The prevalence of cholelithiasis in an adult Western population is between 15 and 20% [1, 2]. In the United States, an estimated 11.8 million people aged between 20 and 74 years have gallbladder (GB) stones . Yearly approximately 1–2% of patients with silent gallstones develop symptoms and require treatment (Figure 1) [4–6]. Acute cholecystitis (AC) accounts for 20% of patients presented to hospital with right upper quadrant pain, which in patients with significant co-morbidities and in the elderly is associated with 2–3% mortality [5, 7, 8]. In the United States, in 2009, cholecystitis was the underlying cause of death in 2009 patients and a contributing cause of death in 3295 patients accounting for a mortality crude rate of 0.7 per 100,000 patients . In the same country, GB disease is one of the most common inpatient diagnosis that accounts for more than 260,000 hospital admissions and annual health care provider expenditure exceeding $3.03 billion .
Laparoscopic cholecystectomy (LC) is considered as the “gold standard” surgical technique for treatment of GB disease . Conversion of LC to open surgery (CTO) is used to prevent intra-abdominal organ injury (IOI), for common bile duct (CBD) exploration and to repair IOI.
The aim of this chapter is to review risk factors and predictive models for CTO, and surgical quality outcome measures.
2. Laparoscopic cholecystectomy
LC is the minimally invasive surgical operation that was introduced in clinical practice by Erich Muehe of Boeblingen, Germany, in 1985 . LC can be performed with the conventional four-port or three-port technique.
Compared to the conventional open cholecystectomy (OC), LC has decreased post-operative pain, decreased respiratory function dysfunction, reduced post-operative ileus, earlier oral fluid and food intake, better cosmesis, reduced patient’s hospital stay, fastened post-operative recovery and lowered morbidity and mortality [12–14].
A meta-analysis comparing LC with the small-incision OC (the length of incision of less than 8 cm) demonstrated that both techniques had similar rates of mortality; intraoperative, minor and severe post-operative complications (without bile duct injuries, BDI); BDI; total complications; and post-operative convalescence . A subgroup analysis of high-quality trials showed a shorter operative time for the small-incision OC than LC (weighted mean difference, random effects 16.4 min, 95% CI: 8.9–23.8 min) . Compared with the small-incision OC, the self-reported quality of life up to 30 days after LC is higher; 2326 (95% CI: 2187–2391) and 2411 (95% CI: 2334–2502), respectively, P = 0.03 .
Main disadvantages of LC compared to the conventional OC and small-incision OC are a lack of a three-dimensional view, narrow field of laparoscopic vision, inconvenience with liver retraction, insufficient tactile sensations due to manipulation with long laparoscopic instruments and difficulties with instruments placement and manoeuvring [17–19].
Another significant limitation of LC is an increased risk of IOI, including bile duct injury (BDI) [20, 21]. In the United States, approximately 750,000 LCs are performed annually . With the incidence of major BDI during LC fluctuating between 0.4 and 2%, it is expected that 3000 to 15,000 patients will suffer from iatrogenic BDI [21, 23, 24]. Major BDI is associated with significant morbidity, mortality and socioeconomic burden .
2.1. Indications for conversion of LC to open surgery
The primary indication for CTO is to prevent IOI. CTO can also be used for CBD exploration, to repair cholecysto-intestinal fistula and to perform an extended OC in patients with gallbladder cancer. In addition, CTO is performed to control haemorrhage and repair single or multiple IOI [25, 26].
Compared to LC, CTO is associated with an increased morbidity and mortality. A clinical audit of 7242 LCs for AC performed in the United States between 2005 and 2011 showed that compared to the LC group, patients who underwent CTO had higher rates of surgical site infection (1.8 versus 9.2%, P < 0.0001), operations for complications (1.4 versus 3.4%, P = 0.001), serious morbidity (3.8 versus 14.9%, P < 0.0001), overall morbidity (6.0 versus 21.8%, P < 0.0001) and mortality (0.5 versus 2.3%, P < 0.0001) . In addition, the CTO group of patients had a longer mean operation time (122.1 ± 51.0 min versus 80.0 ± 42.6 min, P < 0.0001) as well as the length of hospital stay (8.6 ± 13.0 days versus 3.4 ± 6.7 days, P < 0.0001) than LC patients .
