STRUCTURING MEDICAL AGILITY

Christoph J. Stettina, Lucas P. J. Groenewegen and Bernhard R. Katzy

CeTIM / LIACS, Leiden University, Niels Bohrweg 1, 2333 CA Leiden, Netherlands

Keywords:

Clinical pathways, Process modeling, Medical teamwork, Knowledge work, Self-adaptation, Documentation,

Dynamic consistency.

Abstract:

Technology is omnipresent in intervention rooms, potentially having enormous impact on workﬂows and on

ﬂexibility. Despite ICT systems’ more ﬂexible support of medical protocols, the styles of collaborating do not

really match. Intervention support systems are as yet stand alone and not sufﬁciently interoperable. Taking an

ICT perspective, we seek to understand medical work, how to view the way it is organized and how to improve

its integration with intervention support systems, driven by highly dynamic coordination models embedded in

the framework of clinical pathways.

1 INTRODUCTION

New technologies enable new possibilities in medical

operating rooms. They make interventions noninva-

sive and less risky. Technology in fact is omnipresent

in intervention rooms, having enormous impact on

workﬂows and their ﬂexibility. However, despite im-

provements in medical support systems and in medi-

cal protocols the styles of collaborating do not match.

Intervention support systems are as yet stand alone

and not interoperable. They lack the support for col-

laboration agility in the medical domain. CT images

from different vendors require different viewing soft-

ware; they regularly are being transferred on portable

media. Complex patient data rendering requires an

approach tailored towards relevant caregivers, either

present at intervention or being elsewhere.

Within the EDAFMIS project the authors investi-

gate the introduction of state-of-the-art medical inter-

vention support systems in medical operating rooms.

A minimal intervention cockpit is being developed to

support automation and navigation of noninvasive in-

terventions in line with a new generation of imag-

ing systems. The project targets at real-time inter-

operability and user interaction, thereby ﬂexibly inte-

grating medical skills with systems rendering patient

data. Within the part of the project described here, we

want to grasp ﬂexibility in medical work, how it is or

could be organized in view of its integration with ICT,

driven by highly agile coordination models embedded

in clinical pathways.

2 PROCESS MODELING AND

MEDICAL WORK

During a medical intervention, several specialists are

involved in different roles; also, various views on pa-

tient data and on treatment workﬂow are needed. This

is knowledge work, requiring new forms of organiz-

ing. However, the standard concepts of work come

from bureaucratic and hierarchical organizations in

the industrial era and little agreement exists on how

post-bureaucratic organizing could be conceived.

As opposed to industrial work, knowledge work is

valued for the ability to interpret information instead

of performing manual labor. Knowledge work can

especially be found in domains such as health care,

law, education, science or engineering. Information

technology increasingly allows physical operations to

be performed remotely, thus shifting manual work to-

wards knowledge work, done by people governing

machines. This requires new approaches structuring

and studying knowledge work.

Medical doctors and nurses do need agility. Their

tasks include collecting and analyzing patient data in

order to formulate and provide a corresponding treat-

ment, while working under expanded responsibilities

and strict time constrains. Medical agility is difﬁcult

to predict when developing clinical pathways. Quick

reaction to changing information and smooth self-

adaptation is obviously crucial in a life saving envi-

ronment and should be embodied in system as well as

in team processes. To meet clinical effectiveness, ad-

614

J. Stettina C., P. J. Groenewegen L. and R. Katzy B..

STRUCTURING MEDICAL AGILITY.

DOI: 10.5220/0003276406140617

In Proceedings of the International Conference on Health Informatics (HEALTHINF-2011), pages 614-617

ISBN: 978-989-8425-34-8

 2011 SCITEPRESS (Science and Technology Publications, Lda.)

justing has to be done smoothly, quickly and with-

out quiescence. In view of achieving ongoing and

smoothly changing medical collaborations, we want

to apply a new ICT-approach in self-adaptive coordi-

nation.

EDAFMIS aims at extending intervention ﬂexi-

bility by integrating patient speciﬁc data and profes-

sional knowledge with an interoperable visual system.

