Estimation of risk for major bleeding in native kidney biopsies in patients with multiple risk factors

Background and aims Individuals undergoing kidney biopsy are increasingly older and may have concurrent illnesses that cause deranged hematological and renal parameters that are associated with post-biopsy bleeding. We aimed to develop a clinical risk model to quantify bleeding risks in high-risk individuals with multiple risk factors. Methods Single-center retrospective cohort study of consecutive adults with serum creatinine >= 2 mg/dL (176 mu mol/L) and had ultrasound-guided percutaneous native kidney biopsies between June 2011 and July 2015 in our tertiary referral center. The primary outcome was major bleeding, defined as need for red cell transfusion, radiological or surgical intervention, or if bleeding led to death within 7 days after kidney biopsy. Results Among 184 native kidney biopsies with serum creatinine >= 2 mg/dL, median age was 54.1 years and eGFR was 18.8 ml/min/1.73 m(2). Major bleeding occurred in 19 biopsies (10.3%). Multivariate analysis accounting for age, weight, hemoglobin, platelet, prothrombin time and urea found that higher hemoglobin (adjusted OR 0.51, 95% CI 0.33-0.79, p = 0.003) and platelet (adjusted OR 0.99, 95% CI 0.98-0.99, p = 0.01) were independently associated with reduced major bleeding. A risk model that included (1) age >= 62 years old, (2) hemoglobin < 10 g/dL and (3) platelets <= 216 x 10(9)/L as categorical variables predicted major bleeding post-biopsy. Conclusion We developed a risk model that included multiple risk factors to quantify bleeding risks in native kidney biopsies with renal impairment.

Automated summary

The study aimed to develop a clinical risk model to quantify bleeding risks in high-risk individuals undergoing kidney biopsy. Results showed that higher hemoglobin and platelet levels were independently associated with reduced major bleeding, and a risk model including multiple risk factors could predict major bleeding post-biopsy.