Advertisement

Comparing robot-assisted thoracic surgical lobectomy with conventional video-assisted thoracic surgical lobectomy and wedge resection: Results from a multihospital database (Premier)

Published:November 11, 2013DOI:https://doi.org/10.1016/j.jtcvs.2013.09.046

      Background

      Video-assisted thoracic surgical (VATS) lobectomies and wedge resections result in less morbidity and shorter length of stay than resections via thoracotomy. The impact of robot-assisted thoracic surgical (RATS) lobectomy on clinical and economic outcomes has not been examined. This study compared hospital costs and clinical outcomes for VATS lobectomies and wedge resections versus RATS.

      Methods

      Using the Premier hospital database, patients aged ≥18 years with a record of thoracoscopic lobectomy, segmental resection, or excision of a lesion or tissue from the lung between 2009 and 2011 were identified. Procedures using robotic technology were identified if 1 of 2 conditions were met: (1) a robotic International Classification of Diseases, Ninth Revision procedure code or (2) the text fields in the hospital record indicated that the robot was used. Using a propensity score and based on severity and comorbidities, certain demographics and hospital characteristics were matched. The association between VATS or RATS and adverse events, hospital costs, surgery time, and length of stay was examined.

      Results

      Of 15,502 patient records analyzed, 96% (n = 14,837) were performed without robotic assistance. Using robotic assistance was associated with higher average hospital costs per patient. The average cost of inpatient procedures with RATS was $25,040.70 versus $20,476.60 for VATS (P = .0001) for lobectomies and $19,592.40 versus $16,600.10 (P = .0001) for wedge resections, respectively. Inpatient operating times were longer for RATS lobectomy than VATS lobectomy (4.49 hours vs 4.23 hours; P = .0959) and wedge resection (3.26 vs 2.86 hours; P = .0003). Length of stay was similar with no differences in adverse events.

      Conclusions

      RATS lobectomy and wedge resection seem to have higher hospital costs and longer operating times, without any differences in adverse events.

      CTSNet classification

      Abbreviations and Acronyms:

