Methods: Data from the Minimum Data Set 3.0 and Medicare Provider Analysis and Review files for 2011-13 were used to construct 3 NH level measures for the 34 NHs in Monroe County, NY: (1) overall transfers between 5 county hospitals and NHs; (2) transfers to hospitals from each NH; and (3) transfers to each NH from each hospital. Data were linked to Emerging Infections Program CDI population surveillance data, Online Survey, Certification, and Reporting files, and NH Compare for 2011-13. Separate multivariate negative binomial regression models were used to examine the relationship between each of 3 patient transfer variables and NH CDI, controlling for NH covariates and hospital CDI rate.
Results: NH CDI incidence rate is 1.74/10,000 patient-days. From 2011-13, there were 27293 transfers between hospitals and NHs, 14843 transfers from NH to hospital, and 12612 transfers from hospital to NH. On average, when patient transfers increase by 10, NH CDI increases by 1.1% (P=0.020), 2.2%, (P=0.022), and 1.6% (P=0.056) for overall transfers, NH to hospital transfers, and hospital to NH transfers, respectively. As NH size, measured by number of beds, increases, NH CDI decreases (P≤0.030) for all transfer patterns. Increases in health deficiency citations are associated with increases in NH CDI (P≤0.026) for all transfer patterns. In addition, as certified nurse aide hours per resident/day increase, NH CDI decreases (P≤0.032) for all transfer patterns; however, as licensed practical nurse (LPN) hours and registered nurse (RN) hours per resident/day increase, NH CDI increases (P≤0.045) for all transfer patterns.
Conclusion: Overall, higher number of patient transfers is associated with increases in NH CDI. Associations between RN and LPN hours per resident/day and NH CDI may reflect that NHs with more skilled nursing hours often serve higher acuity residents and/or may have better CDI detection protocols in place.
H. Temkin-Greener, None
G. Dumyati, None