Active Feature Acquisition with Control Actions (AFAC) is Active Feature Acquisition (AFA) [1] with additional control actions that affect the state. While AFA allows us to model the medical testing problem, AFAC additionally allows us to model how medical treatments affect those tests. Despite its practical relevance, methods for solving AFAC problems are surprisingly unexplored. Prior work by Li et al introduced a POMDP formulation of AFAC, but their method was only evaluated on two environments, and their experiments lacked important baseline comparisons with off-the-shelf POMDP planners. Furthermore, their paper and code do not provide all hyperparameters, making reproducibility difficult. In this paper we (i) introduce AFAC as a name for a new style of problems and describe them mathematically as AFAC-POMDPs, (ii) present GERI, an acquisition policy based on the idea of myopic performance improvement, and (iii) perform a thorougher benchmarking of AFAC methods, with reproducible results. References: Li et al: "Towards Cost Sensitive Decision Making" http://arxiv.org/abs/2410.03892 [1]: "A Survey on Active Feature Acquisition Strategies" http://arxiv.org/abs/2502.11067