It aims at connecting medical workﬂow and special-

ist skills to the intervention support system. Here we

aim at bridging the gap between medical pathways

and ICT in terms of structured agility. Facing the

challenges in medical work, we have decided to in-

vestigate the applicability of a highly dynamic coor-

dination modeling language in the setting of clinical

pathways in care planning.

2.1 Clinical Pathways

Clinical pathways ﬁrst emerged in the 1980s in the

United States aiming at organizational efﬁciency in

hospitals by uniforming lengths of stay, standardiz-

ing treatment packages and thus enabling predictable

costs (Zander, 1991).

Pathways usually consist of several forms com-

bining the nursing care plan with medical notes, and

some sort of a process diagram visualizing the pro-

cess. The diagrams provide a global view, while the

loose forms guide the respective medical journey and

embed the reports. Implemented through broader na-

tional agendas, pathways aim to improve efﬁciency,

quality and local ﬂexibility (Pinder et al., 2005; DOH,

1997) by continuous improvement based on best prac-

tice while reducing variations (Campbell et al., 1998).

There are two main approaches of pathway proto-

cols in terms of their scope, see Figure 1. From the

broader organizational perspective pathways deﬁne

content such as discharge planning, nutrition man-

agement, pain management and patient education.

Taking the clinical perspective, pathways deal with

the sequencing and timing of care, specifying each

step of a medical intervention (Hunter and Segrott,

2008). From a modeling point of view pathways aim-

ing at the clinical scope are descendants of Critical

Path Methods (CPM) and Program Evaluation Re-

view Techniques (PERT) and Gantt charts (Zander,

1991).

Literature reveals quite some discussion regarding

different pathway approaches and how they should

be designed. Apart from the organizational scope of

pathways and their tightly aligned goals, it is most

notably the level of ﬂexibility in the approach being

discussed, as represented in Figure 2.

We understand such recent variants in path-

Figure 1: Different Scopes of Clinical Pathways.

Figure 2: Aspects of Clinical Pathway Implementations.

way literature regarding nomenclature and deﬁnition

(Hunter and Segrott, 2008; Vanhaecht et al., 2006) as

a pointer at lacking standardization. Within this re-

search we aim to bridge the gap between the existing

medical pathways and ICT solutions, to improve our

understanding of possibilities for integration. Thus,

we hope to contribute to protocol improvement by

incorporating ﬂexibility into the skeleton of clinical

pathways.

2.2 Coordination and Self-adaptation

Dynamic adaptation, consisting of interactive, usu-

ally distributed components, heterogeneous and with

a varying conﬁguration, obviously adds signiﬁcant

complexity to the overall system. Dynamic adaptive

systems (DAS) thus must be safety-critical: no fail-

ure is to be accepted, since it could result in loss of

life. In literature adaptation is viewed conceptually

as a three-layered architecture (Kramer and Magee,

2007; Garlan and Schmerl, 2002).

As many systems today, medical equipment is af-

fected by dynamic changes in its operational environ-

ment. Such systems cannot be simply shutdown to

be changed, updated or upgraded and restarted again.

This is particularly important for a live saving envi-

ronment in which adaptation has to be done smoothly,

quickly and without quiescence to support ongoing

collaboration and meet clinical effectiveness. We ar-

STRUCTURING MEDICAL AGILITY

615

gue that clinical pathways as currently formulated, do

not allow for agility needed within complex medical

interventions supported by ICT; we however conjec-

ture, such pathways can be used as a process skeleton

suited for being enriched with agility.

The coordination modeling language Paradigm

(Andova et al., 2010), as a possible approach, ad-

dresses coordination of collaborating components in

terms of dynamic constraints. Its component McPal

(Andova et al., 2009) allows the addition of new be-

havior, and, subsequently, gradually adapts the sys-

tem dynamics without quiescence. We would like to

model agile patterns observed in medical teamwork in

this manner,thus enriching the existing process skele-

ton as provided by organizational pathways.

3 STRUCTURING AGILITY IN

MEDICAL INTERVENTIONS

As outlined in the introduction, the goal of EDAFMIS

is the development of a well-integrated ICT solution.