      APR-DRG (All Patient Refined Diagnosis Related Groups), ICD-9 (International Classification of Diseases, Ninth Edition), RATS (robot-assisted thoracic surgical), SD (standard deviation), VATS (video-assisted thoracic surgical)
      Lung surgery for the purpose of diagnosis or treatment has evolved in the past 2 decades. Historically, surgery involving the lung was accomplished using 1 of 2 main approaches, depending on the clinical indication: via a thoracoscope inserted using a small incision or by an open thoracotomy, involving a larger incision and rib spreading to improve visibility and access for control of the surgical anatomy.
      Thoracoscopic procedures have been transformed by the ongoing refinement of video-assisted thoracoscopic surgery (VATS) techniques and equipment, particularly high-definition cameras and monitors. For traditional thoracotomy indications, VATS is an evolving technique and is increasingly applied in situations where traditional open thoracotomy has long been used. There is a small but increasing literature to support the growth of VATS in this context. However, VATS is still an evolving phenomenon, with a developing research base,
      • Barker A.
      • Maratos E.C.
      • Edmonds L.
      • Lim E.
      Recurrence rates of video-assisted thoracoscopic versus open surgery in the prevention of recurrent pneumothoraces: a systematic review of randomised and non-randomised trials.
      • Cheng D.
      • Downey R.J.
      • Kernstine K.
      • Stanbridge R.
      • Shennib H.
      • Wolf R.
      • et al.
      Video-assisted thoracic surgery in lung cancer resection: a meta-analysis and systematic review of controlled trials.
      • Sedrakyan A.
      • van der Meulen J.
      • Lewsey J.
      • Treasure T.
      Video assisted thoracic surgery for treatment of pneumothorax and lung resections: systematic review of randomised clinical trials.
      • Whitson B.A.
      • Groth S.S.
      • Duval S.J.
      • Swanson S.J.
      • Maddaus M.A.
      Surgery for early-stage non-small cell lung cancer: a systematic review of the video-assisted thoracoscopic surgery versus thoracotomy approaches to lobectomy.
      and with a variable definition.
      • Sedrakyan A.
      • van der Meulen J.
      • Lewsey J.
      • Treasure T.
      Video assisted thoracic surgery for treatment of pneumothorax and lung resections: systematic review of randomised clinical trials.
      • Whitson B.A.
      • Groth S.S.
      • Duval S.J.
      • Swanson S.J.
      • Maddaus M.A.
      Surgery for early-stage non-small cell lung cancer: a systematic review of the video-assisted thoracoscopic surgery versus thoracotomy approaches to lobectomy.
      • Flores R.M.
      • Alam N.
      Video-assisted thoracic surgery lobectomy (VATS), open thoracotomy, and the robot for lung cancer.
      Variations exist in the number of port incisions and the appropriate incision length.
      The purported benefits of VATS in lung surgery compared with open thoracotomy in the literature published to date includes smaller incisions, less pain, less blood loss, less respiratory compromise, less complications, and faster recovery times, all translating into shorter length of stay and similar survival.
      • McKenna Jr., R.J.
      • Houck W.
      • Fuller C.B.
      Video-assisted thoracic surgery lobectomy: experience with 1,100 cases.
      These findings have been reported in systematic reviews of randomized and nonrandomized clinical trials,
      • Cao C.
      • Manganas C.
      • Ang S.C.
      • Peeceeyen S.
      • Yan T.D.
      Video-assisted thoracic surgery versus open thoracotomy for non-small cell lung cancer: a meta-analysis of propensity score-matched patients.
      and in comparative studies.
      • Cheng D.
      • Downey R.J.
      • Kernstine K.
      • Stanbridge R.
      • Shennib H.
      • Wolf R.
      • et al.
      Video-assisted thoracic surgery in lung cancer resection: a meta-analysis and systematic review of controlled trials.
      In a comparative study of patients with lung cancer, Cajipe and colleagues,
      • Cajipe M.D.
      • Chu D.
      • Bakaeen F.G.
      • Casal R.F.
      • LeMaire S.A.
      • Coselli J.S.
      • et al.
      Video-assisted thoracoscopic lobectomy is associated with better perioperative outcomes than open lobectomy in a veteran population.
      found that there were fewer complications in VATS patients (14 of 46, 30%) than their open counterparts (26 of 45, 58%; P = .009). VATS patients also had a chest tube for a shorter time and shorter length of stay. In multivariate analysis, VATS was associated independently with a reduced risk of complications (odds ratio, 0.359; P = .04).
