- Email: [email protected]
Age-related differences pre-, intra-, and postcholecystectomy: A retrospective cohort study of 6,868 patients
International Journal of Surgery 39 (2017) 119e126
Contents lists available at ScienceDirect
International Journal of Surgery journal homepage: www.journal-surgery.net
Age-related differences pre-, intra-, and postcholecystectomy: A retrospective cohort study of 6,868 patients* Hadia Maqsood a, 1, Kalpesh Patel a, 1, Hamid Ferdosi a, b, 1, Anne M. Sill a, Bin Wu a, Thomas Buddensick a, Amanda Sautter a, Haroon Shaukat a, Gisela Sulkowski a, Dusty Marie Narducci a, Mustafa Siddique a, Farin Kamangar c, Gopal C. Kowdley a, Steven C. Cunningham a, * a
Department of Surgery, Saint Agnes Hospital, Baltimore, MD, USA Department of Epidemiology & Biostatistics, George Washington Univ., Washington, DC, USA c Department of Public Health Analysis, Morgan State Univ., Baltimore, MD, USA b
h i g h l i g h t s Older patients undergoing cholecystectomy (CCY) have distinct perioperative characteristics compared with younger patients. Among older patients, preoperative testing, blood loss, operative times and complications are greater. The risk for developing complications increased by 2% per year of life.
a r t i c l e i n f o
a b s t r a c t
Article history: Received 23 November 2016 Received in revised form 7 January 2017 Accepted 12 January 2017 Available online 16 January 2017
Background: Cholecystectomy (CCY) is increasingly performed in older individuals. We sought to examine age-related differences in pre-, intra-, and postoperative factors at a community hospital, using a very large, single-institution cholecystectomy database. Materials and methods: A retrospective review of 6868 patients who underwent CCY from 2001 to 2013 was performed. ROC analysis identified the optimal age cutoff when complications reached a significant inflection point (<55 and 55 years). Multiple clinical features and outcomes were measured and compared by age. Logistic regression was used to examine how well a set of covariates predicted postoperative complications. Results: Older patients had significantly higher rates of comorbidities and underwent more extensive preoperative imaging. Intraoperatively, older patients had more blood loss, longer operative times, and more open operations. Postoperatively, older patients experienced more complications and had significantly different pathological findings. While holding age and gender constant, regression analyses showed that preoperative creatinine level, blood loss and history of previous operation were the strongest predictors of complications. The risk for developing complications increased by 2% per year of life. Conclusion: Older patients have distinct pre-, intra-, and postoperative characteristics. Their care is more imaging- and cost-intensive. CCY in this population is associated with higher risks, likely due to a combination of comorbidities and age-related worsened physiological status. Pathologic findings are significantly different relative to younger patients. While removing the effect of age, preoperative creatinine levels, blood loss, and history of previous operation predict postoperative complications. Quantifying these differences may help to inform management decisions for older patients. © 2017 IJS Publishing Group Ltd. Published by Elsevier Ltd. All rights reserved.
Keywords: Cholecystectomy Elderly Young Complications Age Adenomyosis Cholesterolosis
* Submitted to International Journal of Surgery (an earlier version of this work was presented at the Americas Hepato-Pancreatico-Biliary Association in Miami, FL, USA 2013). * Corresponding author. Saint Agnes Hospital and Cancer Institute, 900 Caton Avenue, MB 207 Baltimore, MD, 21229, USA. E-mail address: [email protected] (S.C. Cunningham). 1 Maqsood, Patel, and Ferdosi contributed equally to this work.
http://dx.doi.org/10.1016/j.ijsu.2017.01.046 1743-9191/© 2017 IJS Publishing Group Ltd. Published by Elsevier Ltd. All rights reserved.
120
H. Maqsood et al. / International Journal of Surgery 39 (2017) 119e126
1. Introduction Gallbladder disease is one of the most common indications for abdominal surgery. The prevalence of gallbladder disease requiring cholecystectomy (CCY) is highest in older patients, with 8.1% of men, and 16.6% of women aged 60e74 years undergoing CCY [1]. Compounding this problem is the fact that the average age of many global populations is increasing. For example, the United States Census predictions estimate that the percentage of the population 65 years of age and over will increase from 14.8% in 2015 to 20.2% (or an additional 30 million individuals) by the year 2035 [2]. The aim of our study is to critically examine our experience in the treatment of gallbladder disease in younger versus older patients. A sub-aim was to identify risk factors for postoperative complications. We examined differences in pre-, intra-, and postoperative characteristics of older versus younger patients in order to identify areas where improvements may be made to optimize care for older patients undergoing CCY. 2. Methods 2.1. Study population and data collection A recently updated database of patients undergoing CCY at Saint Agnes Hospital from 2000 to 2013 was analyzed. This study was approved by the Institutional Review Board. Patients who underwent CCY as part of another major operation such as a Whipple, liver resection, bariatric operation, or cytoreduction with hyperthermic intraperitoneal chemotherapy, were excluded from the study. Those with a preoperative diagnosis of gallbladder cancer were also excluded. CCY was performed according to individual surgeon preferences, after an evaluation period that was similarly surgeon-dependent, making use of standard laboratory testing and imaging modalities, as dictated by the clinical scenario. Data were obtained from hospital charts, including operative notes, pathology reports, imaging study reports, and lab reports. Standard patient demographic and clinicopathological data were collected, including age, gender, race, ASA score, comorbidities, operative and pathologic details, and postoperative morbidity. Perioperative morbidity was severity-graded based on the ClavienDindo classification system [3]. Data were entered into Excel™ (Redmond, WA) and imported into SPSS™ version 21.0 (Chicago, IL) for statistical analysis. 2.2. Statistical considerations 2.2.1. Sample size This study sample represents a convenience sample of previously collected cases undergoing CCY. Although not formally “powered” to show differences in preselected measures, this exploratory study of 2883 younger and 2166 older patients likely provides adequate sample size to confidently demonstrate true population differences. 2.2.2. Analysis A receiver operator characteristic (ROC) curve identified the point along the age continuum (age 55 years) at which the number of postoperative complications increased significantly. Therefore, patients were stratified by using this age cutoff (<55 and 55 years of age) and these groups served as the main comparison groups throughout the study. Demographic, clinical, diagnostic, pathologic, and perioperative characteristics were compared using appropriate, univariate and bivariate parametric tests (frequencies, student t-test, chi-square test) and by nonparametric methods for ordinal and nominal
variables. Patients with missing data were excluded. Understanding the logical confounders in the relationships between age and number of complications and morbidities, type of operation (eg, open or laparoscopic) and level of urgency (ie, emergent or elective), a binary logistic regression model was built to identify the most prominent predictors of developing 30-day postoperative complications. We controlled for age to determine the ability of the predictors to explain the variance in the outcome, i.e., presence or absence of postoperative complications. Given that 72% of the population was female, we also controlled for gender so as to avoid the overrepresentation of females in the regression analyses. The odds of developing complications and their 95% confidence intervals were also computed. To examine independent predictors of postoperative complications, bivariate analyses of preoperative measures of laboratory markers, preexisting comorbidities and postsurgical pathologic findings were compared for those with and without complications within 30 days of operation. Variables found to be significantly different between those with and without complications (P < 0.05) became candidates for binomial logistic regression. A missingvalues analysis was then conducted to select predictors with the least number of missing values to provide ample sample size for a regression analysis. In an effort to revert the syllogism of the cause and effect of age and complications, we controlled for age in the binomial regression models; gender was also held constant to remove the overrepresentation of females in our population. Covariates were removed in a backward elimination fashion to identify the predictor(s) that gave the most weight in their ability to explain the variance in the outcome (postoperative complications). 3. Results 3.1. Selection Of 6868 patients with varying indications undergoing CCY, 5049 patients were analyzed in this study after excluding cases that had incomplete data, or had the CCY performed secondary to another major operation, as described above. 3.2. Patient characteristics Fifty-seven percent of patients were 55 years old (this cutoff chosen based on the ROC curve, as described above; Table 1). Both age groups contained significantly more females than males. Racial/ ethnic distribution across age groups also varied significantly and appears to be partially explained by a greater number of younger Hispanic patients. The majority (79%) of the younger group had ASA scores of 1e2 while the majority (54%) of older patients had scores of 3e4. The older patients had a significantly higher percentage of comorbidities such as cardiovascular diseases, hypertension, cerebrovascular accident, coronary artery disease, chronic obstructive pulmonary disease, diabetes mellitus, renal failure (all P < 0.001), alcohol abuse, and hepatic cirrhosis (P ¼ 0.05). The only exception was hepatitis infection, which, although higher in younger adults, did not reach statistical significance. 3.3. Imaging studies A significantly higher percentage of patients >55 years old had preoperative investigations such as computed tomography (CT), magnetic resonance imaging (MRI), hepatobiliary iminodiacetic acid (HIDA) scans and endoscopic retrograde cholangiopancreatography (ERCP) performed (Table 2). Only the ultrasound (US) showed significantly larger gallbladder wall thickness
H. Maqsood et al. / International Journal of Surgery 39 (2017) 119e126
121
Table 1 Patient characteristics.
