- Email: [email protected]
Does socioeconomic status predict outcomes after cholecystectomy?
The American Journal of Surgery (2016) -, -–-
Does socioeconomic status predict outcomes after cholecystectomy? Vishnu Ambur, M.D.a,*, Sharven Taghavi, M.D.b, Sagar Kadakia, M.D.a, Senthil Jayarajan, M.D.b, John Gaughan, Ph.D.c, Lars Ola Sjoholm, M.D.a, Abhijit Pathak, M.D.a, Thomas Santora, M.D.a, Joseph Rappold, M.D.a, Amy J. Goldberg, M.D.a a
Temple University Hospital, Department of Surgery, Philadelphia, PA, USA; bBarnes Jewish Hospital/ Washington University in St. Louis, Department of Surgery, St Louis, MO, USA; cTemple University School of Medicine, Biostatistics Consulting Center, Philadelphia, PA, USA
KEYWORDS: Socioeconomic status; Cholecystectomy; Income; Medicaid; National inpatient sample
Abstract BACKGROUND: This study was performed to evaluate the effect of socioeconomic status (SES) on outcomes after cholecystectomy. METHODS: The National Inpatient Sample (NIS) database (2005 to 2011) was queried for patients undergoing cholecystectomy. Clinically relevant variables were used to examine clinical characteristics, postoperative complications, and mortality. SES was investigated by examining income quartile. RESULTS: More than 2 million patients underwent cholecystectomy during this period. They were divided into quartiles by SES. The lowest cohort was younger (50 years, P , .001) and had the lowest Charlson Comorbidity Index (2.08, P , .001). This cohort was more likely African American (15.8%, P , .001) and more likely to have Medicaid (19.2%, P , .001). Using split-sample validation and multivariate analysis, lower SES, Charlson comorbidity Index, and Medicaid recipients were associated with increased mortality. CONCLUSIONS: Patients with Medicaid and lower SES had poorer outcomes after cholecystectomy. Ó 2016 Elsevier Inc. All rights reserved.
Cholecystectomy remains one of the most common surgeries performed in the United States today. It is performed for various indications ranging from benign conditions including biliary colic to acute conditions such as acute cholecystitis and gallstone pancreatitis. Previous There were no relevant financial relationships or any sources of support in the form of grants, equipment, or drugs. The authors declare no conflicts of interest. * Corresponding author. Tel.: 215-707-3633; fax: 215-707-1915. E-mail address: [email protected] Manuscript received October 5, 2015; revised manuscript April 19, 2016 0002-9610/$ - see front matter Ó 2016 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.amjsurg.2016.04.012
studies have shown that gallbladder pathology is more prevalent in patients with lower socioeconomic status.1 However, the etiology of this phenomenon is currently unknown. It is possible that this patient population has poorer access to health care making them more prone to have surgery for an acute process such as acute cholecystitis as it might be cost prohibitive for them to undergo it for a benign condition such as biliary colic. The purpose of this study was to determine the impact of socioeconomic status on outcomes after cholecystectomy using a large national database. We hypothesized that patients with lower socioeconomic status would have worse outcomes after cholecystectomy.
2
The American Journal of Surgery, Vol -, No -, - 2016
Methods
Baseline patient characteristics by socioeconomic status
Database The Healthcare Cost and Utilization Project National Inpatient Sample (NIS) Database was used for this study. The NIS database contains data from approximately 8 million hospital stays each year. The database represents a stratified sample of 20% of the nonfederal United States hospitals. The NIS database is the largest publicly available, all payer inpatient health care database in the United States.2 A selfweighting design reduces the margin of error for estimates and delivers population based estimates. Our statistical analysis was based on this weighting design as established in previous studies.3 The NIS is a publically available deidentified database and was therefore granted exempt status from our IRB committee.
Study population Adult patients that underwent open cholecystectomy (51.21, 51.22) or laparoscopic cholecystectomy (51.23, 51.24) from 2005 to 2011 were initially identified by the International Classification of Disease, 9th Revision (ICD-9) code. Patients were subsequently stratified into quartiles by socioeconomic status as defined by median household income. Diagnoses including cholecystitis, choledocholithiasis, and cholelithiasis as well as complications were determined by ICD-9 procedure codes as defined in previous studies.3
Data and statistical analysis The primary outcome was inpatient mortality after cholecystectomy. Secondary outcomes included in-hospital complications, length of stay, and cost. Common postoperative complications were identified by ICD-9 codes as established in previous studies. Continuous and categorical variables were compared with student t test and chi-square analysis. All continuous variables were presented as mean 6 standard deviation. Weighted frequencies and weighted multiple variable logistic regression analysis using clinically relevant variables were used to examine postoperative complications and mortality. Odds ratios (ORs) with 95% confidence intervals were presented for each covariate. A P value less than .05 was considered statistically significant. The multivariate model was validated by subdividing it into a training group (75%) and a validation group (25%). Each model was fit using the training data set and then assessed by applying the model to the validation set. Data were analyzed using SAS 9.4 software (SAS Institute, Cary, NC).
Results There were 2,058,611 cholecystectomies performed during the study period. Of these, 271,655 (13.2%) were performed open and 1,786,956 (86.8%) were laparoscopic.