      The impact of RATS on clinical and economic outcomes has not been examined, however. This study compared hospital costs and clinical outcomes for VATS lung resection versus RATS lung resection.
      Despite several publications in support of robotic surgery, controversy still remains over limited high-quality evidence of improved clinical outcomes compared with traditional minimally invasive approaches. Clinical outcomes suggest that robotic surgery is equivalent to conventional minimally invasive procedures when important end points such as conversion to open surgery, hospital stay, and recovery time are considered.
      • Deutsch G.B.
      • Sathyanarayana S.A.
      • Gunabushanam V.
      • Mishra N.
      • Rubach E.
      • Zemon H.
      • et al.
      Robotic vs. laparoscopic colorectal surgery: an institutional experience.
      • Masterson T.A.
      • Cheng L.
      • Boris R.S.
      • Koch M.O.
      Open vs. robotic-assisted radical prostatectomy: a single surgeon and pathologist comparison of pathologic and oncologic outcomes.
      The adoption and diffusion of this technology in thoracic surgery, coupled with limited high-quality evidence of improved outcomes compared with traditional VATS procedures, raises important questions about resource allocation. The systems typically cost between $1 million and $2.5 million
      • Turchetti G.
      • Palla I.
      • Pierotti F.
      • Cuschieri A.
      Economic evaluation of da Vinci-assisted robotic surgery: a systematic review.
      plus an additional expensive yearly service contract. Additional direct costs of robotic-assisted surgery include instrument disposables and potentially increased procedure time.
      In this era of comparative effectiveness and health care reform in the United States, and with concerns about resource utilization at the forefront, the trend toward RATS deserves further evaluation. Therefore, this study examined the clinical and economic outcomes (cost and utilization) in patients undergoing VATS lung resection versus RATS lung resection.

      Materials and Methods

       Data Source

      The Premier hospital database was used as the data source for this study.

      Premier, Inc. Available from: http://www.premierinc.com/. Accessed January 12, 2012.

      This database contains complete data on patient billing, hospital costs, and coding histories from more than 600 health care facilities throughout the United States. The data for this study were extracted from more than 25 million inpatient discharges and 175 million hospital outpatient visits from acute care facilities, ambulatory surgery centers, and clinics across the nation.
      A protocol describing the analysis objectives, criteria for patient selection, data elements of interest, and statistical methods was submitted to the New England Institutional Review Board (NEIRB) and exemption was obtained.
      Eligible patients were ≥8 years of age and had undergone a VATS lobectomy or wedge resection between 2009 and 2011, for which the lobectomy or wedge resection was the primary reason for surgery. Patients were categorized according to the following types of VATS procedure: lobectomy (code 32.41), wedge resection (code 32.30 or 32.20).
      Lobectomy and wedge resection procedures using robotic technology were identified if 1 of the 2 following conditions were met: (1) a robotic International Classification of Diseases, Ninth Edition (ICD-9) procedure code accompanied the primary procedure code of interest or (2) text fields were found when mining the hospital charge master file for each patient indicating use of the robot. Procedures that involved conversion from minimally invasive to open or between robotic and VATS approaches were excluded from the analysis dataset.
      For all eligible patients, elements describing hospital cost, surgery time, length of stay, use of robot, type of thoracic procedure, and indication for procedure were obtained from the data. Cost analysis (calculation) reflected the cost of the robotic procedure to the hospital but did not include acquisition or the annual maintenance fee for the da Vinci robot (Intuitive Surgical, Inc, Sunnyvale, Calif). The preoperative All Patient Refined Diagnosis Related Groups (APR-DRG) severity level was used as an index of comorbidity. The 3M APR-DRG Classification System is a widely adopted proprietary risk adjustment classification tool that uses information from routine claims data to produce valid and reliable severity measurement and risk adjustment scores.

