Preventing perioperative adverse events has become the focus of quality improvement efforts, because these complications collectively represent a significant burden of harm with an estimated cost of $25 billion annually.1–6 Minimally invasive surgery for select general surgery and gynecologic procedures is well recognized to reduce perioperative morbidity compared with similar procedures performed by laparotomy.5–12 In the setting of endometrial cancer, the most common gynecologic cancer diagnosed in U.S. women, a minimally invasive surgical approach to hysterectomy is associated with the same stage-specific survival in randomized controlled trials and fewer complications than with open surgery.8–12 Recognizing that nearly 80% of women with endometrial cancer are obese and possess concurrent comorbidities,13 these patients have great potential to derive benefit from the minimally invasive approach.
Recently, the Society of Gynecologic Oncology and the American College of Surgeons' Commission on Cancer jointly approved of two quality measures for endometrial cancer, including utilization of minimally invasive surgery for those with stage I–III disease.14 Despite these goals, little is known about adoption rates, factors associated with minimally invasive utilization, and the cost implications of underutilization. We hypothesized that minimally invasive surgery utilization in endometrial cancer care varies considerably nationwide and that disparities in surgical care may be related to patient and hospital factors. Our primary study aims were to define U.S. hospital-level utilization of minimally invasive hysterectomy in the treatment of early-stage endometrial cancer and to analyze complication rates and costs associated with minimally invasive compared with open abdominal surgery.
MATERIALS AND METHODS
This was an institutional review board-exempt study performed at Johns Hopkins Hospital, Baltimore, Maryland. The study was deemed exempt because the database used in the analysis is publically available, population-based, and deidentified. In this retrospective cohort study, data were obtained from the Nationwide (National) Inpatient Sample under the Agency for Healthcare Research and Quality's Healthcare Cost and Utilization Project.15 The Nationwide Inpatient Sample is an all-payer, inpatient health care database in the United States providing national estimates of hospital inpatient stays and is the largest publicly available database of its kind. The database includes patient-level hospital discharge data from a representative stratified, weighted sample of approximately 8 million hospital stays from 1,051 U.S. hospitals in 45 states. Sampling weights were applied to provide national estimates.16
Women diagnosed with nonmetastatic endometrial cancer who underwent hysterectomy between 2007 and 2011 were included in the analysis. Metastatic disease was ascertained by the presence of concomitant diagnostic codes for metastases; the positive predictive value and overall accuracy of administrative data sets to capture metastatic disease are 75–95%.17 The Elixhauser comorbidity measure was used to analyze how comorbidities affect clinical outcomes in the inpatient setting.18,19 The comorbidities identified by the Elixhauser comorbidity measure are significantly associated with in-hospital mortality and include both acute and chronic conditions.
Minimally invasive surgery was defined as vaginal, laparoscopic, or robot-assisted hysterectomy utilizing International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) procedure codes. Open hysterectomy surgery was defined as subtotal abdominal hysterectomy (ICD-9-CM code 68.3), total abdominal hysterectomy (68.4), radical abdominal hysterectomy (68.6), and other or unspecified hysterectomy (68.9). Minimally invasive surgery was defined as any of the first three listed ICD-9-CM codes if a fourth digit with a value of “1” was present (ie, 68.31, 68.41, 68.61), vaginal hysterectomy (ICD-9-CM codes 68.5, 68.7), or any robotically assisted hysterectomy (any hysterectomy code plus ICD-9-CM code 17.4). The proportion of patients who underwent a concurrent lymphadenectomy was also assessed. Pelvic, iliac, paraaortic lymphadenectomy, or all of these was defined using the following ICD-9-CM codes: 40.10, 40.11, 40.20, 40.29, 40.50, 40.52, and 40.53.
We evaluated factors associated with minimally invasive surgery and with perioperative complications using a generalized linear mixed logistic regression model with a random intercept for hospital to account for the correlation of procedures within hospitals. Demographic patient and hospital data were stratified according to surgical approach (minimally invasive surgery compared with open abdominal) and compared using Student t test and χ2 test. Individual-level predictors included age at diagnosis, indicator for obesity, Elixhauser comorbidity score (a scale from 1 to 10 with “1” representing few comorbidities and “10” representing multiple comorbidities), race, payer type, and year of diagnosis. Hospital-level predictors included type, geographic region, teaching status, location, size, median patient household income by hospital zip code, and average annual volume of patients with endometrial cancer (less than 10, 10 to less than 30, 30 or more). The Nationwide Inpatient Sample database predetermined the category definitions.
