Hospital Star Ratings and Patient Outcomes

Notwithstanding our general grumpy old man bent, every now and then some positive news comes along that we are happy to report on.  Various government-sponsored quality programs have definitely taken their hits, but a new study in the Journal of the American Medical Association Internal Medicine finds that there is an association between hospital star ratings and some patient outcomes, and even better, the association appears to be positive!!   (JAMA Int. Med. Article)   Currently the Centers for Medicare & Medicaid services uses a five-star system for hospitals that is based solely on results from the hospital CAHPS survey.  Eventually CMS will likely include process and outcome measures, but for now there is concern about whether patient experience is linked to patient outcomes.  The researchers examined the correlation between number of stars and hospital readmission and mortality rates, using 2015 data.  The researchers attempted to adjust for several factors, but it does not appear that they fully capture differences is socio-economics or health status of populations.  There were 3076 hospitals in the sample, about 4% had five star ratings, 26% had four stars and 47% had three stars.  The four and five star hospitals were more likely to be smaller and in small rural midwestern towns.

Five star hospitals had the lowest mortality rates, 9.8%, four star facilities were at 10.4%, three star at 10.5%, two star at 10.7% and one-star at 11.2%.  A similar pattern was found in regard to 30-day readmission rates, with one star hospitals at 18.7%, four star at 20.2%, three star at 21%, two star 21.8% and one star at 22.9%.  So there appears to be a linear correlation between number of stars and these outcomes, although these are pretty small differences and there are few hospitals in the extreme star categories.  It seems somewhat likely that the results are also picking up other factors–hospitals in less-populated midwestern towns are themselves small and deal with homogenous, relatively healthy populations.  And being in a large, urban facility likely increases patients’ feelings of anxiety, of not getting individualized attention, and being in a busy, anonymous environment.  The CAHPS survey is not designed to sort out institutional characteristic effects from the individual care delivered.  But at least there is some minimal evidence that there may be a relationship between patient satisfaction and outcomes.