This study contributes to the integration of the ICT

system into medical practice and, more speciﬁcally,

to the agile coordination of a medical team and its

ICT-support. Thus, this research seeks to understand

medical work: how to view the way it is organized

and how to improve its integration with state-of-the-

art medical equipment.

There is ongoing discussion, calling for new con-

cepts regarding working in ﬂexible conformity to the

knowledge dynamics. There is little agreement on

how post-bureaucratic organizing should look like

and we start looking back at work studies of Scientiﬁc

Management (Taylor, 1911) and of detailed ethno-

graphic studies of work (Barley and Kunda, 2001).

Knowledge work is different from Scientiﬁc Manage-

ment where every step of a worker can be measured

according to quantitative criteria. To circumvent difﬁ-

culties in registering and measuring knowledge work

rather strictly, we investigatethe suitability of a highly

ﬂexible coordination modeling language.

Process models for automation of work processes

could be a possible direction for our research. Our

particular goal is understanding ﬂexible knowledge

work were smoothly changing collaborations remain-

ing ongoing. In addition this should be integrated

with likewise smoothly changing as well as ongo-

ing ICT support. Process models and ICT are natu-

ral partners, providing new opportunities to support

knowledge work. Their common strengths, however,

have not yet been fully exploited in practice (Giaglis,

2001). We aim at contributing to research how to

structure knowledge-oriented medical work and its

ICT support, while keeping ﬂexibility via dynami-

cally adaptable coordination models applied.

In the remainder of this section we explain the re-

search method we have chosen to investigate the use

of highly dynamic process models in the medical do-

main.

3.1 Research Method

In order to understand the co-evolutionary impacts of

technology on human teamwork practice, a system-

atic approach needs to be incorporated into this re-

search, capturing human behavior in clinical settings

and its connection to ICT (Morrison et al., 2007).

As it is difﬁcult to predict all possible events within

the initial development phase and as ideas materi-

alize within ongoing development, we organize our

research in an iterative manner. Thus we propose

an exploratory study of ﬁve phases embedded in a

case study design: Orientation, Investigation, Iden-

tiﬁcation, Modeling and Veriﬁcation. We outline the

phases brieﬂy.

Orientation: As a ﬁrst step we discuss the problem

domain. A literature review is conducted to review

the current knowledge. It is important as a prepara-

tion and it helps us to stay focused during the next

steps. Hereby we reﬂect on the current literature in a

structured manner.

Investigation: To study human teams and their in-

teraction with technology different sociological mod-

els and theories have been developed. Literature sug-

gests ethnographic and other qualitative methods for

detailed studies of work. Due to the exploratory na-

ture of this research, we conduct participant observa-

tion accompanied by semi-structured interviews, in

view of exploring the medical work domain on site,

particularly its dynamics in intervention rooms.

Identiﬁcation: In this step we prepare and analyze

the data in order to be able to identify re-occurring

routines. After having cleaned the data, we use it to

ﬁnd appropriate groupings and patterns.

Modeling: We research dynamic coordination

models within a domain governed by tacit knowledge,

and generally not well-aligned to ICT. Concerns of

medical personnel need to be addressed (Lapointe and

Rivard, 2005). Readability of models and ﬂexibility

and self-adaptation as perceived by the medical per-

sonnel, is very important. Here we investigate to what

extent Paradigm can be used or has to be extended.

Veriﬁcation: The process models we develop, are

presented within focussed group sessions of medical

specialist and interviews. The insights gained will be

used to re-evaluate and to improve the models. Possi-

bly, formal analysis might be carried out in combina-

HEALTHINF 2011 - International Conference on Health Informatics

616

tion with Paradigm, thus contributing to veriﬁcation

even further.

4 CONTRIBUTIONS

We contribute to the quality of medical teams by ﬂex-

ibly integrating medical support systems into nonin-

vasive intervention activities.