2.2. Risk factors for conversion of LC to open surgery
As high rate of CTO and IOI can diminish clinical benefits and cost-effectiveness of LC, identification of preoperative and intraoperative patient-dependent and surgeon-related risk factors for CTO can be used for development of risk stratification models and refinement of the management. This will keep CTO at low rates and maintain benefits of minimally invasive GB surgery.
2.2.1. Preoperative patient-related risk factors
Preoperative patient-related risk factors for CTO have been extensively investigated and identified. In previous studies, the advanced age has been shown to be a risk factor for CTO [25, 27, 28]. In a meta-analysis, Yang et al. demonstrated that age >65 years is associated with a twofold increase in CTO rate (odds ratio (OR) = 1.8; 95% confidence interval (CI): 1.4–2.5; P < 0.0001) (Table 1) . These findings can be explained by a higher proportion of severe AC, GB cancer, choledocholithiasis and previous abdominal operations among the older patients compared to their younger counterparts [28, 30].
|Preoperative patient-related risk factors for CTO||OR||95 CI||P value||References|
|Advanced age (>65 years)||1.8||1.4–2.5||<0.0001|||
|Clinical diagnosis of AC||8||6.1–10.5||<0.00005|||
|Duration of AC >72 h||3.1||1.2–7.7||0.0072|||
|Repeated attacks of AC, 2||7.9||1.5–76.8||<0.0052|||
|Obesity, BMI >30 kg/m2||7.6||4.1–14||<0.001|||
|Previous upper abdominal surgery||20.4||2.4–927.4||0.0007|||
|Post-ERC/ES, 16 weeks||3||1.2–7.4||0.009|||
|WCC 11 109/L||4||2.5–6.1||<0.00005|||
|Elevated CRP, 10 mg/L||1.05||1.01–1.09||0.014|||
|Elevated total bilirubin||6.5||4.1–10.2||<0.00005|||
|Alkaline phosphatase (>135 U/L)||7||3.6–14||<0.00001|||
|Gallbladder wall thickness, mm||2||1.7–2.3||<0.00005|||
|Pericholecystic fluid on US||26||5.0–166.1||<0.00005|||
|ASA score >2||2.5||1.3–4.6||0.004|||
Abbreviations: CTO, conversion of laparoscopic cholecystectomy to open surgery; OR, odds ratio; CI, confidence interval; AC, acute cholecystitis; BMI, body mass index; ERC/ES, endoscopic retrograde cholangiography with endoscopic sphincterotomy; WCC, white cell count; CRP, C-reactive protein; U, units; L, litre; ASA, American Society of Anaesthesiologist; US, trans-abdominal ultrasound.
Note: acute hepatitis, other acute abdominal diseases, and chronic cholecystitis should be excluded.
Abbreviations: RUQ, right upper abdominal quadrant; CRP, C-reactive protein; WCC, white cell count; US, ultrasound; MRI, magnetic resonance imaging; CT, computed tomography; Tc-HIDA scan, 99mTc-hepatic iminodiacetic acid cholescintigraphy.
Male gender is a well-recognized risk factor for difficult LC and CTO [18, 25, 28, 29]. Males have more prominent adhesions between the GB and the omentum and surrounding internal organs, have a higher proportion of severe forms of AC on histological examination of the GB, have a higher CTO rate and require a longer operation time . Two-stage LC male patients have a three times higher rate of CTO (OR = 2.8; 95% CI: 1.1–6.6, P = 0.037) than females . This might be due to men’s reluctance to seek medical and surgical help, and their late hospital presentations after several repeated attacks of AC .
Patients with the clinical diagnosis of AC and higher severity grades of AC carry more chances of CTO [25, 33, 34]. Diagnostic criteria for AC are presented in Table 2 . AC patients have eight times higher risk of CTO than patients with uncomplicated gallstone disease (OR = 8.01; 95% CI: 6.1–10.5; P < 0.00005) . Severity grades of AC, as defined by the 2013 Tokyo Guidelines, are associated with an increased rate of CTO [36, 37]. The 2013 Tokyo Guidelines for severity grades of AC are shown in Table 3 . Severe AC makes LC technically more difficult, because AC is accompanied by extensive adhesions around GB and weakness of the GB wall, which preclude its retraction with laparoscopic forceps and cause GB perforation and spillage of infected bile and gallstones into the peritoneal cavity .
