The null findings in the study by Jones et al—evaluating the impact of an early Veterans Affairs oral care initiative on non-ventilator hospital-acquired pneumonia (NV-HAP) across 17 medical centers—cannot be meaningfully interpreted without first establishing whether the intervention was consistently delivered. ^{Reference Jones, Chapman and Ying1} Although the authors acknowledge that oral care data were not captured, this limitation is far more consequential than their discussion conveys. It calls into question a core assumption of the study: that issuing a policy directive translated into actual changes in nursing practice.

This assumption is not supported by the principles of implementation science. Announcing that nurses “should” perform twice-daily oral care does not reliably change practice—particularly for a task historically deprioritized and inconsistently performed. Behavioral change requires structured monitoring, feedback, accountability, and sustained reinforcement. None of these were verified here.

Critically, oral care was not documented in the electronic health record (EHR) in a standardized, trackable way during the study period. Without reliable documentation, it is impossible to determine whether the intervention increased oral care frequency or adherence. The authors are essentially evaluating the effect of an intent to treat—not the treatment itself. A null result under these conditions does not tell the reader whether oral care prevents NV-HAP; it tells us only that an incompletely implemented, unmonitored initiative did not shift outcomes measurable at the population level.

Future studies must incorporate fidelity measurement—EHR-embedded oral care documentation, direct observation, or audit data—as a prerequisite for evaluating clinical impact. A null finding under these conditions tells us only that an unmonitored policy directive did not shift population-level outcomes—not that oral care fails to prevent NV-HAP. Conclusions about the efficacy of oral care as a prevention strategy remain premature until delivery of the intervention can actually be verified.