      Averill RF, Goldfield N, Hughes JS, Bonazelli J, McCullough EC, Steinbeck BA, et al. What are APR-DRGs? An introduction to severity of illness and risk of mortality adjustment methodology. White paper. 2003 Available from: http://solutions.3m.com/3MContentRetrievalAPI/BlobServlet?locale=it_IT&lmd=1218718280000&assetId=1180603360910&assetType=MMM_Image&blobAttribute=ImageFile. Accessed January 12, 2012.

      It is used to account for differences related to an individual’s severity of illness or risk of mortality in large datasets. Comorbid conditions that might influence procedure selection or outcomes of interest, such as the presence of cardiovascular or pulmonary disease, cancer, or diabetes mellitus, were obtained using ICD-9 diagnosis codes. Appendix Table 1 provides a detailed list of all ICD-9 codes for each condition included in the study. Information on sociodemographic characteristics and health insurance status was also included, as were descriptors of the care setting, namely census region, urban or rural setting, teaching hospital status, and facility bed count.
      Adverse events (identified by ICD-9 codes) that occurred intraoperatively and within 30 days postoperatively that included pulmonary complications were flagged and included in the analysis. These were further grouped as major or minor complications for analysis. A detailed list of each event and the corresponding ICD-9 code is found in Appendix Table 2.

       Statistical Analysis

      The study objective was to use the Premier hospital database to compare clinical and economic outcomes in patients undergoing lobectomy or wedge resection using VATS versus RATS. Outcomes of interest included adverse events (minor and major), hospital costs, length of stay, and surgery time. Costs were the actual costs incurred by the hospital for all treatments and services related to the lobectomy and wedge resection and did not include robotic capital or service contracts.
      A quasi-randomization method called propensity scoring was used to create groups of analyzable patients who were well matched. Propensity scores were assigned based on likely predictors of the outcome of interest. Covariates on which to match were selected based on their availability in the Premier database, as well as their general acceptance as factors associated with the outcomes of interest. The goal of this propensity-matching analysis was to find pairs of patients receiving lobectomy or wedge resection via RATS or VATS procedures, who shared like propensities based on the matching variables. An SAS macro from the Mayo Clinic used nearest-neighbor matching on the estimated propensity scores to choose matches for the patients who had a RATS procedure.

      Mayo Clinic. Gmatch macro developed by Erik Bergstralh and Jon Kosanke. 2003. Available from: http://www.mayo.edu/research/departments-divisions/department-health-sciences-research/division-biomedical-statistics-informatics/software/locally-written-sas-macros. Accessed November 20, 2011.

      Propensity scores were calculated for receipt of robotic procedures for each patient included in the analysis based on a nonparsimonious multivariable logistic regression model. Patients were matched on the following characteristics: severity group, age, gender, race, region, size of facility, teaching facility (yes/no), location of facility, insurance group, malignancy status, chronic pulmonary disease, chronic viral hepatitis, connective tissue disease, congestive heart failure, diabetes mellitus, liver disease, myocardial infarction, acute or old, other chronic or unspecified heart failure, peripheral vascular disease, and chronic renal insufficiency. The robotic and nonrobotic patients were randomly ordered and a nonrobotic patient with a propensity score closest to the first robotic patient was chosen. Assessment of residual bias was conducted by evaluating the differences in the distribution of patient characteristics before and after matching.
      To assess the extent to which the propensity matching reduced confounders, the distribution of several variables before and after matching were compared, including age, gender, race, insurance type, health status, malignancy indication, region, location, facility type, facility size, and comorbid conditions among the patients in the cohorts. Group comparisons were made using χ2 tests. Least square means were used to test for differences between the matched cohorts on the 3 continuous variables of interest: hospital cost, surgery time, and length of stay. Logistic regression models were used to test for significant differences between the 2 groups and to generate odds ratios on the following categories of adverse events/complications: major and minor. To further assess any residual bias that may occur because only a few hospitals represent most of the robotic procedures in the sample, sensitivity analysis was conducted by running multiple regression models on 1 of the hospitals in which most of the robotic procedures were performed and a nonrobotic hospital meeting the same hospital characteristics. Ordinary least squares regression models were run for hospital costs and logistic regression models were run for adverse events, both major and minor. Analyses were performed using SAS Version 9.2 (SAS Institute Inc., Cary, NC).

