Journal of Gastrointestinal Surgery

, Volume 17, Issue 2, pp 319–325

The Management of Acute Cholecystitis in Chronic Hemodialysis Patients: Percutaneous Cholecystostomy Versus Cholecystectomy


    • Baskent University
  • Huseyin Yuce Bircan
    • Baskent University
  • Ertan Emek
    • Baskent University
  • Halime Cevik
    • Baskent University
  • Gulum Altaca
    • Baskent University
  • Gokhan Moray
    • Baskent University
Original Article

DOI: 10.1007/s11605-012-2067-3

Cite this article as:
Gunay, Y., Bircan, H.Y., Emek, E. et al. J Gastrointest Surg (2013) 17: 319. doi:10.1007/s11605-012-2067-3



Treatment of acute cholecystitis in chronic hemodialysis (HD) patients still remains controversial. Because of underlying disease that can influence surgical results, less invasive alternative managements have been tried over the last decades. The goal of this study was to analyze the results of cholecystectomy versus percutaneous cholecystostomy for acute cholecystitis (AC) in chronic HD patients.


All patients with end-stage renal disease who were treated for AC were identified retrospectively from our medical records. Between July 2007 and September 2011, 47 patients were treated for AC while they were on chronic HD. The records of these patients were reviewed for documented AC and its treatment.


Of the 47 HD patients, 26 (55.3 %) underwent cholecystectomy (CC), while 21 (44. 7 %) had a percutaneous cholecystostomy (PC) for AC as an initial treatment. The mean length of follow-up was 20.4 ± 16 months in PC and 18 ± 15 months in CC patients. The success rate was higher in CC patients compared to PC patients (92. 3 versus 66.7 %, p = 0.0698). Eleven (52. 4 %) patients who had PC subsequently underwent CC; six open CC and five delayed laparoscopic CC were performed. Of the 26 patients who underwent CC, 18 were performed emergently due to the persistence of AC-related symptoms and gangrenous and perforated gallbladders. Eight were initially treated conservatively and then underwent elective cholecystectomy at an interval of 32 ± 24 (range = 14–59) days following initial treatment. In emergent CC, 10 (55.6 %) were completed laparoscopically, three were open, and five (33.3 %) had conversions. In elective CC patients, two were conversions, but the remainder (75 %) had laparoscopic CC. Readmission rates were higher in the PC group (33.3 versus 12.5 %, p = 0.1732). Although AC-related mortality was higher in PC patients, there was no statistically significant difference in the patient survival rate between the two groups (Kaplan–Meier analysis, Fig. 1, 19 versus 7.7 %; p = 0.4035), and the overall mortality rate was higher in the PC group (33.7 versus 15.7 %, p = 0.2737).


This study confirms that the safety and effectiveness of CC has a higher success rate and lower morbidity and mortality rate compared with percutaneous cholecystostomy for acute cholecystitis in chronic HD patients.


Percutaneous cholecystostomyHemodialysisAcute cholecystitis


The incidence and prevalence of end-stage renal disease (ESRD) have continued to increase globally. However, the demand for kidney transplantation exceeds the supply of transplantable organs, and most patients with ESRD are treated with dialysis.1,2 The length of time the patient is on hemodialysis (HD) increases the rate of complications and chronic diseases in this population, including gallbladder diseases. The prevalence of gallstones in patients with ESRD on HD is higher than that of the general population.3,4 Although cholecystectomy is a definitive treatment for acute cholecystitis in the general population, there are few reports in the literature regarding the best treatment of acute cholecystitis in chronic HD patients.5 Because there are many underlying problems that can influence the outcome of surgery in ESRD patients, less invasive procedures have been considered a first line option in some patients. In order to avoid emergency surgery risks, percutaneous drainage and conservative management of acute cholecystitis (AC) are still reserved for this population. With advanced anesthesia, both perioperative and postoperative patient care have improved in the last decades, and emergent cholecystectomy may become the first option with acceptable results. The goal of our study was to evaluate treatment outcomes of acute cholecystitis in chronic HD patients and identify the morbidity and mortality associated with operative and nonoperative management.