A comparison of patient demographics by socioeconomic status as defined by quartile of median household income is shown in Table 1. The lowest income group was younger (50.5 years; P , .001), fewer were male (31.8%; P , .001), and had lower Charlson Comorbidity Index (2.08; P , .001). They were more likely to be Hispanic (24.7%; P , .001) or African American (15.8%; P , .001) and less likely to be Caucasian (53.9%; P , .001). As expected, they were more likely to have Medicare (32.3%; P , .001), Medicaid (19.2%; P , .001), or be self-payers (11.9%; P , .001) and less likely to have private insurance (31.1%; P , .001). Patients in the lowest income quartile were most likely to have their cholecystectomy in an emergent setting (71.7%, P , .001). In addition, these patients had the highest incidence of open cholecystectomy (14.8%, P , .001). There was a statistically significant difference in presenting diagnosis by income status, but in the authors’ opinion the lowest quartile did not have a clinically significant difference of acute cholecystitis given the small variation in incidence (73.9% vs 73.7% vs 74.2% vs 74.8%, P , .001).
Survival analysis The multivariate analysis for survival is shown in Table 2. Patients with higher income had lower mortality risk (OR .88; 95% CI, .82 to .95; P , .001). Female gender (OR .78; 95% CI, .67 to .90; P , .001), Medicaid status (OR 1.58; 95% CI, 1.31 to 1.91; P , .001), and Charlson Comorbidity index (OR 1.83; 95% CI, 1.74 to 1.93; P , .001) were associated with mortality.
Postoperative outcomes Multivariate analysis by income is shown in Table 3. Income had no association with postoperative complications including pulmonary embolism, acute kidney injury, myocardial infarction, surgical site infection, and deep vein thrombosis. Patients with higher income had shorter length of stay (4.2 vs 4.0 vs 3.9 vs 3.8 days, P , .001) and higher total hospital charges ($33,700 vs $33,300 vs $35,600 vs $35,800, P , .001). Subgroup analysis was performed on patients with Medicaid as they had significantly worse survival (Table 4). As shown, these patients were more likely to have postoperative complications.
Comments There are more than 750,000 cholecystectomies performed yearly in the United States.4 It is important to recognize patient populations that might be at higher risk for postoperative complications and mortality to potentially identify modifiable risk factors. Carbonell et al5 performed
V. Ambur et al. Table 1
Socioeconomic status and cholecystectomy
Baseline patient characteristics by income and operative approach
Age Charlson Comorbidity Index Elective admission (%) Emergent/urgent (%) Male (%) Caucasian (%) Black (%) Hispanic (%) Asian (%) Other (%) Medicare (%) Medicaid (%) Private insurance (%) Self pay (%) Other insurance (%) Teaching hospital Length of stay (days) Total charges (10k) Mortality Laparoscopic cholecystectomy Open cholecystectomy Presenting diagnosis Acute cholecystitis Choledocholithiasis Cholelithiasis Chronic cholecystitis
Quartile 1
Quartile 2
Quartile 3
Quartile 4
P value
50.5 6 44.3 2.08 6 4.4 1,20,463 (21.4) 3,04,250 (71.7) 1,35,020 (31.8) 1,88,349 (53.9) 55,332 (15.8) 86,427 (24.7) 3,594 (1.0) 15,783 (4.6) 1,36,554 (32.3) 81,101 (19.2) 1,31,705 (31.1) 50,177 (11.9) 6,304 (1.5) 1,35,898 (32.0) 4.2 6 9.3 33.7 6 76 2,626 (.47) 4,76,925 (85.2) 82,969 (14.8)
51.9 6 44.1 2.14 6 4.3 1,15,584 (21.3) 2,87,702 (70.1) 1,37,912 (33.8) 2,18,924 (67.7) 27,108 (8.4) 59,758 (18.5) 4,813 (1.5) 12,544 (3.9) 1,33,295 (32.7) 57,554 (14.1) 1,57,461 (38.7) 39,002 (9.6) 4,588 (1.1) 1,22,360 (30.0) 4.0 6 9.2 33.3 6 73.5 2,197 (.40) 4,65,765 (86.4) 73,483 (13.6)
52.3 6 43.3 2.13 6 4.27 1,03,023 (19.9) 2,68,553 (68.9) 1,34,875 (34.6) 2,19,385 (69.2) 21,738 (6.9) 56,273 (17.7) 7,275 (2.3) 12,594 (3.9) 1,21,465 (31.2) 43,378 (11.2) 1,74,891 (45.0) 31,259 (8.0) 4,191 (1.1) 1,24,522 (32.0) 3.9 6 8.6 35.6 6 75.3 2,020 (.39) 4,50,346 (87.4) 65,058 (12.6)
53.4 6 41.8 2.16 6 4.1 85,133 (19.0) 2,23,966 (66.9) 1,22,635 (36.6) 2,15,218 (75.3) 14,798 (5.2) 34,195 (12.0) 11,677 (4.1) 10,133 (3.4) 99,286 (29.7) 21,801 (6.5) 1,83,211 (54.8) 17,783 (5.3) 1768 (.5) 1,19,011 (35.6) 3.8 6 9.0 35.8 6 80.5 1,528 (.34) 3,93,920 (88.7) 50,145 (11.3)
,.001 ,.001 ,.001 ,.001 ,.001 ,.001 ,.001 ,.001 ,.001 ,.001 ,.001 ,.001 ,.001 ,.001 ,.001 ,.001 ,.001 ,.001 ,.001 ,.001 ,.001
4,17,525 15,760 97,560 33,602
4,00,396 15,311 95,325 32,325
3,85,575 14,813 91,465 27,896
3,35,504 12,640 78,070 22,009
,.001 ,.001 ,.001 ,.001
(73.9) (2.8) (17.3) (6.0)
a similar analysis that examined the impact of patient and hospital demographics on cholecystectomy outcomes. They concluded that age, postoperative complications, and emergency surgery were the only predictors of mortality. Interestingly, their study demonstrated a lower complication rate and no mortality risk in those patients in the lowest income bracket. We chose to re-evaluate these outcomes over a longer study period as the previous study was conducted over 1 year and thus potentially more susceptible to bias. In addition, we used variables not present in the data set at the time of their analysis that allowed us to emphasize the impact of socioeconomic status (SES) and presenting diagnosis. The Agency for Healthcare Research and Quality has developed an SES index comprised 7 variables of which median household income had the strongest association Table 2
3
(73.7) (2.8) (17.5) (6.0)
(74.2) (2.8) (17.6) (5.4)
(74.8) (2.8) (17.4) (4.9)
