Kidney biopsy is a very important diagnostic tool for nephrologists and has always been considered a characteristic of nephrologist’s job description. Recently however kidney biopsy has been increasingly reserved to larger hospitals with an experienced nephrologist and pathologist to guarantee maximum procedural safety and diagnostic power or it has been taken over by other specialists. The result is a reduced appeal of Nephrology specialty for young trainees. Eventually young nephrologists miss the technical expertise to perform a renal biopsy due to lack of training. In line with the guidelines given by the American Society of Nephrology our middle sized academic hospital has since 2012 promoted the training of young nephrologists even in biopsy performing.
In our study we evaluated if procedural safety and diagnostic power of the renal biopsy in a middle sized hospital could be comparable to those reported by larger hospitals also assessing procedural safety and adequacy of biopies performed by less experienced operators, including fellows, under tutor supervision.
Our data showed minor complications in 17.3% of biopsies and major complications in 1.2% of cases. This incidence rate was in line with the one reported by larger centers especially if we consider that our definition of minor complications included millimetric sized hematomas which are very often considered an unavoidable consequence of the bioptic procedure just as much as microscopic hematuria [
8]. Furthermore all the biopsies of our study were performed using a 14-gauge needle that has been shown to be associated with a higher rate of complications compared to the most used 16 and 18 gauge needles [
9‐
12]. Manno and colleagues found no differences in complications rate between 14 and 16 gauge needles in their series of biopsies [
13]. However they evaluated biopsies performed in patients with low bleeding risk and this could have blunted the differences between the two needles. It may be difficult to compare differences in complications rates among studies because they can vary substantially due to confounding issues such as the type of the study (retrospective vs. prospective), patient mix, the ultrasound machine used, the needle type or gauge used, and of course the operator performing the biopsy. However one of the largest (1055 patients), prospective, single-center study with PRB performed in adults at an academic institution using real-time ultrasound and 14-gauge needles showed results that were comparable to the results of our study [
14]. Furthermore in the study by Korbert and colleagues the rate of complications was similar and the majority of procedures were performed by nephrology fellows. In our cohort there were no cases of death or nephrectomy and although we described one arteriovenous fistula it did not require any treatment. As reported by several studies [
13,
15] in our cohort female patients were at higher risk of procedural related complications. The increased bleeding risk observed in women has been explained with their different body composition. Women in fact have a greater percentage of fat mass that could increase the bleeding into perirenal tissues [
13]. The role of hypertension as a risk factor for bleeding after a kidney biopsy is still controversial, our study seems to confirm that hypertension is not a major risk factor for postbiopsy complications as reported by previous studies [
8,
15]. Surprisingly postbiopsy complications were more common in the group of patients with normal blood pressure. To note, however, patients in our study were defined as hypertensive based on the need for antihypertensive drugs and moreover all hypertensive patients were treated in order to achieve blood pressure < 140/90 mmHg before the procedure and this could have affected the results reducing the risk of postbiopsy bleeding in hypertensive patients. Our results show that an adequately controlled hypertension, although severe, does not represent a contraindication to the bioptic procedure. We were unable to find any other clinical conditions increasing the incidence of procedural risk such as amyloidosis or advanced renal failure as reported by some studies [
16,
17]. All together our data seems to show that the biopsy procedure can be performed safely in middle sized hospitals as the incidence of complications in our center is equal to the one described in literature. Interestingly, complication rate in our center was equal in the timespan before and after 2012. Before 2012 only one very skilled nephrologist performed kidney biopsies whereas after 2012 several nephrologists, mostly young trainees performed the procedure during their turnation. This very important data suggests that an increase in the number of biopsy performing personnel including nephrology fellows as long as under tutor guide does not imply a reduction of procedural safety. Based on this, the training requirement (ability to independently perform percutaneous kidney biopsy of both native and transplanted kidneys) proposed by the American Board of Internal Medicine (ABIM) and the Accreditation Council for Graduate Medical Education (ACGME) for nephrology fellows appears feasable [