Patients and Methods

This retrospective study was approved by the Institutional Review Board at Baskent University. A search of the institutional electronic medical records and a retrospective review of our dialysis, radiology, and general surgery databases at our institution showed that a total of 47 patients on chronic HD were treated for AC at our hospital between July 2007 and September 2011. The records of these patients were analyzed for documented AC and its treatment. Using electronic medical records, we collected data on patient demographics, age at surgery, gender, reason of kidney failure, length of HD, comorbidities, method of treatment for gallbladder diseases, American Society of Anesthesiology (ASA) classification, operative data, complications, and mortality. The diagnosis of AC was made based on clinical and laboratory findings, pain in the right upper quadrant or a positive Murphy sign, and a sign of systemic inflammation such as fever or elevated infection parameters (C-reactive protein (CRP) and or white blood count (WBC)). The clinical diagnosis was confirmed by either ultrasound or CT scan of the abdomen.


On admission, treatment models were implemented based on the patient’s general condition, including laboratory results, images and comorbidities, response to initial treatment, and the preference of the multidisciplinary team. All 47 patients were deemed to be of high risk due to their poor general condition and severe comorbidities. All patients received initial treatments that included intravenous fluids, nasogastric aspiration, analgesics, and antibiotics. Combinations of antibiotics were started based on the surgeon’s preference, including third-generation cephalosporin, metronidazole, and quinolones. If there was no clinical and laboratory improvement, patients who were considered unfit for surgery underwent percutaneous cholecystostomy (PC). Patients who had gangrenous or perforated gallbladders and had no response to conservative treatment but were considered fit for surgery underwent urgent cholecystectomy. All others were discharged and placed on the schedule for elective surgery. In cholecystectomy (CC) groups, if there was no gallbladder perforation, laparoscopic cholecystectomy was always attempted. Patients who had failed PC and underwent emergent CC were included in the PC group. Patients with choledocholithiasis were treated by endoscopic retrograde cholangiopancreatography (ERCP) following the PC, once the patient stabilized. In the CC group, patients had ERCP prior to surgery.

PC was performed under fluoroscopic and sonographic guidance. The Seldinger technique was the standard procedure for all interventions. An 8-F locking pigtail catheter (Flexima, Boston Scientific, USA) was preferred, but a 10-F catheter was placed if patients had dense purulent material obtained during the initial puncture. The catheters were flushed daily with 5–10 ml of sterile saline to prevent occlusion. All patients underwent regular HD three times weekly prior to AC, and HD was maintained after surgery beginning the first postoperative day.

Statistical Analysis

Data analysis was performed using MedCalc software (Mariakerke, Belgium). Continuous variables were expressed as mean ± SD. All parametric and nonparametric data were analyzed using Student’s t test and chi-squared test. Chi-squared test was used to identify significant differences between proportions and variables. Kaplan–Meier analysis was used to analyze patient survival rate following treatment. A p value <0.05 was considered statistically significant.


Patients’ Characteristics

Of the 47 HD patients, 26 (55.3 %) underwent CC, while 21 (44. 7 %) had a PC for AC as an initial treatment. Both groups were more than half female, and there was no significant difference between the two groups with regard to age (61 ± 15 (range = 36–77) versus 58 ± 18 (range = 29–79) years, p = 0.5442). Although all patients were sick, but there was no significant difference between the two groups with regard to the rate of high-risk patients who were identified by ASA ≥ 3 (76.1 versus 73 %, p = 0.9189) and preoperative Acute Physiology and Chronic Health Evaluation II (APACHE II) scores (20 ± 0.5 versus 19 ± 5.6, p = 0.4199). Moreover, there was no significant difference between two groups in regard to having more than one comorbidity, including (diabetes mellitus, hypertension, heart failure, chronic obstructive pulmonary disease) (42.3 versus 47.6 %, p = 0.9636). The most common cause of renal failure was diabetes mellitus (DM) in the PC group, while glomerulonephritis (GN) was the most common cause in CC patients. The mean time of HD was similar in both groups (9 ± 2.5 (range = 3–10) years versus 11 ± 6.4 (range = 4–16) years, p = 0.1510). Although calculus was the most common cause of AC in both groups, there were acalculus AC and malignancy as well. There were no significant differences in regard to the rate of choledocolithiasis, cholangitis, pancreatitis, and preoperative laboratory values (Table 1). In CC group, five patients (23.8 %) while, in PC group, seven patients (26.9 %) had symptomatic cholelithiasis in their past history. Three patients in the PC group had ERCP following their procedure, while two in the CC had preoperative ERCP. The mean duration of symptoms prior to admission and a mean interval between diagnosis and having the procedure were similar in both groups (p = 0. 3958 and p = 0.6647) (Table 1). All patients had HD following their procedure in 24 h, and seven in the PC group and four in the CC groups needed continuous veno-venous hemodialysis in the intensive care unit (ICU). Five (23.8 %) patients in the PC group and seven (27 %) in the CC group were taken to the ICU following their procedure.
Table 1

