Design for Collaboration in Health Care: Experiences from Highly Specialized Surgical Care in Sweden Jeremiah Scholl

Kristina Groth

Susanne Bødker1

Norwegian Centre for Integrated Care and Telemedicine

CSC, Royal Institute of Technology and CLINTEC, Karolinska Institutet

Department of Computer Science Aarhus University

University Hospital of North Norway

[email protected]

[email protected]

[email protected] ABSTRACT Medical Information Systems often need to be custom designed to fit the organization where they will be implemented. Participatory Design (PD) is a well known method for eliciting the user input that is necessary during this process. Recently it has been suggested that PD as it is often practiced falls short of providing the necessary level of user input that is needed to meet the evolving needs of users over time. It is suggested that we refocus and look at design itself as a socio-technical process within healthcare organizations. By building “socio-technical systems that design themselves” we will better meet these evolving needs. In this paper we present a case study of design as it is practiced at a gastro-surgical department at a University hospital in Sweden. The experiences of the department are used as a framework for discussing this issue and its implications for the CSCW/HCI community.

Categories and Subject Descriptors H.5.3 Group and Organization Interfaces: Computer-supported cooperative work

General Terms Design, Human Factors.

Keywords Participatory Design, Medical Informatics

INTRODUCTION Health care organizations are characterized by a high level of complexity and use of complex work practices [1]. Information Systems for such organizations often need to be custom designed [2]. Participatory Design (PD) is a well known approach for incorporating high levels of user input in these (and other) contexts [3]. A noted challenge with Medical Information Systems (MIS) however is how to incorporate the necessary level of user input needed over time to meet the evolving needs of healthcare organizations [4]. PD is generally reliant on external designers to lead the design process and “…the capacity for external designers to meet all the evolving needs of those inside” seems inadequate in many situations [4]. This capacity includes methods as well as a technical preunderstanding of future technical possibilities and current technical problems [5]. A perspective that has been proposed for dealing with this issue is to start considering MIS design itself as a socio-technical process. Using this perspective socio-technical design experts would then shift their focus and “stop designing technology for people, and [instead] build sociotechnical systems that design themselves” [4]. This would allow

healthcare organizations to better handle the design and iterative redesign of systems in a more independent way. Many tools (theories, artifacts, processes etc.) developed within the HCI/CSCW community may support this process. PD for example offers tools such as cooperative prototyping that have been developed to give users high levels of control over design, and the ability to envision e.g. user interfaces [6]. Adaptability and tailorability of systems for the local context have also been investigated in order to improve understanding of how to give users the possibility to reconfigure systems to meet their own evolving needs over time [7]. Meeting all the evolving needs of health care organizations over time however, will likely require different processes than tailorability of specific systems to the local context. Entirely new ways of organizing work need to be considered for example, and entirely new MIS systems will need to be designed and implemented related to these work practices. Helping healthcare organizations succeed at the range of relevant design tasks seems to create numerous challenges (and opportunities). While these processes are driven by the users, they are more complex, both in terms of technical and organizational issues, than tailorability, and they provide a different role for the professional designer. Knowledge of health care organizations within CSCW/HCI will need to be expanded for example to include issues related to understanding design as a (secondary) health care related work practice. This includes understanding the “decision to design” in these organizations in order to more fully understand unmet and evolving needs [4]. Issues related to developing and maintaining a positive and productive design culture within health care organizations also may need to be explored. This may include reductionist set of design tools that can be communicated and transferred into target organizations. Few studies of design as a work process in healthcare organizations have been conducted. In this paper we present a case study of design as it is performed by a gastro-surgical department at a University hospital in Sweden. The department’s experience related to designing multiple work processes and supporting MIS for at least 5 years are presented. This experience is used to provide examples of evolving needs within the department, challenges in further development of the design process, and implications for future research.

THE CASE The gastro-surgical department is located at Karolinska University Hospital in Stockholm, Sweden. The hospital has approximately 15,000 employees and 1600 patient beds spread out at two different campuses approximately 26 km apart in

1 This work was carried out while the author was on sabbatical at the Royal Swedish Institute of Technology

different sections of the city. The department employs about 80 surgeons at both campuses. 25 of these work with highly specialized care.

Evolving needs and collaboration In 2004 the local county council appointed the department as a centre for highly specialized care for upper abdominal diseases. This resulted in a dramatic increase in the number of patients with difficult and rare cases being referred to the department from hospitals in the surrounding area. Concentrating these cases into a single department provides the advantage of allowing these patients to be treated by a department with significant experience in similar cases, even if the specific case is too rare for a normal hospital to gain this experience. This was a significant structural change and created a number of new needs for the department. Many of these needs are related to collaboration within the department, and between the department and the network of hospitals that refer patients for highly specialized care. One example is that within the department a need to manage a more distributed set of knowledge was created. Different surgeons specialize in different diseases and procedures. Specific surgeons more skilled in specific tasks may need to be called from the operating room for example, so they can assist other surgeons during specific situations. Collaboration between the department and hospitals that refer patients involves a variety of issues that can have a significant impact on the quality of treatment available to the patient. The timing of chemotherapy and quality of diagnostic tests performed at a referring hospital for example may need to be coordinated with the department in order to make sure all the necessary information has been obtained, and the patient is healthy enough, for a surgical procedure to be performed upon arrival at Karolinska.

