9. Various Examples
This section of the class notes considers some topics of current interest, with particular emphasis on ethical issues that are raised.
9.1 Medical Informatics and Bioinformatics
It is interesting to compare the ISHTAR Medical Database Security Guidelines with the material usually found in books on computer security. These very pragmatic guidelines were designed for the administrators of hospital databases in European countries, and are based on real experiences with such databases, often of an unfortunate nature. Encryption is not even mentioned, but several standard software engineering practices are highlighted. This is consistent with the conclusion of our discussion of trust vs. risk, that security as such is not the main issue, and in particular, that encryption is not a panacea.
The paper The Multiple Bodies of the Medical Record, by Marc Berg and Geoff Bowker is a beautiful application of ANT to the work done by medical records in hospitals in the UK national health system (it is also another example of a paper with a bad pun in its title). The paper discusses the work done by the medical record in the production of, and in the multiplicity of, both the human body and the body politic.
A medical record contains at least the following major sections:
It is interesting to contrast the Berg & Bowker paper with the the Data Fusion for the Multi-media Medical Database project of the Fraunhofer Center, which takes an almost totally technology-based approach - the exception is some lip service to meeting its technical goals "in a way that reduces stress, uses the physician's time more effectively, and increases communication with the patient." It is difficult to see how the third goal can be met by imposing yet another layer of computer mediated interaction, and I would predict that this project will encounter (or has already encountered) major problems if its software is deployed in real situations.
Added November 2000: The above paragraph was written in 1999. Sometime during the next year, the project URL was changed to http://www.crcg.edu/projects/medvis/medviswww/mmmdb.html, which is no longer linked from the top level of the institution where it was performed, presumably because the project was not very successful.Although its journalistic style emphasizes some controversial assertions about statistical methods, the reports about the large number of recent failed medical studies in A Bayesian Critique of Statistics in Health, by Robert Matthews, are significant. Actor-network theory (and common sense) would suggest that large pharmaceutical companies are having a significant effect here, because of their huge economic interests in showing positive results.
Ethics of bioinformatics: genomics, cloning, gene manipulation, ..... Safety of medical devices. .....
9.2 Ethics and Communication
Spam; gambling and pornography websites; threats and lies in email and in chatrooms; hate email, hate websites; other bad behavior in chatrooms. .....
One thing that all these issues have in common is that they are in no way new! The fact that a new medium is used for the same bad things that have been going on in other media (such as speech) for millenia is absolutely not a valid reason for shutting that medium down, or even for imposing more draconian regulations on it than on traditional media. Of course, appropriate regulations can improve the situation, just as for traditional media. Moreover, filtering software can be helpful, but it is technological fix that can never be totally successful, and that can be used (abused) in ways that excessively limit free speech.
Data protection standards; ecommerce regulation; ubiquitious computing; cookies; .........