Number of patients Age (mean years, SD) Male Female Caucasian African American Asian Hispanic Other Heart rate at triage in ER ASA score 1 or 2 ASA score 3 or 4 ASA score 5 Congestive heart failure Hypertension Cerebrovascular accident Coronary artery disease COPD Renal failure Diabetes mellitus Alcohol abuse Hepatitis infection Ascites (moderate to severe) Hepatic encephalopathy Cirrhosis Fibrosis
Age Gender Race
Heart Rate ASA
Patient History
55 years N (%)
>55 years N (%)
2883 (57) 39.24 (10.4) 608 (21) 2275 (79) 1853 (64) 866 (30) 16 (0.6) 91 (3.2) 55 (1.9) 80.27 (15.88) 508 (79) 126 (20) 9 (1.4) 38 (1.3) 673 (23) 15 (0.5) 109 (3.8) 61 (2.1) 31 (1.1) 269 (9.3) 356 (12) 106 (3.7) 2829 (98) 65 (17) 20 (0.7) 2 (0.1)
2166 (43) 70.47 (9.91) 766 (35) 1400 (65) 1644 (76) 458 (21) 11 (0.5) 17 (0.8) 33 (1.5) 78.13 (17.17) 180 (45) 216 (54) 2 (0.5) 223 (10) 1340 (62) 56 (2.6) 550 (25) 285 (13) 170 (7.9) 527 (24) 321 (15) 55 (2.6) 3023 (98) 49 (18) 29 (1.3) 4 (0.2)
P-value
<0.001 <0.001 <0.001
0.023
<0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 0.011 0.104 0.120 0.784 0.021 0.239
Abbreviations: COPD, chronic obstructive pulmonary disease; otherwise as per text.
Table 2 Diagnostics tests.
US
CT
MRI
HIDA
ERCP
US performed Gallstones Gallstone size (mm) GBW thickness (mm) CBD diameter (mm) Pericholecystic fluid Calcification CT performed GBW thickness (mm) CBD diameter (mm) Pericholecystic fluid Calcification MRI performed GBW thickness (mm) CBD diameter (mm) Pericholecystic fluid Calcification HIDA performed Symptoms upon CCK injection Ejection fraction (%) Time to GB visualization (min) Time to duodenum visualization (min) Preoperative ERCP Postoperative ERCP
55 years Mean (SD) or N (%)
>55 years Mean (SD) or N (%)
P-value
2190 (70) 1324 (89) 15.42 (8.14) 4.49 (2.61) 4.74 (2.25) 205 (8.1) 91 (3.6) 568 (18) 3.66 (2.28) 11.67 (14.61) 92 (5.9) 27 (1.7) 223 (7.1) 6.36 (4.29) 7.08 (3.14) 57 (4.1) 12 (0.9) 744 (24) 400 (79) 22.64 (23.91) 47.40 (179.23) 58.27 (163.29) 82 (2.60) 56 (1.80)
1676 (69) 1092 (87) 15.13 (7.59) 5.06 (4.38) 5.49 (2.69) 212 (11) 52 (2.8) 816 (34) 4.73 (3.96) 14.61 (2.40) 143 (10.00) 40 (2.8) 256 (11) 4.92 (4.48) 7.90 (3.88) 60 (5.4) 13 (1.2) 644 (27) 380 (84) 24.43 (25.38) 124.21 (547.92) 84.38 (286.18) 145 (6.00) 56 (2.30)
0.737 0.592 0.746 0.006 <0.001 0.001 0.102 <0.001 0.365 0.398 <0.001 0.053 <0.001 0.382 0.085 0.139 0.458 0.013 0.085 0.362 0.002 0.091 <0.001 0.160
Abbreviations: as per text.
and common bile duct diameter in older group but interestingly, CT and MRI failed to show these expected differences. Significantly more older patients had pericholecystic fluid on US and CT than did the younger patients (P < 0.001). Time to visualize the gallbladder at HIDA scan was significantly longer in older patients (P ¼ 0.002). 3.4. Pathologic diagnosis Histopathologic findings (Table 3) showed that rates of acute
(14% versus 24%, P < 0.001), gangrenous (2.7% versus 8.1%, P < 0.001), and hemorrhagic cholecystitis (3.2% versus 5.5%, P < 0.001) were significantly higher in the older CCY patients while younger patients had a significantly higher rate of chronic cholecystitis (86% versus 83%, P < 0.001). Likewise, rates of adenomyosis (0.3% versus 0.9%, P ¼ 0.006), sludge (4.7% versus 7.7%, P < 0.001), GB cancer (0.1% versus 1.1%, P < 0.001) and other histopathologic findings (5.8% versus 8.6%, P < 0.001) were significantly higher in older CCY patients, but younger patients had much higher rates of
122
H. Maqsood et al. / International Journal of Surgery 39 (2017) 119e126
Table 3 Pathology results.
Histopathology findings
Blood and serum chemistry
Acute cholecystitis Chronic cholecystitis Gangrenous cholecystitis Perforated cholecystitis Hemorrhagic cholecystitis Cholelithiasis Adenomyosis Diverticulosis Gallbladder polyp Cholesterolosis Sludge Cancer Other
White blood cells (1000s/uL) Blood urea nitrogen (mg/dL) Creatinine (mg/dL) Glucose (mg/dL) Lactate (mmol/L) Aspartate transferase (U/L) Alanine transferase (U/L) Alkaline phosphatase (U/L) Total bilirubin (mg/dL) Direct bilirubin (mg/dL) Albumin (g/dL) Prothrombin time (s) International normalized ratio
55 years N (%)
>55 years N (%)
P-value
440 (14) 2775 (86) 85 (2.7) 7 (0.2) 103 (3.2) 2424 (76) 10 (0.3) 3 (0.1) 38 (1.2) 743 (23) 149 (4.7) 2 (0.1) 186 (5.8)
589 (24) 2033 (83) 195 (8.1) 11 (0.5) 132 (5.5) 1864 (76) 21 (0.9) 3 (0.1) 26 (1.1) 308 (13) 185 (7.7) 26 (1.1) 207 (8.6)
<0.001 <0.001 <0.001 0.123 <0.001 0.646 0.006 0.734 0.682 <0.001 <0.001 <0.001 <0.001
55 years Mean (SD)
>55 years Mean (SD)
P-value
9.13 (4.19) 11.23 (8.02) 0.88 (0.87) 111.48 38.08) 75.80 (157.56) 69.14 (139.50) 81.16 (153.76) 98.37 (65.54) 0.78 (1.80) 0.27 (0.75) 3.92 (0.63) 11.34 (2.92) 1.03 (0.75)
9.83 (5.67) 18.03 (12.90) 1.19 (1.36) 128.62 (56.93) 37.50 (141.75) 73.28 (138.68) 76.81 (129.35) 125.87 (113.58) 1.19 (2.20) 0.48 (1.27) 3.58 (1.32) 12.55 (5.74) 1.34 (3.36)
<0.001 <0.001 <0.001 <0.001 0.354 0.420 0.414 <0.001 <0.001 <0.001 <0.001 <0.001 0.002
Abbreviations: as per text.
cholesterolosis than older patients (23% versus 13%, respectively, P < 0.001). Serum values for white blood cells, blood urea nitrogen, creatinine, glucose, alkaline phosphatase, total and direct bilirubin, prothrombin time and international normalized ratio were significantly higher in the older group (P < 0.05). There was no significant difference in the serum levels of lactate, aspartate transferase and alanine transferase between the two groups.
3.5. Perioperative course As shown in Table 4, a significantly higher percentage of older patients had a history of prior abdominal operations (78% versus 70%, P < 0.001) and underwent an urgent surgery than did the younger group (42% versus 37%. P < 0.001). Older patients also had a significantly higher estimated blood loss (EBL) during surgery (82.79 mL versus 51.44 mL, P < 0.001) as compared to the younger patients. There was no statistically significant difference in the duration of operation of the two groups. Postoperative hospital length of stay (LOS-PO) (4.3 days versus 2.2 days, P < 0.001) and total hospital length of stay (LOS-T) (6.1 days versus 3.2 days, P < 0.001) were significantly longer for older patients. Chi-square analysis revealed significantly higher rates of conversion from laparoscopic to open CCY procedures in the older patients (5.2 versus 1.6%, respectively, P < 0.001). Also of note is the significantly higher rate of primarily open CCY in the older group. When comparing age-related differences in postoperative complication rates by type of approach (laparoscopic, open, and laparoscopic converted to open), complication rates were significantly highest among laparoscopic-converted-to-open CCY, but rates of complications were nearly double in the older group across all three types of approach (Table 4; P < 0.001). Older patients suffered significantly more postoperative complications which were generally of a higher severity (Clavien
grading system [3]) compared with their younger counterparts. Those who experienced postoperative complications within 30 days of surgery were significantly older than those with no complications (Table 5). Significant increases in several laboratory parameters were also observed for those who experienced postoperative complications with the exception of albumin, which was significantly lower in those with complications, and in the elderly group (all P < 0.001). Significantly more patients with postoperative complications had a history of one or more comorbidities, previous abdominal surgeries, and intraoperative finding of gallbladder cancer (Table 6). While holding age and gender constant, a binomial logistic regression modeling showed the predictive value of each candidate covariate in step 1, of which 5 covariates independently explained the majority of the variance of postoperative complications (P ¼ 0.163). Further tightening of the model left 3 covariates (preoperative creatinine levels, estimated blood loss and previous operation) to predict the occurrence of postoperative complications (Table 7). The odds of a postoperative complication increased by1.02 (95% CI ¼ 1.02e1.33) per year of life.
4. Discussion As global populations continue to age, the particular characteristics of older patients versus younger patients gains much importance, especially when prognosticating outcomes for common procedures such as cholecystectomies. In the current study, using one of the largest single-institution cholecystectomy databases in the world, we elucidated several important aspects of this population. The use of preoperative imaging studies was significantly higher among older patients and these older patients were significantly more likely to undergo an urgent (as opposed to an elective)
H. Maqsood et al. / International Journal of Surgery 39 (2017) 119e126
123
Table 4 Perioperative data.
