A nationwide evaluation of robotic ventral hernia surgery

Accepted Manuscript A nationwide evaluation of robotic ventral hernia surgery Kathleen M. Coakley, Stephanie M. Sims, Tanushree Prasad, Amy E. Lincourt, Vedra A. Augenstein, Ronald F. Sing, B. Todd Heniford, Paul D. Colavita PII:

S0002-9610(17)30427-0

DOI:

10.1016/j.amjsurg.2017.08.022

Reference:

AJS 12491

To appear in:

The American Journal of Surgery

Received Date: 11 March 2017 Revised Date:

31 July 2017

Accepted Date: 5 August 2017

Please cite this article as: Coakley KM, Sims SM, Prasad T, Lincourt AE, Augenstein VA, Sing RF, Heniford BT, Colavita PD, A nationwide evaluation of robotic ventral hernia surgery, The American Journal of Surgery (2017), doi: 10.1016/j.amjsurg.2017.08.022. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

Background: The purpose of this study was to examine outcomes of robotic ventral hernia repair(RVHR) versus laparoscopic ventral hernia repair(LVHR). Methods:

RI PT

The Nationwide Inpatient Sample was queried from October 2008 to December 2013 for ventral hernia repairs. Demographics, morbidity, mortality, and charges were compared between RVHR and LVHR. Results:

M AN U

SC

From 2008-2013, 149,622 ventral hernia surgeries were identified; 117,028 open, 32,243 laparoscopic, and 351 robotic. Open repairs were excluded. RVHR rose annually with 2013 containing 47.9% of all RVHRs. RVHR patients were more likely to be older and have more chronic conditions. There was no difference between length of stay. Pneumonia rates were higher with RVHR; however, after controlling for confounding variables, there was no difference in pneumonia rates. Mortality and other major complications were similar. Total charges were increased for RVHR in univariate and multivariate analysis. RVHR was more common in teaching hospitals and wealthier zip codes. Conclusion

AC C

EP

TE D

RVHR demonstrates comparable safety to the laparoscopic technique, with increased charges and increased volume in urban teaching hospitals and patients from areas of higher median income.

ACCEPTED MANUSCRIPT

RI PT

A Nationwide Evaluation of Robotic Ventral Hernia Surgery

M AN U

SC

Kathleen M. Coakley DO, Stephanie M. Sims MD, Tanushree Prasad MA, Amy E. Lincourt PhD MBA, Vedra A. Augenstein MD, Ronald F. Sing DO, B. Todd Heniford MD, Paul D. Colavita MD

Carolinas Medical Center

Division of Gastrointestinal and Minimally Invasive Surgery, Department of Surgery, Carolinas Medical Center, Charlotte, NC

TE D

Disclosures: Drs. Heniford and Augenstein have previously been awarded surgical research and education grants from W.L. Gore, Ethicon, Bard, and LifeCell Inc. Dr. Colavita has received honoraria for speaking from W.L. Gore. All other authors confirm they have no financial and personal relationships that could potentially and inappropriately influence this work or its conclusions.

AC C

EP

Accepted for oral presentation at the 69th annual Southwestern Surgical Congress, Maui, Hi April 1-5, 2017

Corresponding author: Paul D. Colavita, MD Carolinas Medical Center 1025 Morehead Medical Drive, Suite 300 Charlotte, NC 28204 Office: (704) 355-3168 Fax: (704) 355-4117 [email protected]

Keywords: Ventral Hernia, Robotics Trends, National Inpatient Sample (NIS)

ACCEPTED MANUSCRIPT

Abstract Background: The purpose of this study was to examine outcomes of robotic ventral hernia repair(RVHR) versus laparoscopic ventral hernia repair(LVHR).

RI PT

Methods:

The Nationwide Inpatient Sample was queried from October 2008 to December 2013 for ventral hernia repairs. Demographics, morbidity, mortality, and charges were compared between RVHR and LVHR.

SC

Results:

M AN U

From 2008-2013, 149,622 ventral hernia surgeries were identified; 117,028 open, 32,243 laparoscopic, and 351 robotic. Open repairs were excluded. RVHR rose annually with 2013 containing 47.9% of all RVHRs. RVHR patients were more likely to be older and have more chronic conditions. There was no difference between length of stay. Pneumonia rates were higher with RVHR; however, after controlling for confounding variables, there was no difference in pneumonia rates. Mortality and other major complications were similar. Total charges were increased for RVHR in univariate and multivariate analysis. RVHR was more common in teaching hospitals and wealthier zip codes. Conclusion

AC C

EP

TE D

RVHR demonstrates comparable safety to the laparoscopic technique, with increased charges and increased volume in urban teaching hospitals and patients from areas of higher median income.