with SES.6 For this reason, we used income quartile as our marker for SES. Our study demonstrated that increasing SES was associated with survival (OR .88; 95% CI, .82 to .95; P , .001). Socioeconomic status has been demonstrated to be a risk factor for postoperative mortality in numerous other studies.7–9 Bennett et al10 included more complex cases such as cardiac, vascular, and complex gastrointestinal procedures, whereas other studies have found it is not significant in the trauma population. Patients in the lowest SES quartile had longer length of stay (4.2 vs 4.0 vs 3.9 vs 3.8 days, P , .001) and lower total hospital charges ($33,700 vs $33,300 vs $35,600 vs $35,800, P , .001). The length of stay is likely clinically significant when comparing the lowest and highest income quartiles where the difference is .4 days. Patients with lower SES had the longest length of stay but lower total
Multivariate analysis of survival
Covariate
Odds ratio
95% Confidence interval
P value
Age African American Income Female Medicaid Charlson Comorbidity Index Teaching hospital Emergency surgery
1.09 1.2 .88 .78 1.58 1.83 1.11 .98
.99–1.02 .93–1.55 .82–.95 .67–.90 1.31–1.91 1.74–1.93 .95–1.3 .84–1.14
.08 .16 ,.001 .001 ,.001 ,.001 .17 .77
The American Journal of Surgery, Vol -, No -, - 2016
4 Table 3
Multivariate analysis by income
Covariate
Odds ratio
95% Confidence interval
P value
Mortality Pulmonary embolism Pneumonia Acute kidney injury Myocardial infarction Surgical site infection Deep vein thrombosis
.88 .95 .96 .99 .98 .99 1.06
.82–.94 .83–1.08 .93–.99 .96–1.03 .89–1.1 .94–1.05 .92–1.2
,.01 .43 .03 .60 .56 .78 .43
hospital charges and no increased rate of postoperative complications which is corroborated by the findings from Carbonell et al.5 These patients were also more likely to be African American (15.8%) or Hispanic (24.7%), although race was not found to be associated with mortality. These 2 groups were more likely to have cholecystectomy for acute cholecystitis. Hispanic patients had the shortest length of stay (3.7 days), lowest incidence of conversion to open cholecystectomy (10.2%), and lowest Charlson Comorbidity Index (1.4). These findings confirm those by Carbonell et al although the exact association is unclear and needs to be investigated further. Patients with Medicaid also had an increased mortality risk (OR 1.58; 95% CI, 1.31 to 1.91; P , .001). They were significantly more prone to postoperative complications including pneumonia, surgical site infection, and acute kidney injury (Table 4). Subgroup analysis revealed that these patients did not have a higher rate of open cholecystectomy (10.5%) or higher Charlson Comorbidity Index (1.04 6 1.45). Clearly, the Medicaid group and lower income group include similar patients. As expected given their increased age, patients with Medicare had the longest length of stay (5.4 days vs 3.7 days), highest total hospital charges ($42,000 vs $34,000), highest mortality (1% vs .2%), and highest Charlson Comorbidity Index (4.2 vs 1.0). Other factors associated with mortality were male sex and Charlson Comorbidity Index, which have been demonstrated in prior studies.11,12 As stated previously, socioeconomic status has been demonstrated in prior studies to be associated with increased postoperative mortality. Our study builds on the current literature by establishing this increased mortality risk in patient with lower SES after cholecystectomy, a
Table 4
more common procedure. Prior studies have speculated on the proposed mechanisms which include a combination of hospital and/or patient related factors.13–15 Hospital-related factors include hospital demographics, health care provider, and system bias. The current health care system may limit access to care and predispose a greater percentage of these patients to be cared for at institutions with less resources. In addition, there might be a system bias toward referring these patients to lessexperienced surgeons.16 However, previous studies have shown that hospital bed size and location are not associated with increased mortality.5 Patient-related factors include more severe comorbidities and disease severity at presentation secondary to limitations in access to care and patient education. Comorbidities do not appear to be a likely cause as patients with lower SES had the lowest Charlson Comorbidity Index (2.08, P , .001). Unfortunately, disease severity cannot be assessed using this database which is a limitation of the study. Interestingly, they were not more likely to have cholecystectomy for acute cholecystitis. However, they did have the highest incidence of emergent/urgent cholecystectomy suggesting some combination of poorer access to care to have elective surgery (21.4% vs 19.0%) and possibly advanced disease which explains the higher incidence of open cholecystectomy among this group. We were unable to determine the exact etiology of this discrepancy in operative approach due to limitations of the data provided in this administrative database. We recommend that health care providers continue to strive to deliver equivalent care independent of socioeconomic status. Emphasis needs to be placed on patient education and effective communication during patient encounters, which is corroborated by previous studies examining the doctor-patient relationship in patients with lower
Multivariate analysis by Medicaid
Covariate
Odds ratio
95% Confidence interval
P value
Pulmonary embolus Pneumonia Respiratory failure Acute kidney injury Myocardial infarction Surgical site infection Deep vein thrombosis
