IntroductionVeteran satisfaction of care within the Veterans Affairs is typically very high. Yet recommendation ratings of VA medical center (VA) hospitals as measured by Hospital Consumer Assessment of Healthcare Providers and Systems are generally lower than non-VA hospitals.Therefore, it was our objective to assess Veteran satisfaction and recommendation scores and then examine whether satisfaction correlates to recommendation. MethodsWe identified all acute care VAs as our primary analytic cohort. As a comparator group, we also included all acute care academic hospitals (non-VAs), as designated by the Centers for Medicare Services. Using data from Hospital Consumer Assessment of Healthcare Providers and Systems and Strategic Analytics for Improvement and Learning (SAIL) Value Model, we collated patient satisfaction scores, as well as markers of surgical safety from Hospital Compare. We then analyzed the correlation within VAs and non-VAs, primarily focusing the relationship between the “would you recommend Hospital Rating” and subdomains of the “Overall Hospital Rating,” as well as a composite score of patient safety. ResultsA total of 133 VAs and 1116 non-VAs were identified. Among VAs, the “Would you Recommend” hospital rating was significantly and positively correlated with markers of patient satisfaction including care transitions (Pearson's r = 0.59, P = 0.03), Nursing communication (Pearson's R 0.79, P = 0.001), and percent of primary care provider wait times less than 30 min (Pearson's r = 0.25, P = 0.01). VA-recommended scores were negatively correlated with factors such as time to emergency department discharge, and the “leaving the emergency department before being evaluated.” When looking at non-VAs, correlation directions were similar, albeit with stronger associations at almost every metric. While recommended scores correlated strongly to overall hospital ratings for both groups, VAs had no significant correlation between “would you recommend” and patient safety. However, there was a slight negative correlation between patient safety and “recommend” among non-VAs. ConclusionsAlthough satisfiers and dissatisfiers of care appear similar between VAs and non-VAs, “would you recommend” is a far weaker marker of patient perceptions of safety and quality. These seemingly empathetic markers such as “would you recommend” should be used with caution as they may not address the fundamental question being asked.