The longer the duration of untreated AC, the greater is the risk for CTO . Asai et al. showed that 12 of 29 (41.4%) patients in the CTO group and 36 of 196 (18.4%) patients in LC group had symptoms of AC for longer than 72 h (P = 0.0004) . Therefore, untreated AC lasting more than 2 days increases the risk for CTO by three times (OR 3.1; 95% CI: 1.2–7.7; P = 0.0072).
Repeated attacks of AC are associated with difficult LC and CTO [38–40]. When compared to AC patients with less than two previous episodes, patients with at least two past attacks of AC have an eightfold increase in the difficulty of LC (OR = 7.9; 95% CI: 3.4–18.2; P < 0.00005) and the rate of CTO (OR = 7.9; 95% CI: 1.5–76.8; P = 0.0052) . Compared to patients with one previous attack of AC, having at least three attacks of AC escalates LC difficulty by a factor of 28 (OR = 28.3; 95% CI: 7.4–127.6; P < 0.00005) and increases the rate of CTO a 14-fold (OR = 14.1; 95% CI: 2.1–153.2; P = 0.0021) .
Diabetes mellitus (DM) has been consistently shown to be associated with CTO [18, 25, 42]. Diabetics undergoing LC have a 2.5 times higher risk for CTO than nondiabetic patients (OR = 2.5; 95% CI: 1.3–4.4, P = 0.003) . This might be because diabetic patients have a threefold greater risk for the development of AC than non-diabetics (OR 2.7; 95% CI: 1.8–4.2; P < 0.00005) . Also, diabetics, especially if they are on insulin, have an 85% increased risk of preoperative GB perforation (adjusted OR = 1.85; 95% CI: 1.38–2.48; P < 0.001) . Poor glycaemic control and presence of diabetic microagiopathy and autonomic neuropathy, as well as frequent bactibilia, are important conditions that predispose diabetics to advanced forms of AC and infective complications [40, 46–49].
Obesity (body mass index (BMI) > 30 kg/m2) is not only a risk factor for CTO but also associated with major BDI [18, 28, 50, 51]. Obese patients undergoing LC have an eightfold higher risk of CTO than non-obese patients (OR = 7.6; 95% CI: 4.1–14; P < 0.001) . Compared to the non-obese patients, class I and class II–III obese patients have a two- and threefold increase in the risk of CTO (OR = 1.8; 95% CI: 1.1–2.8; P = 0.0105) and (OR = 2.7; 95% CI: 1.5–4.6; P = 0.0006), respectively .
Previous surgery above the umbilicus is a risk factor for CTO [52, 54]. In Lee’s study, 7 (17%) of 41 patients from the CTO group and 1 (1%) of 100 patients from the LC group had a history of previous upper abdominal surgery . This estimates the risk for CTO for patients with the past history of upper abdominal surgery 20 times higher than for those without previous surgery above the umbilicus (OR = 20.4, 95% CI: 2.4–927.4; P = 0.0007).
The risk of CTO is also higher in patients following endoscopic retrograde cholangiography with sphincterotomy (ERC/ES) for CBD stone clearance [54–56]. A two-stage LC after 15 weeks following ERC/ES increases the rate of CTO three times (RR = 2.7, 95% CI: 1.4–5.5, P = 0.004) and major BDI 10-fold (RR = 10.2, 95% CI: 1.1–95.7, P = 0.043) . Boerma et al. compared the two-stage LC conducted within 6 weeks after ERC with that performed after 6 weeks and found that the latter was more technically demanding and associated with a threefold increase in the CTO rate (RR = 2.7, 95% CI: 1.3–3.4, P = 0.01) . An association between preoperative ERC/ES and difficult LC can be explained by bactibilia-related inflammation of bile ducts with desmoplastic changes around the Calot’s triangle and shrinking of the GB [59, 60].