2

ACCEPTED MANUSCRIPT

Introduction In the United States, an estimated 350,000 ventral hernia repairs (VHR) are performed annually [1]. Laparoscopic ventral hernia repair (LVHR) was first described in 1993[2], and has been

RI PT

shown to be both safe and effective [3]. When compared to open VHR, LVHR has been shown to be associated with shorter length of hospital stay (LOS) and lower infection rates [4-8]. Like many intraabdominal surgical operations, ventral hernia has had a high rate of laparoscopic

SC

adoption [9], but the adoption of laparoscopy for ventral hernia peaked at around 25 % of all VHR [10-12]. Robotic-assisted laparoscopic surgery introduced advantages compared to

M AN U

standard laparoscopy, including three-dimensional imaging, superior ergonomics, and precise control that facilitate suturing and mesh placement in planes that exploit the layers of the abdominal wall [13]. The utilization of robotic-assisted surgery for hernia surgery remains unclear. The purpose of this study is to compare outcomes, utilization, and charges between

TE D

robotic ventral hernia (RVHR) and LVHR in a large national database, containing a representative sample of the U.S. population. Materials and Methods

EP

Nationwide Inpatient Sample

The Healthcare Cost and Utilization Project’s (HCUP) National Inpatient Sample (NIS) database

AC C

was queried. The NIS is the largest all-payer inpatient database in the United States, containing data on more than seven million hospital stays [14]. The NIS does not contain patient identifying information, and IRB approval was not required for this study. Admissions reported are from a 20 % sample of non-federal hospitals. Hospitals were classified by HCUP variables including region (Northeast, Midwest, South, West), hospital ownership (non-federal government, private non-profit, and investor-owned private), hospital characteristics (bed size, urban, region of country, and teaching status), and location (rural and urban)[14]. Each inpatient record in the 3

ACCEPTED MANUSCRIPT

NIS represents a single hospital admission and includes demographics, admission type, up to 15 individual primary and secondary diagnoses and procedures for the hospitalization (based on ICD-9, Clinical Modification [ICD-9-CM] and Procedure Related [ICD-9-PR] coding), total

RI PT

hospital charges, LOS, payer status, in patient mortality, and patient’s zip code median income. In this study, only actual patient data within the database were used and the NIS-implemented

SC

weighting strategy was not applied.

ICD-9 codes were used to define postoperative complications. These codes were classified into

M AN U

minor and major complications, similar to previously published work [15], and defined in Table 1.

To measure the burden of comorbid disease between RVHR and LVHR, ICD-9 CM coding was used to calculate the Charlson comorbidity index (CCI) for each patient. The CCI was initially designed to calculate mortality risk of 22 weighted comorbidities [16]. Using administrative data,

TE D

the CCI has been shown to be a better predictor than individual comorbidities for mortality [17]. Several algorithms have been developed to extract the CCI using ICD-9 CM coding [16-18]. The

EP

method of Quan et al. [18] was used in the present study. In addition to CCI, the number of comorbidities was also reported for each patient, a total figure of each contributing disease to the

AC C

CCI.

Inclusion and Exclusion criteria LVHRs and RVHRs were identified from October 2008 to December 2013 (Table 2). ICD-9 CM codes for robotic procedures were introduced in October 2008 [19]. To determine the total number of patients undergoing elective LVHRs and RVHRs, all patients with an elective admission and procedure codes from the ICD-9 CM for ventral hernia were included (Table 2). To reduce the risk of patients having robotic coding associated with other common robotic 4

ACCEPTED MANUSCRIPT

procedures, all patients with any ICD-9-PR codes for cardiothoracic, urologic, obstetric and gynecologic, or orthopedic operations during their admission were excluded. After isolating the population of adult patients undergoing elective ventral hernia during their

RI PT

hospitalization, all patients undergoing RVHR were determined based on the presence of ICD-9PR coding for robotic-assisted procedures. Patients undergoing LVHR were compared with RVHR for differences in demographics, outcomes, charges, and region. Patients younger than

SC

age 18 years and those who also underwent resection of any portion of the gastrointestinal tract were excluded.

M AN U

Statistical Analysis

LVHRs and RVHRs were compared for primary outcomes (total hospital charges, postoperative major and minor complications, and inpatient mortality) in univariate analysis. The incidence of pre-operative variable, as well as unadjusted outcomes, were compared using the Wilcoxon-

TE D

Mann-Whitney test for continuous and ordinal variables and Pearson’s chi-square and Fisher’s exact tests for categorical variables as appropriate. Frequencies of categorical variables are expressed as a percentage of the group of origin, and continuous variables are reported as means

EP

± standard deviation.