Based on the joint guideline issued by the Society of Gynecologic Oncology and the American Cancer Society's Commission on Cancer (ie, that all women with stage I–III endometrial cancer should undergo minimally invasive surgery), we also analyzed hospital factors associated with performance of at least 80% minimally invasive surgery in nonmetastatic endometrial cancer. A conservative estimate of 80% was chosen because we hypothesized that up to 20% of women with endometrial cancer may not be ideal candidates for minimally invasive surgery as a result of uterine size or features, patient comorbidities, abdominal surgery history, or evidence of metastatic disease requiring laparotomy.
Perioperative complications of interest were defined as those that occurred in the inpatient setting and included overall surgical complication rates, surgical site infection, pneumonia, venous thromboembolic disease (defined as pulmonary embolus, stroke, or deep venous thrombosis), sepsis, wound or fascial dehiscence, vascular injury (laceration), major blood loss (defined as acute posthemorrhagic anemia), hospital stay greater than 2 days, required aftercare, and death. Analyses were performed using SAS.
Total costs of inpatient hospitalization for each patient were calculated by converting the charge data into cost through cost-to-charge ratio, both of which are available in the Nationwide Inpatient Sample database.20 Cost data were inflation-adjusted to 2011 U.S. dollars using the Gross Domestic Product Implicit Price Deflator.21 To account for the skewed nonnormal distribution of health care cost, we used a generalized linear model with log link and gamma distribution for cost, which also adjusted for the patient and health care provider characteristics as was done for the other two outcomes described previously.21 Our cost analysis accounted for the complex survey design used to collect Nationwide Inpatient Sample data and adjusted for clustering at the hospital level. Significant differences were determined by 95% confidence interval (CI) of the difference between minimally invasive surgery and patients undergoing open hysterectomy. The survey module of Stata 13 was used to analyze cost.
We identified 51,700 patients diagnosed with endometrial cancer identified in the database during the study period (2007–2011). After excluding patients classified as having metastatic endometrial cancer, those who did not undergo hysterectomy, and patients with missing age, there were a total of 32,560 unique hospitalizations included in the final analysis (Fig. 1), which represents a population of approximately 161,000 endometrial patients undergoing hysterectomy after sampling weights were applied to provide national estimates. Patient characteristics, stratified by surgical approach, are detailed in Table 1. There was no significant difference in the proportion of patients who underwent surgical staging overall (when analyzing all ICD-9-CM codes for lymphadenectomy) based on surgical approach (minimally invasive surgery=31.2% compared with open=30.7%, P=.210).
Overall, 33.6% underwent a minimally invasive hysterectomy and 66.4% underwent an open abdominal hysterectomy during the study period. The proportion of patients who underwent minimally invasive surgery increased significantly during the study period (Fig. 2) from 22.0% in 2007 compared with 50.8% in 2011 (P<.000). This was primarily a function of increasing robotic surgery rates during the time period (P<.000).
Several patient and hospital factors were associated with receipt of minimally invasive hysterectomy. Older patients (P<.000) and those diagnosed more recently were more likely to undergo minimally invasive surgery: (adjusted odds ratio [OR] 6.35 for 2011 compared with 2007; Table 2). Obese patients (adjusted OR 0.88, 95% CI 0.85–0.91) and those with a higher Elixhauser comorbidity score (adjusted OR 0.82, 95% CI 0.81–0.83) were less likely to undergo minimally invasive surgery, although this association was not observed at the high-volume (greater than 30 cases annually) endometrial cancer centers. Variation in surgical approach was observed across racial groups with black (adjusted OR 0.43, 95% CI 0.41–0.46), Hispanic (adjusted OR 0.77, 95% CI 0.72–0.82), and women of “other race” (adjusted OR 0.74, 95% CI 0.67–0.80) being less likely to undergo minimally invasive surgery than white patients with endometrial cancer. Patients with Medicaid (adjusted OR 0.67, 95% CI 0.62–0.72) or self-pay status (adjusted OR 0.77, 95% CI 0.70–0.85) were also less likely to undergo minimally invasive surgery compared with patients with private insurance. Additionally, those treated at a teaching hospital were more likely to undergo a minimally invasive procedure (OR 1.58, 95% CI 1.12–2.23).
