This chapter explored key decisions that designers and developers of a biosurveillance system should make before system acquisition and/or development. This set of considerations are in addition to those discussed in Chapters 32 (Standards) and 33 (Architecture).

Perhaps the most critical decision from the perspective of project success and cost is whether an organization should attempt to develop and operate a biosurveillance system itself, use a professional ASP, or join forces with other organizations (e.g., health departments) to form a regional system. Because most biosurveillance systems must be available 24 hours a day seven days a week, a physical plant that is capable of supporting reliable operation is necessary. If uninterrupted operation of the system in the face of complete loss of the facility is a requirement (the Wall Street Standard), then a second site is mandatory. Because of cost considerations, these requirements should bias decision makers toward regional joint projects or use of ASP service providers.

The choice of software is also important to project success. De novo development of software is to be avoided if possible, owing to the cost and risk of failure. When selecting software, it is important to review the track record of the provider as well as the functionality of the actual software, not a prototype. Decisions should be biased toward open source because of expected lower cost and the expectation that you will modify the software considerably over time. Although ideally software products are so well designed that they can be configured by end-users to meet their needs without the need for custom programming, software for biosurveillance is still so immature that no product will be that configurable. If you elect to use a product for which source code is not available, pay attention to whether the vendor will modify the software to your specifications in a timely and affordable way, and consider your options should the provider stop supporting the product.

Additional key decisions relate to the design of specific elements in a biosurveillance system such as methods of data acquisition and exchange and persistent storage of data. Many data transmission methods exist for collecting data, and you may have to use more than one to be able to work with different data providers. Because of a strong direction that industry is taking, we expect data providers to move toward Web services as the prime method for data transmission. For data storage, the sizing of a DBMS is a commonly overlooked area, as is the need for caching to optimize the speed at which data can be retrieved by end-users and programs. An almost universally overlooked requirement is the need to obtain identifiers with routine surveillance data to enable look-back. This seemingly small detail involves considerably technical and administrative attention owing to the complex interplay of individual and societal interests and the technical work that data providers may need to perform to enable look-back. Finally, data utilities exist, and you should take into account their data transmission methods when designing or acquiring your system.

Was this article helpful?

0 0

Post a comment