Evolving work practices and design The department began to design a number of new work practices supported by MIS related to these needs. One such example is a dedicated “Coordinator” role. The role consists of an experienced surgeon (rotated on a three week basis) that works full time coordinating care between the referring hospitals and the department, and taking care of related administrative issues together with the secretaries and nurses at the care planning unit. Another example was the creation of multi-disciplinary treatment and decision conferences. The conferences are conducted over video conferencing (VC) and contain members of different medical specialties from the two campuses, and referring hospitals that have a patient scheduled for surgery in the coming week. Establishing the conferences required an investigation into the appropriate technology and work process. A team was organized to investigate and “own” the overall system design and work process regarding these meetings (a design team). The team included two surgeons, one radiologist, two engineers from the biomedical department and a video communication expert from the project assistance department at the hospital. An iterative design process was developed by the team in 2004. It began with a pilot study using a point-to-point connection between the radiological departments at the two Karolinska campuses. The pilot was used to investigate issues such as how to support the necessary resolution and interactivity for supporting medical imaging, and communication for conducting successful meetings. After six months when the financing on a county

council level was secured (including purchasing VC technology for the participating hospitals), four hospitals within the region were connected. These four hospitals had been conducting surgery of patients with upper abdominal diseases that were now referred to Karolinska. Later on the team developed a region-wide technological infrastructure and moved towards full scale use of the system involving the last two hospitals within the region. This was done in parallel with connecting some hospitals outside the region who now participate in conferences and also refer patients for highly specialized treatment. The department is also involved in design activities to support other work practices. One example is related to a patient monitoring system for the intermediary care ward (IMA). The system automatically and manually gathers patient data and visualizes it through different graphs. A similar application is currently used within the ICU but is generally considered to have usability problems. This caused the department to initiate a process to develop new features into the application with the vendor in parallel with implementing it in the department. The surgeon in charge of IMA has been working closely with a developer at the software vendor in order to deal with this issue. The system also required changes in the work process at the department including installing a computer by each bed. A pilot project was initiated by the IMA manager, where a trolley with a computer was installed at one bed. The manager originally planned to have the nurses keep unstructured notes about their experience. Later he consulted with an HCI expert [j2] and decided to adopt a more structured PD-oriented method of having the nurses keep a diary. Each day they answer a couple of questions regarding the use of the bedside computer, describing how often they used it, what worked well and what worked less well.

Challenges with further development Although the department has undertaken significant changes in order to design new systems that can improve collaboration and the care process there seems to be significant challenges regarding the further development of the care process over time. A key challenge is related to traditions and perceptions in the medical field in general. Many of the staff explained that the department is “unique in the whole world” in the way that it has focused on investigating improved ways to support the care process. While the costs of supporting such collaboration are easy to identify, when looking in from the outside, it is not clear at all what the benefits are. It also seems difficult to communicate these benefits to medical professionals outside the department. One of the most experienced surgeons in the department expressed for example that the Coordinator role is extremely important for ensuring the quality of patient treatment. The same surgeon however was highly skeptical of the role when he was hired. He explained: “When I was hired into this department 3.5 years ago I thought they were crazy. Here they have this Coordinator role where you take an experienced surgeon that could be performing procedures and he ends up with an ear that is big and red from talking on the phone all day.” Over time however he realized the benefits of the role. Even though it requires an experienced surgeon to be kept out of the operating room for weeks at a time he explained that it may provide overall economic benefits. He explained: “Surgery is

expensive, right? Wrong! Surgery is not expensive, complications are expensive. The Coordinator may save money because a strong investigation to make sure everything is taken care of before the patient arrives reduces complications.”

Figure 1 An illustration created to emphasize the connection between advancements in the care process and advancements in surgical care. The head of the department has considered this issue and is working on developing a vision and process for increasing activity around developing the care process. He described this vision at a meeting in the beginning of 2008 at a hospital in Norway. The meeting was conducted to initiate a joint research project between the two hospitals and was attended by researchers, practitioners from two hospitals and representatives from industry. His vision describes the situation regarding surgical care as a circle in two halves (see Figure 1). The lower part of the circle refers to knowledge related to surgical care including surgical procedures and techniques, methods for related treatment, and surgical devices. The upper part of the circle refers to the care situation including the meeting between the patient and the doctor and medical collaboration, and how this can be performed optimally from a medical perspective, from the patient’s perspective, and from a social perspective. The halves of the circle are connected in that new possibilities created by improvements in the care situation open up new possibilities for improved surgical care, and the development of new surgical care knowledge can require further improvements on the care process. He explained that an existing infrastructure and support are in place, including funding sources and years of tradition in research, for generating new surgical knowledge. These traditions include a driving force to arrange meetings, disseminate knowledge and support the global development of the field. Similar methods and traditions also exist within the drug and biomedical industries to develop new methods and technologies. A problem exists with respect to knowledge generation to improve the care situation

however. There is a lack of infrastructure, money and traditions within medicine. This poses a challenge because a gap exists between what may be possible to support, and the resources that exist for investigating and developing this support. Figure 2 shows an illustration of this gap.