Notably, patients treated at medium- (adjusted OR 2.64, 95% CI 0.36–0.83) or high-volume endometrial cancer centers (adjusted OR 4.22, 95% CI 2.15–8.27) were more likely to undergo minimally invasive surgery than patients who were treated at low-volume hospitals. Approximately one of four patients undergoing surgery for nonmetastatic endometrial cancer surgeries (25.7%) in the United States were performed at low-volume hospitals during the study period, which demonstrated the lowest minimally invasive surgery rates (23.9%; P<.001). Furthermore, patients treated at hospitals located at a zip code within the highest quartile for median household income (ie, hospitals located in more affluent areas) were more likely to undergo minimally invasive surgery (adjusted OR 1.08, 95% CI 1.02–1.13).
In 2007, only 1.5% of hospitals performed minimally invasive surgery in 80% or more of nonmetastatic endometrial cancer cases; by 2011, this increased to 18.9%. Hospital factors independently associated with an 80% (or higher) minimally invasive surgery rate on multivariate analysis included medium- and high- compared with low-volume endometrial cancer centers (adjusted OR 1.28, 95% CI 1.21–1.35), having surgery in the West compared with the Northeast (adjusted OR 1.61, 95% CI 1.51–1.71), private insurance (reference) compared with Medicaid (adjusted OR 0.41, 95% CI 0.33–0.52), and Medicare (adjusted OR 0.86, 95% CI 0.76–0.96) and private hospitals (reference) compared with government (adjusted OR 0.32, 95% CI 0.29–0.36) or nonprofit (adjusted OR 0.57, 95% CI 0.53–0.61) health care facilities.
In all, 5,257 (16.1%) patients experienced a major perioperative complication during their hospitalization. After controlling for obesity, age, Elixhauser comorbidity score, race, payer status, lympadenectomy, and other variables (Table 3), open abdominal surgery was associated with a greater than twofold increased risk of a complication (adjusted OR 2.33, 95% CI 2.23–2.44) compared with patients who underwent minimally invasive surgery. Specifically, open hysterectomy was associated with a higher rate of an intraoperative surgical complication (adjusted OR 2.08, 95% CI 1.98–2.18), surgical site infection (adjusted OR 6.21, 95% CI 5.11–7.54), pneumonia (adjusted OR 2.36, 95% CI 2.06–2.71), venous thromboembolic disease (adjusted OR 3.65, 95% CI 3.12–4.27), major blood loss (adjusted OR 2.92, 95% CI 2.72–3.13), and greater than 2 days' hospital stay (adjusted OR 34.49, 95% CI 37.72–41.35). Hysterectomy approach was not associated with sepsis, myocardial infarction, or death. Additionally, given that the rate of lymphadenectomy was not significantly different between the open and minimally invasive cohorts, the inclusion of lymphadenectomy in the multivariable model did not affect the results (Appendix 1, available online at http://links.lww.com/AOG/A726).
Adjusted mean inpatient costs for open and minimally invasive hysterectomies are shown in Table 4. Overall, laparoscopic hysterectomy without complications ($10,809) and open hysterectomy ($10,935) were associated with the lowest adjusted mean costs compared with of all robotic-assisted hysterectomy (adjusted mean cost $12,408; P<.001). Compared with hysterectomy cases without complications, cases with complications resulted in a significantly higher mean inpatient hospitalization costs (an additional +$7,358 for laparoscopy, $7,444 for open, and $8,447 for robotic surgery; P<.000) for all hysterectomy procedure types. As seen in Table 4, costs associated with robotic cases were significantly higher than open (by $1,473, P<.001) and laparoscopy cases (by $1,600, P<.001) without complications. Similar trends were also observed for cases with complications; robotic cases were approximately $2,476 (P<.001) costlier than open cases and robotic cases were $2,689 costlier than laparoscopy cases (P<.001). Open and laparoscopy cases were not statistically different in terms of their costs.