.92 1.36 1.22 1.26 1.12 1.21 1.26
.62–1.38 1.23–1.50 1.10–1.35 1.14–1.40 .88–1.45 1.02–1.43 .84–1.89
.69 ,.001 ,.001 ,.001 .36 .03 .27
V. Ambur et al.
Socioeconomic status and cholecystectomy
SES.17,18 Further investigation to identify potential policy interventions that can ameliorate the mortality effect of socioeconomic status are vital. This study does have several limitations. Major limitations are inherent to our use of a large administrative database such as the NIS. This database does not include outpatient cholecystectomies which are performed in many patients with benign gallbladder disease. We could not control for all confounders as certain variables are not available in the database. Examples of such variables include disease severity, preoperative physiologic and functional status. Furthermore, we used quartile of median household income by zip code to designate socioeconomic status. It is also important to note that even small deviations were statistically significant in sample sizes this large. The authors’ determined whether these differences or relationships were clinically significant given the degree of deviation and the current knowledge on the topic. This is a major limitation as some of the outcomes are open to subjective clinical interpretation. In conclusion, our study confirms the negative impact of socioeconomic status on postoperative outcomes after cholecystectomy. These patients are more likely to have surgery on an emergent basis, which may be due to a lack of access to elective surgery. Being insured with Medicaid was also associated with increased postoperative mortality. We recommend that the health care system continues to make efforts to limit disparities in access to health care and that providers continue to emphasize communication and patient education during encounters. Further studies are needed to clarify the etiology behind these findings and to identify additional policy interventions that can ameliorate this mortality effect.
Acknowledgments The authors have no acknowledgements to make.
References 1. Diehl AK, Rosenthal M, Hazuda HP, et al. Socioeconomic status and the prevalence of clinical gallbladder disease. J Chronic Dis 1985;38:1019–26.
5 2. HCUP-US NIS Overview [Internet]. [cited 2014 Jul 5]. Available at: http://www.hcup-us.ahrq.gov/nisoverview.jsp. Accessed February 2, 2015. 3. Kilic A, Sheer A, Shah A, et al. Outcomes of cholecystectomy in US heart transplant recipients. Ann Surg 2013;258:312–7. 4. Proceedings of the NIH Consensus Development Conference on gallstones and laparoscopic cholecystectomy. Bethesda, Maryland, September 14-16, 1992. Am J Surg 1993;165:387–548. 5. Carbonell AM, Lincourt AE, Kercher KW, et al. Do patient or hospital demographics predict cholecystectomy outcomes? A nationwide study of 93,578 patients. Surg Endosc 2005;19:767–73. 6. Bonito A, Bann C, Eicheldinger C, et al. Creation of New Raceethnicity Codes and Socioeconomic Status (SES) Indicators for Medicare Beneficiaries. Rockville, MD: RTI International, US Department of Health and Human Services, Agency for Health Research and Quality; 2008. 7. Bennett KM, Scarborough JE, Pappas TN, et al. Patient socioeconomic status is an independent predictor of operative mortality. Ann Surg 2010;252:552–7; discussion 557–8. 8. Birkmeyer NJO, Gu N, Baser O, et al. Socioeconomic status and surgical mortality in the elderly. Med Care 2008;46:893–9. 9. Reames BN, Birkmeyer NJO, Dimick JB, et al. Socioeconomic disparities in mortality after cancer surgery: failure to rescue. JAMA Surg 2014;149:475–81. 10. Taghavi S, Jayarajan SN, Duran JM, et al. Does payer status matter in predicting penetrating trauma outcomes? Surgery 2012;152:227–31. 11. Giger UF, Michel J-M, Opitz I, et al. Risk factors for perioperative complications in patients undergoing laparoscopic cholecystectomy: analysis of 22,953 consecutive cases from the Swiss association of laparoscopic and thoracoscopic surgery database. J Am Coll Surg 2006;203:723–8. 12. Grossman R, Mukherjee D, Chang DC, et al. Preoperative Charlson comorbidity score predicts postoperative outcomes among older intracranial meningioma patients. World Neurosurg 2011;75:279–85. 13. Kim C, Diez Roux AV, Hofer TP, et al. Area socioeconomic status and mortality after coronary artery bypass graft surgery: the role of hospital volume. Am Heart J 2007;154:385–90. 14. Hasnain-Wynia R, Baker DW, Nerenz D, et al. Disparities in health care are driven by where minority patients seek care: examination of the hospital quality alliance measures. Arch Intern Med 2007;167:1233–9. 15. Skinner J, Chandra A, Staiger D, et al. Mortality after acute myocardial infarction in hospitals that disproportionately treat black patients. Circulation 2005;112:2634–41. 16. Greenstein AJ, Moskowitz A, Gelijns AC, et al. Payer status and treatment paradigm for acute cholecystitis. Arch Surg 2012;147:453–8. 17. Bernheim SM, Ross JS, Krumholz HM, et al. Influence of patients’ socioeconomic status on clinical management decisions: a qualitative study. Ann Fam Med 2008;6:53–9. 18. Willems S, De Maesschalck S, Deveugele M, et al. Socio-economic status of the patient and doctor-patient communication: does it make a difference? Patient Educ Couns 2005;56:139–46.
