Interfacultaire de Technology Assessment (CITA), University of Namur, 21, rue Grandgagnage, 5000 Namur,

Belgium

Anne Rousseau

Centre d’Innovation par les Technologies de l’Innovatio (CITI), Centre de Recherche Public Henri Tudor, 29 Avenue JF

Kennedy , 1855 Luxembourg

Keywords: Information systems, organizational change, methodology, constructivist perspective

Abstract: In this article, we develop a method of Requirements Engineering (RE), based on the contributions of the

theories of the organizations. This method was developed within the framework of an IS project led in three

Hospital Emergency Departments (HEDs) in Belgium. The method intends to take into account as well the

functional aspects of the specifications as their political and cultural dimensions. The method that we

propose is built on three theoretical levels: the first is mechanical or functional and allows us to describe the

existing organizational structure; the second is relational or political and allows us to understand this

structure in action; finally, the third is cultural or symbolic system and opens to us with the values which

guide the actions in organization.

1 INTRODUCTION

Empirical studies relating to the development of

Information Systems (IS) shows that the

Requirements Engineering (RE) is one of the most

critical phase of the development of an IS project

and its success (Boehm, 1991, Chaos, 1994-2001,

ESPITTI, 1996). Generally, one considers that RE

rests on two principal sources of information : the

description of the field of application of the project

and the requirements and expectations of the

different categories of actors, direct or indirect users

of the future system. On the basis of this

information, the conceivers draw up then the

specifications of the future system that is to say the

expected functionality's and behavior of the future

IS.

The quality of the specifications is thus directly

related to the quality of the information collected on

the application field and the requirements of the

future users. And the quality of this collect of

information depends on the theoretical model used

to understand the existing organization and the

requirements of the various actors. Very often, the

models stick to a functional and rational vision of

the organization, leaving on side the less rational

aspects of the human behavior in organization. The

importance of these less rational or functional

dimensions in RE appeared to us shouting within the

framework of an IS project developed in 3 HEDs

Initially, this project aimed at improving the

efficiency of the coordination of the care and the

management of flows of patients within these

services. Presented like such, the project seems very

mechanical directing the glance of the analysts

towards the workflow and the structures which

underlie the activity of these services.

But the things are never as simple. Very quickly

the first confrontations with the grounds of the

HEDs showed us how much the relatively intangible

and abstract elements such as the equity of treatment

of the patients, the quality of the care, the global

approach of the patient, the assumption of each

An inter-university project (UCL, FPTMS, ULg, FUNDP)

financed by the Walloon Region in Belgium (2000-2003).

390

De Vos A., Lobet-Maris C. and Rousseau A. (2005).

WHAT CAN ORGANIZATIONAL ANALYS GIVE TO REQUIREMENT ANALYS? - Developing an Information System in Hospital Emergency Departments.

In Proceedings of the Seventh International Conference on Enterprise Information Systems - DISI, pages 390-393

 SciTePress

responsibility, the management of the stress, etc.

were dimensions important to take into account in

RE although the traditional methods of engineering

are not prepared to. To capture these more intangible

elements, it appeared necessary to enrich the

organizational visions which generally chair RE. It

is from this point of view that we worked out a

model based on three dimensions.

2 FROM CO-ORDINATION TO

LOYALTY: THE CONCEPTUAL

FRAMEWORK

The central assumption on which our model rests is

that a relevant representation of organizational

reality concerns three dimensions.

The first dimension is purely mechanical or

rational. Resting on classic work, such those of

March and Simon (1958), Thompson (1967),

Galbraith (1973) and Mintzberg (1979), it considers

the organization under the angle of the structures

implemented to back the activity. From a pure

mechanical point of view, the structure aims to co-

ordination and puts back on two contradictory

orders, namely, on a side the differentiation which

consists in simplifying the activity and, on the other

the integration which aims to connect it. The

efficiency of a structure depends on balance to find

between these two orders. In RE, the approach by

the existing structures is our first entrance point.

This first approach does not say anything on the

way in which the actors make function these

structures. All occurs in the mechanical approach as

if the structures functioned all alone. However, the

operation of a structure of organization depends on

the behavior of the actors. It is them which, through

their actions, produce these structures and contribute

to transform them. But this “acting together” is far

from natural... It however puts in the presence of the

actors with various interests which must together

manage their interdependencies to make "turn" the

organization. This idea brings us to the second

dimension of our model, in which the analyst will

try to understand the way in which the actors grasp

the existing structure, operate with it and transform

it to regulate their necessary co-operation. (Crozier,

1963; Crozier, Friedberg, 1977; Friedberg 1988;

Friedberg 1993).

