Factors that can minimize bleeding complications after renal biopsy

Blood pressure was significantly higher in patients with complications, and it was regarded as a risk factor for bleeding complications independently of all other variables by logistic multivariate analysis [25]. The study of Shidham et al. [1] showed that the risk of bleeding increased with high systolic blood pressure (SBP), diastolic blood pressure (DBP) or mean arterial pressure (MAP). For example, the risk of bleeding was 10.7 % when SBP >160 mmHg compared with 5.3 % when the SBP <160 mmHg (P = 0.03). Similarly, the risk was 12.5 % when the MAP >120 mmHg compared with 5.1 % when the MAP <120 mmHg (P = 0.009).

The serum creatinine also affected the bleeding complications after renal biopsy. Patients with major complications had higher serum creatinine levels when compared with those who did not. In the study of Whittier et al. [10], multivariate analysis using logistic regression showed that serum creatinine at baseline predictive of a complication and patients with a serum creatinine >5.0 mg/dl were 2.3 times as likely to have a complication (odds ratio, 2.3; 95 % CI 1.3–4.1; P < 0.005). In another study, for patients with creatinine >2 mg/dl, the risk ratio for post-biopsy bleeding (PBB) was 5.89 when compared with patients with creatinine <2 mg/dl, and logistic regression analysis showed that serum creatinine of >2.0 mg/dl was associated with an odds ratio of 2.56 (CI 1.48–4.42, P = 0.001) for PBB [1]. All these similar results showed that the risk of bleeding postoperatively increases with worsening levels of renal insufficiency.

The role of prolonged bleeding time was controversial. Previous studies on renal biopsy did not demonstrate a firm relationship between the bleeding time and bleeding complications. Eiro et al. [25] performed a research that pro-thrombin time, activated partial thromboplastin time and bleeding time were measured routinely in all patients before performing renal biopsy, and patients who had at least one abnormal value were regarded as contraindicated. In the study of Manno et al. [31], baseline partial thromboplastin time was significantly higher in patients who developed post-biopsy bleeding complications compared with those who did not (102.7 ± 11.8 vs. 100.1 ± 10.0 %, P = 0.013). However, in the study by Mackinnon et al. [12], the correction of prolonged bleeding time with pro-coagulants might not significantly reduce the risk of clinically important bleeding and they warned us that the administration of a pro-coagulant might increase the risk of a thrombotic vascular event, and thus, the practice of administering pro-coagulants routinely to correct a prolonged bleeding time should be reassessed. In the study of van den Hoogen et al. [13], the platelet function analyzer (PFA) had a higher positive and similar negative predictive value compared to the bleeding time. When a screening of the primary hemostasis was performed prior to a renal biopsy, they recommended using the PFA instead of the bleeding time (BT).

The relationship between the histological diagnosis and the complications of post-renal biopsy had been studied. Fisi et al. [14] performed a research which showed that in patients with the diagnosis of diabetic nephropathy or acute tubular necrosis, the overall complication prevalence was significantly lower compared to others, while complication rate was the highest in patients who suffered from thin basement membrane syndrome, vasculitis, rapidly progressive glomerulonephritis (RPGN) or acute interstitial nephritis. However, in the latter patients, the higher rate of complications may be related to the possibility of low glomerular filtration rate (GFR) because of acute kidney injury in many of these instances. In the meta-analysis by Corapi et al. [7], acute kidney injury was related to significantly higher rates of need for transfusion after renal biopsy (1.1 vs. 0.04 %; P < 0.001). The result of another study by Tondel et al. [9] also showed that low estimated GFR (estimated GFR = 30–59 ml/min per 1.73 m² [OR = 4.90 (1.60–14.00)] and estimated GFR <30 ml/min per 1.73 m² [OR 15.50 (5.60–43.00)]) had higher odds in adjusted analysis for risk factors for major complication after renal biopsy. So perhaps the low GFR was a confounding factor which would make it look like that certain diseases were linked with higher complication rates. Still, in the latter patients with a higher risk, stricter post-biopsy monitoring might be necessary. Amyloidosis was also regarded as an important risk factor for bleeding complication after renal biopsy in some studies [25]. To assess its risk, Soares et al. [15] performed a large retrospective study and their result showed that incidences of bleeding after kidney biopsy were similar between the amyloidosis and control groups (9.9 vs. 10.6 %). They suggested that the doctrine that patients with amyloid were at a greater risk of hemorrhage after kidney biopsy appeared to be the result of reporting bias. In fact, since patients with any systemic amyloidosis could have widespread small-vessel fragility by direct vessel infiltration and acquired hemostatic abnormalities caused by coagulation factor deficiency, it was reasonable for us to undertake the procedure in such patients with great care. So it was reasonable for us to check the hemostatic condition and whether hypertension existed in the patients with amyloidosis before the renal biopsy was done.

Other risk factors included that there was a significantly greater proportion of women compared with men who developed post-biopsy bleeding complications. In the study by Manno et al. [31], the increased risk of post-biopsy bleeding in women might be explained by their different body composition. The greater percentage of fat mass in women might be responsible for a tendency for hematoma to expand in the peri-renal fatty tissue. There was also a variable distribution in the risk of post-biopsy bleeding according to age, as the incidence of bleeding complication was more in older patients than in younger ones [16, 25]. A hypothesis was made in the study by Kohli et al. [16] that the higher incidence of gross hematuria in the elderly might have been due to the arterial wall changes related to aging. Due to the near ubiquitous presence of arteriosclerosis in the elderly, the ability of severed small vessels to undergo vasoconstriction might be impaired resulting in hematuria. Also, in the study of Whittier et al. [10], patients with a major complication after biopsy were older (53 ± 17 vs. 43 ± 18 years; P < 0.05).