Duration of operation (min) EBL (mL) Urgency (urgent/emergent/inpatient) LOS-T (days) LOS-PO (days) Laparoscopic cholecystectomy only Laparoscopic cholecystectomy with IOC Open cholecystectomy (begun open) Open cholecystectomy (begun open) with IOC Laparoscopic converted to open cholecystectomy Laparoscopic converted to open cholecystectomy with IOC Conversion rates (laparoscopic to open CCY)
Type of operation
Any complication
Grade of complication
1 2 3a 3b 4a 4b 5 Infectious Cardiovascular Pulmonary Renal Gastrointestinal Hemorrhage Other None One or more
Type of complications
Previous abdominal incisions
55 years Mean (SD)
>55 years Mean (SD)
P-value
133.04 (1203.15) 51.44 (129.91) 1168 (37.1) 3.18 (4.10) 2.20 (3.66) 2410 (85.0) 301 (11) 45 (1.6) 14 (0.5) 40 (1.4) 24 (0.8) 64 (1.6)
102.12 (130.06) 82.79 (159.47) 1020 (42.3) 6.07 (8.78) 4.31 (7.59) 1616 (73.5) 252 (12) 153 (7.0) 41 (1.9) 84 (3.8) 53 (2.4) 137 (5.2)
0.523 <0.001 <0.001 <0.001 <0.001 <0.001
150 (4.9)
248 (10)
<0.001
55 years N (%)
>55 years N (%)
P-value
0 61 (1.9) 31 (1.0) 14 (0.4) 0 0 2 (0.1) 19 (0.6) 3 (0.1) 5 (0.2) 2 (0.1) 0 3 (0.1) 26 (0.8) 963 (64.6) 528 (35.4)
0 109 (4.4) 54 (2.2) 12 (0.5) 17 (0.7) 4 (0.2) 11 (0.4) 43 (1.7) 27 (1.1) 25 (1.0) 3 (0.1) 1 (0.0) 6 (0.2) 37 (1.5) 1728 (52.1) 1591 (47.9)
<0.001
<0.001
<0.001
<0.001
Abbreviations: IOC, intraoperative cholangiogram; otherwise as per text.
Table 5 Complications by age group by CCY approach. 55 years
Type of CCY
Laparoscopic CCY Open CCY Converted lap / open Total
N % N % N % N %
>55 years
P-value
No complications
Complications
No complications
Complications
2561 95.6 53 88.3 53 82.8 2667 95.1
119 4.4 7 11.7 11 17.2 137 4.9
1715 92.0 153 76.1 102 73.4 1970 89.3
150 8.0 48 23.9 37 26.6 235 10.7
operation. Although our retrospective study cannot demonstrate causality, it is likely that these two findings are causally related, in that patients presenting to the emergency room with abdominal pain are very likely to undergo, for instance, a CT in addition to the US that would typically be performed for complaints of right upper quadrant pain in the outpatient setting. Indeed, older patients in this study were significantly more likely to undergo each and every one of the five preoperative imaging studies in Table 2, and nearly 2-times as likely to undergo a CT in particular. This is likely due to a combination of their higher incidence of comorbidities, and a higher index of suspicion on the part emergency-room physicians when presented with older patients. Once the decision is made to proceed to cholecystectomy, the operation in older patients is associated with significantly greater estimated blood loss in our study, but the fact that this difference is < 30 mL highlights the point that statistical significance does not necessarily imply clinical significance. Much more important
<0.000 <0.000 <0.000 <0.001
clinically is the increased number of open versus laparoscopic operations in the older patients, both conversions from laparoscopic to open, and primarily open operations. This may be partially attributed to the higher incidence of complicated cholecystitis such as acute and gangrenous cholecystitis, as previously reported [4e6], and to the higher incidence of previous operations in the older patients. However, there was no statistically significant difference in operative times of the two groups. The complication rate and length of stay were both increased in the older group. With each year of increasing age the odds for developing postoperative complications in our study increased by 2%. While some studies have similarly found that older patients have increased morbidity, and a longer hospital stay, when compared with the rates of younger patients [7e14], others have not [15,16]. Whereas several of these prior studies used a relatively arbitrary age cutoff, such as 80 or 65 [812], we chose our cutoff between older and younger groups based on the ROC analysis.
124
H. Maqsood et al. / International Journal of Surgery 39 (2017) 119e126
Table 6 Candidate predictors of postoperative complications.
Age White blood cells (1000s/uL) Blood urea nitrogen (mg/dL) Creatinine (mg/dL) Glucose (mg/dL) Alkaline phosphatase (U/L) Total bilirubin (mg/dL) Albumin (g/dL) EBL
Complications
N
Mean
Std. Deviation
P-value
None One or None One or None One or None One or None One or None One or None One or None One or None One or
5055 398 3306 333 3223 334 3219 331 3230 328 2681 257 2676 260 2615 255 4171 325
51.81 60.73 9.30 10.91 13.91 18.23 0.98 1.41 118.01 128.86 107.55 138.61 0.91 1.52 3.79 3.51 58.75 142.54
18.25 19.64 4.84 5.54 10.35 16.22 0.92 2.35 47.72 51.89 83.37 144.84 1.82 2.72 1.01 0.93 122.71 289.37
<0.001
more more more more more more more more more
Complications N ¼ 5045
Comorbidities PCCF-US Any cholecystitis Cholesterolosis Postop diagnosis of GBCa Previous operation
<0.001 <0.001 0.001 <0.001 0.001 <0.001 <0.001 <0.001
Complications N ¼ 398
P-value
N
%
N
%
2610 366 4619 935 6 3665
51.7 9.1 91.6 18.7 0.1 72.5
281 39 369 43 3 348
70.7 12.3 92.7 10.9 0.8 87.7
<0.001 0.059 0.422 <0.001 0.003 <0.001
Abbreviations: as per text.
Table 7 Logistic regression model to predict complications at 30 days postoperatively. Step 1 N ¼ 250 (All significant variables with <59% Missing data)
Age (held constant) Gender (held constant) Comorbidities PCCF-US Any cholecystitis Cholesterolosis Postop diagnosis of GBCa White blood cells (1000s/uL) Creatinine (mg/dL) Glucose (mg/dL) Alkaline phosphatase (U/L) Total bilirubin (mg/dL) Albumin (g/dL) EBL Previous operation Constant
Step 2 N ¼ 2037 (Removing nonsignificant variables: P > 0.1
Step 3 N ¼ 2749 Final model
Adjusted OR (95% CI), P-value
Adjusted OR (95% CI), P-value
Adjusted OR (95% CI), P-value
1.014 1.627 1.025 0.846 1.121 0.667 5.859 1.012 1.101 1.000 1.001 1.031 0.838 1.002 1.580 0.027
1.02 (1.01e1.03), 0.000 1.42 (1.09e1.86), 0.009
1.02 (1.01e1.03), 0.000 1.44 (1.10e1.87), 0.007
(1.00e1.02, 0.005 (1.17e2.27), 0.004 (0.90e1.16), 0.698 (0.54e1.13), 0.459 (0.50e2.5), 0.784 (0.38e1.18), 0.163* (0.60e57.2), 0.128* (0.99e1.04), 0.328 (1.0e1.2), 0.027 (0.99e1.00), 0.891 (0.99e1.00), 0.482 (0.97e1.09), 0.291 (0.66e1.06), 0.291 (1.00e1.010), 0.000 (0.99e2.51), 0.000
0.741 (0.481e1.14), 0.174 1.90 (0.31e11.64), 0.486 1.42 (1.09e1.86), 0.009
1.12 (1.03e1.21), 0.005
1.00 (1.00e1.01),0.000 1.73 (1.17e2.56), 0.006
1.00 (1.00e1.01), 0.000 1.81 (1.23e2.67), 0.003
* Accepted P < 0.15 as candidates for model entry. Table 7 illustrates the three iterative steps in which covariates were removed when the probability to enter the model was set at P < 0.05 and the probability to exit the model was set at 0.15. Abbreviations: as per Table 6 and text.