Thickened gallbladder wall on ultrasound (US) of the upper abdomen is associated with CTO [25, 28]. The risk of CTO doubles with every millimetre increase in gallbladder wall thickness (OR = 2; 95% CI: 1.7–2.3; P < 0.001) . Patients with GB wall thickness >5 mm on trans-abdominal US have a 16 times higher risk of CTO than those with GB wall thickness 3–5 mm (OR = 16.3; 95% CI: 8.1–33.3; P < 0.00005) . GB wall thickness >4 mm on US is not only a radiological marker for AC but also associated with greater operational difficulty [40, 61].
The presence of pericholecystic fluid on imaging of the abdomen increases the risk of CTO by 26 times (OR = 26; 95% CI: 5.0–166.1; P < 0.00005) . This radiological sign has the sensitivity of 70% in predicting CTO, the specificity of 92%, the positive predictive value of 33% and the negative predictive value of 98% .
An elevated white cell count (WCC) is a predictor of CTO . Compared to the LC group, the CTO group had a higher proportion of patients with leucocytosis, defined as WCC ≥ 11 × 109/L, 161 (12.7%) of 1265 patients versus 41 (36.6%) of 112 patients, respectively, P < 0.001 . This estimates a fourfold higher risk of CTO among patients with WCC ≥ 11 × 109/L (OR = 4.0; 95% CI: 2.5–6.1) compared with leucopenic and patients with normal WCC (WCC ≤ 11 × 109/L), P < 0.00005. In another study, Nidoni et al. showed that the sensitivity of WCC ≥ 11 ×109/L in predicting CTO is 80%, the specificity 83.5%, the positive and negative predictive values 22.2 and 98.6%, respectively .
Abbreviations: CRP, C-reactive protein; WCC, white cell count; AC, acute cholecystitis; AUC, the area under receiver operating characteristic curve.
An elevated C-reactive protein (CRP) level also predicts CTO . Every 10 mg/L increase in CRP concentration, the rate of CTO increases by 5% (OR = 1.05; 95% CI: 1.01–1.09; P = 0.014) . The cut-off point of CRP > 115 mg/L predicts CTO with the sensitivity of 79% and specificity of 57%. CRP as a predictor of CTO has the area under receiver operating characteristic curve (AUC) of 67% . An association between CTO and CRP concentration can be explained by severity of AC. The cut-off point of CRP concentration increases with more advanced histological forms of AC and its severity grades (Table 4) .
Deranged liver function tests predict CTO [28, 64]. Patients with elevated concentrations of total bilirubin and alkaline phosphatase (>135 U/L) have seven times higher risk of CTO, OR = 6.5 (95% CI: 4.1–10.2; P < 0.00005) and OR = 7.0 (95% CI: 3.6–14; P < 0.00001), respectively [25, 52]. Elevated levels of total bilirubin and alkaline phosphatase are both independently associated with choledocholithiasis . Open CBD exploration is one of the reasons for CTO.
An increasing ASA score has been shown in multiple studies to be an independent risk factor for CTO [25, 27, 28, 52, 62, 66]. Patients with ASA score of 3 have 2.5 times odds of CTO (OR = 2.5; 95% CI: 1.3–4.6) than those with ASA score of 1 (P = 0.004) .
2.2.2. Intraoperative patient-related risk factors
To prevent IOI during LC, when the surgeon encounters dense intra-abdominal adhesions, extensive inflammatory changes around the gallbladder, haemorrhage, inability to grasp and retract a friable gallbladder with forceps, CTO is advised .
Severe intra-abdominal adhesions make laparoscopic dissection very difficult and are associated with a fivefold increase in CTO risk (OR = 5.2; 95% CI: 1.9–14.4; P = 0.002) . Some authors report their institutional policy of avoidance of difficult dissections during LC and making a decision of CTO if there is a lack of dissection progress for 15–30 min .
Intense inflammatory infiltrate in the Calot’s triangle makes identification of the cystic duct and cystic artery very challenging predisposing patients to iatrogenic BDI and uncontrollable bleeding [32, 35]. To prevent BDI, CTO is directed when one of three fundamentals of the critical view of safety cannot be ascertained [30, 68]. These essentials include the clearance of the Calot’s triangle from adipose and fibrous tissue, detachment of the lowest part of the gallbladder from the GB bed and identification of the cystic duct and cystic artery going into the gallbladder (Figure 2) . Alternatively, when the surgeon encounters severe inflammatory or desmoplastic reaction in the Calot’s triangle laparoscopic subtotal cholecystectomy can be performed [70–72].