Baseline demographics and clinical characteristics of patients in HD


PC (n = 21)

CC (n = 26)

p value

Age (years)

61 ± 15 (36–77)

58 ± 18 (29–79)


Gender (F)

15 (71.4 %)

15 (57.6 %)


ASA grade ≥3

16 (76.1 %)

19 (73 %)



20 ± 0.5

19 ± 5.6


Cause of ESRD


8 (38 %)

5 (19 %)



4 (19 %)

7 (27 %)



5 (23.8 %)

9 (34.6 %)



4 (19 %)

5 (19 %)


Duration of HD (years)

9 ± 2.5 (3–10)

11 ± 6.4 (4–16)


Symptomatic cholelithiasis

5 (23.8 %)

7 (26.9 %)


Cause of AC


11 (52.4 %)

19 (73 %)



8 (38 %)

7 (30.4 %)



2 (9.5 %)




3 (14.3 %)

4 (15.4 %)



3 (14.3 %)

1 (3.8 %)



3 (14.3 %)

1 (3.8 %)


Duration of symptoms (days)

2 ± 0.9 (1–4)

1.8 ± 0.7 (1–3)




12.3 ± 3.1

14 ± 4.2



120 ± 93

132 ± 72


 Total bilirubin

7 ± 9.8

4 ± 6.4



136 ± 47

140 ± 32



165 ± 53

143 ± 71



4 ± 3.8

3.65 ± 2.7


Interval between diagnose and procedure (days)

1 ± 3.4 (0–2)

1.3 ± 0.9 (0–3)


Percutaneous Cholecystostomy

The transhepatic route was used more than the transperitoneal for placing catheters (76.2 versus 23.8 %, respectively). PC was successfully performed at the first attempt in all patients. The mean catheterization was 44 ± 17 (range = 16–67) days. The reasons for placement of PC were sepsis (n = 5), pneumonia (n = 2), poor general condition (n = 8), and having severe comorbidities (n = 6). Three patients were in the ICU when AC was diagnosed and were considered for PC. Of them, 14 responded to PC treatment: clinical and laboratory improvement was shown in 10 (47.6 %) patients in 24 h and four in 72 hours following the procedure. Eleven (52.4 %) patients who had PC underwent CC: four emergent CC and seven delayed CC were performed.


Based on clinical and laboratory improvement following conservative treatment and surgeon evaluation, the decision was made to either perform urgent or delayed CC. The mean time between admission and having CC was 1.3 ± 0.9 (0–3) days. Of the 26 patients who underwent CC, 18 were performed emergently after AC-related symptoms and gangrenous and perforated gallbladder symptoms did not resolve. Eight had an initial conservative treatment and then underwent elective cholecystectomy at an interval of 32 ± 24 (range = 14–59) days following initial treatment. Of emergent CC, 10 (55.6 %) were completed laparoscopically, three were open, and five (33.3 %) were conversions. In elective CC patients, two had conversions, but the remainder (75 %) had laparoscopic CC. The decision to perform either laparoscopic or open was made based on the patient’s general condition, ultrasound finding or computerized tomography, and surgeon preference. Multiple previous abdominal surgeries and perforated or gangrenous AC are the two main reasons for choosing open surgery. Of the three patients who had open surgery, two had gangrenous and perforated AC and one had multiple previous abdominal surgeries. Conversion occurred due to severe adhesions and prolonged operative time. Two patients in the emergent group, but five in the elective group, had a preoperative ERCP. Two common bile duct explorations were performed during an emergent open CC. Mean operative time was numerically higher in the emergent group, but this was not statistically significant (90 ± 35 min, range 55–130 versus 72 ± 30 min, range 40–105 min; p = 0. 2636). The postoperative stay at the hospital was statistically longer in the emergent group (2.3 ± 1.2 versus 3.5 ± 2.1, p = 0.0243), but the length of overall hospital stay was longer and statistically significant in elective CC patients (5 ± 3.5 versus 2.6 ± 2.8, p = 0.0057).