The choice of surgical approach in the care of women with early-stage endometrial cancer largely remains the result of surgeon preference after discussion with the patient. Given the compelling level I evidence (including eight randomized controlled trials and a Cochrane database review) supporting the use of minimally invasive surgery in the management of early-stage endometrial cancer, it is our hope these data and that of the current study will inform these decisions in favor of minimally invasive approaches.8–11 The current variation in practice patterns identified in our study of the Nationwide Inpatient Sample brings new attention on minimally invasive surgery as an intervention to improve quality and safety in health care.13,23–29 Data from our study suggest that fewer than half of 32,560 patients with nonmetastatic endometrial cancer underwent minimally invasive hysterectomy during a contemporary treatment period. Although not every patient is a candidate for minimally invasive hysterectomy, this marked underutilization represents an opportunity to address the nearly twofold higher rate of perioperative complications associated with open surgery and the higher costs related to complications.
In the current study, patients of black or Hispanic race, those insured by Medicaid, and patients treated at hospitals with low endometrial cancer volume were less likely to receive minimally invasive surgery. Comorbid, minority, and underserved health care populations are at great risk for poorer outcomes after surgery.12,13,30 The use of laparotomy resulted in a sixfold increase in the rate of surgical site infection, three times the rate of venous thromboembolic events, and a more than doubling in pneumonia rate. By considering a 50% increase in minimally utilization nationally for a variety of surgical procedures, this would potentially result in 3,578 fewer complications, 144,853 fewer hospital days, and $288 million in savings annually.23 Because most women with endometrial cancer are obese and possess comorbidities,24,30 this cohort is at a particularly high risk of perioperative morbidity and would be among the populations most likely to benefit from a nationwide effort to reduce complications and health care costs with better utilization of minimally invasive surgery.
In the current study, by 2011, only 18.9% of hospitals reported a minimally invasive surgery rate for endometrial cancer care in greater than 80%. Improvement in this quality measure will be challenging, recognizing that one of every four patients with endometrial cancer care receive surgery at low-volume cancer centers with exceptionally low minimally invasive surgery rates. However, it is encouraging that the overall utilization of minimally invasive surgery increased from 22.0% in 2007 to 50.8% in 2011. It is possible that the higher utilization of minimally invasive approaches at high-volume centers is related to the availability of a gynecologic oncologist. Variability in residency and fellowship training may also partially explain present-day underutilization of minimally invasive surgery.29,31,32 Notably, in the current study, surgeons at teaching hospitals were more likely to perform minimally invasive hysterectomy than surgeons at nonteaching facilities (OR 1.58; Table 2). Universal opportunities for minimally invasive training in gynecology residency and fellowship programs and nationwide standardization of surgical practices may help alleviate these disparities.
Mean adjusted procedural costs for laparoscopic hysterectomy and open hysterectomy were lower than robotic hysterectomy costs. However, the mean difference of $1,483 for the robotic hysterectomy cost is only an 11.8% increase in the mean cost of open hysterectomy—a relatively modest increase that may be considered a good value given the associated decrease in complications and hospital stay observed with minimally invasive approaches and that the increase in minimally invasive surgery rates during the study period was largely attributed to robotic surgery. Additionally, the costs associated with perioperative complications are likely to be higher than reported in the Nationwide Inpatient Sample,33 because data are not captured on readmissions, aftercare, follow-up visits, or on complications that occur after discharge.
This study has limitations inherent to administrative claims databases. Claims data may have incomplete coding and undercapture of select variables. However, our study results are corroborated by National Surgical Quality Improvement Program data analyzed during a similar study period12 and represent one of the largest and most contemporary studies regarding minimally invasive surgery utilization in endometrial cancer care. An additional strength is the application of sampling weights methodology. It is not merely a measure for increasing the patient number to provide data on national estimates, but its nonutilization can lead to erroneous national estimates and conclusions.16 Our data are based on current practice until 2011; however, a time lag in data cannot be avoided and is common in national databases. Because there was an observed increase in the use of minimally invasive surgical approaches during the study years of 2007–2011, it is hoped that it continued to increase thereafter. Additionally, our study does not capture hysterectomies that were performed with same-day discharge. However, the Nationwide Inpatient Sample captures all procedures with an overnight stay and, therefore, should capture the vast majority of women undergoing surgery for early-stage endometrial cancer. Also lacking are data on physician factors such as laparoscopic training and surgical experience, both of which may influence the choice of surgical procedure. Finally, selection bias is not completely accounted for and may detract from the observed difference in surgical complications.