Unclear biliary anatomy is another reason for CTO . Instead, intraoperative cholangiography (IOC) can be used to prevent misidentification of the cystic duct and prevent BDI [73, 74]. Failure of the contrast to opacify the common hepatic duct and the right and left hepatic ducts would signal the surgeon that the CBD, not the cystic duct, has been cannulated . The use of IOC is associated with a 62% reduction in CTO rate (OR = 0.38, 95% CI: 0.17–0.94; P = 0.04) . If IOC does not facilitate unmistakeable biliary ducts recognition, then CTO is indicated .
Abbreviations: CTO, conversion of laparoscopic cholecystectomy to open surgery; CBD, common bile duct.
CTO is performed when open CBD exploration and stone clearance is required provided CBD stones cannot be removed laparoscopically or post-operatively by ERC/ES [20, 76, 77]. Furthermore, CTO is advised for an open repair of intraoperatively diagnosed IOI [20, 25]. Infrequently, CTO is necessary when the patient is not able to tolerate 12 mm Hg intraperitoneal carbon dioxide insufflation pressure and develops premature ventricular beats and bradycardia with hypotension .
Intraoperative patient-related risk factors that may require conversion of laparoscopic cholecystectomy to open surgery are presented in Table 5.
2.2.3. Surgeon-related risk factors
Surgeon’s knowledge, laparoscopic fellowship training, operative experience and skills in laparoscopic surgery play an important role in timely recognition of the need for CTO and are important predictors for CTO [49, 74, 75, 78]. Surgical registrars (postgraduate year 4–5) have a twofold higher rate of CTO (OR = 1.7; 95% CI: 1.1–2.5; P = 0.0067) than surgical consultants . Compared to general surgical registrars without fellowship training, completion of the surgical fellowship program is associated with 92% reduction in the rate of CTO (OR = 0.08; 95% CI: 0.02–0.32, P < 0.0001) .
Importantly, there is statistically significant inverse correlation between surgeons’ LC volume and the rate of CTO (P = 0.03) . The surgeons with more than 5 years of experience in independent practice, who performed at least 100 LC or more than 75 cases annually are considered to be experts, and the surgeons with personal records of less than 100 LC as the first surgeons to be non-experts [81, 82]. In addition, surgeon’s specialization influences the incidence of CTO. The expertise of upper gastrointestinal (UGI) surgeons is recognized to be the standard in LC against which the other surgeons’ capability can be measured. Compared to the UGI surgeons, non-UGI surgeons have a two times higher incidence of CTO (OR = 2.1; 95% CI: 1.1–3.7; P = 0.0122) .
2.3. Predictive models for conversion of LC to open surgery
Few predictive models have been developed to help the surgeon make an early CTO decision. Lipman et al. found that variables such as male gender, WCC ≥ 11 × 109/L, low albumin, pericholecystic fluid on US, the presence of diabetes mellitus and elevated total bilirubin independently predict CTO. These risk factors were included into the model which has the AUC of 83%. The authors showed that if none of these risk factors were present, the risk of conversion is 2%, but when six risk factors were present, the risk of conversion escalated to 90% .
Kama et al. presented a CTO risk scoring model consisting of a constant (-20) and six variables with their coefficients, age ≥60 years (coefficient of 5), male gender (11), previous upper abdominal surgery (8), abdominal tenderness (9), thickness of gallbladder wall >4 mm on US (13) and the clinical diagnosis of AC (15) . The final risk score for CTO (RSCLO) is the sum of the constant and coefficients of the risk factors that are present in an individual patient. The RSCLO can take a value between −20 and 41. An operation with RSCLO exceeding −3 is considered difficult .
Goonawardena et al. proposed a CTO prediction model that is constructed on five independent variables, previous upper abdominal surgery, obesity (BMI > 30 kg/m2) and the presence of choledocholithiasis, impacted stone at the Hartmann’s pouch and GB wall thickening on the trans-abdominal US .
Sugrue et al. developed an intraoperative 10-point scoring system for an assessment of the difficulty in LC . A score of <2 indicates mild degree difficulty, 2–4 moderate, 5–7 severe and 8–10 the extreme difficulty of LC .