      Results

      A total of 15,502 patient records from 305 hospitals were analyzed. The patient attrition process is shown in Table 1. Ninety-six percent of these thoracic procedures were traditional VATS (n = 14,837); lung resection was performed with RATS in 665 procedures, 4% of the total. In the robotic group, 335 lobectomies and 330 wedge resections were performed.
      Table 1Attrition
      DescriptionNumber of patients remainingNumber of patients dropped for this reason
      Total number of patients in Premier database 2009 Q1 to 2011 Q2102,914,774
      Patients with a primary procedure code for lobectomy or wedge resection (32.20, 32.30, 32.41)15,965102,898,809
      Patients 18 y or older at date of procedure15,596369
      Patients with inpatient visits only15,50294
      Before matching, patients undergoing RATS lobectomy had similar distributions for age, gender, and health status. Those undergoing RATS resection had similar distributions for age, gender, and insurance type compared with those undergoing VATS lung resection (Table 2). Furthermore, few differences in comorbidities were noted between robot and nonrobot groups. Characteristics of the hospitals showed notable differences for region, teaching versus nonteaching, and bed count. For teaching versus nonteaching and bed count, most of the robotic procedures were performed in teaching hospitals (292 of 335 lobectomies, 281 of 330 wedge resections) with more than 200 beds (334 of 335 lobectomies, 324 of 330 wedge resections), compared with nonrobotic procedures, with almost half coming from nonteaching hospitals (44% lobectomy, 48% wedge resection) with greater variation in bed size (Table 3).
      Table 2Patient demographics—before match
      CategoryLobectomyWedge resection
      RATSVATSP valueRATSVATSP value
      NPercentNPercentNPercentNPercent
      Total number335100381810033010011,019100
      Age (mean)66.3666.27.885661.8159.56.0126
      Age group.9089.1606
       18-40 y72.09832.17329.7144313.1
       41-50 y257.462436.36288.48130211.82
       51-60 y6017.9173819.337522.73226320.54
       61-70 y10932.54128833.739829.7295926.85
       71-80 y10230.45114529.997522.73235921.41
       >80 y329.553218.41226.676936.29
      Gender.4484.7783
       Female17652.54208854.6916951.21545149.47
       Male15947.46173045.3116148.79556450.49
       Unknown00000040.04
      Insurance<.0001.7399
       Government19558.21249065.2218656.36597354.21
       Managed care12437.0197725.5910331.21361732.83
       Other164.783519.194112.42142912.97
      Race<.0001.0011
       White24071.64301278.8926480802172.79
       African American144.183278.56216.3610459.48
       Hispanic4011.94601.57164.853453.13
       Other4112.2441910.97298.79160814.59
      Health status
      APR-DRG severity level was used as an index of preoperative comorbidity.
      .1353.001
       APR-DRG severity level (1, 2)26779.7290576.0927382.73823674.74
       APR-DRG severity level (3, 4)6820.391323.915717.27278325.26
      Malignancy indication.0093.0006
       No lung cancer4814.333769.8516850.91676161.36
       Primary neoplasm of the lung26679.4326885.5911635.15306427.81
       Metastases other than lung216.271744.564613.94119410.84
      Comorbid conditions
       Myocardial infarction, acute or old308.963589.38.7994309.098257.49.2767
       Congestive heart failure226.572356.16.7641247.279008.17.5580
       Chronic or unspecified heart failure61.79581.52.698530.912872.6.0544
       Peripheral vascular disease308.9639610.37.4125257.587987.24.8178
       Dementia92.69541.41.067851.521391.26.6849
       Chronic pulmonary disease17151.04191750.21.769416349.39564951.27.5026
       Connective tissue disease82.391493.9.1634185.455494.98.6980
       Liver disease185.371844.82.6514236.976475.87.4043
       Chronic viral hepatitis00330.86.087651.521131.03.3876
       Renal insufficiency, chronic226.572707.07.7291185.458017.27.2094
       Diabetes mellitus8224.4877020.17.06115115.45216119.61.0603
      RATS, Robot-assisted thoracic surgery; VATS, video-assisted thoracic surgery; APR-DRG, all patient refined diagnosis related group.
      APR-DRG severity level was used as an index of preoperative comorbidity.
      Table 3Hospital demographics based on patient counts
      CategoryLobectomyWedge resection
      RATSVATSP valueRATSVATSP value
      NPercentNPercentNPercentNPercent
      Total number335100381810033010011,019100
      Census region<.0001<.0001
       Northeast5215.5290823.787723.33285825.94
       West0060315.79226.67175515.93
       South25375.52169344.3420060.61416037.75
       Midwest308.9661416.08319.39224620.38
      Location.1851.0539
       Urban31594.03365095.631996.6710,37594.16
       Not urban205.971684.4113.336445.84
      Type<.0001<.0001
       Teaching29287.16213255.8428185.15574452.13
       Nonteaching4312.84168644.164914.85527547.87
      Bed count.0023.0585
       <5000000070.06
       51–10000150.3900400.36
       101–20010.31413.6961.825154.67
       >20033499.7366295.9132498.1810,45794.9
      RATS, Robot-assisted thoracic surgery; VATS, video-assisted thoracic surgery.
      To balance cohorts and mitigate the possibility of confounders and the large discrepancy in sample size between robotic and nonrobotic procedures, patients were matched using a propensity score, as described earlier, on certain demographic and hospital characteristics. After matching, a total of 1240 patients remained; 590 lobectomies and 650 wedge resections in each group from 132 hospitals; of these, there were 40 hospitals (30%) where robotic procedures were being performed. Patient characteristics, comorbid conditions, and hospital characteristics after matching are represented in Table 4. After matching, patients were balanced with respect to demographics, comorbid conditions, and hospital characteristics, with the exception of region for the lobectomy cohort and bed count for the wedge resection cohort.
      Table 4Patient demographics—matched cohorts
      CategoryLobectomyWedge resection
      RATSVATSP valueRATSVATSP value
      NPercentNPercentNPercentNPercent
      Total number295100295100325100325100
      Age (mean)66.4366.54.911261.7461.50.8403
      Age group.9596.9307
       18–40 y72.3751.69329.853410.46
       41-50 y217.12186.1288.62329.85
