**5. Complications**

According to SIR, the reported success rate for percutaneous nephrostomy is 98-99%, and this is defined as successful placement of catheter of sufficient size to allow for adequate drainage of the urinary tract or to allow successful tract dilatation for further interventional procedure. The success rates have been reported to be lower in cases of non-dilated collecting system or complex calculus disease (e.g. staghorn calculus) where a success rate of about 85% was reported (Ramchandani, Cardella et al. 2003). Despite the high success rates however, complications are frequently encountered, be it minor or major, with a reported incidence of approximately 10% of cases (Ramchandani, Cardella et al. 2003).

Several factors are associated with increased risk of complications. Patients at the extremes of age may develop complications from the procedure itself or even that related to the use of general anaesthesia, should this become necessary particularly in young children. Patient's coexisting comorbidities such as obesity, scoliosis, hepatomegaly and extremely mobile kidneys may necessitate greater manipulation, resulting in a technically challenging and thereby risky procedure. Further, in patients with chronic lung diseases and poor respiratory reserve such as emphysema, particular attention should be paid to the use of a subcostal approach to minimize risk of respiratory complications such as pneumothorax.

#### **5.1 Minor complications**

Minor complications are defined as complications occurring in relation to the procedure that are of no consequence and can be managed conservatively, or those requiring nominal therapy with no consequences (Ramchandani, Cardella et al. 2003). These patients may still require overnight hospitalization for observation. According to published reports, minor complications have been reported to occur in the range of 15-28% of cases (Stables 1982; Lee, Smith et al. 1987; Dyer, Regan et al. 2002).

Post-procedure bleeding varies in severity, and may range from simple transient hematuria to severe hemorrhage requiring transfusion or intervention. Minor bleeding complications include transient hematuria, which occurs in virtually all patients, and small perirenal hematomas that can resolve on conservative management. Transient hematuria occurs very

major.

below 4%.

**5.2.1 Hemorrhage** 

Percutaneous Nephrostomy 307

increase in level of care and prolonged hospitalization of more than 48 hours. Complications with permanent adverse sequelae or those that result in death are certainly considered to be

Hemorrhage requiring transfusion with or without radiological or surgical intervention is uncommon but is certainly a dreaded complication that carries a mortality risk. A number of published case series have reported major bleeding following percutaneous nephrostomy or percutaneous nephrolithotomy, and this occurs in the range of 1-4% (Lee, Smith et al. 1987; von der Recke, Nielsen et al. 1994; Dyer, Regan et al. 2002; Ramchandani, Cardella et al. 2003). This may manifest in prolonged hematuria, hemodynamic instability and perirenal hematomas. Hemorrhage may be related to vascular injury during the procedure, whether a normal vessel that are inadvertently injured, or it may be related to underlying vascular malformations or aneurysm. Hemorrhage could also be attributable to an underlying coagulopathy or bleeding diathesis. The guideline for quality improvement by SIR recommended a threshold rate of hemorrhage requiring blood transfusion to be kept

Significant bleeding during the procedure may be controlled by a tamponade applied to the tract with a nephrostomy catheter or balloon dilatation catheter in larger tracts. If at any point that this fails, or if the patient develops subsequent significant blood loss after the procedure, angiographic evaluation would be indicated for identification of abnormal vascularity or major vascular injury with possible need for embolization. Surgical intervention may later become necessary if poor bleeding control is achieved. By this way, injured vessels may be ligated to arrest the bleeding, or as a last resort, partial or total nephrectomy may have to be performed. Lee reported 4 cases of persistent bleeding which were arrested by successful embolization, while a further 2 cases had to undergo nephrectomy or partial nephrectomy following failed embolization (Lee, Smith et al. 1987). Cowan reported 7 cases of persistent bleeding in a series of 3100 patients following percutaneous nephrostomy, and these were found to be secondary to underlying arteriovenous aneurysms that were treated successfully with embolization (Cowan 2008).