In conclusion, we observed that most U.S. hospitals up to 2011 have low utilization of minimally invasive surgery for the treatment of nonmetastatic endometrial cancer with resultant increases in perioperative morbidity and cost. The Society of Gynecologic Oncology and American College of Surgeons' Commission on Cancer recognize the importance of minimally invasive surgery to patient safety and outcome in this setting. The American College of Obstetricians and Gynecologists may also wish to consider endorsing minimally invasive hysterectomy for the definitive treatment of several benign and malignant gynecologic conditions as part of the American Board of Internal Medicine Foundation's Choosing Wisely initiative. Utilization of minimally invasive surgery is likely a meaningful process measure in health care to complement existing and evolving outcome measures of surgical cancer care. As gynecologic cancer treatment becomes more specialized, the importance of training and strategic placement of gynecologic oncologists should be a priority to narrow the current health care disparity.
1. Mangano DT. Perioperative medicine: NHLBI working group deliberations and recommendations. J Cardiothorac Vasc Anesth 2004;18:1–6.
2. Sauerland S, Jaschinski T, Neugebauer EA. Laparoscopic versus open surgery for suspected appendicitis. The Cochrane Database of Systematic Reviews 2010, Issue 10. Art. No: CD001546. DOI: 10.1002/14651858.CD001546.pub3.
3. Clinical Outcomes of Surgical Therapy Study Group. A comparison of laparoscopically assisted and open colectomy for colon cancer. N Engl J Med 2004;350:2050–9.
4. Schwenk W, Haase O, Neudecker J, Müller JM. Short term benefits for laparoscopic colorectal resection. The Cochrane Database of Systematic Reviews 2005, Issue 3. Art. No: CD003145. DOI: 10.1002/14651858.CD003145.pub2.
5. Aimaq R, Akopian G, Kaufman HS. Surgical site infection rates in laparoscopic versus open colorectal surgery. Am Surg 2011;77:1290–4.
6. Kiran RP, El-Gazzaz GH, Vogel JD, Remzi FH. Laparoscopic approach significantly reduces surgical site infections after colorectal surgery: data from national surgical quality improvement program. J Am Coll Surg 2010;211:232–8.
7. Nieboer TE, Johnson N, Lethaby A, Tavender E, Curr E, Garry R, et al.. Surgical approach to hysterectomy for benign gynecological disease. The Cochrane Database of Systematic Reviews 2009, Issue 3. Art. No: CD003677. DOI: 10.1002/14651858.CD003677.pub4.
8. Galaal K, Bryant A, Fisher AD, Al-Khaduri M, Kew F, Lopes AD. Laparoscopy versus laparotomy for the management of early stage endometrial cancer. The Cochrane Database of Systematic Reviews 2012, Issue 9. Art. No: CD006655. DOI: 10.1002/14651858.CD006655.pub2.
9. Walker JL, Piedmonte MR, Spirtos NM, Eisenkop SM, Schlaerth JB, Mannel RS, et al.. Laparoscopy compared with laparotomy for comprehensive surgical staging of uterine cancer: Gynecologic Oncology Group Study LAP2. J Clin Oncol 2009;27:5331–6.
10. Walker JL, Piedmonte MR, Spirtos NM, Eisenkop SM, Schlaerth JB, Mannel RS, et al.. Recurrence and survival after random assignment to laparoscopy versus laparotomy for comprehensive surgical staging of uterine cancer: gynecologic Oncology Group LAP2 Study [published erratum appears in J Clin Oncol 2012;30:1570]. J Clin Oncol 2012;30:695–700.
11. Palomba S, Falbo A, Mocciaro R, Russo T, Zullo F. Laparoscopic treatment for endometrial cancer: a meta-analysis of randomized controlled trials (RCTs). Gynecol Oncol 2009;112:415–21.
12. Scalici J, Laughlin BB, Finan MA, Wang B, Rocconi RP. The trend towards minimally invasive surgery (MIS) for endometrial cancer: an ACS-NSQIP evaluation of surgical outcomes. Gynecol Oncol 2015;136:512–5.
13. Fader AN, Arriba LN, Frasure HE, von Gruenigen VE. Endometrial cancer and obesity: epidemiology, biomarkers, prevention and survivorship. Gynecol Oncol 2010;114:121–7.
14. Society of Gynecologic Oncology. Available at: https://www.sgo.org/newsroom/sgo-issues/sgo-issues-may-15-2014/
. Retrieved November 15, 2015.
