Right TraC™ Post-Hospitalization Care Transitions Program to Reduce Readmissions for Hemodialysis Patients

Abstract

Background: Hemodialysis (HD) patients have high hospitalization rates. This nonrandomized trial tested the effect of a bundle of renal-specific “Right TraC™” strategies on 30-day all-cause readmission rates and, secondarily, 90-day readmissions and overall admissions among HD patients. Methods: Twenty-six Fresenius clinics in West Virginia, Ohio, and Kentucky participated in the interventions. Eighteen matched clinics served as controls; intervention clinics also served as their own controls. We deployed the intervention in 3 incremental phases focused on patient information exchange, post-hospital follow-up, and telephonic case management. Thirty-day hospital readmissions per patient year (ppy) were calculated by dividing the total number of readmissions within 30 days of index admission by the total number of patient-years in baseline (2012) and remeasurement (2014) periods. We also compared readmission rates from 2010 to 2015. We used repeated measures Poisson regression to compare outcomes between groups and time periods. Results: From 2012 to 2014, 30-day all-cause readmissions ppy declined for Right TraC clinics (from 0.88 to 0.66 [p < 0.001]; for controls, from 0.73 to 0.61 [p = 0.16]). Difference in change between groups was nonsignificant (p = 0.26). Overall admissions ppy declined: for Right TraC clinics from 2.51 to 1.97 (p < 0.001); for controls from 2.14 to1.92 (p = 0.21); difference in change between groups was significant (p = 0.01). For 2010, 2011, and 2012, Right TraC clinic 30-day readmissions ppy were unchanged: 0.89, 1.00, 0.88 (p = 0.61 and p = 0.49); they declined to 0.66 (p < 0.001) in 2014 (intervention year); rose to 0.70 (p = 0.06) in 2015 (interventions discontinued). Conclusion: We conclude that Right TraC interventions may have been helpful in reducing hospital readmission rates.

On the Right Track: Implementing Interventions to Reduce Readmissions in Dialysis Patients

“…In this issue of the American Journal of Nephrology , Wingard et al [3] report the results of a multi-component, facility-level, pragmatic trial designed to reduce 30-day readmissions among hemodialysis patients. Overall, 26 intervention and 18 matched control facilities were targeted in the West Virginia/Ohio/Kentucky region.…”

“…Wingard et al [3] noted that their results could not be generalized to other dialysis patient populations, particularly to urban dialysis patients who may have differences in terms of demographics, education, and access to care. The authors also noted that, in their pragmatic design, the effects of individual components of the intervention could not be teased apart.…”

“…Indeed, the primary limitation of the study by Wingard et al [3] was not that the design was pragmatic but that the intervention was not randomized. While facilities were matched on a few facility characteristics and adjustment for patient demographics and select comorbid conditions was performed, some level of residual confounding by facility and patient factors is likely.…”

“…While facilities were matched on a few facility characteristics and adjustment for patient demographics and select comorbid conditions was performed, some level of residual confounding by facility and patient factors is likely. Other limitations may have led to the observed non-statistically significant results [3]. First, the number of facilities was small and this limited the power to detect the effect of this facility-level intervention.…”

“…Third, potential secular trends of declining readmissions and contamination across clinics by hospital and dialysis providers cannot be ruled out. Fourth, fidelity was potentially suboptimal for some components of the intervention: e.g., fewer than half of patients were ever reached by telephone in the 30 days post-discharge by the care coordinator [3]. Finally, some of the intervention components may have been least likely to reach the patients who were most likely to be readmitted.…”