The performing physician should therefore be aware of the risk of severe blood loss, and the patient should be counseled appropriately regarding this risk during consent taking prior to the procedure. However, there are steps that can be taken during the procedure to minimise risk of hemorrhage. As described above, particular attention should be paid to the coagulation profile prior to the procedure, and any significant abnormality should be corrected accordingly. The renal vascular anatomy should be reviewed and taken into consideration when planning for puncture site and approach. The kidney should be punctured along the Brodel's avascular line as described above, and similarly, punctures too close to the inferior surface of a rib run the risk of injury to the intercostal vessels. The uses of fine needles and small-bore catheters have been associated with smaller risk of severe hemorrhage. A two-wall puncture of the renal pelvis should also be avoided to minimize risk of injury to the anterior segmental renal artery. As an additional support measure, highrisk patients should be prepared with support from the blood transfusion services should

blood transfusion becomes necessary during or after the procedure.

frequently that some authors do not regard it as a complication (Stables 1982). Clinically silent perinephric hematoma have been reported to occur fairly frequently, and is found in up to 13% of cases following percutaneous nephrostomy (Cowan 2008). These can resolve spontaneously without necessitating further interventions, leaving no serious consequences to the patient in the majority of cases. Stables also observed that in 79% of these cases, no significant renal alteration was seen. However, the presence of prolonged hematuria with or without hemodynamic instability should alert the physician for possible continuing bleeding as well as vascular injury.

Catheter-related complications such as kinking, obstruction or dislodgement may frequently be encountered and may require further intervention in 14% of cases (Cowan 2008). Published reports quoted varying rates of catheter dislodgement, from 4.8% - 11.6%. The use of larger bore catheters (for example 14Fr catheter) may reduce this rate to 1% (Cowan 2008). Stables recommended advancement of the catheter well into the renal pelvis or calyces to minimize risk of dislodgement (Stables, Ginsberg et al. 1978). However care should be taken to avoid obstructing the ureter particularly if a large bore catheter is used. A dislodged tube may have to be replaced and the new catheter may have to be inserted by creating a new tract unless if the previous tract has been well established.

To reduce the rate of catheter obstruction, routine irrigation with normal saline solution should be performed after the procedure, although the use of larger-bore catheters may reduce the rate further while providing good drainage. Debris may also be removed by manipulation with a guide wire. Occasionally however, if catheter obstruction persists despite conventional measures, catheter replacement may be necessary.

Urine leak is known to occur following percutaneous nephrostomy, with a rate of approximately 7-7.2% (Lee, Smith et al. 1987; Moskowitz, Lee et al. 1989). This is frequently minor, and contrast extravasation during or immediately after the procedure may or may not indicate ensuing complication. Also, most small leaks and tears resolve spontaneously with adequate urinary drainage or ureteral catheter insertion (Lee, Smith et al. 1987). Urine leak can also be controlled by using a larger bore catheter (Cowan 2008). However urine leak may occasionally be prolonged (lasting more than a week) and the ensuing urinoma may be large enough to require surgical intervention.

Other minor complications that may be seen following the procedure may include pain and fever. While fever can be worrisome for ensuing sepsis with potential of shock, febrile patients may require nothing more than conservative management with or without antipyretics. Lee reported 23% of raised temperature of more than 38.5 degrees Celsius in his published series of 582 patients who underwent percutaneous nephrolithotomy (Lee, Smith et al. 1987). These were attributed to retrograde urine flow as well as the use of irrigation fluid during the procedure. Minor wound infection has also been reported (von der Recke, Nielsen et al. 1994; Kaskarelis, Papadaki et al. 2001). These may be related to prolonged catheter use, and the use of sutures to secure the catheters to the skin (Kaskarelis, Papadaki et al. 2001). Pneumonia and atelectasis have been reported in a minority of cases, but is usually managed conservatively with antibiotics.

#### **5.2 Major complications**

Major complications are defined by SIR as complications that require therapy or minor hospitalization of up to 48 hours, as well as those that require major therapy, unplanned

frequently that some authors do not regard it as a complication (Stables 1982). Clinically silent perinephric hematoma have been reported to occur fairly frequently, and is found in up to 13% of cases following percutaneous nephrostomy (Cowan 2008). These can resolve spontaneously without necessitating further interventions, leaving no serious consequences to the patient in the majority of cases. Stables also observed that in 79% of these cases, no significant renal alteration was seen. However, the presence of prolonged hematuria with or without hemodynamic instability should alert the physician for possible continuing

Catheter-related complications such as kinking, obstruction or dislodgement may frequently be encountered and may require further intervention in 14% of cases (Cowan 2008). Published reports quoted varying rates of catheter dislodgement, from 4.8% - 11.6%. The use of larger bore catheters (for example 14Fr catheter) may reduce this rate to 1% (Cowan 2008). Stables recommended advancement of the catheter well into the renal pelvis or calyces to minimize risk of dislodgement (Stables, Ginsberg et al. 1978). However care should be taken to avoid obstructing the ureter particularly if a large bore catheter is used. A dislodged tube may have to be replaced and the new catheter may have to be inserted by creating a new