AC C

Multivariate linear regression models were created to estimate the rate of post-operative pneumonia and adjusted total charges between RVHR and LVHR when controlling for confounding variables. Multivariate analyses were performed controlling for age and CCI when evaluating post-operative pneumonia rates between RVHR and LVHR. Age, CCI, as well as hospital geographic region, public versus private hospital type, urban versus rural teaching status, and patient’s zip code median income were included in the multivariate model to evaluate the impact of robotic technique on total charges. All tests and 95 % confidence

5

ACCEPTED MANUSCRIPT

intervals were two-sided with a p ≤ 0.05 considered statistically significant. All statistical analyses were conducted using Statistical Analysis Software, version 9.4 (SAS Institute, Inc.,

RI PT

Cary, NC).

Results

From 2008-2013, 149,622 ventral hernia surgeries were identified; 117,028 open VHR, 32,243 LVHR and 351 RVHR. Open VHR patients were excluded from further analysis. Throughout

M AN U

study period performed in 2013 (Figure 1).

SC

the study timeframe, the number of RVHR rose with almost half (47.9%) of all RVHRs for the

Demographics and comorbidities for RVHR versus LVHR can be found in Table 3. RVHR patients were more likely to be older (59.4 ±14.6 years vs 57.4±14.9 years; p=0.01) and less likely to be female (52% vs 57%; p=0.04). CCI scores were comparable in RVHR and

TE D

LVHR (1.1±1.7 vs 0.83±1.3; p= 0.074), but RVHR patients had more comorbidities (3.8±2.7 vs 3.4±2.6; p=0.007). No difference in race or individual comorbidities was found between RVHR and LVHR.

EP

Hospital characteristics, regional details, and primary payer comparisons between RVHR and LVHR are demonstrated in Table 4. When comparing RVHR and LVHR, there was no

AC C

difference in hospital bed size distribution (p=0.092), with the majority of RVHR (62.5%) and LVHR (61.5%) being performed in large bed (>450) hospitals. When examining teaching status, 64.63% of RVHRs were performed in teaching hospitals, whereas only 47.1% of LVHR were conducted in teaching hospitals (p= 0.001). 98.8% of RVHR were performed in an urban location, compared to 89.1% of LVHR (p=0.005). With respect to hospital ownership, 19.5% of RVHR were performed in privately owned hospitals versus 12.9% of LVHR (p=0.001).

6

ACCEPTED MANUSCRIPT

A similar geographic pattern was seen in both RVHR and LVHR, with the highest proportion of RVHR (41.1%) and LVHR (38.1%) being performed in the South. Examining payer status differences between RVHR and LVHR, Medicaid insurance was less common in

RI PT

patients undergoing RVHR (6.3% vs 9.6%; p=0.038). There was no significant difference in rates of private insurance between RVHR and LVHR (48.0% vs 43.3%; p= 0.079).

When comparing RVHR and LVHR, the distribution amongst patient’s median income,

SC

as determined by home zip code, is contained in Table 4. RVHR occurred most frequently in patients from zip codes with median income greater than $64,000 (28.6 %) and least frequently

M AN U

in patients from the lowest income zip codes (median income less than $37,999; 22.6%). These differences were not statistically significant (global p = 0.09). However, 28.6% of RVHR took place in patients from zip codes with median income greater than $64,000, compared to 23.0% for LVHR (p = 0.013).

TE D

Table 5 demonstrates hospitalization outcome differences between surgical techniques. When comparing RVHR versus LVHR, there was no difference between LOS (3.5±3.6 vs 3.4±2.6, p=0.2). Routine discharge to home rates were similar between RVHR (88.3%) and

EP

LVHR (90.8%) (p=0.117). There was no difference in overall minor complication rates (10.5% vs 10.8%; p=0.858), but post-operative pneumonia rates were higher in the RVHR group (4.3%

AC C

vs 2.4%; p=0.02). However, after controlling for age and CCI there was no difference seen in pneumonia rates between RVHR and LVHR (Odds Ratio 1.587; 95% Confidence Interval 0.9352.692). There was no difference in mortality or major complications: acute CHF, acute stroke, acute renal failure, MI, wound dehiscence, PE, ARDS, and postoperative shock (all p>0.05). Regarding hospital charges, RVHR cases were found to have increased total charges compared to LVHR ($61,274 ± $42,666 vs $38,715 ± $28,533; p<0.0001).