Interestingly, two of the most recent studies on CCY for acute cholecystitis in older patients reached disparate conclusions, one recommending “no hesitation” for otherwise healthy octogenarians [12] and the other recommending medical management, with interval CCY only for recurrent acute cholecystitis, due to significant postoperative morbidity [9]. Indeed, even for much larger operations with much higher overall complication rates, such as pancreaticoduodenectomy, some analyses have shown that comorbidities, and not age per se, correlate with complications [17]. Similarly, others have found that frailty itself predicts complications better than either age or comorbidities (as approximated by American Society of Anesthesiology [ASA] score) among
predominantly septuagenarians undergoing elective laparoscopic CCY [18]. Accordingly, many tools and scoring systems exist to assess functional status among the elderly being evaluated for operation [19]. Our findings, however, suggest only a modest increase in comorbidities-related complications among older patients, which may be due to difficulties in obtaining complete and accurate past medical histories inherent in a retrospective analysis. Our rate of postoperative complications (10% in the older group and 4.9% among their younger counterparts) is consistent with previous studies reporting a rate of postoperative complications ranging from 5% to 18% in older patients [14,20,21]. Similar to our study, Lill et al. [13] reported increased hospitalization time and
H. Maqsood et al. / International Journal of Surgery 39 (2017) 119e126
complication rate in older patients and did not find any significant difference in mean duration of operations. However, unlike our study, they reported no difference in conversion rate to open surgery between age groups. Following cholecystectomy, several pathologic diagnoses were age-specific. As alluded to above, acute and gangrenous acute cholecystitis were significantly more common in the older group. The nearly 3-fold higher rate of gangrenous cholecystitis in older patients in this series, 8% of patients aged 65 years or older, is concordant with other reports reported in the literature [4,5,22e25]. In addition, older patients were 3-fold more likely to have the uncommon diagnosis adenomyomatosis (P ¼ 0.006, Table 3), which is almost always benign, but has been associated with an often overlooked risk of developing cancer [26]. By contrast, however, older patients were significantly less likely to have cholesterolosis of the gallbladder as a final pathologic diagnosis. Notably, cholesterolosis was the only diagnosis occurring less frequently in older patients (the one exception being chronic cholecystitis, which may simply be a marker of biliary dyskinesia in the database, since this diagnosis frequently accompanies gallbladder dysfunction in younger patients, even in grossly normal gallbladders [27]). This finding of a lower rate of cholesterolosis is relatively novel [28] in that other, smaller studies found merely a nonsignificant trend in the same directions (older patients having a lower rate of cholesterolosis) [29]. In our series, however, the difference was rather striking, with a highly significant (P < 0.001), 2fold difference between younger and older patients (Table 3, and reference [28]). This study has several limitations. Most importantly, the retrospective nature of the study carries the potential for selection bias and systematic error. By definition, more than 5% of all multiple comparisons will randomly achieve statistical significance when P is set 0.05. However, as discussed, and as seen in the tables, all of these significant differences are clinically plausible. We acknowledge the possibility for residual confounding from both measured and unmeasured variables. For example, given that data collection is imperfect, not every comorbidity for every patient can be captured. If such capturing were perfect, perhaps comorbidities would have a greater effect on complications in the analysis. Finally, we analyzed only short-term outcomes, but long-term functional outcomes might be better indicators of health-related quality of life after a cholecystectomy. 5. Conclusion Our results support the bulk of the literature that some combination of age and the accompanying comorbidities, and other factors, is strongly associated with acceptable, but significantly worse outcomes, a finding which urges caution when proposing, performing, and following up on, cholecystectomy in this population. In addition, we have identified and better characterized important age-related differences in healthcare utilization and pathologic findings following cholecystectomy. Preoperative creatinine levels, estimated blood loss, and a history of previous operation may predict the development of postoperative complications in older patients. Ethical approval Yes; Saint Agnes Institutional Review Board; 2011-026. Funding None.
125
Author contribution Cunningham and Kowdley designed the research; Wu, Buddensick, Sautter, Shaukat, Sulkowski, Narducci, Siddique, Kamangar, Kowdley, and Cunningham performed the research; Sill and Kamangar contributed analytic tools; Sill, Kamangar, Ferdosi, Kowdley, and Cunningham analyzed the data; Maqsood, Patel, Ferdosi, Sill, and Cunningham wrote the paper, and all authors drafted the article or revised it critically for important intellectual content, and approved the current version. Conflicts of interest None. Research registration unique identifying number (UIN) researchregistry1885. Guarantor Steven Clark Cunningham. References [1] J.E. Everhart, M. Khare, M. Hill, K.R. Maurer, Prevalence and ethnic differences in gallbladder disease in the United States, Gastroenterology 117 (3) (Sep 1999) 632e639. [2] U.S. Census Bureau PD, Percent Distribution of the Projected Population by Selected Age Groups and Sex for the United States: 2015 to 2016 (NP2012e13), December 2012. [3] D. Dindo, N. Demartines, P.A. Clavien, Classification of surgical complications: a new proposal with evaluation in a cohort of 6336 patients and results of a survey, Ann. Surg. 240 (2) (Aug 2004) 205e213. [4] B. Wu, T.J. Buddensick, H. Ferdosi, et al., Predicting gangrenous cholecystitis, HPB Oxf. 16 (9) (Sep 2014) 801e806. [5] W.N. Yacoub, M. Petrosyan, I. Sehgal, Y. Ma, P. Chandrasoma, R.J. Mason, Prediction of patients with acute cholecystitis requiring emergent cholecystectomy: a simple score, Gastroenterol. Res. Pract. 2010 (2010) 901739. [6] G. Conzo, C. Mauriello, C. Gambardella, S. Napolitano, F. Cavallo, E. Tartaglia, L. Santini, Int. J. Surg. Case Rep. 4 (2013) 316e318. [7] T.H. Magnuson, L.E. Ratner, M.E. Zenilman, J.S. Bender, Laparoscopic cholecystectomy: applicability in the geriatric population, Am. Surg. 63 (1) (Jan 1997) 91e95. [8] J.G. Maxwell, B.A. Tyler, B.G. Maxwell, C.C. Brinker, D.L. Covington, Laparoscopic cholecystectomy in octogenarians, Am. Surg. 64 (9) (Sep 1998) 826e831. [9] E.A. McGillicuddy, K.M. Schuster, K. Barre, et al., Non-operative management of acute cholecystitis in the elderly, Br. J. Surg. 99 (9) (Sep 2012) 1254e1261. [10] M. Nikfarjam, D. Yeo, M. Perini, et al., Outcomes of cholecystectomy for treatment of acute cholecystitis in octogenarians, ANZ J. Surg. 84 (12) (Dec 2013) 943e948. [11] S. Kuy, J.A. Sosa, S.A. Roman, R. Desai, R.A. Rosenthal, Age matters: a study of clinical and economic outcomes following cholecystectomy in elderly Americans, Am. J. Surg. 201 (6) (Jun 2011) 789e796. [12] L.B. Nielsen, K.M. Harboe, L. Bardram, Cholecystectomy for the elderly: no hesitation for otherwise healthy patients, Surg. Endosc. 28 (1) (Jan 2014) 171e177. [13] S. Lill, A. Rantala, T. Vahlberg, J.M. Gronroos, Elective laparoscopic cholecystectomy: the effect of age on conversions, complications and long-term results, Dig. Surg. 28 (3) (2011) 205e209. [14] L.M. Brunt, M.A. Quasebarth, D.L. Dunnegan, N.J. Soper, Outcomes analysis of laparoscopic cholecystectomy in the extremely elderly, Surg. Endosc. 15 (7) (Jul 2001) 700e705. [15] I. Yetim, A. Dervisoglu, O. Karakose, Y. Buyukkaraba-Cak, Y. Bek, K. Erzurumlu, Is advanced age a significant risk factor for laparoscopic cholecystectomy? Minerva Chir. 65 (5) (Oct 2010) 507e513. [16] A.H. Kwon, Y. Matsui, Laparoscopic cholecystectomy in patients aged 80 years and over, World J. Surg. 30 (7) (Jul 2006) 1204e1210. [17] M.A. Makary, J.M. Winter, J.L. Cameron, et al., Pancreaticoduodenectomy in the very elderly, J. Gastrointest. Surg. 10 (3) (Mar 2006) 347e356. [18] K. Lasithiotakis, J. Petrakis, M. Venianaki, et al., Frailty predicts outcome of elective laparoscopic cholecystectomy in geriatric patients, Surg. Endosc. 27 (4) (Apr 2013) 1144e1150. [19] G.C. Kowdley, N. Merchant, J.P. Richardson, J. Somerville, M. Gorospe, S.C. Cunningham, Cancer surgery in the elderly, ScientificWorldJournal 2012 (2012) 303852.
126
H. Maqsood et al. / International Journal of Surgery 39 (2017) 119e126
[20] J. Bingener, M.L. Richards, W.H. Schwesinger, W.E. Strodel, K.R. Sirinek, Laparoscopic cholecystectomy for elderly patients: gold standard for golden years? Arch. Surg. 138 (5) (May 2003) 531e535 discussion 535e536. [21] P. Pessaux, N. Regenet, J.J. Tuech, C. Rouge, R. Bergamaschi, J.P. Arnaud, Laparoscopic versus open cholecystectomy: a prospective comparative study in the elderly with acute cholecystitis, Surg. Laparosc. Endosc. Percutan Tech. 11 (4) (Aug 2001) 252e255. [22] S.P. Fagan, S.S. Awad, K. Rahwan, et al., Prognostic factors for the development of gangrenous cholecystitis, Am. J. Surg. 186 (5) (Nov 2003) 481e485. [23] D.R. Hunt, F.C. Chu, Gangrenous cholecystitis in the laparoscopic era, Aust. N. Z. J. Surg. 70 (6) (Jun 2000) 428e430. [24] L.T. Merriam, S.A. Kanaan, L.G. Dawes, et al., Gangrenous cholecystitis: analysis of risk factors and experience with laparoscopic cholecystectomy, Surgery 126 (4) (Oct 1999) 680e685 discussion 685e686. [25] M. Nikfarjam, V. Niumsawatt, A. Sethu, et al., Outcomes of contemporary
[26]
[27]
[28]
[29]
management of gangrenous and non-gangrenous acute cholecystitis, HPB Oxf. 13 (8) (Aug 2011) 551e558. L. Howe, G. Kowdley, S.C. Cunningham, Gallbladder adenomyomatosis: not always benign, HPB (2017), http://dx.doi.org/10.1016/j.hpb.2016.12.006 (In Press), PMID: 28100404. N. Goussous, G.C. Kowdley, N. Sardana, E. Spiegler, S.C. Cunningham, Gallbladder dysfunction: how much longer will it be controversial? Digestion. 90 (3) (2014) 147e154. S. Dairi, A. Demeusy, A.M. Sill, S.T. Patel, G.C. Kowdley, S.C. Cunningham, Implications of gallbladder cholesterolosis and cholesterol polyps? J. Surg. Res. 200 (2) (2016 Feb) 467e472. F. Yaylak, A. Deger, B.I. Ucar, Y. Sonmez, Z. Bayhan, F. Yetisir, Cholesterolosis in routine histopathological examination after cholecystectomy: what should a surgeon behold in the reports? Int. J. Surg. 12 (11) (Nov 2014) 1187e1191.
