Criteria for initiation of highly active antiretroviral treatments (HAART) in HIV-infected patients remain a matter of debate world-wide because short-term benefits have to be balanced with costs of these therapies, and restrictions placed on future treatment options if resistant viral strains develop. On the other hand, postponing the introduction of HAART may involve a therapeutic opportunity cost if a patient's health is allowed to deteriorate to such an extent of becoming unable to benefit from new treatments currently under development when they become available. We introduce a two period model where period one treatment adoption is an irreversible act with future, but uncertain, consequences. New information, both endogenous and exogenous, arises over time and shapes the conditions surrounding the second period therapeutic decision. A surprising result is that, under conditions that appear close to those surrounding the HAART debate, the magnitude of the feared resistance effect has no effect on leaves the optimal treatment decision as far as it is high enough.