The evolution of computer and communications systems in the past decad
e brings new opportunities for increased efficiency and accuracy of ob
servational studies. In ongoing, large scale research studies, the pro
blem of bridging organizational and analytic methods of the past to mo
dern methods of data structure and control can consume considerable ef
fort. A model for information flow in an observational study is sketch
ed, and the flow is found inherently complex. This complexity and corr
esponding managerial demands are compounded by database file prolifera
tion and by evolution of system hardware and software. Interactive, ne
twork-based database mapping and documentation tools are described as
currently implemented in an SAS-based system for the management and an
alysis of several large epidemiologic studies. (C) 1997 by John Wiley
& Sons, Ltd.