Successful treatment was identified as clinical and laboratory improvement following initial treatment and no need for further and alternative treatment during their hospital stay. Death related to AC due to sepsis was accounted as an unsuccessful result of the procedure. The success rate was numerically higher in CC patients than that in PC (92.3 versus 66.7 %, p = 0.0698) (Table 2). Of the 21 PC patients, clinical symptoms did not improve following PC in four (19 %), and they had to undergo emergent open CC (one had gallbladder perforation, two had gangrenous and pericholecystitis abscesses that were discovered in the operation, and the remainder of the situation deterioriated following PC).
Table 2

Outcomes of treatment for AC in HD patients

Variable (n = 21)


CC (n = 26)

p value


20.4 ± 16 (3–44)

18 ± 15 (4–30)


Success rate

14 (66.7 %)

24 (92.3 %)



4/12 (33.3 %)

1/8 (12.5 %)


Hospital stay

19.7 ± 24.6 (4–52)

3.7 ± 2.9 (1–9)



❑ Early

9 (42.8 %)

8 (30.8 %)


❑ Late

6(50 %)




❑ Procedure related




❑ AC-related

4 (19 %)

2 (7.7 %)


❑ Overall

7(33.3 %)

4 (15.4 %)


The overall hospital stay was statistically longer in PC patients (19.7 ± 24, 6 days versus 3.7 ± 2.9; p = 0.0019). During the hospital stay, although there were no statistically significantly differences in complication rates between the PC and CC groups (42.8 versus 30.8 %, p = 0.5851), the PC complications required interventions described as major complications (23.8 versus 3.8 %, p = 0.1086). In PC patients, two liver lacerations, one abdominal wall hematoma, and one bleed in the gallbladder occurred, and all were treated conservatively. In the PC group, three catheter obstructions and two dislodgements occurred, and all were managed by replacing the catheter. In the CC group, one patient was taken back to surgery due to common bile duct injury in patients with laparoscopic CC and repaired with T-tube placement. In the latter group, three wound infections occurred in open CC, and two fluid collections in the GB bed and two pneumonias were all treated conservatively.

The mean length of follow-up was numerically higher in PC patients, but no statistically significant difference was seen between the two groups (20.4 ± 16 versus 18 ± 15 months, p = 0.6587). During this period, some patients were readmitted due to AC and previous treatment. The readmission rate was numerically greater in the PC group (33.3 versus 12.5 %, p = 0.1732). Of the PC patients who were discharged following a successful procedure, two were readmitted for recurrent AC and both underwent open CC. Four catheter obstructions and two catheter slippages occurred, and all six were treated by replacing the catheter. Of these six patients, two patients were readmitted, while others were treated as outpatients. In the CC group that had been treated conservatively, one was readmitted for recurrent AC and treated conservatively and then scheduled for elective CC. In the 14 patients who had a successful PC, seven (50 %) underwent subsequent cholecystectomy: two were emergent and five were elective with a mean interval of 7 ± 3.5 months after discharge. In PC patients, those who had malignancy, one of whom who had unresectable distal common bile duct cancer underwent hepaticojejunostomy and cholecystectomy. Another patient had pancreas cancer and died with a PC catheter in place.

