When examining the impacts of exposure to air pollution on health outcomes, researchers usually carry out "placebo tests" to provide evidence in support of their identification assumption. In general, this exercise targets health conditions seemingly unrelated to air pollution. In this study, we argue that one should proceed with caution when running such falsification tests. If healthcare infrastructure is limited, when we observe health shocks such as those driven by air pollution, the infrastructure needs to be adjusted to meet the increased demand by canceling or rescheduling elective and non-urgent procedures, for example. As a result, even health conditions seemingly unrelated to air pollution may be indirectly affected by pollution.