These predictive models have limitations. They have not been tested on an independent sample. Therefore, their real-life predictive ability is unknown. In addition, these models excluded surgeon-related risk factors. Thus, it is difficult to tailor management of high-risk for CTO patients according to an available hepato-biliary expertise.
3. Outcomes and quality indicators in laparoscopic cholecystectomy
Time to operation theatre (TTO) is the period between hospital admission and the start of operation. TTO is correlated with the length of the index hospital admission and associated with an increased expenditure for health care provider. Fry et al. showed that the length of preoperative hospitalization exceeding 3 days (96 h) incurs an added cost of 2011 US$ 7584 .
The CTO rate is the proportion of cases of LC converted to open surgery to the total number of intended LC. Mueller et al. consider CTO as a quality variable, because CTO is associated with a worse outcome . Other authors do not regard CTO as a complication, but rather as a “sign of experience” or “mature judgment” to prevent IOI [20, 28]. We classify all CTO into three groups: (1) the safety CTO when conversion is performed to prevent IOI; (2) conversion of LC to extended surgery is implemented for an open CBD exploration, for closure of cholecysto-intestinal fistula, or for an extended OC in case of GB cancer; and (3) the emergency CTO is performed to stop bleeding or repair IOI detected intraoperatively. In this view, safety CTO and extended surgery CTO are not complications, but carefully thought-out surgical strategies.
Laparoscopic BDI rate is the proportion of injury to bile ducts that occurred during LC to the total number of LCs. Commonly, BDIs are specified according to the Strasberg’s classification .
IOI rate is the proportion of iatrogenic injury to one or more intra-abdominal organs, including the liver, bile ducts, small and large bowel, stomach, spleen, pancreas, mesentery and vascular injury accompanied with extensive haemorrhage to the total number of cholecystectomies.
An association between CTO and IOI is a surgical quality indicator that shows that CTO is used predominantly as an emergency surgical strategy to control haemorrhage and repair IOI diagnosed intraoperatively .
Procedures or operations performed for treatment of complications are procedures performed for complications during the index hospital admission and readmissions.
Severity of post-operative complications can be graded using the Dindo-Clavien classification . This classification does not specify the hospital admission during which they developed. We suggest using the Dindo-Clavien classification to grade post-operative complications that arose during the index hospital admission.
The length of stay (LOS) is the duration of the index hospital admission.
Readmission within 30 or 90 days from the index hospital discharge is the readmission to any acute care hospital with a condition that could be considered an outcome of the procedure or operation . To define readmissions period, some authors use a 30-day period after the index hospital admission discharge, the others a 90-day period [86, 91, 92]. We think that the 90-day post-index hospital stay readmission period is more meaningful, because it is more likely to capture an adverse event due to a retained CBD stone rather than a 30-day period.
The post-operative length of stay (LOS) below 2 days (72 h) with no hospital readmission is an indicator of an uncomplicated cholecystectomy .
The prolonged post-operative LOS exceeding 3 days (96 h) is an indicator of protracted admission after LC.
The total length of hospital stay includes the length of the index hospital admission and duration of readmissions within 90-day post-index hospital stay discharge.
Thirty-day or 90-day postdischarge death is the death that occurs within 30 days or 90 days from either the index hospital admission or hospital readmission discharge.
Patient-reported outcome measures including generic (the Short Form 36 (SF-36) Health Survey, Nottingham Health Profile), preference-based (European Quality of Life Questionnaire, EQ-5D) and condition-specific instruments (Otago Gallstones Condition-Specific Questionnaire, Gastrointestinal Quality of Life Index, Abdominal Surgery Impact Scale and Gallstone Impact Checklist) as well as economic evaluations (cost–minimization analysis, cost–consequence analysis, cost–effectiveness analysis, cost–utility analysis and cost–benefit analysis) can also be used as outcome measures.
LC is the treatment of choice for symptomatic GB disease, which in some patients requires CTO. CTO risk stratification based on patient- and surgeon-dependent variables may allow a better patient’s management to keep CTO at low rates and maintain benefits of minimally invasive GB surgery. The absence of an association between CTO and IOI is an important surgical safety indicator that demonstrates that CTO is used as a safety strategy rather than an emergency measure to repair iatrogenic IOI and control haemorrhage.