Figure 2 An illustration created to describe the gap that exists with respect to needs and resources for improving the care situation. He suggested that to solve this problem we need to decrease this gap and move more resources closer to investigating and developing the care situation (this is illustrated by the vertical green arrow). This will require collaboration over different medical disciplines with the entire care process being turned into a “living lab”. Researchers and developers with experience and competence that can contribute to this process will also need to be moved closer to the health care settings in which the work is performed. Today, the head of the department is focusing on creating a physical place where researchers and developers can meet the clinical work practices.

DISCUSSION One salient feature related to the “decision to design” new work processes, and MIS by the department is that it seems to be a fairly broad and deep process. It takes place on several levels and seems to have broad implications. At the political level the local county council saw a need to redesign the care process in order to better serve patients with rare and difficult cases. This created a socio-technical gap related to new needs for inter- and intradepartment collaboration at the department after it was named a center for highly specialized care. This in turn lead to a number of design related activities impacting the department, referring hospitals and others. Design is also performed by the individual health workers as they seek to contribute to future systems that will impact their day-to-day work. A lack of traditions and infrastructure within the health field for investigating improvements to the care situation also seems to be an issue impacting the design process. This has lead the department to initiate its own efforts to “redesign the system” for these investigations by bringing medical work closer to

1 This work was carried out while the author was on sabbatical at the Royal Swedish Institute of Technology

individuals and organizations with expertise and resources that may help improve this process. This suggests that fairly large changes on a societal level may in fact be necessary in order for the department to fully investigate the issues uncovered during its design process. This in turn may lead to a design-oriented positive feedback loop by requiring further political decisions to be made in order to further support design processes focused on improving the care situation. This has already occurred on at least one occasion, as the county council was convinced to fund the necessary equipment for the creation of VC networks between hospitals that participate in the highly specialized care process. Design as it is performed in the department also illustrates a key difference between design performed by health care organizations and design performed by CSCW/HCI professionals. Design conducted as a work process in health care has the explicit goal of improving the overall situation for the organization performing the design. At its very core however CSCW/HCI is concerned with the development of tools that help designers contribute to systems for people that are inherently different from the designer. These different people are referred to as “users” and terms such as “user-centered design” are adopted in order to emphasize the non self-centered aspect of the design process. Shifting focus to developing practices that help users succeed at “self-centered design” seems to entail a somewhat different perspective, and likely requires the development of a new set of tools within the field. PD for example was created based on the premise that designers run the design process, and elicit participation from users in order to create better design with these users. Selfcentered design on the other hand turns this around, as the design professional then becomes the participant in the users’ design process. In this case an HCI practitioner [j2] has been incorporated into the department’s design process in a fairly adhoc way. A key question for healthcare organizations then becomes if, when and how they should elicit participation from outsiders without losing control over the ability to meet their own evolving needs independently? A similarly obvious and key

question for HCI is what tools and techniques designers should be equipped with in this facility?

ACKNOWLEDGMENTS This research is supported by the Research Council of Norway, grant no. 176852/S10. We wish to thank the gastro-surgical department at Karolinska University Hospital, Johan Permert for reading and commenting on our paper and VINNOVA for funding the Funk-IS and HDViper projects.

REFERENCES [1] Berg, M. Health Information Management: Integrating Information and Communication Technology in Health Care Work. Routledge London, (2003). [2] Grudin, J. Groupware and social dynamics : eight challenges for developers. Communications of the ACM, 37, 1 (1994), 92 105. [3] Muller, M. J. Participatory design: The third space in HCI. The human-computer interaction handbook: fundamentals, evolving technologies and emerging applications, Lawrence Erlbaum, Hillsdale, NJ, (2002), 1051 – 1068. [4] Coiera, E. Putting the technical back into socio-technical systems research. International Journal of Medical Informatics, 76, S1 (2007), S98–S103. [5] Kyng, M. Keynote address, Participatory Design Conference. Chapel Hill, North Carolina, (1994). [6] Bødker, S. and Grønbæk, K. Design in action: from prototyping by demonstration to cooperative prototyping. L. Erlbaum Associates, Hillsdale, NJ, (1992). [7] Trigg, R. H. and Bødker, S. From implementation to design: tailoring and the emergence of systematization in CSCW. In Proceedings of the 1994 ACM conference on Computer supported cooperative work, Chapel Hill, North Carolina, United States, ACM Press, (1994).