The third dimension intends to understand the

principles and the values on which the actors support

their actions and decisions in organization.

According to Boltanski and Thevenot (1999), the

problem of acting together in an organization is not

only a problem of co-operation between various and

personal interests but it is also a problem of

convention to define between various

representations of values and principles that should

guide the actions and decisions in organization. With

this third dimension, it is the loyalty of actors which

is questioned or the way they can find and define a

community of values, legitimate for each of them

and giving an acceptable meaning of their work.

3 FROM THE CONCEPTUAL

FRAME TO THE METHODS

PROPOSALS

In this section, we will present the research methods

that we use to describe the domain and to define the

requirement in HEDs. The initial system

requirements must allow patients to be tracked

continuously from first contact to the time they are

discharged, and also allow data to be shared among

several responders who may be miles from the

scene.

The methodology used has two main

characteristic features: first of all, it is multi-

disciplinary and combines concepts and techniques

from two disciplines, namely Organizational

Sciences and Software Requirements Engineering;

and secondly, it is participative and aims at strongly

involving the actors concerned about the

computerization of the organization.

To collect the relevant information for the RE,

we chose to follow the principles of qualitative

research based on case studies as defined by Yin

(1990) : because it places emphasis on the dynamics

at work in a fully contextualized perspective. This

focus is justified by the complexity of the HEDs

fields.

We proceeded to two periods of observation at

each site in order to better capture the structure and

the dynamics of action in the three HEDS. This first

range of information was completed by interviews

with relevant actors (nurses, doctors and

administrative staff) in each of the considered

HEDs.

Let us explain now the method followed to apply

our model to RE in the 3 HEDs.

3.1 The organizational context: the

co-ordination approach

The first dimension of the analysis aims to

understand the current state of the organization, in a

pure mechanical analysis.

WHAT CAN ORGANIZATIONAL ANALYSIS GIVE TO REQUIREMENT ANALYS? Developing an Information

System in Hospital Emergency Departments

391

To understand the structure of the HEDs, we

have initially to analyze their external context in

order to locate the various pressures and constraints

coming from their environment. Then, we have

focused on the objectives which structure these 3

HEDs. To do that, we have applied the typology of

Mintzberg concerning goals.

To analyze the organizational structure, we

identified principles of the division of tasks,

specialization's and the system of roles, and we

attempted to determine the organigram of the

organizational unity analyzed. In the same manner,

we paid quite close attention to mechanisms of

integration. Such mechanisms can be of many kinds,

involving existing hierarchies, norms and

procedures, values.

At the end of this stage, we were able to proceed

to an initial organizational diagnosis of the structure

and dynamic of the organizational units under

consideration.

An organizational diagnosis of this existing

situation aims at evaluating how satisfactorily the

identified objectives are met. This led us to the

identification of certain functional problems linked

to the handling of information and the management

of interdependent relationships between various

emergency response actors, having to do with

information.

This initial diagnosis also allowed us to build

different scenarios that illuminate the extent of the

field of possibilities regarding the future information

system.

3.2 The map of actors and their

interdependence: the co-operation

approach

The second dimension focus on actors and the way

they manage their necessary co-operation.

Depending on the particular information system

one is attempting to design, the identification of

actors is done in a contingent manner, through

determination of the various parties who are to

participate in the design of the system and its future

use.

We have therefore set up what we call a map of

actors, their identification and their relationships

which connect them. We were particularly interested

in exchanges of information between actors, since

this is an essential and important power resource

(Crozier, Friedberg, 1977). The nature of such

information and the mode of exchanging it were

noted in specific observations.

After having identified the actors and their

exchanges of information, we have focused on the

way in which each one perceived its role, its

resources and its interests in this network of

interdependences. We also considered the way in

which the actors perceived the role of the future IS,

expectations and resistance's to which this one gave

place. This collect of information can only be

accomplished through in depth interviews of actors

involved in the system being constructed.