Contents lists available at ScienceDirect
International Journal of Surgery journal homepage: www.journal-surgery.net
Age-related differences pre-, intra-, and postcholecystectomy: A retrospective cohort study of 6,868 patients* Hadia Maqsood a, 1, Kalpesh Patel a, 1, Hamid Ferdosi a, b, 1, Anne M. Sill a, Bin Wu a, Thomas Buddensick a, Amanda Sautter a, Haroon Shaukat a, Gisela Sulkowski a, Dusty Marie Narducci a, Mustafa Siddique a, Farin Kamangar c, Gopal C. Kowdley a, Steven C. Cunningham a, * a
Department of Surgery, Saint Agnes Hospital, Baltimore, MD, USA Department of Epidemiology & Biostatistics, George Washington Univ., Washington, DC, USA c Department of Public Health Analysis, Morgan State Univ., Baltimore, MD, USA b
h i g h l i g h t s Older patients undergoing cholecystectomy (CCY) have distinct perioperative characteristics compared with younger patients. Among older patients, preoperative testing, blood loss, operative times and complications are greater. The risk for developing complications increased by 2% per year of life.
a r t i c l e i n f o
a b s t r a c t
Article history: Received 23 November 2016 Received in revised form 7 January 2017 Accepted 12 January 2017 Available online 16 January 2017
Background: Cholecystectomy (CCY) is increasingly performed in older individuals. We sought to examine age-related differences in pre-, intra-, and postoperative factors at a community hospital, using a very large, single-institution cholecystectomy database. Materials and methods: A retrospective review of 6868 patients who underwent CCY from 2001 to 2013 was performed. ROC analysis identified the optimal age cutoff when complications reached a significant inflection point (<55 and 55 years). Multiple clinical features and outcomes were measured and compared by age. Logistic regression was used to examine how well a set of covariates predicted postoperative complications. Results: Older patients had significantly higher rates of comorbidities and underwent more extensive preoperative imaging. Intraoperatively, older patients had more blood loss, longer operative times, and more open operations. Postoperatively, older patients experienced more complications and had significantly different pathological findings. While holding age and gender constant, regression analyses showed that preoperative creatinine level, blood loss and history of previous operation were the strongest predictors of complications. The risk for developing complications increased by 2% per year of life. Conclusion: Older patients have distinct pre-, intra-, and postoperative characteristics. Their care is more imaging- and cost-intensive. CCY in this population is associated with higher risks, likely due to a combination of comorbidities and age-related worsened physiological status. Pathologic findings are significantly different relative to younger patients. While removing the effect of age, preoperative creatinine levels, blood loss, and history of previous operation predict postoperative complications. Quantifying these differences may help to inform management decisions for older patients. © 2017 IJS Publishing Group Ltd. Published by Elsevier Ltd. All rights reserved.
Keywords: Cholecystectomy Elderly Young Complications Age Adenomyosis Cholesterolosis
* Submitted to International Journal of Surgery (an earlier version of this work was presented at the Americas Hepato-Pancreatico-Biliary Association in Miami, FL, USA 2013). * Corresponding author. Saint Agnes Hospital and Cancer Institute, 900 Caton Avenue, MB 207 Baltimore, MD, 21229, USA. E-mail address: [email protected] (S.C. Cunningham). 1 Maqsood, Patel, and Ferdosi contributed equally to this work.
http://dx.doi.org/10.1016/j.ijsu.2017.01.046 1743-9191/© 2017 IJS Publishing Group Ltd. Published by Elsevier Ltd. All rights reserved.
120
H. Maqsood et al. / International Journal of Surgery 39 (2017) 119e126
1. Introduction Gallbladder disease is one of the most common indications for abdominal surgery. The prevalence of gallbladder disease requiring cholecystectomy (CCY) is highest in older patients, with 8.1% of men, and 16.6% of women aged 60e74 years undergoing CCY [1]. Compounding this problem is the fact that the average age of many global populations is increasing. For example, the United States Census predictions estimate that the percentage of the population 65 years of age and over will increase from 14.8% in 2015 to 20.2% (or an additional 30 million individuals) by the year 2035 [2]. The aim of our study is to critically examine our experience in the treatment of gallbladder disease in younger versus older patients. A sub-aim was to identify risk factors for postoperative complications. We examined differences in pre-, intra-, and postoperative characteristics of older versus younger patients in order to identify areas where improvements may be made to optimize care for older patients undergoing CCY. 2. Methods 2.1. Study population and data collection A recently updated database of patients undergoing CCY at Saint Agnes Hospital from 2000 to 2013 was analyzed. This study was approved by the Institutional Review Board. Patients who underwent CCY as part of another major operation such as a Whipple, liver resection, bariatric operation, or cytoreduction with hyperthermic intraperitoneal chemotherapy, were excluded from the study. Those with a preoperative diagnosis of gallbladder cancer were also excluded. CCY was performed according to individual surgeon preferences, after an evaluation period that was similarly surgeon-dependent, making use of standard laboratory testing and imaging modalities, as dictated by the clinical scenario. Data were obtained from hospital charts, including operative notes, pathology reports, imaging study reports, and lab reports. Standard patient demographic and clinicopathological data were collected, including age, gender, race, ASA score, comorbidities, operative and pathologic details, and postoperative morbidity. Perioperative morbidity was severity-graded based on the ClavienDindo classification system [3]. Data were entered into Excel™ (Redmond, WA) and imported into SPSS™ version 21.0 (Chicago, IL) for statistical analysis. 2.2. Statistical considerations 2.2.1. Sample size This study sample represents a convenience sample of previously collected cases undergoing CCY. Although not formally “powered” to show differences in preselected measures, this exploratory study of 2883 younger and 2166 older patients likely provides adequate sample size to confidently demonstrate true population differences. 2.2.2. Analysis A receiver operator characteristic (ROC) curve identified the point along the age continuum (age 55 years) at which the number of postoperative complications increased significantly. Therefore, patients were stratified by using this age cutoff (<55 and 55 years of age) and these groups served as the main comparison groups throughout the study. Demographic, clinical, diagnostic, pathologic, and perioperative characteristics were compared using appropriate, univariate and bivariate parametric tests (frequencies, student t-test, chi-square test) and by nonparametric methods for ordinal and nominal
variables. Patients with missing data were excluded. Understanding the logical confounders in the relationships between age and number of complications and morbidities, type of operation (eg, open or laparoscopic) and level of urgency (ie, emergent or elective), a binary logistic regression model was built to identify the most prominent predictors of developing 30-day postoperative complications. We controlled for age to determine the ability of the predictors to explain the variance in the outcome, i.e., presence or absence of postoperative complications. Given that 72% of the population was female, we also controlled for gender so as to avoid the overrepresentation of females in the regression analyses. The odds of developing complications and their 95% confidence intervals were also computed. To examine independent predictors of postoperative complications, bivariate analyses of preoperative measures of laboratory markers, preexisting comorbidities and postsurgical pathologic findings were compared for those with and without complications within 30 days of operation. Variables found to be significantly different between those with and without complications (P < 0.05) became candidates for binomial logistic regression. A missingvalues analysis was then conducted to select predictors with the least number of missing values to provide ample sample size for a regression analysis. In an effort to revert the syllogism of the cause and effect of age and complications, we controlled for age in the binomial regression models; gender was also held constant to remove the overrepresentation of females in our population. Covariates were removed in a backward elimination fashion to identify the predictor(s) that gave the most weight in their ability to explain the variance in the outcome (postoperative complications). 3. Results 3.1. Selection Of 6868 patients with varying indications undergoing CCY, 5049 patients were analyzed in this study after excluding cases that had incomplete data, or had the CCY performed secondary to another major operation, as described above. 3.2. Patient characteristics Fifty-seven percent of patients were 55 years old (this cutoff chosen based on the ROC curve, as described above; Table 1). Both age groups contained significantly more females than males. Racial/ ethnic distribution across age groups also varied significantly and appears to be partially explained by a greater number of younger Hispanic patients. The majority (79%) of the younger group had ASA scores of 1e2 while the majority (54%) of older patients had scores of 3e4. The older patients had a significantly higher percentage of comorbidities such as cardiovascular diseases, hypertension, cerebrovascular accident, coronary artery disease, chronic obstructive pulmonary disease, diabetes mellitus, renal failure (all P < 0.001), alcohol abuse, and hepatic cirrhosis (P ¼ 0.05). The only exception was hepatitis infection, which, although higher in younger adults, did not reach statistical significance. 3.3. Imaging studies A significantly higher percentage of patients >55 years old had preoperative investigations such as computed tomography (CT), magnetic resonance imaging (MRI), hepatobiliary iminodiacetic acid (HIDA) scans and endoscopic retrograde cholangiopancreatography (ERCP) performed (Table 2). Only the ultrasound (US) showed significantly larger gallbladder wall thickness
H. Maqsood et al. / International Journal of Surgery 39 (2017) 119e126
121
Table 1 Patient characteristics.