Medical pathways for noninvasive interventions

exist today and we extend their organizational skele-

ton with more detailed, but ﬂexible descriptions, as

provided by the Paradigm-McPal approach for self-

adaptive coordination. Thereby we follow the itera-

tive set-up of an ethnographic study, to improve our

understanding of applying highly agile coordination

embedded in clinical pathways. We conduct a de-

tailed study of medical work and use recorded rou-

tines to gather data about and acquire insight into

structuring medical knowledge work. Our ﬁndings

will be used to reﬁne the coordination approach to-

wards highly agile medical team work, to be carried

out in conformity to relevant clinical pathways and

supported by well-integrated ICT systems.

In addition, this is related to coordination of

knowledge work in general, so it will lead to new

insight into knowledge work. We consider medi-

cal workﬂows and clinical pathways as a good em-

pirical case and starting point to address highly dy-

namic knowledge work modeling. By establishing

a bridge-head for understanding and supporting ﬂex-

ibility within medical team work, we see great op-

portunities for generalizing such insight towards non-

medical ﬁelds.

ACKNOWLEDGEMENTS

This research has been kindly supported by the

EDAFMIS project in the framework of ITEA2.

REFERENCES

Andova, S., Groenewegen, L. P. J., and de Vink, E. P.

(2010). Dynamic consistency in process algebra:

From Paradigm to ACP. Science of Computer Pro-

gramming. 45pp, to appear.

Andova, S., Groenewegen, L. P. J., Staﬂeu, J., and de Vink,

E. P. (2009). Formalizing adaptation on-the-ﬂy. Elec-

tron. Notes Theor. Comput. Sci., 255:23–44.

Barley, S. R. and Kunda, G. (2001). Bringing work back in.

ORGANIZATION SCIENCE, 12(1):76–95.

Campbell, H., Hotchkiss, R., Bradshaw, N., and Porteous,

M. (1998). Integrated care pathways. British Medical

Journal, 316(7125):133–137.

DOH (1997). The New NHS: Modern, Dependable. HMSO,

London.

Garlan, D. and Schmerl, B. (2002). Model-based adaptation

for self-healing systems. In WOSS ’02: Proceedings

of the ﬁrst workshop on Self-healing systems, pages

27–32, New York, NY, USA. ACM.

Giaglis, G. M. (2001). A taxonomy of business process

modeling and information systems modeling tech-

niques. International Journal of Flexible Manufac-

turing Systems, 13(2):209–228.

Hunter, B. and Segrott, J. (2008). Re-mapping client jour-

neys and professional identities: a review of the liter-

ature on clinical pathways. International Journal of

Nursing Studies, 45(4):608–25.

Kramer, J. and Magee, J. (2007). Self-managed systems:

an architectural challenge. In FOSE ’07: 2007 Fu-

ture of Software Engineering, pages 259–268, Wash-

ington, DC, USA. IEEE Computer Society.

Lapointe, L. and Rivard, S. (2005). A multilevel model of

resistance to information technology implementation.

MIS Quarterly, 29(3):461–492.

Morrison, R., Balasubramaniam, D., Kirby, G., Mickan,

K., Warboys, B., Greenwood, R., Robertson, I., and

Snowdon, B. (2007). A framework for supporting dy-

namic systems co-evolution. Automated Software En-

gineering, 14:261–292.

Pinder, R., Petchey, R., Shaw, S., and Carter, Y. (2005).

What’s in a care pathway? Towards a cultural car-

tography of the new NHS. Sociology of Health and

Illness, 27(6):759–779.

Taylor, F. W. (1911). The Principles of Scientiﬁc Manage-

ment. Harper and Brothers, New York.

Vanhaecht, K., Bollmann, M., Bower, K., Gallagher,

C., Gardini, A., Guezo, J., Jansen, U., Massoud,

R., Moody, K., Sermeus, W., Zelm, R. V., Whit-

tle, C., Yazbeck, A., Zander, K., and Panella, M.

(2006). Prevalence and use of clinical pathways in

23 countries-an international survey by the european

pathway association. Journal of Integrated Care Path-

ways, 10:28–34.

Zander, K. (1991). What’s new in managed care and case

management. The New Deﬁnition, 6:1–2.

STRUCTURING MEDICAL AGILITY

617