RESUMO
BACKGROUND: Renal biopsy (RB) is necessary for the diagnosis, prognosis, and therapy guidance of native kidney diseases. Few studies have compared outcomes of RB procedures. We retrospectively compared the safety and efficiency of five biopsy procedures. METHODS: The number of glomeruli on light microscopy (LM) and on immunofluorescence (IF) and serious adverse events following RB performed in five nephrology units (C1-C5) were collected. C1 performed ultrasound (US) assessment before RB and used a 14-gauge core-cutting needle biopsy gun, C2 US guidance and a 14-gauge needle, C3 tomodensitometry guidance and a 14-gauge needle, C4 US guidance and a 16-gauge needle, and C5 tomodensitometry guidance and a 16-gauge needle. RESULTS: RB was performed in 943 adults between January 2006 and July 2010. Serious adverse events occurred in 1.5% of biopsies. The complication rate was not different between nephrology units. The mean number of glomeruli on biopsy was 14.2 ± 8.6 with LM and 4.4 ± 3.3 on IF. It was different according to the nephrology unit for LM (p = 0.01) and for IF (p < 0.001). The number of failed biopsies was influenced by the nephrology unit and radiological guidance technique, favoring real-time US guidance. Failed biopsies using US or tomodensitometry assessment before RB was certainly due to kidney imprecise localization since it was often non-renal tissue sampling. At least 10 glomeruli were found in 69% of biopsies on LM. This rate varied according to the nephrology unit (p = 0.004) and was higher when 14-gauge needles were used in comparison with 16-gauge needles. CONCLUSION: RB is safe regardless of the technical procedure, but radiological guidance and needle size influence the efficiency of biopsies.