Percutaneous kidney biopsy: Can we predict which patients are at risk of bleeding complications?

Abstract

Aim: To determine the prevalence and predictors of bleeding complications of percutaneous kidney biopsies (PKB).

Background: PKB are the gold standard for investigating kidney parenchymal diseases. Although considered generally safe, PKB have been associated with bleeding complications. Predictors of bleeding events, however, remain poorly characterized.

Materials and methods: Patients who underwent PKB performed by Nephrologists/Advanced Trainees from January 2017 to March 2020 in a tertiary referral centre were retrospectively studied. Data were extracted from medical records and laboratory database. Minor bleeding was defined as the development of either haematuria or haematoma, and major bleeding as the need for either blood products, radiological intervention or admission to ICU.

Results: 285 PKB (231 Native, 54 Allograft) were performed during the study period. The prevalence of bleeding events was 9% (5% minor and 4% major events). No major bleeding was observed post‐allograft PKB. Female gender, older age, DDAVP use, dialysis and lower pre‐procedure haemoglobin were significantly associated with major bleeding events in univariate analysis. After adjustment, only female gender remained significantly associated with major bleeding events (adjusted OR 8.71, CI 1.06‐71.58). Female gender also had a significant independent association with minor bleeding events (adjusted OR 4.78, CI 1.32‐17.34). No association with bleeding was noted for needle size (16 vs. 18 gage), number of passes, coagulation profile, hypertension, serum creatinine, urine albumin: creatinine ratio or inpatient (vs outpatient) PKB.

Conclusions: We found no independent association between plausible clinical predictors and bleeding events post‐PKB. The independent association between female gender and bleeding events warrants prospective validation. The association of bleeding with the use of DDAVP is possibly related to selection bias, as patients at higher risk are likely to receive DDAVP.