Does socioeconomic status predict outcomes after cholecystectomy? Vishnu Ambur, M.D.a,*, Sharven Taghavi, M.D.b, Sagar Kadakia, M.D.a, Senthil Jayarajan, M.D.b, John Gaughan, Ph.D.c, Lars Ola Sjoholm, M.D.a, Abhijit Pathak, M.D.a, Thomas Santora, M.D.a, Joseph Rappold, M.D.a, Amy J. Goldberg, M.D.a a
Temple University Hospital, Department of Surgery, Philadelphia, PA, USA; bBarnes Jewish Hospital/ Washington University in St. Louis, Department of Surgery, St Louis, MO, USA; cTemple University School of Medicine, Biostatistics Consulting Center, Philadelphia, PA, USA
KEYWORDS: Socioeconomic status; Cholecystectomy; Income; Medicaid; National inpatient sample
Abstract BACKGROUND: This study was performed to evaluate the effect of socioeconomic status (SES) on outcomes after cholecystectomy. METHODS: The National Inpatient Sample (NIS) database (2005 to 2011) was queried for patients undergoing cholecystectomy. Clinically relevant variables were used to examine clinical characteristics, postoperative complications, and mortality. SES was investigated by examining income quartile. RESULTS: More than 2 million patients underwent cholecystectomy during this period. They were divided into quartiles by SES. The lowest cohort was younger (50 years, P , .001) and had the lowest Charlson Comorbidity Index (2.08, P , .001). This cohort was more likely African American (15.8%, P , .001) and more likely to have Medicaid (19.2%, P , .001). Using split-sample validation and multivariate analysis, lower SES, Charlson comorbidity Index, and Medicaid recipients were associated with increased mortality. CONCLUSIONS: Patients with Medicaid and lower SES had poorer outcomes after cholecystectomy. Ó 2016 Elsevier Inc. All rights reserved.
Cholecystectomy remains one of the most common surgeries performed in the United States today. It is performed for various indications ranging from benign conditions including biliary colic to acute conditions such as acute cholecystitis and gallstone pancreatitis. Previous There were no relevant financial relationships or any sources of support in the form of grants, equipment, or drugs. The authors declare no conflicts of interest. * Corresponding author. Tel.: 215-707-3633; fax: 215-707-1915. E-mail address: [email protected] Manuscript received October 5, 2015; revised manuscript April 19, 2016 0002-9610/$ - see front matter Ó 2016 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.amjsurg.2016.04.012
studies have shown that gallbladder pathology is more prevalent in patients with lower socioeconomic status.1 However, the etiology of this phenomenon is currently unknown. It is possible that this patient population has poorer access to health care making them more prone to have surgery for an acute process such as acute cholecystitis as it might be cost prohibitive for them to undergo it for a benign condition such as biliary colic. The purpose of this study was to determine the impact of socioeconomic status on outcomes after cholecystectomy using a large national database. We hypothesized that patients with lower socioeconomic status would have worse outcomes after cholecystectomy.
2
The American Journal of Surgery, Vol -, No -, - 2016
Methods
Baseline patient characteristics by socioeconomic status
Database The Healthcare Cost and Utilization Project National Inpatient Sample (NIS) Database was used for this study. The NIS database contains data from approximately 8 million hospital stays each year. The database represents a stratified sample of 20% of the nonfederal United States hospitals. The NIS database is the largest publicly available, all payer inpatient health care database in the United States.2 A selfweighting design reduces the margin of error for estimates and delivers population based estimates. Our statistical analysis was based on this weighting design as established in previous studies.3 The NIS is a publically available deidentified database and was therefore granted exempt status from our IRB committee.
Study population Adult patients that underwent open cholecystectomy (51.21, 51.22) or laparoscopic cholecystectomy (51.23, 51.24) from 2005 to 2011 were initially identified by the International Classification of Disease, 9th Revision (ICD-9) code. Patients were subsequently stratified into quartiles by socioeconomic status as defined by median household income. Diagnoses including cholecystitis, choledocholithiasis, and cholelithiasis as well as complications were determined by ICD-9 procedure codes as defined in previous studies.3
Data and statistical analysis The primary outcome was inpatient mortality after cholecystectomy. Secondary outcomes included in-hospital complications, length of stay, and cost. Common postoperative complications were identified by ICD-9 codes as established in previous studies. Continuous and categorical variables were compared with student t test and chi-square analysis. All continuous variables were presented as mean 6 standard deviation. Weighted frequencies and weighted multiple variable logistic regression analysis using clinically relevant variables were used to examine postoperative complications and mortality. Odds ratios (ORs) with 95% confidence intervals were presented for each covariate. A P value less than .05 was considered statistically significant. The multivariate model was validated by subdividing it into a training group (75%) and a validation group (25%). Each model was fit using the training data set and then assessed by applying the model to the validation set. Data were analyzed using SAS 9.4 software (SAS Institute, Cary, NC).
Results There were 2,058,611 cholecystectomies performed during the study period. Of these, 271,655 (13.2%) were performed open and 1,786,956 (86.8%) were laparoscopic.