Actor’s interests in a situation in which a new

information system is about to be introduced can

never be predicted because the situation is always

quite fluid. The basic idea is that these interests do

not necessarily coincide with those of the

organization such as these might be identified

through a mechanical analysis such as was presented

above.

Collective action always depends on

participation by members, and this participation is

always negotiated through leader, even implicitly.

The terms of this kind of negotiation are exactly

what we were trying to identify.

The understanding of those terms allowed us to

narrow down the field of possibilities identified in

the initial diagnosis, yielding specifications which

were certainly less optimal than those which would

have been produced by the mechanical analysis,

though more satisfying or practical regarding the

actors and their positions.

3.3 The “common higher principle”

identification: the cultural

approach

With the cultural approach we wanted to take into

account the forms of justification (Boltanski,

Thevenot, 1999) which are mobilized by actors in

order to legitimate their actions and their perceptions

of the information system which was being

constructed. For this, we based ourselves on the

typology of “ Economics of Worth”. We sought to

identify in the 3 HEDs. We founded on

- the “common higher principle” held to by

those at the scene, principle which could

serve as a basis for constructing agreements

regarding the the system to be designed ;

- and on, the “states of worth” which

characterize what is “the greater” or “the

smaller” with regard to the higher principles

which were identified. These permanent

states allowed us to determine what is

legitimate within the organization but also

regarding the future IS.

We obtained those data from specific documents,

as rules and charters defining the missions and the

ICEIS 2005 - INFORMATION SYSTEMS ANALYSIS AND SPECIFICATION

392

codes of conduct to be applied in the HEDs. But this

stage of the analysis can also be assisted by

discussion sessions with the actors so that changes

which are acceptable can emerge from dialogue. Our

recourse to a symbolic level allows us to understand

the norms and values which help to regulate the

behavior of the HEDs’professionals. It also allows

us to test different possible ways of constructing the

IS, these possibilities having different meanings and

legitimacy for various actors.

4 CONCLUSION

This article underlines the interest of bringing

together Organization Sciences and Requirements

Engineering to build an information system which is

a success i.e. which is adapted to users' expectations

and who is finally used.

If the approach seems profitable, it remains to

convince the world of the IS in developing some

formal tools able to represent those intangible

dimensions of the organizations.

REFERENCES

Boehm, B., 1981. Software Engineering Economics,

Prentice-Hall.

Boltanski, L., Thevenot, L., 1999. De la justification. Les

économies de la grandeur. Gallimard. Paris.

Crozier, M., 1963. Le phénomène bureaucratique, Seuil.

Paris.

Crozier, M., Friedberg, E.,1977. L’acteur et le système.

Seuil. Paris.

European Software Institute, 1996. ESPITI: European

User Survey Analysis, Retrieved October 10, 2004

http://www.esi.es/ESSI/Reports/All/11000/Download.

html

Friedberg, E. 1988. L’analyse sociologique des

organisations. Privat. Paris.

Friedberg, E., 1993. Le pouvoir et la règle. Dynamiques de

l’action organisée. Seuil. Paris.

Galbraith, J. R., 1973. Designing complex organization.

Reading, Mass. : Addison-Wedley.

March, J.G. & Simon, H., 1958. Organization. Wiley.

New York.

Mintzberg, H., 1979. The structuring of Organizations : a

synthesis of the research. Englewood Cliffs. NJ :

Prentice-Hall.

Orlikowski, W.J., 1992, "The duality of technology :

rethinking the concept of technology in organizations",

Organization Science, vol. 3, n°3, pp. 398-427.

Orlikowski, W.J. & Barley, S. R., 2001, "Technology and

Institutions : What Can Research on Information

Technology and Research on Organizations Learn

from Each Other ?", MIS Quarterly, vol. 25, pp. 146-

165.

Standish Group Int'l, 1994, The chaos report, Retrieved

October 10, 2004

HTTP://WWW1.STANDISHGROUP.COM/

Thompson, J. D., 1967. Organizations in actions.

McGraw-Hill. New York

Yin, R., 1990. Case Study Research : Design and

Methods. Sage. London.

Zan, S., Ferrante, M., 1996. Le phénomène

organisationnel. Paris, L’Harmattan.

WHAT CAN ORGANIZATIONAL ANALYSIS GIVE TO REQUIREMENT ANALYS? Developing an Information

System in Hospital Emergency Departments

393