Post-biopsy complications may be even less frequent with the use of smaller gauge needles. In the study of Kim et al. [24], the use of 14-gauge versus 18-gauge in native kidney renal biopsies had been compared in a randomized trial that showed a greater incidence of complications in manual biopsy with 14-gauge needle compared with automated renal biopsy using 18-gauge guns. Though their study had been flawed as it may reflect the impact of biopsy technique rather than needle size, similar results could be seen in the meta-analysis of Corapi et al. [7], which was that significantly higher rates of transfusion after renal biopsy were related to the needle size: 14-gauge compared with smaller needles (2.1 % vs. 0.5 %; P = 0.009). The study of Tondel et al. [9] showed that the dominance o f 16- and 18-G needles may result in higher focus on minimizing risk factors and the acceptance of less tissue per needle pass. In their study, the median number of glomeruli per subject was comparable with other studies, while the percentage of biopsies characterized as representative tissue was in the same level irrespective of needle size.

The study of Eiro et al. [25] showed that the frequency of the puncture was not significantly different between moderate complications and no or mild complications; however, by logistic multivariate analysis, the frequency of the puncture was an independent risk factor. The relationship between the bleeding complications and the depth of needle insertion was studied. Pasquariello et al. [26] performed a research that if the trigger was pushed exactly at the depth previously calculated by a mathematical formula: BW/H less 0.5 (body weight expressed in hectograms divided by patient height expressed in centimeters), it would be extremely useful to reduce the incidence of bleeding complications and allowed an adequate sampling for diagnostic evaluation in all cases. Another study also found some positive relationship between hemoglobin decrease and depth of needle insertion and warned that nephrologists should be cautious of depth of needle insertion to avoid major hemorrhage complication [4].

Some studies showed that there was no inevitable relationship between the experience of the operator and the incidence of post-biopsy bleeding. In the study of Maya et al. [5], the complications rate was low despite their performance by first-year nephrology fellows. The reason may be that the use of real-time ultrasound guidance permitted 91 % of the patients to require just one or two needle passes, which likely contributed to the low major complication rate. Similar results could be seen in another study, which showed that the use of the real-time ultrasound-guided technique minimized the risk of major complications even in the hands of inexperienced operators [27]. In summary, the real-time ultrasound guidance makes this operation safer.

Fisi et al. [14] in their study showed that after the renal biopsy, patients remaining in bed lying on their backs all the time with a sandbag under the site of puncture for 4 h was a way to reduce the complications and a fluid intake of at least 3L was recommended in this period. In fact, this was also a common measure taken to deal with post-renal biopsy complications in some Chinese hospitals, for example, a sandbag or abdominal compression belts were used to press the puncture site for 6 h with blood pressure monitoring and the urinary catheter might be inserted in order to reduce the patients’ movements, which may cause bleeding [21, 28, 29]. Few studies showed the influence of such procedure on the incidence rate of post-renal biopsy bleeding; however, it had been debated that since kidney was deep in the body and the pressure of the sandbag worked on the surface of the back, maybe the hemostatic effect would not be as perfect as imagine. Besides, the sandbag would cause discomfort when placed on the back of the patients on bed. In these aspects, the effectiveness of this method and its application needs more clinical observations.

Whittier et al. [10] performed a study on the timing of the complications and showed that complications happened in 67 % patients in 8 h, and in 89 % patients at ≤24 h, which meant an observation time of up to 24 h was optimal because an observation period of <8 h meant missing about 22 % of complications. In another research, if the patient showed no signs of bleeding complications 18 h after the biopsy, there was little probability for them to grow thereafter [25]. These results presented the importance of the observation of monitoring after the biopsy. As the utility of early post-renal biopsy ultrasound was useful in predicting the risk of major bleeding complications, some studies showed that the ultrasound findings 1 h post-biopsy were clinically helpful in predicting the bleeding complication. While the presence of a hematoma at 1 h post-biopsy was not predictive of a complicated post-biopsy course, the absence of a hematoma 1 h post-biopsy was highly predictive of an uncomplicated post-biopsy course [2]. In the retrospective study by Ishikawa et al. [4], perirenal hematoma >2 cm immediately after biopsy was the strongest predictor of more severe anemia the morning after biopsy. Their findings were similar with other reports of progressive anemia with larger post-biopsy hematoma [30]. This indicated that ultrasonographic evaluation of hematoma size immediately after renal biopsy was useful in predicting potentially severe blood loss. A conclusion was also made in the study by Daram et al. [8] that the degree of decline in Hct at 6 h was predictive of the degree of decline at 18–24 h. But this result still needs to be validated in larger prospective studies.

With the new technologies of real-time ultrasonography for guiding the procedure and the use of automatic biopsy needles, percutaneous renal biopsies are well established as a safe and effective technique for obtaining samples of renal parenchyma, and it has improved the rate of successful diagnosis in over 95 % of cases [32]. However, absolute and relative contraindications for the percutaneous approach do exist. When renal histology is necessary for clinical management but percutaneous biopsy is contraindicated or unsuccessful, other methods of renal biopsy by experienced physicians may be attempted.