A comparison of patient demographics by socioeconomic status as defined by quartile of median household income is shown in Table 1. The lowest income group was younger (50.5 years; P , .001), fewer were male (31.8%; P , .001), and had lower Charlson Comorbidity Index (2.08; P , .001). They were more likely to be Hispanic (24.7%; P , .001) or African American (15.8%; P , .001) and less likely to be Caucasian (53.9%; P , .001). As expected, they were more likely to have Medicare (32.3%; P , .001), Medicaid (19.2%; P , .001), or be self-payers (11.9%; P , .001) and less likely to have private insurance (31.1%; P , .001). Patients in the lowest income quartile were most likely to have their cholecystectomy in an emergent setting (71.7%, P , .001). In addition, these patients had the highest incidence of open cholecystectomy (14.8%, P , .001). There was a statistically significant difference in presenting diagnosis by income status, but in the authors’ opinion the lowest quartile did not have a clinically significant difference of acute cholecystitis given the small variation in incidence (73.9% vs 73.7% vs 74.2% vs 74.8%, P , .001).
Survival analysis The multivariate analysis for survival is shown in Table 2. Patients with higher income had lower mortality risk (OR .88; 95% CI, .82 to .95; P , .001). Female gender (OR .78; 95% CI, .67 to .90; P , .001), Medicaid status (OR 1.58; 95% CI, 1.31 to 1.91; P , .001), and Charlson Comorbidity index (OR 1.83; 95% CI, 1.74 to 1.93; P , .001) were associated with mortality.
Postoperative outcomes Multivariate analysis by income is shown in Table 3. Income had no association with postoperative complications including pulmonary embolism, acute kidney injury, myocardial infarction, surgical site infection, and deep vein thrombosis. Patients with higher income had shorter length of stay (4.2 vs 4.0 vs 3.9 vs 3.8 days, P , .001) and higher total hospital charges ($33,700 vs $33,300 vs $35,600 vs $35,800, P , .001). Subgroup analysis was performed on patients with Medicaid as they had significantly worse survival (Table 4). As shown, these patients were more likely to have postoperative complications.
Comments There are more than 750,000 cholecystectomies performed yearly in the United States.4 It is important to recognize patient populations that might be at higher risk for postoperative complications and mortality to potentially identify modifiable risk factors. Carbonell et al5 performed
V. Ambur et al. Table 1
Socioeconomic status and cholecystectomy
Baseline patient characteristics by income and operative approach
Age Charlson Comorbidity Index Elective admission (%) Emergent/urgent (%) Male (%) Caucasian (%) Black (%) Hispanic (%) Asian (%) Other (%) Medicare (%) Medicaid (%) Private insurance (%) Self pay (%) Other insurance (%) Teaching hospital Length of stay (days) Total charges (10k) Mortality Laparoscopic cholecystectomy Open cholecystectomy Presenting diagnosis Acute cholecystitis Choledocholithiasis Cholelithiasis Chronic cholecystitis
Quartile 1
Quartile 2
Quartile 3
Quartile 4
P value
50.5 6 44.3 2.08 6 4.4 1,20,463 (21.4) 3,04,250 (71.7) 1,35,020 (31.8) 1,88,349 (53.9) 55,332 (15.8) 86,427 (24.7) 3,594 (1.0) 15,783 (4.6) 1,36,554 (32.3) 81,101 (19.2) 1,31,705 (31.1) 50,177 (11.9) 6,304 (1.5) 1,35,898 (32.0) 4.2 6 9.3 33.7 6 76 2,626 (.47) 4,76,925 (85.2) 82,969 (14.8)
51.9 6 44.1 2.14 6 4.3 1,15,584 (21.3) 2,87,702 (70.1) 1,37,912 (33.8) 2,18,924 (67.7) 27,108 (8.4) 59,758 (18.5) 4,813 (1.5) 12,544 (3.9) 1,33,295 (32.7) 57,554 (14.1) 1,57,461 (38.7) 39,002 (9.6) 4,588 (1.1) 1,22,360 (30.0) 4.0 6 9.2 33.3 6 73.5 2,197 (.40) 4,65,765 (86.4) 73,483 (13.6)
52.3 6 43.3 2.13 6 4.27 1,03,023 (19.9) 2,68,553 (68.9) 1,34,875 (34.6) 2,19,385 (69.2) 21,738 (6.9) 56,273 (17.7) 7,275 (2.3) 12,594 (3.9) 1,21,465 (31.2) 43,378 (11.2) 1,74,891 (45.0) 31,259 (8.0) 4,191 (1.1) 1,24,522 (32.0) 3.9 6 8.6 35.6 6 75.3 2,020 (.39) 4,50,346 (87.4) 65,058 (12.6)
53.4 6 41.8 2.16 6 4.1 85,133 (19.0) 2,23,966 (66.9) 1,22,635 (36.6) 2,15,218 (75.3) 14,798 (5.2) 34,195 (12.0) 11,677 (4.1) 10,133 (3.4) 99,286 (29.7) 21,801 (6.5) 1,83,211 (54.8) 17,783 (5.3) 1768 (.5) 1,19,011 (35.6) 3.8 6 9.0 35.8 6 80.5 1,528 (.34) 3,93,920 (88.7) 50,145 (11.3)
,.001 ,.001 ,.001 ,.001 ,.001 ,.001 ,.001 ,.001 ,.001 ,.001 ,.001 ,.001 ,.001 ,.001 ,.001 ,.001 ,.001 ,.001 ,.001 ,.001 ,.001
4,17,525 15,760 97,560 33,602
4,00,396 15,311 95,325 32,325
3,85,575 14,813 91,465 27,896
3,35,504 12,640 78,070 22,009
,.001 ,.001 ,.001 ,.001
(73.9) (2.8) (17.3) (6.0)
a similar analysis that examined the impact of patient and hospital demographics on cholecystectomy outcomes. They concluded that age, postoperative complications, and emergency surgery were the only predictors of mortality. Interestingly, their study demonstrated a lower complication rate and no mortality risk in those patients in the lowest income bracket. We chose to re-evaluate these outcomes over a longer study period as the previous study was conducted over 1 year and thus potentially more susceptible to bias. In addition, we used variables not present in the data set at the time of their analysis that allowed us to emphasize the impact of socioeconomic status (SES) and presenting diagnosis. The Agency for Healthcare Research and Quality has developed an SES index comprised 7 variables of which median household income had the strongest association Table 2
3
(73.7) (2.8) (17.5) (6.0)
(74.2) (2.8) (17.6) (5.4)
(74.8) (2.8) (17.4) (4.9)