No procedure-related death occurred in either group. Although AC-related mortality was higher in PC patients, there was no statistically significant difference in the patient survival rate between the two groups (Kaplan–Meier analysis, Fig. 1, 19 versus 7.7 %; p = 0. 4035), and overall mortality rate was higher in the PC group (33.7 versus 15.7 %, p = 0.2737). Sepsis occurred in the two emergent open CC patients and caused death. Two CC died of unrelated causes. Three deaths occurred in PC patients due to sepsis and multiorgan failure while they were in the ICU, one died after an unsuccessful PS and underwent subsequent open CC, and three died from unrelated reasons during follow-up. Two patients in CC group received renal transplantation during this study.
Fig. 1

Survival curve of patients following PC and CC for AC in HD patients


In the general population, the incidence of cholelithiasis varies between 10 and 20 % depending on risk factors.6,7 Kazama et al. showed that the incidence of cholelithiasis increases in chronic end-stage renal disease.4 As the incidence of cholelithiasis increases in chronic HD, the risk of AC increases in this population. However, the method of choice for treatment of AC in chronic HD patients still remains controversial. Although there are many reports with regard to AC treatment in critically ill patients in the general population,8,9 there is limited evidence in the literature to compare PC and CC for AC in chronic HD patients. We analyzed all treatment options for AC in HD patients at our hospital.

Since emergent abdominal surgery carries a high risk of morbidity and mortality in hemodialysis patients,10,11 the less invasive procedure has became an option for AC in patients who are not fit for surgery. PC has been used in critically ill patients for AC, and its safety and less invasiveness brought it to the surgeon’s attention and were recommended for AC treatment in the general population.12,13 Patients with ESRD are always considered challenging patients for urgent surgery due to underlying disease and severe comorbidities. In light of the results in general critically ill patients, PC may be recommended as a palliative option for AC in patients with ESRD, who are unfit for surgery and do not respond to conservative treatment. However, with the advances in anesthesiology and laparoscopic surgery technique and postoperative care, the role of CC for AC in critically ill patient with ESRD has been increasingly gaining popularity. In the literature, it was reported that urgent cholecystectomy has higher rates of success in AC compared with PC.9 Because of the obvious advantages of delayed laparoscopic cholecystectomy, Miura et al. recommended that initial management with PC followed by elective CC is the preferable approach for AC in critically ill patients.14 However, it was reported that in critically ill patients, two-stage approaches increase morbidity and mortality rates.15

Many studies showed the efficacy and safety of PC with a high rate of success for AC in critically ill patients in the general population. In fact, our data showed relatively lower success rates of PC than those that have previously been reported in the general population. The success rate of PC was 66.7 % in this study, whereas it has been reported as 92.6, 78, and 100 % in different studies.1618 One of the reasons of the lower success rate in our patients could be because of our definition of success. We defined successful procedures as relieving the patient of all clinical and laboratory values consistent with AC and no need for alternative treatment during their hospital stay; patients were also discharged in a good condition. Moreover, death related to AC was classified as an unsuccessful procedure. Another reason for the lower success rates could be chronic renal failure. The success rate in patients with no ESRD was higher than that of patients on chronic HD in this study (66.7 versus 82.3 %, p = 0.3045). Moreover, a recurrence rate of 10–20 % of AC following PC is reported in the literature19 supported by our data (14.3 %). Although PC still remains an option in AC, almost half of patients eventually undergo subsequent CC (40 %),20 and this result was supported by our study (52.4 %). The mortality rate was higher in chronic HD than patients without ESRD following PC (19 versus 8.1 %). Moreover, the mortality rate was higher than that previously reported in the literature (0–11 %).21,22

