|
|
Bulletin of Applied Computing and Information Technology |
 |
 |
 |
 |
 |
|
SoDIS in London |
|
02:02 |
|||
|
Lesley Smith Smith, L. (2004, June), SoDIS in London. Bulletin of Applied Computing and Information Technology Vol. 2, Issue 2. ISSN 1176-4120. Retrieved from ABSTRACTIn November 2003, an opportunity arose to present SoDIS to a group of medical researchers at the Royal Free Hospital in London. This article provides some background to the presentation and a brief discussion of the event and its KeywordsSoDIS, Royal Free Hospital, medical ethics. 1. BACKGROUNDThe Oncology Department at London’s Royal Free Hospital is at the forefront of global cancer research. As one of several teaching hospitals associated with University College London, the Royal Free supports a range of cancer research groups covering topics such as Drug-DNA Interactions, Targeting and Imaging, and Bioinformatics. The research groups focus on developing new therapies that target cancer with fewer side effects than conventional treatments. The Bioinformatics team, led by Dr Sylvia Nagl, uses mathematical and computer tools to model cellular pathways and tumour biology, with the goal of identifying potential targets for antigen therapy. Dr Nagl’s interests also include ethical and social perspectives of cancer bioinformatics, so when SoDIS was mentioned in casual conversation while we discussed my plans to visit London she was immediately interested. Familiar with the concepts of medical ethics, a tool that exposed ethical issues in a software engineering context was intriguing. The following weeks saw a plan hatched to present SoDIS to a seminar of interested staff and students of the Oncology Department in December 2003. 2. SoDISDon Gotterbarn and Simon Rogerson developed SoDIS in response to a need for a tool that could provide an effective stakeholder analysis during software development. It has been applied in many undergraduate projects with beginning students, and has also been used successfully in commercial applications. The use of SoDIS in a medical environment was new and we were very interested to have the opportunity to discuss its applications with the group at the Royal Free Hospital. 3. PLANNINGOn my arrival in London I was delighted to make the acquaintance of the renowned Professor Simon Rogerson (from the Centre for Computing and Social Responsibility at de Montfort University in Leicester). This name had been familiar to me from the error message on my SoDIS screen suggesting that I contact him to obtain a legal license for the product. Over a Leicester pub lunch we discussed options for the proposed seminar, and decided to present a 45-minute introduction, followed by questions and a workshop session for enthusiasts. One issue debated was the suitability of the London Ambulance Service case study, widely used by both Simon and Don in SODIS workshops. Could we assume that the London medical community was familiar with the outcomes of this software development disaster from the early 1990’s? This would lessen the impact of the SoDIS analysis, which requires an open mind on the possible outcomes of the case. A few subtle enquiries suggested that they had no more than a vague awareness of the issues – a couple of staff remembered that there had been a problem with the ambulance service – some sort of political row, perhaps. 4. PRESENTATIONPresenting to the research group posed quite different challenges to those faced in teaching SoDIS to a second year software engineering class. The software engineering students first need to be convinced of the need for ethical practices, and gradually be introduced to the concept of responsibility for ethical risks in a software development environment. In contrast, the Royal Free group was well aware of the ethical dilemmas posed by the medical issues in their field of work. The fusion of ethics with software development was a novel concept, which generated a good discussion. Simon presented the issues raised in a recent SoDIS analysis of electronic voting options in the UK, in which security of data and confidence in the outcomes were major concerns. The researchers suggested parallel concerns in the area of electronic patient records, which is currently being trailed at Royal Free Hospital. The broad definition of stakeholders, as used in SoDIS, was also considered. Dr Nagl reported back that “they particularly valued the idea of being prompted to think very widely about stakeholders and potential consequences.” Although the software developed by the Bioinformatics team was typically small-scale models, the information generated could ultimately be used to guide clinical practice. If errors had been inadvertently been introduced in the original model, the consequences could be devastating. Dr Nagl elaborates: “The most immediate challenge is the biological complexity of the processes and the potential risks of modeling from insufficient understanding, oversimplification or overlooking important parameters.” A thorough consideration of the immediate and ongoing effects on stakeholders would result from SoDIS analysis of the modeling tools. The SoDIS analysis of e-voting had produced a large number of concerns, and Simon introduced a process that he is developing to manage concerns after an analysis. The concepts can be grouped into clusters on the basis of similar themes. Each cluster then can be addressed with appropriate risk management strategies, rather than each concern needing to be solved separately. An ongoing dialogue will ensure that identified concerns are not lost. 5. WORKSHOPThe workshop session was run with a smaller group, consisting mostly of the members of the bioinformatics computing group. These students enthusiastically installed the SoDIS software and launched into the issues posed by the LAS case. We observed animated discussions on the traffic problems of inner London and the risks of assuming perfect data gathering in an imperfect world. Their local knowledge of London conditions assisted with their decisions – one student observed that an ambulance staffed with ethnic minorities would be at risk if it answered a call to Brixton (a suburb notorious for racial tension). One student, who specializes in visualization software and is very experienced in UML, was interested in exploring a visual version of SoDIS. He drew a parallel between the stakeholder in SoDIS and the actor represented in a UML activity diagram. He felt that if the concepts could be expressed visually, it would make SoDIS a simpler tool to learn and use. One notable impression of the afternoon was the similarity between the responses of this group of students and my own Dunedin students. The same concerns were identified, similar questions asked, the same complaints made about the interface. 6. FEEDBACKThe SoDIS lecture and workshop generated a lot of interest in the Department, and was discussed further in a staff meeting. "Researchers thought that a decision support tool of this kind would be very valuable for their research… We also discussed briefly the ways in which biomedical software needs to be sensitive to different cultural backgrounds and values, a multicultural dimension, and whether this could be incorporated in a tool like SoDIS. This is a very involved and complex issue, and I think this would be important to explore further". (Nagl, personal communication). As an outcome of the seminar an ongoing dialogue has been established with the Oncology Department of the Royal Free Hospital. This will bring benefits in several areas - the possibility of developing a customised SoDIS tool has been discussed, and the sharing of medical and ethical concerns. 7. ACKNOWLEDGEMENTSThanks are due to Dr Sylvia Nagl for arranging the presentation at the Royal Free Hospital, and to Prof. Simon Rogerson for his support and input. 8. REFERENCESUCL (2003). Department of Oncology. Retrieved March 02, 2004, from http://www.ucl.ac.uk/oncology/. Copyright © 2004 Lesley Smith |
|||||
|
|
|
|
|
|
|
|
Copyright © 2004 NACCQ. All rights reserved. |
|||||
