Kidney Week

Abstract: FR-PO1191

Center-Level Variation in the Association of Clinical Factors with Use of an Early Steroid Withdrawal Regimen in Kidney Transplant Recipients

Session Information

• Transplantation: Clinical - Immunosuppression, Adherence, Outcomes
  November 08, 2019 | Location: Exhibit Hall, Walter E. Washington Convention Center
  Abstract Time: 10:00 AM - 12:00 PM

Category: Transplantation

• 1902 Transplantation: Clinical

Authors

• Bae, Sunjae, Johns Hopkins University, Baltimore, Maryland, United States

Sunjae Bae,

• McAdams-DeMarco, Mara, Johns Hopkins, Baltimore, Maryland, United States

Mara McAdams-DeMarco,

• Garonzik wang, Jacqueline, Johns Hopkins, Baltimore, Maryland, United States

Jacqueline Garonzik wang,

• Massie, Allan, Johns Hopkins School of Medicine, Baltimore, Maryland, United States

Allan Massie,

• Coresh, Josef, Welch Center for Prevention, Epidemiology & Clinical Research, Baltimore, Maryland, United States

Josef Coresh,

• Segev, Dorry L., Johns Hopkins University, Baltimore, Maryland, United States

Dorry L. Segev,

Background

Early steroid withdrawal (ESW) may confer net benefit to low-risk kidney transplant (KT) recipients. However, there is limited evidence on what clinical factors constitute the “low-risk” status that favors ESW, possibly resulting in heterogeneous, and even contradicting, practices across KT centers. Quantifying this heterogeneity using real-world data is also crucial for making unbiased inferences on ESW. We aimed to characterize the center-level variation in how clinical factors influence the selection of ESW.

Methods

Using SRTR data, we studied 210,133 KT recipients in 2002-2017, after excluding who did not receive tacrolimus and mycophenolate for maintenance immunosuppression (n=47,756). ESW was defined as withdrawal of steroid by the time of discharge from KT admission. We quantified the center-level variation in the associations of 74 variables with ESW, via the standard deviation (SD) of the random slope terms in multilevel logistic models.

Results

We identified 10 variables with greater variation (Figure). Factors such as recipient hypertension and pulsatile perfusion were associated with ESW in opposing directions at different centers. For example, the center-specific odds ratio (OR) of ESW for recipient hypertension was <0.8 at 110 (39.6%) centers, but >1.25 at 63 (22.7%) centers. On the other hand, factors such as increased PRA and longer cold ischemic time were associated with lower odds of ESW at most centers, but to substantially varying degrees. For example, while high PRA (80-100 vs 0-9) was associated with lower odds of ESW in the entire population (OR=_0.440.50_0.58), this association was particularly stronger at some centers [eg, OR<0.2 at 29 (10.4%) centers].

Conclusion

Our findings suggest a substantial discordance among KT centers on what clinical factors indicate ESW and how important each factor is.

Distribution of center-specific odds ratios of early steroid withdrawal

Funding

• NIDDK Support