Although PC with antibiotics can convert sepsis into nonseptic conditions in patients with AC, unfortunately, it may not help in patients with gangrenous and severe sepsis, and eventually, CC becomes necessary in those patients. Because the sepsis is caused by the gallbladder, the best treatment is to remove the source of infection rather than leave it in place. However, PC only removes the infected bile. Therefore, CC has been accepted as the first line treatment for AC in the general population, and in the last decade, LCC also became a preferable option as well.23,24 Open cholecystectomy is associated with a high risk of morbidity with postoperative pain and prolonged hospital stay.25 However, it is necessary for gangrenous, perforated gallbladders, malignancy, and other difficulties that increase conversion rates in laparoscopic procedures.26 Therefore, open CC can be performed for AC treatment at any time. Because laparoscopic surgery is less stressful than laparotomy, it may be advantageous in this population. Moreover, many studies have proven that laparoscopic CC is safe and effective in AC.24,25,27 Although laparoscopic surgery has become an acceptable procedure in AC, the conversion rate is still as high as 11–28 % in the general population24,28 and up to 39 % in those with advanced cholecystitis.29 However, some studies pointed out that there are no significant differences in conversion and morbidity rates between urgent and delayed laparoscopic CC.25 In our series, the conversion rate was higher in urgent laparoscopic cholecystectomy than that in delayed (33.3 versus 25 %, p = 0.7634). In this study, the conversion rate of laparoscopic cholecystectomy was 14.6 % in urgent and 5.8 % in delayed cholecystectomy in patients with no ESRD. Although common bile duct injury has been expressed as a concern in urgent laparoscopic cholecystectomy, this study was not consistent with that result.30 One common bile duct injury occurred in early laparoscopic surgery, but no injuries were seen in delayed patients. Another important issue is recurrent AC and other complications of gallstones.25 In this study, one (12.5 %) patient developed recurrent AC in CC and two in PC patients (14.3 %) while waiting for delayed laparoscopic cholecystectomy. This finding showed that delayed laparoscopic CC is the preferable choice to early laparoscopic surgery for AC in patients with ESRD due to higher rates of conversion and complications of urgent laparoscopic cholecystectomy.

The comparison in this study between PC and CC suggested that CC can be implemented for AC in chronic HD patients with a high success rate (92.8 versus 66.6 %). Our results proved that PC could be a palliative option in some patients for AC, but has higher rates of failure and complication with comparing CC. CC is a one-stage but PC is a two-stage treatment. Two-stage approaches carry high morbidity and mortality.15 Readmission is another issue in health economics, and the rate of readmission was reported as 5.6–23 % in the literature following PC.3133 In this study, the readmission rate was 33.3 % in PC and 12.5 % in CC patients. Moreover, the complication rate that required intervention was also higher in patients who had PC versus CC (23.8 versus 3.8 %, p = 0.1086). However, the mortality rate is predominantly related to underlying disease, comorbidity, and age rather than surgery or PC procedure. Reports in the literature showed that the mortality rate in PC is higher than that of CC (15.4 versus 4.5 %). In this study, AC-related mortality was consistent with the literature, and it was higher in the PC group than CC patients (19 versus 7.7 %). In addition, the overall hospital stay was statistically longer in PC patients (19.7 ± 24, 6 days versus 3.7 ± 2.9; p = 0.0019).

Cholecystectomy in prerenal transplant candidates with asymptomatic cholelithiasis is still controversial. When performed in patients who had a previous transplant, the risk of cholecystectomy complications increases, especially when performed as an emergency.34 Patients who submitted to prophylactic pretransplant cholecystectomy had lower operative morbidity and mortality. By contrast, patients who did not undergo pretransplant cholecystectomy and subsequently developed biliary complications requiring urgent operation were subject to higher morbidity and mortality rates.34,35 Therefore, we recommend a careful screening in this population and pretransplant cholecystectomy if cholelithiasis is discovered during transplant evaluation. We do not recommend routine screening and surgical treatment of asymptomatic cholelithiasis in patients with ESRD if they are not interested in renal transplantation.

In conclusion, this study confirms the safety and effectiveness CC for initial treatment of patients with acute cholecystitis in chronic HD patients. It seems that CC is a better choice than PC for AC in chronic HD patients in terms of morbidity and mortality for patients. This study showed that in the absence of a perforated or gangrenous gallbladder, laparoscopic cholecystectomy may be attempted in all cases, but delayed laparoscopic CC is recommended. Based on our experience, ESRD should not be an absolute contraindication for laparoscopic CC. Therefore, based on our experience, we recommend medical treatment of patients. If there is response to medical treatment, we recommend delayed laparoscopic CC.

This study had some shortcomings. CC was performed by different surgeons; therefore, the complication and conversion rates may have been affected by the surgeon’s experience. In order to identify the best treatment of AC in chronic HD patients, a randomized controlled trial is needed.

Copyright information

© The Society for Surgery of the Alimentary Tract 2012