Number of patients Age (mean years, SD) Male Female Caucasian African American Asian Hispanic Other Heart rate at triage in ER ASA score 1 or 2 ASA score 3 or 4 ASA score 5 Congestive heart failure Hypertension Cerebrovascular accident Coronary artery disease COPD Renal failure Diabetes mellitus Alcohol abuse Hepatitis infection Ascites (moderate to severe) Hepatic encephalopathy Cirrhosis Fibrosis
Age Gender Race
Heart Rate ASA
Patient History
55 years N (%)
>55 years N (%)
2883 (57) 39.24 (10.4) 608 (21) 2275 (79) 1853 (64) 866 (30) 16 (0.6) 91 (3.2) 55 (1.9) 80.27 (15.88) 508 (79) 126 (20) 9 (1.4) 38 (1.3) 673 (23) 15 (0.5) 109 (3.8) 61 (2.1) 31 (1.1) 269 (9.3) 356 (12) 106 (3.7) 2829 (98) 65 (17) 20 (0.7) 2 (0.1)
2166 (43) 70.47 (9.91) 766 (35) 1400 (65) 1644 (76) 458 (21) 11 (0.5) 17 (0.8) 33 (1.5) 78.13 (17.17) 180 (45) 216 (54) 2 (0.5) 223 (10) 1340 (62) 56 (2.6) 550 (25) 285 (13) 170 (7.9) 527 (24) 321 (15) 55 (2.6) 3023 (98) 49 (18) 29 (1.3) 4 (0.2)
P-value
<0.001 <0.001 <0.001
0.023
<0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 0.011 0.104 0.120 0.784 0.021 0.239
Abbreviations: COPD, chronic obstructive pulmonary disease; otherwise as per text.
Table 2 Diagnostics tests.
US
CT
MRI
HIDA
ERCP
US performed Gallstones Gallstone size (mm) GBW thickness (mm) CBD diameter (mm) Pericholecystic fluid Calcification CT performed GBW thickness (mm) CBD diameter (mm) Pericholecystic fluid Calcification MRI performed GBW thickness (mm) CBD diameter (mm) Pericholecystic fluid Calcification HIDA performed Symptoms upon CCK injection Ejection fraction (%) Time to GB visualization (min) Time to duodenum visualization (min) Preoperative ERCP Postoperative ERCP
55 years Mean (SD) or N (%)
>55 years Mean (SD) or N (%)
P-value
2190 (70) 1324 (89) 15.42 (8.14) 4.49 (2.61) 4.74 (2.25) 205 (8.1) 91 (3.6) 568 (18) 3.66 (2.28) 11.67 (14.61) 92 (5.9) 27 (1.7) 223 (7.1) 6.36 (4.29) 7.08 (3.14) 57 (4.1) 12 (0.9) 744 (24) 400 (79) 22.64 (23.91) 47.40 (179.23) 58.27 (163.29) 82 (2.60) 56 (1.80)
1676 (69) 1092 (87) 15.13 (7.59) 5.06 (4.38) 5.49 (2.69) 212 (11) 52 (2.8) 816 (34) 4.73 (3.96) 14.61 (2.40) 143 (10.00) 40 (2.8) 256 (11) 4.92 (4.48) 7.90 (3.88) 60 (5.4) 13 (1.2) 644 (27) 380 (84) 24.43 (25.38) 124.21 (547.92) 84.38 (286.18) 145 (6.00) 56 (2.30)
0.737 0.592 0.746 0.006 <0.001 0.001 0.102 <0.001 0.365 0.398 <0.001 0.053 <0.001 0.382 0.085 0.139 0.458 0.013 0.085 0.362 0.002 0.091 <0.001 0.160
Abbreviations: as per text.
and common bile duct diameter in older group but interestingly, CT and MRI failed to show these expected differences. Significantly more older patients had pericholecystic fluid on US and CT than did the younger patients (P < 0.001). Time to visualize the gallbladder at HIDA scan was significantly longer in older patients (P ¼ 0.002). 3.4. Pathologic diagnosis Histopathologic findings (Table 3) showed that rates of acute
(14% versus 24%, P < 0.001), gangrenous (2.7% versus 8.1%, P < 0.001), and hemorrhagic cholecystitis (3.2% versus 5.5%, P < 0.001) were significantly higher in the older CCY patients while younger patients had a significantly higher rate of chronic cholecystitis (86% versus 83%, P < 0.001). Likewise, rates of adenomyosis (0.3% versus 0.9%, P ¼ 0.006), sludge (4.7% versus 7.7%, P < 0.001), GB cancer (0.1% versus 1.1%, P < 0.001) and other histopathologic findings (5.8% versus 8.6%, P < 0.001) were significantly higher in older CCY patients, but younger patients had much higher rates of
122
H. Maqsood et al. / International Journal of Surgery 39 (2017) 119e126
Table 3 Pathology results.
Histopathology findings
Blood and serum chemistry
Acute cholecystitis Chronic cholecystitis Gangrenous cholecystitis Perforated cholecystitis Hemorrhagic cholecystitis Cholelithiasis Adenomyosis Diverticulosis Gallbladder polyp Cholesterolosis Sludge Cancer Other
White blood cells (1000s/uL) Blood urea nitrogen (mg/dL) Creatinine (mg/dL) Glucose (mg/dL) Lactate (mmol/L) Aspartate transferase (U/L) Alanine transferase (U/L) Alkaline phosphatase (U/L) Total bilirubin (mg/dL) Direct bilirubin (mg/dL) Albumin (g/dL) Prothrombin time (s) International normalized ratio
55 years N (%)
>55 years N (%)
P-value
440 (14) 2775 (86) 85 (2.7) 7 (0.2) 103 (3.2) 2424 (76) 10 (0.3) 3 (0.1) 38 (1.2) 743 (23) 149 (4.7) 2 (0.1) 186 (5.8)
589 (24) 2033 (83) 195 (8.1) 11 (0.5) 132 (5.5) 1864 (76) 21 (0.9) 3 (0.1) 26 (1.1) 308 (13) 185 (7.7) 26 (1.1) 207 (8.6)
<0.001 <0.001 <0.001 0.123 <0.001 0.646 0.006 0.734 0.682 <0.001 <0.001 <0.001 <0.001
55 years Mean (SD)
>55 years Mean (SD)
P-value
9.13 (4.19) 11.23 (8.02) 0.88 (0.87) 111.48 38.08) 75.80 (157.56) 69.14 (139.50) 81.16 (153.76) 98.37 (65.54) 0.78 (1.80) 0.27 (0.75) 3.92 (0.63) 11.34 (2.92) 1.03 (0.75)
9.83 (5.67) 18.03 (12.90) 1.19 (1.36) 128.62 (56.93) 37.50 (141.75) 73.28 (138.68) 76.81 (129.35) 125.87 (113.58) 1.19 (2.20) 0.48 (1.27) 3.58 (1.32) 12.55 (5.74) 1.34 (3.36)
<0.001 <0.001 <0.001 <0.001 0.354 0.420 0.414 <0.001 <0.001 <0.001 <0.001 <0.001 0.002
Abbreviations: as per text.
cholesterolosis than older patients (23% versus 13%, respectively, P < 0.001). Serum values for white blood cells, blood urea nitrogen, creatinine, glucose, alkaline phosphatase, total and direct bilirubin, prothrombin time and international normalized ratio were significantly higher in the older group (P < 0.05). There was no significant difference in the serum levels of lactate, aspartate transferase and alanine transferase between the two groups.
3.5. Perioperative course As shown in Table 4, a significantly higher percentage of older patients had a history of prior abdominal operations (78% versus 70%, P < 0.001) and underwent an urgent surgery than did the younger group (42% versus 37%. P < 0.001). Older patients also had a significantly higher estimated blood loss (EBL) during surgery (82.79 mL versus 51.44 mL, P < 0.001) as compared to the younger patients. There was no statistically significant difference in the duration of operation of the two groups. Postoperative hospital length of stay (LOS-PO) (4.3 days versus 2.2 days, P < 0.001) and total hospital length of stay (LOS-T) (6.1 days versus 3.2 days, P < 0.001) were significantly longer for older patients. Chi-square analysis revealed significantly higher rates of conversion from laparoscopic to open CCY procedures in the older patients (5.2 versus 1.6%, respectively, P < 0.001). Also of note is the significantly higher rate of primarily open CCY in the older group. When comparing age-related differences in postoperative complication rates by type of approach (laparoscopic, open, and laparoscopic converted to open), complication rates were significantly highest among laparoscopic-converted-to-open CCY, but rates of complications were nearly double in the older group across all three types of approach (Table 4; P < 0.001). Older patients suffered significantly more postoperative complications which were generally of a higher severity (Clavien
grading system [3]) compared with their younger counterparts. Those who experienced postoperative complications within 30 days of surgery were significantly older than those with no complications (Table 5). Significant increases in several laboratory parameters were also observed for those who experienced postoperative complications with the exception of albumin, which was significantly lower in those with complications, and in the elderly group (all P < 0.001). Significantly more patients with postoperative complications had a history of one or more comorbidities, previous abdominal surgeries, and intraoperative finding of gallbladder cancer (Table 6). While holding age and gender constant, a binomial logistic regression modeling showed the predictive value of each candidate covariate in step 1, of which 5 covariates independently explained the majority of the variance of postoperative complications (P ¼ 0.163). Further tightening of the model left 3 covariates (preoperative creatinine levels, estimated blood loss and previous operation) to predict the occurrence of postoperative complications (Table 7). The odds of a postoperative complication increased by1.02 (95% CI ¼ 1.02e1.33) per year of life.
4. Discussion As global populations continue to age, the particular characteristics of older patients versus younger patients gains much importance, especially when prognosticating outcomes for common procedures such as cholecystectomies. In the current study, using one of the largest single-institution cholecystectomy databases in the world, we elucidated several important aspects of this population. The use of preoperative imaging studies was significantly higher among older patients and these older patients were significantly more likely to undergo an urgent (as opposed to an elective)
H. Maqsood et al. / International Journal of Surgery 39 (2017) 119e126
123
Table 4 Perioperative data.