       51-60 y5117.295819.667422.776319.38
       61-70 y9933.569833.229529.239830.15
       71-80 y8829.838829.837523.087824
       >80 y299.83289.49216.46206.15
      Gender.4574.0698
       Female15452.216355.2516751.3819058.46
       Male14147.813244.7515848.6213541.54
       Unknown
      Insurance.6862.7658
       Government17358.6418061.0218356.3117453.54
       Managed care10635.9310334.9210131.0810632.62
       Other165.42124.074112.624513.85
      Race.4626.9712
       White23579.6624984.4126481.2326481.23
       African American144.75113.73216.46216.46
       Hispanic72.3772.37113.38134
       Other3913.22289.49298.92278.31
      Health status
      APR-DRG severity level was used as an index of preoperative comorbidity.
      .7597.1921
       APR-DRG severity level (1, 2)23378.982368026982.7728186.46
       APR-DRG severity level (3, 4)6221.0259205617.234413.54
      Malignancy indication.9706.5428
       No lung cancer4013.563812.8816751.3816952
       Primary neoplasm of the lung23880.6824081.3611334.7712036.92
       Metastases other than lung175.76175.764513.853611.08
      Census region<.0001.9672
       Northeast5016.959431.867723.698225.23
       West00237.8226.77237.08
       South21572.8812341.691956019058.46
       Midwest3010.175518.64319.54309.23
      Location.4915.2464
       Urban27593.2227994.5831496.6230894.77
       Not urban206.78165.42113.38175.23
      Type.39741.0000
       Teaching25285.4225987.827684.9227684.92
       Nonteaching4314.583612.24915.084915.08
      Bed count.5627.0117
       <500000
       51–10000000010.31
       101–20010.3420.6861.85206.15
       >20029499.6629399.3231998.1530493.54
      Comorbid conditions
       Myocardial infarction, acute or old268.81248.14.7675298.92288.62.8897
       Congestive heart failure217.12165.42.3959226.77206.15.7497
       Chronic or unspecified heart failure62.0382.71.588530.9230.921.0000
       Peripheral vascular disease258.47279.15.7715257.69247.38.8819
       Dementia82.7131.02.128151.5430.92.4768
       Chronic pulmonary disease14850.1715351.86.680516149.5416550.77.7537
       Connective tissue disease82.7182.711.0000185.54175.23.8620
       Liver disease165.4293.05.1525226.77154.62.2360
       Chronic viral hepatitis0020.68.156651.5430.92.4768
       Renal insufficiency, chronic217.12103.39.0424185.54175.23.8620
       Diabetes mellitus6823.056020.34.42425015.385216.8292
      RATS, Robot-assisted thoracic surgery; VATS, video-assisted thoracic surgery; APR-DRG, all patient refined diagnosis related group.
      APR-DRG severity level was used as an index of preoperative comorbidity.
      After matching, cohorts were tested for differences in average hospital costs, operating room time, and length of stay (Table 5). The average cost of inpatient procedures with/without robotic assistance was $25,040.70 versus $20,476.58 (P = .0001) for lobectomies and $19,592.42 versus $16,600.13 (P = .0001) for wedge resections, respectively. Operating room times were longer, although not statistically different for robotic lobectomy (4.49 vs 4.23 hours; P = .0959). Operating room times were increased for robotic wedge resection versus nonrobotic wedge resection (3.26 vs 2.86 hours; P = .0003). Average length of stay of both cohorts was not statistically different for lobectomy (6.07 vs 5.83 days; P = .6131) and wedge resection (5.23 vs 5.38 days; P = .7188).
      Table 5Length of stay, hospital costs, and surgery time after matching
      LobectomyWedge resection
      RATSVATSP valueRATSVATSP value
      Length of stay (d)
       Median4444
       Mean6.075.83.61315.235.380.7188
       SD6.445.035.185.27
      Total hospital costs ($)
       Median21,833.3418,080.1117,341.3313,640.52
       Mean25,040.7020,476.58<.000119,592.4216,600.13.0001
       SD13,164.0110,977.679,293.6410,367.82
      Operating room time (h)
       Median4.2542.932.5
       Mean4.494.23.09593.262.86.0003
       SD1.981.731.411.31
      RATS, Robot-assisted thoracic surgery; VATS, video-assisted thoracic surgery; SD, standard deviation.
      Adverse events occurring in the postoperative period, up to 30 days after discharge, were tabulated and grouped into major and minor. Complications rates between robotic and nonrobotic surgery cohorts, regardless of whether they were examined within a perioperative 30-day period or only within the original perioperative hospital stay (Table 6) were reported. The odds of an event occurring were not significantly different for major and minor events in either time period for lobectomy or wedge resection (Table 7).
      Table 6Adverse events among matched data by analysis groups
      CategoryLobectomyWedge resection
      RATSVATSRATSVATS
      DuringDuring or afterDuringDuring or afterDuringDuring or afterDuringDuring or after
      NPercentNPercentNPercentNPercentNPercentNPercentNPercentNPercent
      Total number295100295100295100295100325100325100325100325100
      Any major event4013.565016.955016.955618.985316.317121.854112.625115.69
       Acute respiratory failure165.42186.1206.78227.46216.46329.85206.15226.77
       Empyema10.3420.6810.3420.6810.3130.9241.2341.23
       Bronchopleural fistula51.6951.6920.6831.0230.9230.9210.3110.31
       Pneumonia186.1258.47279.15299.83288.623310.15164.92247.38
      Any minor event10033.9010936.9510134.2411338.3110030.7710833.2310632.6211435.08
       Air leak and other pneumothorax7023.737525.426522.037023.7378248024.6278248124.92
       Atelectasis/pulmonary collapse3010.173311.193612.24314.58195.85257.69247.38288.62
       Cellulitis0010.34000000000010.31
       Chylothorax10.3410.3431.0231.02
       Spontaneous tension pneumothorax000010.3420.6830.9230.9241.2341.23
      RATS, Robot-assisted thoracic surgery; VATS, video-assisted thoracic surgery; SD, standard deviation.
      Table 7Adverse events after matching
      LobectomyWedge resection
      Odds ratio estimateLower CIUpper CIP valueOdds ratio estimateLower CIUpper CIP value
      Original hospital stay or within 30 d of follow-up
       Major0.8640.5371.389.54581.240.8071.905.3265
       Minor1.0150.7221.427.93080.9710.6951.357.8647
      Original hospital stay only
       Major0.7890.4731.316.36451.2240.7631.963.4013
       Minor1.0320.7281.462.85890.9270.6591.304.6639
      CI, Confidence interval.