7

ACCEPTED MANUSCRIPT

Using multivariate regression (Table 5), when controlling for age, CCI, hospital geographic region, public versus private hospital type, urban versus rural, teaching status, and zip code

coefficient $19,294, Standard Error $1431, p<0.0001. Discussion

RI PT

median income, robotic repair remained an independent predictor of increased charges (beta

SC

While open VHR remains the most common technique for ventral hernia repair, with 117,028 of the 149,622 hernias for the study period performed in an open manner, minimally

M AN U

invasive techniques account for over 20% of ventral hernia repairs and, particularly robotic techniques, are increasing in popularity (Figure 1). Potential benefits have been proposed for the use of robotic assistance when repairing ventral hernias. Robotic technology addresses the limitations of laparoscopy, improving dexterity and ergonomics[20], leading to growth in the use

TE D

of robotics in general surgery [11]. Schluender et al. first described RVHR in a porcine model using circumferential suture fixation of mesh [21], revealing a relative ease in intracorporeal suturing compared to standard laparoscopy. Where LVHR typically involves mesh overlap

EP

without closure of the fascial defect, which may lead to eventration of the mesh through the defect [22], RVHR can facilitate intracorporeal closure through the dexterity of robotic

AC C

instruments [23]. Ease of fascial closure is of interest since some publications have shown defects in the abdominal wall alter compliance, where defect closure reduces local stress, allowing for better distribution of tension across the repaired abdominal wall [13, 24, 25], leading in turn to decreased recurrence [26]. Gonzalez et al, compared RVHR with closure of fascial defect to LVHR without fascial defect closure and showed decreased recurrence rates and complications in the RVHR group [23]. Robotic proponents purport the ability to perform true abdominal wall reconstruction with myofascial release to offset tension of midline fascial 8

ACCEPTED MANUSCRIPT

closure, while keeping SSI rates low [27]. Additionally, the tremorless precision of the robotic technology allows ease in performing more challenging repairs, such as pre-peritoneal mesh placement techniques, allowing isolation of mesh from the abdominal viscera, in turn decreasing

RI PT

the risk of mesh related complications, such as enteral adhesions or mesh fistula. Despite the proposed advantages of RVHR, and while this study does shows increasing incidence, RVHR accounts for only 1% of all minimally invasive ventral hernias for the study period (Figure 1).

SC

Despite the excitement for the potential of robotics in VHR, the impact on clinical

M AN U

outcomes have yet to be well demonstrated [28]. The lack of strong clinical evidence and the high associated expense of the technology may represent a limiting factor in the adoption of RVHR. There have been some case series demonstrating the application of robotic technology to VHR is safe and feasible [13, 29]. Gonzalez et al. [29] presented 368 patients who underwent a RVHR with total complication rate of 8.4%, and the mean operative time was 102 min. Only a

TE D

few small retrospective analyses comparing LVHR and RVHR have been published [27, 30], with Chen et. al. examining small ventral hernias only and seeing no significant difference in outcomes. Warren et al. compared 103 LVHR with 53 RVHR, utilizing a retro muscular mesh

EP

placement in the RVHR arm [27]. Warren achieved fascial closure more often with RVHR (96.2

AC C

vs. 50.5 %; p < 0.001), aided by myofascial release in 43.4 % of cases. The average defect size was small in this study, but still larger at 3.07 cm for the robot group than 2.02cm in the laparoscopic group (p < 0.0001). Mesh was placed in an intraperitoneal position in 90.3 % of LVHR and extraperitoneal in 96.2 % of RVHR. These studies are promising but limited in design, outcome, and size. There is a great need for controlled, unbiased clinical trials, with long term follow up. Evaluating the results of experienced RVHR surgeons, may help address the cost-effectiveness between techniques.

9

ACCEPTED MANUSCRIPT

This analysis adds to the growing body of literature finding RVHR equally effective when compared to LVHR. Both groups had similar in-hospital complications, LOS, disposition, and mortality. There was higher rate of postoperative pneumonia in the RVHR group in

RI PT

univariate analysis, with no effect demonstrated after controlling for confounding variables. Total charges were higher in the RVHR group. While operative time is not captured in NIS, this is certainly a potential factor impacting the increased charges. Increased operative time is seen

SC

in robotic cases in general [31, 32]. This has also been demonstrated when comparing LVHR

RVHR compared to LVHR.

