An information science simulator for epidemic disease

The information science simulator for epidemic disease described below is a first step toward the total use of all relevant information to account for disease spread. The relevant information includes descriptive details over time, to any degree of resolution, about the envirm~nent, disease agents and persons involved in a particular epidemic; the potentially large volume of such detail requires the use of an information system. The relevant information also includes knowledge of the disease agent spread and disease manifestations, i.e., the interactions of agent, host and environment; this theory and associated interrelationships is generally represented algorithmically in the simulator and programmed to operate on the data base of observable details. In operation, the system produces realistic descriptive results which unfold in strict accord with the consequences of the assembled knowledge and detail. In th~s way simulations are obtained to symbol ically describe any aspect such as the health states of each person for each time period. The symbolic results may be summarized numerically or transformed to provide answers or evaluations to specific questions. By producing simulated results which correspond to actual outcome observations, the predictive validity of the system may be assessed by comparison of the abstract simulated results to the real results. Upon validation the system can be used to test or evaluate conditions and factors in disease spread by analytically replacing the real data or knowledge with experimental versions. Thus, information science simulation has the potential of adding a very powerful tool in exploration and analytical experimentation in realms of complex phenomena. Management sci~ence-operations research (MS-OR) modeling and simulation was orginated and has been further developed to provide a quantitative basis for operational decision making (i). Its distinguishing features include a focus on decision making, effectiveness measured by cost and the use of formal mathematical models. This approach has been used in health care delivery research and in such other health areas as finding optimal vaccination strategies (2). The mathematical model, or derivations of it, used in the vaccine studies is widely known as the Reed-Frost model (3). Objections have been raised to the Reed-Frost and related models where predictions are required because they are tautological (4). In the exploratory situations which precede decision making and in which the economics or relative costs are unknown, MS-OR is not especially useful. Certainly new relationships, theory and general knowledge developed by information science simulation could become useable subsequently by MS-OR for exploitation; hence, these two approaches are supplemental and not antagonistic. While the following description will also show their differences, it should be emphasized that the information science goal is understanding; it depends on an information system of details and, while formal mathematical models may be useful, they do not characterize this approach; and, finally, the results are descriptive and symbolic but capable of numerical summarization.