       Sensitivity Analysis

      In the propensity model, the following hospital characteristics were adjusted: location (urban vs rural), region, teaching status, and bed size. However, more hospitals contributed to our sample of nonrobotic procedures than robotic procedures. Of the 132 hospitals in our sample, only 40 had robotic procedures. Therefore, to further assess any residual bias that may occur because only a few hospitals represented most of the robotic procedures in the sample, the 1 hospital in which 126 lobectomies and 71 wedge resections were performed was matched to a nonrobotic hospital meeting the same hospital characteristics: southern, urban, teaching with a bed count of 500. Only the patients from these 2 hospitals were analyzed. Results of the ordinary least squares cost model revealed significantly higher differences in costs for RATS lobectomy versus VATS lobectomy ($30,365 vs $20,238) and RATS wedge resection versus VATS wedge resection ($27,969 vs $17,887). Regarding adverse events, results were confirmatory for major events in both lobectomy and wedge resection with no statistical differences. Regarding minor events, there were no differences for patients undergoing wedge resection. However, there was a significant difference for lobectomies. Patients who underwent lobectomy with the robot were 4.24 (odds ratio) times more likely to have a minor event then their nonrobotic counterparts (P < .0001).

      Discussion

      Based on our matched cohorts, there seems to be no differences between robot-assisted lobectomy and wedge resection procedures when considering intraoperative and postoperative complications. After matching and creating 2 balanced cohorts, robot-assisted procedures were associated with higher hospital costs, with an average incremental cost per procedure of $4565 for lobectomy and $2992 for wedge resection. Robot-assisted wedge procedures were also associated with longer operating room times.
      The findings related to higher hospital costs associated with robotic surgery are consistent with similar studies in the literature evaluating minimally invasive surgical procedures
      • Weissman J.S.
      • Zinner M.
      Comparative effectiveness research on robotic surgery.
      • Wright J.D.
      • Ananth C.V.
      • Lewin S.N.
      • Burke W.M.
      • Lu Y.S.
      • Neugut A.I.
      • et al.
      Robotically assisted vs laparoscopic hysterectomy among women with benign gynecologic disease.
      ; robotic surgery is consistently more expensive.
      Another important consideration is that the costs associated with robotic surgery in these studies do not account for substantial acquisition costs for the robot. The robotic unit costs between $1 million and $2.5 million, and is associated with annual maintenance costs of $100,000 to $180,000.

      Intuitive Surgical Investor Presentation Q4 2009. Available from: http://investor.intuitivesurgical.com/phoenix.zhtml?c=122359&p=irol-IRHome. Accessed January 12, 2012.

      The combination of limited high-quality clinical evidence comparing traditional VATS and RATS lobectomy and wedge resection to date and relatively high costs raises questions about the cost-effectiveness of this technology. In addition, robotic surgery ideally requires a dedicated operating team and an additional surgeon at the table, who are not generally required for standard VATS procedures. This cost is difficult to assess but certainly needs to be considered.
      Important strengths of this analysis included the prospectively developed protocol that directed the analysis, the quasi-randomization propensity scoring methodology that was used, the broad geographic and demographic representation of US hospitals included in the sample, and the fact that these data are relatively recent and represent a national setting. This study also had some noteworthy limitations. Because the data were mined from a hospital administrative database used for billing purposes, certain data points could not be captured or could not be clearly identified. Examples include patient body mass index and patient behavior such as smoking habits. An analysis of pain scores, quality of life, morbidity, and time to return to work would be interesting contributions to the literature, but these data are not available in the database used in this analysis. This analysis was limited to patients who underwent VATS or RATS procedures only, and excluded those procedures that were converted from RATS to VATS or to open thoracotomy. Furthermore, data regarding the precision of robotic versus nonrobotic procedures, including surgical margins and adequacy of lymph node dissection, could not be evaluated. The dataset provides operating room time and is not suited to further exploration regarding time allotted to set-up, bronchoscopy, take down of adhesions, preresection examination or other subcategories. The analysis was limited to a 30-day perioperative period, which limits analysis related to long-term survival or potential long-term complications. However, these limitations are inherent to the data source and could be rationalized to affect both cohorts similarly; therefore, the risk of bias in 1 cohort is lessened. The surgeon and institutional learning curve for using robotic technology could not be evaluated. Because robotic lung surgery is a relatively new procedure, there may be future potential efficiencies in both cost and time related to increasing familiarity with this technology.

      Conclusions

      This study represents the most up-to-date and expansive analysis of cost and effectiveness outcomes associated with RATS and VATS procedures in a national setting. These findings reveal little clinical differences in perioperative adverse events. Coupled with the increased cost of the robot per case and increased operative times for robotic lobectomy and wedge resections, results suggest that further consideration is warranted before using this technology when standard VATS seems to provide better results. Future studies that evaluate cost relative to robotic-assisted case volume and prospective randomized controlled studies focusing on comparative effectiveness between traditional and robotic-assisted thoracic procedures are needed.
      Appendix Table 1Comorbid conditions
      ConditionICD-9 code
      Myocardial infarction, acute or old410.xx, 412
      Congestive heart failure428.0
      Other chronic or unspecified heart failure428.20, 428.22, 428.30, 428.32, 428.40, 428.42, 428.9
      Peripheral vascular disease440.xx, 443.8x, 443.9
      Dementia290.xx, 294.xx, 331.0, 331.11, 331.19, 331.2, 331.7, 331.82
      Chronic pulmonary disease490.xx-494.xx, 495.x, 496, 500-505
      Connective tissue disease710.xx, 714.xx
      Liver disease571.x, 572.x, 573.xx
      Chronic viral hepatitis070.22-070.23, 070.32-070.33, 070.44, 070.54
      Renal insufficiency, chronic585.xx
      Diabetes mellitus249.xx, 250.xx
      Hemiplegia342.xx
       Due to cerebral palsy 343.1, 343.4
       Due to previous cardiovascular accident 438.2x
      ICD, International Classification of Diseases.
      Appendix Table 2Postoperative procedure-specific complications
      CategoryPulmonaryICD-9 code
      MajorAcute respiratory failure518.81, 518.84 , 518.5
      MajorEmpyema510.9
      MajorBronchopleural fistula510.0
      MajorPneumonia480.x to 486, 507.0
      MinorSpontaneous tension pneumothorax512.0
      MinorAtelectasis/pulmonary collapse518.0
      MinorAir leak and other pneumothorax512.1, 512.8
      MinorChylothorax457.8
      MinorCellulitis682.2
      ICD, International Classification of Diseases.