M AN U

and RVHR. Chen et al [30] and Warren et al [27] both demonstrated increased operative time in

The low adoption rate of robotics in VHR demonstrated in the present study may be associated with the increased expense. When new modes of treatment are embraced, particularly before weaknesses, equality, or superiority are fully understood, there is a concern for needless

TE D

increased expense [33]. The results of this NIS analysis show significantly higher hospital charges associated with the use of RVHR when controlling for both patient- and hospital-specific factors. This study, and others [34-36] utilizing the NIS, have demonstrated the increased total

EP

charges of robotic techniques compared to laparoscopic counterparts. The NIS dataset does not

AC C

supply detailed contribution to the increased charges, however the robotic platform alone has an initial acquisition cost of over two million dollars and an additional $100,000 in yearly maintenance fees[37]. Barbash examined the cost studies of robot-assisted procedures and found that, on average, across a full range of 20 types of surgical procedures, the additional cost of using a robot was about $1600. When the cost of the robot itself was included, the additional cost rose to $3200 [33]. A review of 11 studies performed on robotic cost-analysis demonstrated that robotic was significantly more expensive if the purchase and maintenance costs of the robot

10

ACCEPTED MANUSCRIPT

system are included [32]. The charges for robotic equipment in the NIS is unknown, and cost data is unavailable in the database, limiting the analysis of cost and charges in the current study.

RI PT

Only comparisons of charges between laparoscopic and robotic surgeries can be performed. Some propose robotic prices may decline when competition enters the market [38, 39]. A recent 14-year, single center, retrospective study by Hagen et al [40] demonstrated decreased cost for robotic gastric bypass over laparoscopic when balancing the greater overhead costs of

SC

the robot with savings associated with avoiding stapler use and anastomotic complications. In

M AN U

fact, Warren et al, showed direct hospital cost was similar with LVHR compared with RVHR [27]. However, a series of 110 robot-assisted hand-sewn gastrojejunal anastomoses demonstrated significantly longer average OR time and higher cost than a matched group undergoing laparoscopic-only, without significant differences in complications [41]. As the need

TE D

for healthcare cost containment becomes paramount, comparative trials must be undertaken. The use of health care resources in the United States is highly localized, with access to advancing technologies and variation in utilization and expense of care being well proven

EP

between hospitals and regions [42-46]. Across the US, rates of hip replacement, coronary artery bypass surgery, prostatectomy, and many other major procedures vary up to fivefold across

AC C

hospital referral region[47]. Not surprisingly, this analysis confirmed a significant amount of geographic impact associated with RVHR, with most procedures being performed at urban teaching hospitals, in the South, and in areas of higher median income. This is similar to the findings in robotic-assisted colorectal and urologic surgery [35, 48] and may be impacted by access to surgeons with robotic expertise or hospital systems with robotic technology. The main limitations of this study are inherent in the NIS database, as the in-hospital admission data lacks comprehensive clinical information, as well as excludes outpatient ventral 11

ACCEPTED MANUSCRIPT

hernias. The dataset does not afford follow-up of the patient outcomes following discharge, quality of life, hernia recurrence, and out of hospital mortality. Operative time and conversion rates are unavailable in the NIS, as well as hernia specific details such as defect size, mesh, or

RI PT

operative technique details, making it impossible to evaluate technical differences between the procedures. The charge data evaluates overall hospital charges and not cost. Hospital charges are traditionally significantly higher than cost, and cost to charge ratios vary between hospital

SC

system. The contribution of initial capital investment of the robotic system and maintenance are not elucidated in the total hospital charges variable supplied in the NIS dataset. Finally, the

M AN U

small sample of RVHR in this study introduces some biases, as the RVHR evaluated may represent a subpopulation of specialized surgeons compared with laparoscopy. Conclusion The application of RVHR in the US is low but increasing. When comparing RVHR and LVHR, no major differences were found in perioperative outcomes, but increased charges and

TE D

regionalization to urban teaching hospitals and in patients from areas of higher median income were seen for RVHR. The results of this nationwide evaluation of RVHR substantiate its safe

EP

role in the treatment of ventral hernia, but continue to raise concerns about its higher expense compared with conventional laparoscopy.

1. 2.

3. 4. 5.

AC C

References

Poulose, B.K., et al., Epidemiology and cost of ventral hernia repair: making the case for hernia research. Hernia, 2012. 16(2): p. 179-83. LeBlanc, K.A. and W.V. Booth, Laparoscopic repair of incisional abdominal hernias using expanded polytetrafluoroethylene: preliminary findings. Surg Laparosc Endosc, 1993. 3(1): p. 3941. Heniford, B.T., et al., Laparoscopic repair of ventral hernias: nine years' experience with 850 consecutive hernias. Ann Surg, 2003. 238(3): p. 391-9; discussion 399-400. Colavita, P.D., et al., Prospective, long-term comparison of quality of life in laparoscopic versus open ventral hernia repair. Ann Surg, 2012. 256(5): p. 714-22; discussion 722-3. Rosen, M.J., et al., Laparoscopic versus open-component separation: a comparative analysis in a porcine model. Am J Surg, 2007. 194(3): p. 385-9. 12

ACCEPTED MANUSCRIPT

12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23.

24. 25. 26. 27. 28. 29.

RI PT

11.

SC

10.

M AN U

9.

TE D

8.

EP

7.