with SES.6 For this reason, we used income quartile as our marker for SES. Our study demonstrated that increasing SES was associated with survival (OR .88; 95% CI, .82 to .95; P , .001). Socioeconomic status has been demonstrated to be a risk factor for postoperative mortality in numerous other studies.7–9 Bennett et al10 included more complex cases such as cardiac, vascular, and complex gastrointestinal procedures, whereas other studies have found it is not significant in the trauma population. Patients in the lowest SES quartile had longer length of stay (4.2 vs 4.0 vs 3.9 vs 3.8 days, P , .001) and lower total hospital charges ($33,700 vs $33,300 vs $35,600 vs $35,800, P , .001). The length of stay is likely clinically significant when comparing the lowest and highest income quartiles where the difference is .4 days. Patients with lower SES had the longest length of stay but lower total
Multivariate analysis of survival
Covariate
Odds ratio
95% Confidence interval
P value
Age African American Income Female Medicaid Charlson Comorbidity Index Teaching hospital Emergency surgery
1.09 1.2 .88 .78 1.58 1.83 1.11 .98
.99–1.02 .93–1.55 .82–.95 .67–.90 1.31–1.91 1.74–1.93 .95–1.3 .84–1.14
.08 .16 ,.001 .001 ,.001 ,.001 .17 .77
The American Journal of Surgery, Vol -, No -, - 2016
4 Table 3
Multivariate analysis by income
Covariate
Odds ratio
95% Confidence interval
P value
Mortality Pulmonary embolism Pneumonia Acute kidney injury Myocardial infarction Surgical site infection Deep vein thrombosis
.88 .95 .96 .99 .98 .99 1.06
.82–.94 .83–1.08 .93–.99 .96–1.03 .89–1.1 .94–1.05 .92–1.2
,.01 .43 .03 .60 .56 .78 .43
hospital charges and no increased rate of postoperative complications which is corroborated by the findings from Carbonell et al.5 These patients were also more likely to be African American (15.8%) or Hispanic (24.7%), although race was not found to be associated with mortality. These 2 groups were more likely to have cholecystectomy for acute cholecystitis. Hispanic patients had the shortest length of stay (3.7 days), lowest incidence of conversion to open cholecystectomy (10.2%), and lowest Charlson Comorbidity Index (1.4). These findings confirm those by Carbonell et al although the exact association is unclear and needs to be investigated further. Patients with Medicaid also had an increased mortality risk (OR 1.58; 95% CI, 1.31 to 1.91; P , .001). They were significantly more prone to postoperative complications including pneumonia, surgical site infection, and acute kidney injury (Table 4). Subgroup analysis revealed that these patients did not have a higher rate of open cholecystectomy (10.5%) or higher Charlson Comorbidity Index (1.04 6 1.45). Clearly, the Medicaid group and lower income group include similar patients. As expected given their increased age, patients with Medicare had the longest length of stay (5.4 days vs 3.7 days), highest total hospital charges ($42,000 vs $34,000), highest mortality (1% vs .2%), and highest Charlson Comorbidity Index (4.2 vs 1.0). Other factors associated with mortality were male sex and Charlson Comorbidity Index, which have been demonstrated in prior studies.11,12 As stated previously, socioeconomic status has been demonstrated in prior studies to be associated with increased postoperative mortality. Our study builds on the current literature by establishing this increased mortality risk in patient with lower SES after cholecystectomy, a
Table 4
more common procedure. Prior studies have speculated on the proposed mechanisms which include a combination of hospital and/or patient related factors.13–15 Hospital-related factors include hospital demographics, health care provider, and system bias. The current health care system may limit access to care and predispose a greater percentage of these patients to be cared for at institutions with less resources. In addition, there might be a system bias toward referring these patients to lessexperienced surgeons.16 However, previous studies have shown that hospital bed size and location are not associated with increased mortality.5 Patient-related factors include more severe comorbidities and disease severity at presentation secondary to limitations in access to care and patient education. Comorbidities do not appear to be a likely cause as patients with lower SES had the lowest Charlson Comorbidity Index (2.08, P , .001). Unfortunately, disease severity cannot be assessed using this database which is a limitation of the study. Interestingly, they were not more likely to have cholecystectomy for acute cholecystitis. However, they did have the highest incidence of emergent/urgent cholecystectomy suggesting some combination of poorer access to care to have elective surgery (21.4% vs 19.0%) and possibly advanced disease which explains the higher incidence of open cholecystectomy among this group. We were unable to determine the exact etiology of this discrepancy in operative approach due to limitations of the data provided in this administrative database. We recommend that health care providers continue to strive to deliver equivalent care independent of socioeconomic status. Emphasis needs to be placed on patient education and effective communication during patient encounters, which is corroborated by previous studies examining the doctor-patient relationship in patients with lower
Multivariate analysis by Medicaid
Covariate
Odds ratio
95% Confidence interval
P value
Pulmonary embolus Pneumonia Respiratory failure Acute kidney injury Myocardial infarction Surgical site infection Deep vein thrombosis