To reduce the rate of catheter obstruction, routine irrigation with normal saline solution should be performed after the procedure, although the use of larger-bore catheters may reduce the rate further while providing good drainage. Debris may also be removed by manipulation with a guide wire. Occasionally however, if catheter obstruction persists

Urine leak is known to occur following percutaneous nephrostomy, with a rate of approximately 7-7.2% (Lee, Smith et al. 1987; Moskowitz, Lee et al. 1989). This is frequently minor, and contrast extravasation during or immediately after the procedure may or may not indicate ensuing complication. Also, most small leaks and tears resolve spontaneously with adequate urinary drainage or ureteral catheter insertion (Lee, Smith et al. 1987). Urine leak can also be controlled by using a larger bore catheter (Cowan 2008). However urine leak may occasionally be prolonged (lasting more than a week) and the ensuing urinoma

Other minor complications that may be seen following the procedure may include pain and fever. While fever can be worrisome for ensuing sepsis with potential of shock, febrile patients may require nothing more than conservative management with or without antipyretics. Lee reported 23% of raised temperature of more than 38.5 degrees Celsius in his published series of 582 patients who underwent percutaneous nephrolithotomy (Lee, Smith et al. 1987). These were attributed to retrograde urine flow as well as the use of irrigation fluid during the procedure. Minor wound infection has also been reported (von der Recke, Nielsen et al. 1994; Kaskarelis, Papadaki et al. 2001). These may be related to prolonged catheter use, and the use of sutures to secure the catheters to the skin (Kaskarelis, Papadaki et al. 2001). Pneumonia and atelectasis have been reported in a minority of cases, but is

Major complications are defined by SIR as complications that require therapy or minor hospitalization of up to 48 hours, as well as those that require major therapy, unplanned

bleeding as well as vascular injury.

tract unless if the previous tract has been well established.

may be large enough to require surgical intervention.

usually managed conservatively with antibiotics.

**5.2 Major complications** 

despite conventional measures, catheter replacement may be necessary.

increase in level of care and prolonged hospitalization of more than 48 hours. Complications with permanent adverse sequelae or those that result in death are certainly considered to be major.

#### **5.2.1 Hemorrhage**

Hemorrhage requiring transfusion with or without radiological or surgical intervention is uncommon but is certainly a dreaded complication that carries a mortality risk. A number of published case series have reported major bleeding following percutaneous nephrostomy or percutaneous nephrolithotomy, and this occurs in the range of 1-4% (Lee, Smith et al. 1987; von der Recke, Nielsen et al. 1994; Dyer, Regan et al. 2002; Ramchandani, Cardella et al. 2003). This may manifest in prolonged hematuria, hemodynamic instability and perirenal hematomas. Hemorrhage may be related to vascular injury during the procedure, whether a normal vessel that are inadvertently injured, or it may be related to underlying vascular malformations or aneurysm. Hemorrhage could also be attributable to an underlying coagulopathy or bleeding diathesis. The guideline for quality improvement by SIR recommended a threshold rate of hemorrhage requiring blood transfusion to be kept below 4%.

Significant bleeding during the procedure may be controlled by a tamponade applied to the tract with a nephrostomy catheter or balloon dilatation catheter in larger tracts. If at any point that this fails, or if the patient develops subsequent significant blood loss after the procedure, angiographic evaluation would be indicated for identification of abnormal vascularity or major vascular injury with possible need for embolization. Surgical intervention may later become necessary if poor bleeding control is achieved. By this way, injured vessels may be ligated to arrest the bleeding, or as a last resort, partial or total nephrectomy may have to be performed. Lee reported 4 cases of persistent bleeding which were arrested by successful embolization, while a further 2 cases had to undergo nephrectomy or partial nephrectomy following failed embolization (Lee, Smith et al. 1987). Cowan reported 7 cases of persistent bleeding in a series of 3100 patients following percutaneous nephrostomy, and these were found to be secondary to underlying arteriovenous aneurysms that were treated successfully with embolization (Cowan 2008).