      References

        • Barker A.
        • Maratos E.C.
        • Edmonds L.
        • Lim E.
        Recurrence rates of video-assisted thoracoscopic versus open surgery in the prevention of recurrent pneumothoraces: a systematic review of randomised and non-randomised trials.
        Lancet. 2007; 370: 329-335
        • Cheng D.
        • Downey R.J.
        • Kernstine K.
        • Stanbridge R.
        • Shennib H.
        • Wolf R.
        • et al.
        Video-assisted thoracic surgery in lung cancer resection: a meta-analysis and systematic review of controlled trials.
        Innovations (Phila). 2007; 2: 261-292
        • Sedrakyan A.
        • van der Meulen J.
        • Lewsey J.
        • Treasure T.
        Video assisted thoracic surgery for treatment of pneumothorax and lung resections: systematic review of randomised clinical trials.
        BMJ. 2004; 329: 1008
        • Whitson B.A.
        • Groth S.S.
        • Duval S.J.
        • Swanson S.J.
        • Maddaus M.A.
        Surgery for early-stage non-small cell lung cancer: a systematic review of the video-assisted thoracoscopic surgery versus thoracotomy approaches to lobectomy.
        Ann Thorac Surg. 2008; 86 (discussion 16-18): 2008-2016
        • Flores R.M.
        • Alam N.
        Video-assisted thoracic surgery lobectomy (VATS), open thoracotomy, and the robot for lung cancer.
        Ann Thorac Surg. 2008; 85: S710-S715
        • McKenna Jr., R.J.
        • Houck W.
        • Fuller C.B.
        Video-assisted thoracic surgery lobectomy: experience with 1,100 cases.
        Ann Thorac Surg. 2006; 81 (discussion 5-6): 421-425
        • Cao C.
        • Manganas C.
        • Ang S.C.
        • Peeceeyen S.
        • Yan T.D.
        Video-assisted thoracic surgery versus open thoracotomy for non-small cell lung cancer: a meta-analysis of propensity score-matched patients.
        Interact Cardiovasc Thorac Surg. 2013; 16: 244-249
        • Cajipe M.D.
        • Chu D.
        • Bakaeen F.G.
        • Casal R.F.
        • LeMaire S.A.
        • Coselli J.S.
        • et al.
        Video-assisted thoracoscopic lobectomy is associated with better perioperative outcomes than open lobectomy in a veteran population.
        Am J Surg. 2012; 204: 607-612
        • Deutsch G.B.
        • Sathyanarayana S.A.
        • Gunabushanam V.
        • Mishra N.
        • Rubach E.
        • Zemon H.
        • et al.
        Robotic vs. laparoscopic colorectal surgery: an institutional experience.
        Surg Endosc. 2012; 26: 956-963
        • Masterson T.A.
        • Cheng L.
        • Boris R.S.
        • Koch M.O.
        Open vs. robotic-assisted radical prostatectomy: a single surgeon and pathologist comparison of pathologic and oncologic outcomes.
        Urol Oncol. 2013; 31: 1043-1048
        • Turchetti G.
        • Palla I.
        • Pierotti F.
        • Cuschieri A.
        Economic evaluation of da Vinci-assisted robotic surgery: a systematic review.
        Surg Endosc. 2012; 26: 598-606
      1. Premier, Inc. Available from: http://www.premierinc.com/. Accessed January 12, 2012.

      2. Averill RF, Goldfield N, Hughes JS, Bonazelli J, McCullough EC, Steinbeck BA, et al. What are APR-DRGs? An introduction to severity of illness and risk of mortality adjustment methodology. White paper. 2003 Available from: http://solutions.3m.com/3MContentRetrievalAPI/BlobServlet?locale=it_IT&lmd=1218718280000&assetId=1180603360910&assetType=MMM_Image&blobAttribute=ImageFile. Accessed January 12, 2012.

      3. Mayo Clinic. Gmatch macro developed by Erik Bergstralh and Jon Kosanke. 2003. Available from: http://www.mayo.edu/research/departments-divisions/department-health-sciences-research/division-biomedical-statistics-informatics/software/locally-written-sas-macros. Accessed November 20, 2011.

        • Weissman J.S.
        • Zinner M.
        Comparative effectiveness research on robotic surgery.
        JAMA. 2013; 309: 721-722
        • Wright J.D.
        • Ananth C.V.
        • Lewin S.N.
        • Burke W.M.
        • Lu Y.S.
        • Neugut A.I.
        • et al.
        Robotically assisted vs laparoscopic hysterectomy among women with benign gynecologic disease.
        JAMA. 2013; 309: 689-698
      4. Intuitive Surgical Investor Presentation Q4 2009. Available from: http://investor.intuitivesurgical.com/phoenix.zhtml?c=122359&p=irol-IRHome. Accessed January 12, 2012.