Asencio, F., et al., Open randomized clinical trial of laparoscopic versus open incisional hernia repair. Surg Endosc, 2009. 23(7): p. 1441-8. Pring, C.M., et al., Laparoscopic versus open ventral hernia repair: a randomized controlled trial. ANZ J Surg, 2008. 78(10): p. 903-6. Itani, K.M., et al., Comparison of laparoscopic and open repair with mesh for the treatment of ventral incisional hernia: a randomized trial. Arch Surg, 2010. 145(4): p. 322-8; discussion 328. Nguyen, N.T., et al., Use of laparoscopy in general surgical operations at academic centers. Surg Obes Relat Dis, 2013. 9(1): p. 15-20. Colavita, P.D., et al., Laparoscopic versus open hernia repair: outcomes and sociodemographic utilization results from the nationwide inpatient sample. Surg Endosc, 2013. 27(1): p. 109-17. Tsui, C., R. Klein, and M. Garabrant, Minimally invasive surgery: national trends in adoption and future directions for hospital strategy. Surg Endosc, 2013. 27(7): p. 2253-7. Funk, L.M., et al., Current national practice patterns for inpatient management of ventral abdominal wall hernia in the United States. Surg Endosc, 2013. 27(11): p. 4104-12. Kudsi, O.Y., et al., Robotic Repair of Ventral Hernias: Preliminary Findings of a Case Series of 106 Consecutive Cases. Am J Robot Surg, 2015. 2(1): p. 22-26. Healthcare Cost and Utilization Project HCUP in National Inpatient Sample (NIS)2013, Agency for Healthcare Research and Quality: Rockville, MD. Ross, S.W., et al., National outcomes of laparoscopic Heller myotomy: operative complications and risk factors for adverse events. Surg Endosc, 2015. 29(11): p. 3097-105. Charlson, M.E., et al., A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis, 1987. 40(5): p. 373-83. Quan, H., G.A. Parsons, and W.A. Ghali, Validity of information on comorbidity derived rom ICD9-CCM administrative data. Med Care, 2002. 40(8): p. 675-85. Quan, H., et al., Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data. Med Care, 2005. 43(11): p. 1130-9. Association, A.H., AHA coding clinic for ICD-9-CM, (2008): Chicago. Mack, M.J., Minimally invasive and robotic surgery. JAMA, 2001. 285(5): p. 568-72. Schluender, S., et al., Robot-assisted laparoscopic repair of ventral hernia with intracorporeal suturing. Surg Endosc, 2003. 17(9): p. 1391-5. Carter, S.A., et al., Recurrence and pseudorecurrence after laparoscopic ventral hernia repair: predictors and patient-focused outcomes. Am Surg, 2014. 80(2): p. 138-48. Gonzalez, A.M., et al., Laparoscopic ventral hernia repair with primary closure versus no primary closure of the defect: potential benefits of the robotic technology. Int J Med Robot, 2015. 11(2): p. 120-5. Rea, R., et al., Laparocopic ventral hernia repair with primary transparietal closure of the hernial defect. BMC Surg, 2012. 12 Suppl 1: p. S33. Tandon, A., et al., Meta-analysis of closure of the fascial defect during laparoscopic incisional and ventral hernia repair. Br J Surg, 2016. 103(12): p. 1598-1607. Chelala, E., et al., Long-term outcomes of 1326 laparoscopic incisional and ventral hernia repair with the routine suturing concept: a single institution experience. Hernia, 2016. 20(1): p. 101-10. Warren, J.A., et al., Standard laparoscopic versus robotic retromuscular ventral hernia repair. Surg Endosc, 2017. 31(1): p. 324-332. Szold, A., et al., European Association of Endoscopic Surgeons (EAES) consensus statement on the use of robotics in general surgery. Surg Endosc, 2015. 29(2): p. 253-88. Gonzalez, A., et al., Robotic-assisted ventral hernia repair: a multicenter evaluation of clinical outcomes. Surg Endosc, 2017. 31(3): p. 1342-1349.

AC C

6.

13

ACCEPTED MANUSCRIPT

36. 37. 38. 39. 40. 41. 42.

43. 44.

45.

46. 47. 48.

RI PT

35.

SC

34.

M AN U

33.

TE D

32.

EP

31.