15. Overview of the Nationwide Inpatient Sample (NIS). Healthcare Cost and Utilization Project (HCUP). 2013. Available at: www.hcup-us.ahrq.gov/nisoverview.jsp
. Retrieved November 15, 2015.
16. Changes in the NIS sampling and weighting strategy for 1998. Available at: https://www.hcup-us.ahrq.gov/db/nation/nis/reports/Changes_in_NIS_Design_1998.pdf
. Retrieved November 15, 2015.
17. Whyte JL, Engel-Nitz NM, Teitelbaum A, Gomez Rey G, Kallich JD. An Evaluation of algorithms for identifying metastatic breast, lung, or colorectal cancer in administrative claims data. Med Care 2015;53:e49–57.
18. Elixhauser A, Steiner C, Harris DR, Coffey RM. Comorbidity measures for use with administrative data. Med Care 1998;36:8–27.
19. Overview of the Agency for Healthcare Research and Quality patient safety indicators. 2013. Available at: www.qualityindicators.ahrq.gov/Modules/psi_resources.aspx
. Retrieved November 15, 2015.
20. Cost-to-charge ratio files: 2011 Nationwide Inpatient Sample (NIS) User Guide. Rockville (MD): Agency for Healthcare Quality and Research. Available at: https://www.hcup-us.ahrq.gov/db/state/CCR2011NISUserGuide.pdf
. Retrieved March 26, 2015.
21. Medical Expenditure Panel Survey: using appropriate price indices for analyses of health care expenditures or income across multiple years. Agency for Healthcare Research and Quality; 2013. Available at: http://meps.ahrq.gov/mepsweb/about_meps/Price_Index.shtml
. Retrieved March 26, 2015.
22. Makary MA. A minimally-invasive approach to healthcare reform. Wall Street J 2014.
23. Xu T, Hutfless SM, Cooper MA, Zhou M, Massie AB, Makary MA. Hospital cost implications of increased use of minimally invasive surgery. JAMA Surg 2014;150:489–90.
24. Wysham WZ, Kim KH, Roberts JM, Sullivan SA, Campbell SB, Roque DR, et al.. Obesity and perioperative pulmonary complications in robotic gynecologic surgery. Am J Obstet Gynecol 2015;213:33.e1–7.
25. Chan JK, Gardner AB, Taylor K, Blansit K, Thompson CA, Brooks R, et al.. The centralization of robotic surgery in high-volume centers for endometrial cancer patients—a study of 6560 cases in the U.S. Gynecol Oncol 2015;138:128–32.
26. Chan JK, Sherman AE, Kapp DS, Zhang R, Osann KE, Maxwell L, et al., Influence of gynecologic oncologists on the survival of patients with endometrial cancer. J Clin Oncol 2011;29:832–8.
27. Mahdi H, Lockhart D, Moslemi-Kebria M, Rose PG. Racial disparity in the 30-day morbidity and mortality after surgery for endometrial cancer. Gynecol Oncol 2014;134:510–5.
28. Conrad LB, Ramirez PT, Burke W, Naumann RW, Ring KL, Munsell MF, et al.. Role of Minimally Invasive Surgery in Gynecologic Oncology: An Updated Survey of Members of the Society of Gynecologic Oncology. Int J Gynecol Cancer 2015;25:1121–7.
29. Kroft J, Li Q, Saskin R, Elit L, Bernardini MQ, Gien LT. Trends over time in the use of laparoscopic hysterectomy for the treatment of endometrial cancer. Gynecol Oncol 2015;138:536–41.
30. Cooper MA, Hufless S, Segev DL, Ibrahim A, Lyu H, Makary MA. Hospital level under-utilization of minimally invasive surgery in the United States: retrospective review. BMJ 2014;349:g4198.
31. Stovall DW, Fernandez AS, Cohen SA. Laparoscopy training in United States obstetric and gynecology residency programs. JSLS 2006;10:11–5.
32. Sfakianos GP, Frederick PJ, Kendrick JE, Straughn JM, Kilgore LC, Huh WK. Robotic surgery in gynecologic oncology fellowship programs in the USA: a survey of fellows and fellowship directors. Int J Med Robot 2014;6:405–12.
33. Dimick JB, Chen SL, Taheri PA, Henderson WG, Khuri SF, Campbell DA Jr. Hospital costs associated with surgical complications: a report from the private-sector National Surgical Quality Improvement Program. J Am Coll Surg 2004;199:531–7.