Duration of operation (min) EBL (mL) Urgency (urgent/emergent/inpatient) LOS-T (days) LOS-PO (days) Laparoscopic cholecystectomy only Laparoscopic cholecystectomy with IOC Open cholecystectomy (begun open) Open cholecystectomy (begun open) with IOC Laparoscopic converted to open cholecystectomy Laparoscopic converted to open cholecystectomy with IOC Conversion rates (laparoscopic to open CCY)
Type of operation
Any complication
Grade of complication
1 2 3a 3b 4a 4b 5 Infectious Cardiovascular Pulmonary Renal Gastrointestinal Hemorrhage Other None One or more
Type of complications
Previous abdominal incisions
55 years Mean (SD)
>55 years Mean (SD)
P-value
133.04 (1203.15) 51.44 (129.91) 1168 (37.1) 3.18 (4.10) 2.20 (3.66) 2410 (85.0) 301 (11) 45 (1.6) 14 (0.5) 40 (1.4) 24 (0.8) 64 (1.6)
102.12 (130.06) 82.79 (159.47) 1020 (42.3) 6.07 (8.78) 4.31 (7.59) 1616 (73.5) 252 (12) 153 (7.0) 41 (1.9) 84 (3.8) 53 (2.4) 137 (5.2)
0.523 <0.001 <0.001 <0.001 <0.001 <0.001
150 (4.9)
248 (10)
<0.001
55 years N (%)
>55 years N (%)
P-value
0 61 (1.9) 31 (1.0) 14 (0.4) 0 0 2 (0.1) 19 (0.6) 3 (0.1) 5 (0.2) 2 (0.1) 0 3 (0.1) 26 (0.8) 963 (64.6) 528 (35.4)
0 109 (4.4) 54 (2.2) 12 (0.5) 17 (0.7) 4 (0.2) 11 (0.4) 43 (1.7) 27 (1.1) 25 (1.0) 3 (0.1) 1 (0.0) 6 (0.2) 37 (1.5) 1728 (52.1) 1591 (47.9)
<0.001
<0.001
<0.001
<0.001
Abbreviations: IOC, intraoperative cholangiogram; otherwise as per text.
Table 5 Complications by age group by CCY approach. 55 years
Type of CCY
Laparoscopic CCY Open CCY Converted lap / open Total
N % N % N % N %
>55 years
P-value
No complications
Complications
No complications
Complications
2561 95.6 53 88.3 53 82.8 2667 95.1
119 4.4 7 11.7 11 17.2 137 4.9
1715 92.0 153 76.1 102 73.4 1970 89.3
150 8.0 48 23.9 37 26.6 235 10.7
operation. Although our retrospective study cannot demonstrate causality, it is likely that these two findings are causally related, in that patients presenting to the emergency room with abdominal pain are very likely to undergo, for instance, a CT in addition to the US that would typically be performed for complaints of right upper quadrant pain in the outpatient setting. Indeed, older patients in this study were significantly more likely to undergo each and every one of the five preoperative imaging studies in Table 2, and nearly 2-times as likely to undergo a CT in particular. This is likely due to a combination of their higher incidence of comorbidities, and a higher index of suspicion on the part emergency-room physicians when presented with older patients. Once the decision is made to proceed to cholecystectomy, the operation in older patients is associated with significantly greater estimated blood loss in our study, but the fact that this difference is < 30 mL highlights the point that statistical significance does not necessarily imply clinical significance. Much more important
<0.000 <0.000 <0.000 <0.001
clinically is the increased number of open versus laparoscopic operations in the older patients, both conversions from laparoscopic to open, and primarily open operations. This may be partially attributed to the higher incidence of complicated cholecystitis such as acute and gangrenous cholecystitis, as previously reported [4e6], and to the higher incidence of previous operations in the older patients. However, there was no statistically significant difference in operative times of the two groups. The complication rate and length of stay were both increased in the older group. With each year of increasing age the odds for developing postoperative complications in our study increased by 2%. While some studies have similarly found that older patients have increased morbidity, and a longer hospital stay, when compared with the rates of younger patients [7e14], others have not [15,16]. Whereas several of these prior studies used a relatively arbitrary age cutoff, such as 80 or 65 [812], we chose our cutoff between older and younger groups based on the ROC analysis.
124
H. Maqsood et al. / International Journal of Surgery 39 (2017) 119e126
Table 6 Candidate predictors of postoperative complications.
Age White blood cells (1000s/uL) Blood urea nitrogen (mg/dL) Creatinine (mg/dL) Glucose (mg/dL) Alkaline phosphatase (U/L) Total bilirubin (mg/dL) Albumin (g/dL) EBL
Complications
N
Mean
Std. Deviation
P-value
None One or None One or None One or None One or None One or None One or None One or None One or None One or
5055 398 3306 333 3223 334 3219 331 3230 328 2681 257 2676 260 2615 255 4171 325
51.81 60.73 9.30 10.91 13.91 18.23 0.98 1.41 118.01 128.86 107.55 138.61 0.91 1.52 3.79 3.51 58.75 142.54
18.25 19.64 4.84 5.54 10.35 16.22 0.92 2.35 47.72 51.89 83.37 144.84 1.82 2.72 1.01 0.93 122.71 289.37
<0.001
more more more more more more more more more
Complications N ¼ 5045
Comorbidities PCCF-US Any cholecystitis Cholesterolosis Postop diagnosis of GBCa Previous operation
<0.001 <0.001 0.001 <0.001 0.001 <0.001 <0.001 <0.001
Complications N ¼ 398
P-value
N
%
N
%
2610 366 4619 935 6 3665
51.7 9.1 91.6 18.7 0.1 72.5
281 39 369 43 3 348
70.7 12.3 92.7 10.9 0.8 87.7
<0.001 0.059 0.422 <0.001 0.003 <0.001
Abbreviations: as per text.
Table 7 Logistic regression model to predict complications at 30 days postoperatively. Step 1 N ¼ 250 (All significant variables with <59% Missing data)
Age (held constant) Gender (held constant) Comorbidities PCCF-US Any cholecystitis Cholesterolosis Postop diagnosis of GBCa White blood cells (1000s/uL) Creatinine (mg/dL) Glucose (mg/dL) Alkaline phosphatase (U/L) Total bilirubin (mg/dL) Albumin (g/dL) EBL Previous operation Constant
Step 2 N ¼ 2037 (Removing nonsignificant variables: P > 0.1
Step 3 N ¼ 2749 Final model
Adjusted OR (95% CI), P-value
Adjusted OR (95% CI), P-value
Adjusted OR (95% CI), P-value
1.014 1.627 1.025 0.846 1.121 0.667 5.859 1.012 1.101 1.000 1.001 1.031 0.838 1.002 1.580 0.027
1.02 (1.01e1.03), 0.000 1.42 (1.09e1.86), 0.009
1.02 (1.01e1.03), 0.000 1.44 (1.10e1.87), 0.007
(1.00e1.02, 0.005 (1.17e2.27), 0.004 (0.90e1.16), 0.698 (0.54e1.13), 0.459 (0.50e2.5), 0.784 (0.38e1.18), 0.163* (0.60e57.2), 0.128* (0.99e1.04), 0.328 (1.0e1.2), 0.027 (0.99e1.00), 0.891 (0.99e1.00), 0.482 (0.97e1.09), 0.291 (0.66e1.06), 0.291 (1.00e1.010), 0.000 (0.99e2.51), 0.000
0.741 (0.481e1.14), 0.174 1.90 (0.31e11.64), 0.486 1.42 (1.09e1.86), 0.009
1.12 (1.03e1.21), 0.005
1.00 (1.00e1.01),0.000 1.73 (1.17e2.56), 0.006
1.00 (1.00e1.01), 0.000 1.81 (1.23e2.67), 0.003
* Accepted P < 0.15 as candidates for model entry. Table 7 illustrates the three iterative steps in which covariates were removed when the probability to enter the model was set at P < 0.05 and the probability to exit the model was set at 0.15. Abbreviations: as per Table 6 and text.