18]. However, although the Board of Internal Medicine requires that nephrology fellows be trained to perform percutaneous kidney biopsy, the Renal Pathology Society and some authors underline that a kidney biopsy should only be done by someone skillful in performing the procedure [
19,
20]. Dawoud and colleagues even proposed a simulation tool that mimics biopsy conditions in human patients to improve confidence and procedural skill competence of trainees [
21]. In their study individual fellows practiced repeatedly ultrasound-guided renal biopsy procedures using a porcine kidney/turkey breast phantom until they attained reasonable accuracy and gained confidence. They were guided by an experienced operator. The effect of this simulation training on trainees’ procedural competence was evaluated by comparing outcomes of renal biopsies performed by fellows before and after the implementation of the simulation training. The study showed that the implementation of the simulation training reduced the severity of biopsy-associated bleeding complications. The use of simulation tools could be implemented in the curriculum of nephrology fellows to improve the accuracy and confidence of nephrology fellows not only for renal biopsy procedure but also for arterio-venous fistula and central catheter placement. Our results confirm those obtained by Chung and colleagues [
22] who evaluated the safety of kidney biopsy according to practitioner and ultrasound technique. In their study the authors compared complication rate of kidney biopsy performed by nephrologists and radiologists. Nephrologists in their study were first and second year fellows. Their results show that percutaneous renal biopsy performed by young trainees was not inferior to that performed by expert ultrasound radiologists. Furthermore the mean number of glomeruli in renal tissue obtained by nephrologists was significantly higher than that obtained by radiologists. In our study the number of adequate biopsies was similar before and after 2012 but the number of passes was significantly lower after 2012 demonstrating, as shown in the study by Chung, that young nephrology trainees can obtain a better biopsy core. To date, however, our study is the first one comparing kidney biopsies performed by a skilled nephrologist to those performed by unskilled nephrologists and trainees. Furthermore according to a study by Corapi and coworkers [
8] our cohort of patients biopsied after 2012 cannot be considered low risk patients since mean age was quite high (59.24 ± 15.6 years) much more than any previous study, a large percentage of patients had a reduced renal function (creatinine >2 mg/dl, 36.7% of patients) and many presented with acute kidney failure. Another essential data of our study is the one concerning the diagnostic power of renal biopsy. Our study points out adequacy of procedural diagnostic power in our Hospital. Of the 337 biopsies evaluated in fact, 97.3% proved adequate, allowing diagnosis, whereas only 9 did not allow diagnosis. Interestingly we show that starting from 2012 the number of non diagnostic biopsies decreased from 7 (4.1%) to 2 (1.2%). Although not statistically significant this data, which is probably secondary to the practice of observing the freshly obtained sample by light microscopy to evaluate adequacy of the bioptic specimen, once more shows that diagnostic power of the bioptic procedure does not change if it is performed by less expert operators. In the present paper we focused on the ability of young fellows to perform kidney biopsies under the guidance of an experienced nephrologist. However to perform kidney biopsy is only part of the task. Glomerular diseases are rare and it is mandatory, in order to make a diagnosis, to have the biopsy evaluated by an experience pathologist who can describe the biopsy as soon as possible to guide the treatment.
We acknowledge that the present study has many limitations. First of all it is a retrospective study and the biopsies considered were performed in two rather long periods of time. We did not find a significant dfference in outcomes however many variables could have affected outcomes such as different patients characteristics, different technical approach and different quality of ultrasound machines. Patients undergoing kidney biopsy in the two study periods had similar characteristics and no differences were found in kidney diseases. We used the same ultrasound machine throughout the period examined, a machine that was only older when used by young trainees. The only remarkable difference in the techique used to perform the kidney biopsy was the analysis of the tissue by light microscopy to check for glomeruli.
Since a16G needle is perfectly adequate for obtaining good histology sample it could be used when a more inexperienced operator is performing the procedure, increasing the safety of the procedure.