.92 1.36 1.22 1.26 1.12 1.21 1.26
.62–1.38 1.23–1.50 1.10–1.35 1.14–1.40 .88–1.45 1.02–1.43 .84–1.89
.69 ,.001 ,.001 ,.001 .36 .03 .27
V. Ambur et al.
Socioeconomic status and cholecystectomy
SES.17,18 Further investigation to identify potential policy interventions that can ameliorate the mortality effect of socioeconomic status are vital. This study does have several limitations. Major limitations are inherent to our use of a large administrative database such as the NIS. This database does not include outpatient cholecystectomies which are performed in many patients with benign gallbladder disease. We could not control for all confounders as certain variables are not available in the database. Examples of such variables include disease severity, preoperative physiologic and functional status. Furthermore, we used quartile of median household income by zip code to designate socioeconomic status. It is also important to note that even small deviations were statistically significant in sample sizes this large. The authors’ determined whether these differences or relationships were clinically significant given the degree of deviation and the current knowledge on the topic. This is a major limitation as some of the outcomes are open to subjective clinical interpretation. In conclusion, our study confirms the negative impact of socioeconomic status on postoperative outcomes after cholecystectomy. These patients are more likely to have surgery on an emergent basis, which may be due to a lack of access to elective surgery. Being insured with Medicaid was also associated with increased postoperative mortality. We recommend that the health care system continues to make efforts to limit disparities in access to health care and that providers continue to emphasize communication and patient education during encounters. Further studies are needed to clarify the etiology behind these findings and to identify additional policy interventions that can ameliorate this mortality effect.
Acknowledgments The authors have no acknowledgements to make.
References 1. Diehl AK, Rosenthal M, Hazuda HP, et al. Socioeconomic status and the prevalence of clinical gallbladder disease. J Chronic Dis 1985;38:1019–26.
5 2. HCUP-US NIS Overview [Internet]. [cited 2014 Jul 5]. Available at: http://www.hcup-us.ahrq.gov/nisoverview.jsp. Accessed February 2, 2015. 3. Kilic A, Sheer A, Shah A, et al. Outcomes of cholecystectomy in US heart transplant recipients. Ann Surg 2013;258:312–7. 4. Proceedings of the NIH Consensus Development Conference on gallstones and laparoscopic cholecystectomy. Bethesda, Maryland, September 14-16, 1992. Am J Surg 1993;165:387–548. 5. Carbonell AM, Lincourt AE, Kercher KW, et al. Do patient or hospital demographics predict cholecystectomy outcomes? A nationwide study of 93,578 patients. Surg Endosc 2005;19:767–73. 6. Bonito A, Bann C, Eicheldinger C, et al. Creation of New Raceethnicity Codes and Socioeconomic Status (SES) Indicators for Medicare Beneficiaries. Rockville, MD: RTI International, US Department of Health and Human Services, Agency for Health Research and Quality; 2008. 7. Bennett KM, Scarborough JE, Pappas TN, et al. Patient socioeconomic status is an independent predictor of operative mortality. Ann Surg 2010;252:552–7; discussion 557–8. 8. Birkmeyer NJO, Gu N, Baser O, et al. Socioeconomic status and surgical mortality in the elderly. Med Care 2008;46:893–9. 9. Reames BN, Birkmeyer NJO, Dimick JB, et al. Socioeconomic disparities in mortality after cancer surgery: failure to rescue. JAMA Surg 2014;149:475–81. 10. Taghavi S, Jayarajan SN, Duran JM, et al. Does payer status matter in predicting penetrating trauma outcomes? Surgery 2012;152:227–31. 11. Giger UF, Michel J-M, Opitz I, et al. Risk factors for perioperative complications in patients undergoing laparoscopic cholecystectomy: analysis of 22,953 consecutive cases from the Swiss association of laparoscopic and thoracoscopic surgery database. J Am Coll Surg 2006;203:723–8. 12. Grossman R, Mukherjee D, Chang DC, et al. Preoperative Charlson comorbidity score predicts postoperative outcomes among older intracranial meningioma patients. World Neurosurg 2011;75:279–85. 13. Kim C, Diez Roux AV, Hofer TP, et al. Area socioeconomic status and mortality after coronary artery bypass graft surgery: the role of hospital volume. Am Heart J 2007;154:385–90. 14. Hasnain-Wynia R, Baker DW, Nerenz D, et al. Disparities in health care are driven by where minority patients seek care: examination of the hospital quality alliance measures. Arch Intern Med 2007;167:1233–9. 15. Skinner J, Chandra A, Staiger D, et al. Mortality after acute myocardial infarction in hospitals that disproportionately treat black patients. Circulation 2005;112:2634–41. 16. Greenstein AJ, Moskowitz A, Gelijns AC, et al. Payer status and treatment paradigm for acute cholecystitis. Arch Surg 2012;147:453–8. 17. Bernheim SM, Ross JS, Krumholz HM, et al. Influence of patients’ socioeconomic status on clinical management decisions: a qualitative study. Ann Fam Med 2008;6:53–9. 18. Willems S, De Maesschalck S, Deveugele M, et al. Socio-economic status of the patient and doctor-patient communication: does it make a difference? Patient Educ Couns 2005;56:139–46.