The performing physician should therefore be aware of the risk of severe blood loss, and the patient should be counseled appropriately regarding this risk during consent taking prior to the procedure. However, there are steps that can be taken during the procedure to minimise risk of hemorrhage. As described above, particular attention should be paid to the coagulation profile prior to the procedure, and any significant abnormality should be corrected accordingly. The renal vascular anatomy should be reviewed and taken into consideration when planning for puncture site and approach. The kidney should be punctured along the Brodel's avascular line as described above, and similarly, punctures too close to the inferior surface of a rib run the risk of injury to the intercostal vessels. The uses of fine needles and small-bore catheters have been associated with smaller risk of severe hemorrhage. A two-wall puncture of the renal pelvis should also be avoided to minimize risk of injury to the anterior segmental renal artery. As an additional support measure, highrisk patients should be prepared with support from the blood transfusion services should blood transfusion becomes necessary during or after the procedure.

Percutaneous Nephrostomy 309

Significant infection and sepsis following percutaneous nephrostomy is an important and well-recognised complication, with several published reports documenting its occurrence. According to SIR, sepsis related complications have been reported to occur in 1-9% of cases (Ramchandani, Cardella et al. 2003). There is a wide spectrum of severity of infection, but major sepsis may be defined as cases of septicemia requiring escalation in hospital care and longer use of antibiotics therapy, with or without shock. Transient fever is common following the procedure, occurring in almost all patients, the majority of which may settle in less than 6 hours (Lee, Patel et al. 1994). However, persistent fever with chills and signs of hemodynamic instability are worrisome signs and should be identified and treated accordingly. Septic shock is a serious complication, and has been reported to be a contributing factor towards patients mortality in some published case series. In 318 patients who underwent percutaneous nephrostomy, Lewis reported sepsis as the most common major complication, occurring in 2.2% of patients in his published case series, and it is the most severe complication necessitating intensive care (Lewis and Patel 2004). Sepsis was also considered to be a contributing factor in the death of two of these cases. Moskowitz further reported 2 cases of septicemia with shock in 11 cases of severe sepsis in his published case series of patients who underwent percutaneous nephrolithotomy (Moskowitz, Lee et al. 1989). SIR recommended a rate of septic shock of less than 4%, and a

rate of less than 10% for cases of septic shock in the setting of pyonephrosis.

irrigation fluid have also been found to be contributory (Lee, Patel et al. 1994).

the results of urine culture and sensitivity.

**5.2.3 Pleural complications** 

Several factors have been found to contribute towards increased risk of sepsis, and this includes the duration of the procedure itself, urine bacterial load, severity of urinary tract obstruction as well as presence of bacteria within the calculus (Mariappan, Smith et al. 2006). Mariappan also reported higher risk of upper urinary tract infections in patients with calculi larger than 20mm or a dilated pelvicalyceal system. Further, the puncture itself, and even the removal of calculus may reactivate underlying pre-existing infection within the urinary tract with release of bacteria into the system. Therefore, care should be taken to avoid over-distension of the renal collecting system during puncture, as this may result in bacterial reflux into the peripapillary plexus. Urine extravasation and absorption of

The use of prophylactic antibiotics is therefore recommended in high-risk patients, and this has been shown to be of some benefit as reported in a prospective controlled study by Mariappan, who prescribed one week of ciprofloxacin to patients prior to percutaneous nephrolithotomy (Mariappan, Smith et al. 2006). Patients who received prophylactic ciprofloxacin were reported to have significantly reduced incidence of upper urinary tract infection as compared to the control group, with three times less risk of developing systemic inflammatory response syndrome. Antibiotics therapy may be further escalated in patients with evidence of urosepsis following the procedure, and this is best adjusted according to

Pleural complications such as pneumo-, hydro-, or hemothorax and empyema are uncommon but have been known to occur from percutaneous nephrostomy, with a reported rate of 0.1-0.3% (Dyer, Regan et al. 2002; Ramchandani, Cardella et al. 2003). The risk of

**5.2.2 Sepsis** 

Fig. 8. An 8 Fr Navarre catheter in place with its loop apparently sited within the renal pelvis. Figures 8 to 11 are of the same patient.

Fig. 9. Nephrogram showing contrast leakage into peripelvic fat due to transgression of the pelvicalyceal system. The patient subsequently developed flank pain and hypotension, indicating concomitant vascular injury. Sonography confirmed the presence of a perinephric hematoma.

Fig. 10. Renal angiography demonstrated bleeding from a branch of the inferior segmental artery. Super-selective injection of contrast via a microcatheter (m) into the bleeding artery showed extravasation into the renal pelvis (p).

Fig. 11. Successful embolization with microcoils (c).
