Policy modeling to support administrative decisionmaking on the New York state HIV testing law

A recent New York law requires medical providers to offer HIV tests as part of routine care. We developed a system dynamics simulation model of the HIV testing and care system to help administrators understand the law's potential epidemic impact, resource needs, strategies to improve implementation, and appropriate outcome indicators for future policy evaluations once postlaw data become available. Policy modeling allowed us to synthesize information from numerous sources including quantitative administrative data sets and practitioners' content expertise, structure the information to be viewed both numerically and visually, and organize consensus for decisionmaking purposes. This case illustrates how policy modeling can provide an integrated framework for administrators to examine policy problems in complex systems, particularly when data time lags limit pre--post comparisons and key outcomes cannot be measured directly.