Chen, Y.J., et al., Outcomes of robot-assisted versus laparoscopic repair of small-sized ventral hernias. Surg Endosc, 2017. 31(3): p. 1275-1279. Spinoglio, G., et al., Robotic colorectal surgery: first 50 cases experience. Dis Colon Rectum, 2008. 51(11): p. 1627-32. Turchetti, G., et al., Economic evaluation of da Vinci-assisted robotic surgery: a systematic review. Surg Endosc, 2012. 26(3): p. 598-606. Barbash, G.I. and S.A. Glied, New technology and health care costs--the case of robot-assisted surgery. N Engl J Med, 2010. 363(8): p. 701-4. Anderson, J.E., et al., The first national examination of outcomes and trends in robotic surgery in the United States. J Am Coll Surg, 2012. 215(1): p. 107-14; discussion 114-6. Yu, H.Y., et al., Use, costs and comparative effectiveness of robotic assisted, laparoscopic and open urological surgery. J Urol, 2012. 187(4): p. 1392-8. Yu, H.Y., et al., Hospital volume, utilization, costs and outcomes of robot-assisted laparoscopic radical prostatectomy. J Urol, 2012. 187(5): p. 1632-7. Higgins, R.M., et al., Cost analysis of robotic versus laparoscopic general surgery procedures. Surg Endosc, 2017. 31(1): p. 185-192. Hottenrott, C., Robotic versus laparoscopic surgery for rectal cancer and cost-effectiveness analysis. Surg Endosc, 2011. 25(12): p. 3954-6; author reply 3957-8. Wilson, E.B., The evolution of robotic general surgery. Scand J Surg, 2009. 98(2): p. 125-9. Hagen, M.E., et al., Reducing cost of surgery by avoiding complications: the model of robotic Roux-en-Y gastric bypass. Obes Surg, 2012. 22(1): p. 52-61. Moon, R.C., et al., Robotic Roux-en-Y Gastric Bypass, is it Safer than Laparoscopic Bypass? Obes Surg, 2016. 26(5): p. 1016-20. Aboagye, J.K., H.E. Kaiser, and A.J. Hayanga, Rural-Urban Differences in Access to Specialist Providers of Colorectal Cancer Care in the United States: A Physician Workforce Issue. JAMA Surg, 2014. 149(6): p. 537-43. Alnasser, M., et al., National disparities in laparoscopic colorectal procedures for colon cancer. Surg Endosc, 2014. 28(1): p. 49-57. Ancona, C., et al., Coronary artery bypass graft surgery: socioeconomic inequalities in access and in 30 day mortality. A population-based study in Rome, Italy. J Epidemiol Community Health, 2000. 54(12): p. 930-5. Carney, M.J., et al., Trends in open abdominal surgery in the United States-Observations from 9,950,759 discharges using the 2009-2013 National Inpatient Sample (NIS) datasets. Am J Surg, 2017. Huntington, C.R., et al., Nationwide variation in outcomes and cost of laparoscopic procedures. Surg Endosc, 2016. 30(3): p. 934-46. The Dartmouth Atlas. http://www.dartmouthatlas.org/data/region/, 2014. Halabi, W.J., et al., Robotic-assisted colorectal surgery in the United States: a nationwide analysis of trends and outcomes. World J Surg, 2013. 37(12): p. 2782-90.

AC C

30.

14

ACCEPTED MANUSCRIPT

Tables and Figures Table 1: Classification of Postoperative Inpatient Morbidity by ICD-9 Billing Codes Cellulitis

486.0, 507.0, 997.31, 997.32, 997.39

Paralytic ileus DVT Postoperative infection Postoperative infected seroma Nonhealing surgical wound Wound dehiscence (Minor)

560.1 453.40, 453.41, 453.42, 453.83 998.59, 998.9 998.51

M AN U

998.83 998.32

SC

Pneumonia

RI PT

Minor Complications 682.9

Major Complications Acute congestive heart failure

410.9 411.89 4280 428.1 997.1 434.11, 430.0, 431.0, 432.0, 432.1, 432.9, 434.01, 434.91 584.9, 997.5

Acute stroke

Myocardial infarction

410.00-410.02, 410.10-410.12, 410.20410.22, 410.30-410.32, 410.40-410.42, 410.50-410.52, 410.60-410.62, 410.70410.72, 410.80-410.82, 410.90-410.92

TE D

Acute renal failure

998.31, 998.32 415.1, 415.11, 415.12, 415.19

Pulmonary insufficiency following trauma or surgery, including ARDS

518.5

Postoperative shock

998

AC C

EP

Wound dehiscence (Major) Pulmonary embolism

DVT: Deep venous thromboembolism. ARDS: Adult Respiratory Distress Syndrome.