Interestingly, two of the most recent studies on CCY for acute cholecystitis in older patients reached disparate conclusions, one recommending “no hesitation” for otherwise healthy octogenarians [12] and the other recommending medical management, with interval CCY only for recurrent acute cholecystitis, due to significant postoperative morbidity [9]. Indeed, even for much larger operations with much higher overall complication rates, such as pancreaticoduodenectomy, some analyses have shown that comorbidities, and not age per se, correlate with complications [17]. Similarly, others have found that frailty itself predicts complications better than either age or comorbidities (as approximated by American Society of Anesthesiology [ASA] score) among
predominantly septuagenarians undergoing elective laparoscopic CCY [18]. Accordingly, many tools and scoring systems exist to assess functional status among the elderly being evaluated for operation [19]. Our findings, however, suggest only a modest increase in comorbidities-related complications among older patients, which may be due to difficulties in obtaining complete and accurate past medical histories inherent in a retrospective analysis. Our rate of postoperative complications (10% in the older group and 4.9% among their younger counterparts) is consistent with previous studies reporting a rate of postoperative complications ranging from 5% to 18% in older patients [14,20,21]. Similar to our study, Lill et al. [13] reported increased hospitalization time and
H. Maqsood et al. / International Journal of Surgery 39 (2017) 119e126
complication rate in older patients and did not find any significant difference in mean duration of operations. However, unlike our study, they reported no difference in conversion rate to open surgery between age groups. Following cholecystectomy, several pathologic diagnoses were age-specific. As alluded to above, acute and gangrenous acute cholecystitis were significantly more common in the older group. The nearly 3-fold higher rate of gangrenous cholecystitis in older patients in this series, 8% of patients aged 65 years or older, is concordant with other reports reported in the literature [4,5,22e25]. In addition, older patients were 3-fold more likely to have the uncommon diagnosis adenomyomatosis (P ¼ 0.006, Table 3), which is almost always benign, but has been associated with an often overlooked risk of developing cancer [26]. By contrast, however, older patients were significantly less likely to have cholesterolosis of the gallbladder as a final pathologic diagnosis. Notably, cholesterolosis was the only diagnosis occurring less frequently in older patients (the one exception being chronic cholecystitis, which may simply be a marker of biliary dyskinesia in the database, since this diagnosis frequently accompanies gallbladder dysfunction in younger patients, even in grossly normal gallbladders [27]). This finding of a lower rate of cholesterolosis is relatively novel [28] in that other, smaller studies found merely a nonsignificant trend in the same directions (older patients having a lower rate of cholesterolosis) [29]. In our series, however, the difference was rather striking, with a highly significant (P < 0.001), 2fold difference between younger and older patients (Table 3, and reference [28]). This study has several limitations. Most importantly, the retrospective nature of the study carries the potential for selection bias and systematic error. By definition, more than 5% of all multiple comparisons will randomly achieve statistical significance when P is set 0.05. However, as discussed, and as seen in the tables, all of these significant differences are clinically plausible. We acknowledge the possibility for residual confounding from both measured and unmeasured variables. For example, given that data collection is imperfect, not every comorbidity for every patient can be captured. If such capturing were perfect, perhaps comorbidities would have a greater effect on complications in the analysis. Finally, we analyzed only short-term outcomes, but long-term functional outcomes might be better indicators of health-related quality of life after a cholecystectomy. 5. Conclusion Our results support the bulk of the literature that some combination of age and the accompanying comorbidities, and other factors, is strongly associated with acceptable, but significantly worse outcomes, a finding which urges caution when proposing, performing, and following up on, cholecystectomy in this population. In addition, we have identified and better characterized important age-related differences in healthcare utilization and pathologic findings following cholecystectomy. Preoperative creatinine levels, estimated blood loss, and a history of previous operation may predict the development of postoperative complications in older patients. Ethical approval Yes; Saint Agnes Institutional Review Board; 2011-026. Funding None.
125
Author contribution Cunningham and Kowdley designed the research; Wu, Buddensick, Sautter, Shaukat, Sulkowski, Narducci, Siddique, Kamangar, Kowdley, and Cunningham performed the research; Sill and Kamangar contributed analytic tools; Sill, Kamangar, Ferdosi, Kowdley, and Cunningham analyzed the data; Maqsood, Patel, Ferdosi, Sill, and Cunningham wrote the paper, and all authors drafted the article or revised it critically for important intellectual content, and approved the current version. Conflicts of interest None. Research registration unique identifying number (UIN) researchregistry1885. Guarantor Steven Clark Cunningham. References [1] J.E. Everhart, M. Khare, M. Hill, K.R. Maurer, Prevalence and ethnic differences in gallbladder disease in the United States, Gastroenterology 117 (3) (Sep 1999) 632e639. [2] U.S. Census Bureau PD, Percent Distribution of the Projected Population by Selected Age Groups and Sex for the United States: 2015 to 2016 (NP2012e13), December 2012. [3] D. Dindo, N. Demartines, P.A. Clavien, Classification of surgical complications: a new proposal with evaluation in a cohort of 6336 patients and results of a survey, Ann. Surg. 240 (2) (Aug 2004) 205e213. [4] B. Wu, T.J. Buddensick, H. Ferdosi, et al., Predicting gangrenous cholecystitis, HPB Oxf. 16 (9) (Sep 2014) 801e806. [5] W.N. Yacoub, M. Petrosyan, I. Sehgal, Y. Ma, P. Chandrasoma, R.J. Mason, Prediction of patients with acute cholecystitis requiring emergent cholecystectomy: a simple score, Gastroenterol. Res. Pract. 2010 (2010) 901739. [6] G. Conzo, C. Mauriello, C. Gambardella, S. Napolitano, F. Cavallo, E. Tartaglia, L. Santini, Int. J. Surg. Case Rep. 4 (2013) 316e318. [7] T.H. Magnuson, L.E. Ratner, M.E. Zenilman, J.S. Bender, Laparoscopic cholecystectomy: applicability in the geriatric population, Am. Surg. 63 (1) (Jan 1997) 91e95. [8] J.G. Maxwell, B.A. Tyler, B.G. Maxwell, C.C. Brinker, D.L. Covington, Laparoscopic cholecystectomy in octogenarians, Am. Surg. 64 (9) (Sep 1998) 826e831. [9] E.A. McGillicuddy, K.M. Schuster, K. Barre, et al., Non-operative management of acute cholecystitis in the elderly, Br. J. Surg. 99 (9) (Sep 2012) 1254e1261. [10] M. Nikfarjam, D. Yeo, M. Perini, et al., Outcomes of cholecystectomy for treatment of acute cholecystitis in octogenarians, ANZ J. Surg. 84 (12) (Dec 2013) 943e948. [11] S. Kuy, J.A. Sosa, S.A. Roman, R. Desai, R.A. Rosenthal, Age matters: a study of clinical and economic outcomes following cholecystectomy in elderly Americans, Am. J. Surg. 201 (6) (Jun 2011) 789e796. [12] L.B. Nielsen, K.M. Harboe, L. Bardram, Cholecystectomy for the elderly: no hesitation for otherwise healthy patients, Surg. Endosc. 28 (1) (Jan 2014) 171e177. [13] S. Lill, A. Rantala, T. Vahlberg, J.M. Gronroos, Elective laparoscopic cholecystectomy: the effect of age on conversions, complications and long-term results, Dig. Surg. 28 (3) (2011) 205e209. [14] L.M. Brunt, M.A. Quasebarth, D.L. Dunnegan, N.J. Soper, Outcomes analysis of laparoscopic cholecystectomy in the extremely elderly, Surg. Endosc. 15 (7) (Jul 2001) 700e705. [15] I. Yetim, A. Dervisoglu, O. Karakose, Y. Buyukkaraba-Cak, Y. Bek, K. Erzurumlu, Is advanced age a significant risk factor for laparoscopic cholecystectomy? Minerva Chir. 65 (5) (Oct 2010) 507e513. [16] A.H. Kwon, Y. Matsui, Laparoscopic cholecystectomy in patients aged 80 years and over, World J. Surg. 30 (7) (Jul 2006) 1204e1210. [17] M.A. Makary, J.M. Winter, J.L. Cameron, et al., Pancreaticoduodenectomy in the very elderly, J. Gastrointest. Surg. 10 (3) (Mar 2006) 347e356. [18] K. Lasithiotakis, J. Petrakis, M. Venianaki, et al., Frailty predicts outcome of elective laparoscopic cholecystectomy in geriatric patients, Surg. Endosc. 27 (4) (Apr 2013) 1144e1150. [19] G.C. Kowdley, N. Merchant, J.P. Richardson, J. Somerville, M. Gorospe, S.C. Cunningham, Cancer surgery in the elderly, ScientificWorldJournal 2012 (2012) 303852.
126
H. Maqsood et al. / International Journal of Surgery 39 (2017) 119e126
[20] J. Bingener, M.L. Richards, W.H. Schwesinger, W.E. Strodel, K.R. Sirinek, Laparoscopic cholecystectomy for elderly patients: gold standard for golden years? Arch. Surg. 138 (5) (May 2003) 531e535 discussion 535e536. [21] P. Pessaux, N. Regenet, J.J. Tuech, C. Rouge, R. Bergamaschi, J.P. Arnaud, Laparoscopic versus open cholecystectomy: a prospective comparative study in the elderly with acute cholecystitis, Surg. Laparosc. Endosc. Percutan Tech. 11 (4) (Aug 2001) 252e255. [22] S.P. Fagan, S.S. Awad, K. Rahwan, et al., Prognostic factors for the development of gangrenous cholecystitis, Am. J. Surg. 186 (5) (Nov 2003) 481e485. [23] D.R. Hunt, F.C. Chu, Gangrenous cholecystitis in the laparoscopic era, Aust. N. Z. J. Surg. 70 (6) (Jun 2000) 428e430. [24] L.T. Merriam, S.A. Kanaan, L.G. Dawes, et al., Gangrenous cholecystitis: analysis of risk factors and experience with laparoscopic cholecystectomy, Surgery 126 (4) (Oct 1999) 680e685 discussion 685e686. [25] M. Nikfarjam, V. Niumsawatt, A. Sethu, et al., Outcomes of contemporary
[26]
[27]
[28]
[29]
management of gangrenous and non-gangrenous acute cholecystitis, HPB Oxf. 13 (8) (Aug 2011) 551e558. L. Howe, G. Kowdley, S.C. Cunningham, Gallbladder adenomyomatosis: not always benign, HPB (2017), http://dx.doi.org/10.1016/j.hpb.2016.12.006 (In Press), PMID: 28100404. N. Goussous, G.C. Kowdley, N. Sardana, E. Spiegler, S.C. Cunningham, Gallbladder dysfunction: how much longer will it be controversial? Digestion. 90 (3) (2014) 147e154. S. Dairi, A. Demeusy, A.M. Sill, S.T. Patel, G.C. Kowdley, S.C. Cunningham, Implications of gallbladder cholesterolosis and cholesterol polyps? J. Surg. Res. 200 (2) (2016 Feb) 467e472. F. Yaylak, A. Deger, B.I. Ucar, Y. Sonmez, Z. Bayhan, F. Yetisir, Cholesterolosis in routine histopathological examination after cholecystectomy: what should a surgeon behold in the reports? Int. J. Surg. 12 (11) (Nov 2014) 1187e1191.