This manuscript extends the classical models of the value of information to ask whether a hospital's net financial return is ever maximized by a cost-effectiveness analysis of retrospective data when watchful waiting and a full randomized clinical trial are alternative methodologies. The manuscript demonstrates that (1) some small-scale retrospective analyses may negatively affect net income and (2) under some conditions, larger-scale retrospective analyses may maximize net income. The manuscript also suggests that risk aversion increases the value of information and therefore the optimum expenditure on a project evaluation.