15

ACCEPTED MANUSCRIPT

Table 2: Inclusion, Exclusion Criteria and ICD-9 CM Codes Inclusion Criteria

ICD-9 CM Diagnosis and/or Procedure Codes

Oct. 2008 – Dec. 2013 Elective admissions Robotic ICD9 PR codes Ventral Hernia Repair Laparoscopic VHR

RI PT

17.41, 17.42, 17.43, 17.44, 17.45, 17.49 53.41, 53.49, 53.51, 53.59, 53.61, 53.69 53.42, 53.43, 53.62, 53.63

Exclusion Criteria Age < 18

AC C

EP

TE D

M AN U

cardiothoracic, urologic, gynecologic, or orthopedic

42.4x, 43.5x–43.9x, 44.3x, 45.5x–45.8x, 48.5x–48.6x, 50.3x–50.5x, 52.5x, 52.6, 52.7, 52.8x 30–39.x, 55–64.x, 65–75.x, 76–84.x

SC

Concomitant GI tract resections

16

ACCEPTED MANUSCRIPT

Table 3: Laparoscopic versus Robotic Ventral Hernias: Demographics and Comorbidities

75.3 10.5 10.1 4.1

73.0 11.3 9.5 6.2

RI PT

59.4 ±14.6 52%

p value

3.4±2.6 3.7 17.8 25.3 3.8 20.3 1.9 4.1 0.1 0.83±1.3

0.019 0.041 0.371

SC

57.4±14.9 57%

TE D

Comorbidity (%) Number of Chronic Diseases Congestive heart failure Chronic pulmonary disease Obese Mild liver disease Diabetes without complications Diabetes with complications Renal disease AIDS/HIV Charlson Comorbidity Index

RVH (n = 351)

M AN U

Demographic (%) Age (yrs) Female Race Caucasian African American Hispanic Other

LVH (n = 32243)

3.8±2.7 4.3 17.1 20.5 3.4 17.4 0.6 5.7 0.0 1.1±1.7

0.006 0.567 0.728 0.041 0.736 0.178 0.074 0.130 0.640 0.074

AC C

EP

AIDS/HIV - Acquired immunodeficiency syndrome/Human immunodeficiency virus

17

ACCEPTED MANUSCRIPT

LVH (n = 32243)

RVH (n = 351)

RI PT

Table 4: Laparoscopic versus Robotic Ventral Hernia: Hospital Characteristics and Regionalization

p value

9.2 28.4 62.5

Teaching status (%)

47.1

64.6

Urban (%)

89.1

98.8

Region (%) Northeast Midwest South West

20.8 23.4 38.1 17.7

TE D

Primary payer (%) Medicare Medicaid Private, including HMO Self-pay Other

EP

38.3 9.6 43.3 4.5 3.6

AC C

Median houshold income (%) † $1 - $37,999 25.4 $38,000 - $47,999 25.6 $48,000 - $63,999 26.0 $64,000 and above 23.0

0.001 0.005

M AN U

12.8 25.6 61.5

SC

0.092

Hospital bedsize (%) Small Medium Large

17.4 18.5 45.0 19.1

0.017 0.119 0.0323 0.0084 0.502

41.1 6.3 48.0 1.4 2.6

0.011 0.280 0.038 0.079 0.005 0.295

22.6 25.1 23.7 28.6

0.089 0.226 0.860 0.324 0.013

† Median household income is based on patient zip code HMO - Health maintenance organization

18

ACCEPTED MANUSCRIPT

Table 5: Laparoscopic versus Robotic: Outcomes and Cost RVH (n = 351)

p value

3.4 ± 2.6

3.5 ± 3.6

0.211

Disposition (%) Routine

90.8

88.3

Mortality (%)

0.1

0

Major complication Minor Complications Wound (mechanical) Infection MI PE Renal Failure Pneumonia CHF Paralytic Ileus DVT

7.93 10.8 0.07 0.47 0.2 0.16 3.44 2.4 4.14 8.34 0.11

9.7 10.54 0.28 0.85 0.28 0.57 3.99 4.27 5.41 5.41 0.28

0.226 0.858 0.128 0.234 0.729 0.057 0.611 0.026 0.234 0.048 0.307

$61,274.47 ± 42666

<0.0001

TE D

$38,715.00 ± 28533

Mean total charges ($) Standard deviation

0.117

0.478

SC

M AN U

Length of stay (days)

RI PT

LVH (n = 32243)

¶

AC C

EP

$41,911.00 $61,205.00 <0.0001 Adjusted mean charges ($) Standard Error ± 258 ± 1431 ¶ controlling for age, Charlson Comorbidity Index (CCI), hospital geographic region, public versus private hospital type, urban versus rural teaching status, and zip code

19

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

Figure 1

20

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

Disclosures: Drs. Heniford and Augenstein have previously been awarded surgical research and education grants from W.L. Gore, Ethicon, Bard, and LifeCell Inc. Dr. Colavita has received honoraria for speaking from W.L. Gore. All other authors confirm they have no financial and personal relationships that could potentially and inappropriately influence this work or its conclusions.