Factors associated with postoperative costs following anatomic lung resections without major complications

Transcription

1 European Journal of Cardio-Thoracic Surgery 51 (2017) doi: /ejcts/ezw307 Advance Access publication 24 October 2016 ORIGINAL ARTICLE Cite this article as: Brunelli A, Drosos P, Ismail H, Pompili C, Bassi V. Factors associated with postoperative costs following anatomic lung resections without major complications. Eur J Cardiothorac Surg 2017; 51: Factors associated with postoperative costs following anatomic lung resections without major complications Alessandro Brunelli a, *, Polyvios Drosos a, Haaris Ismail b, Cecilia Pompili c and Vinod Bassi b a b c Department of Thoracic Surgery, St James s University Hospital, Leeds, UK Costing Team, Finance Department, Leeds Teaching Hospitals NHS Trust, Leeds, UK Cost team and Finance Department, Leeds Institute of Cancer and Pathology, University of Leeds, Leeds, UK * Corresponding author. Department of Thoracic Surgery, St. James s University Hospital, Leeds, UK. Tel: ; fax: ; brunellialex@gmail.com (A. Brunelli). Received 12 June 2016; received in revised form 6 July 2016; accepted 11 July 2016 Abstract OBJECTIVES: To detect factors associated with costs of anatomic lung resection without major complications. METHODS: Two hundred and fifty consecutive patients submitted to anatomic lung resection (185 by VATS) in 1 fiscal year (1 April March 2015) were included. Thoracic Morbidity and Mortality (TMM) system was used to grade the severity of complications. Two hundred and ten patients who did not develop major complications (TMM < 3) were analysed. Postoperative costs were retrieved from the Financial Department through a Patient Level Information and Costing System. Multivariable regression and bootstrap analyses were used to test the association of several baseline patient characteristics with costs and obtain an aggregate scoring system to estimate postoperative costs. RESULTS: Among the 210 patients, 117 (56%) did not develop any complication and 93 (44%) had minor complications. Their average postoperative cost was 4040e, significantly lower than the one observed in patients with major complications (13 156e, P < ). Multivariable regression revealed that open thoracotomy (P = 0.01), carbon monoxide lung diffusion capacity (DLCO) < 60% (P = 0.001) and coronary artery disease (CAD) (P = 0.009) were associated with postoperative costs. Open thoracotomy would increase the cost by 648e, DLCO < 60% by 935e and CAD by 1043e. If all three factors were present, they would cause an increase of postoperative costs from 3592e to 6219e. CONCLUSION: We were able to identify clinical factors associated with postoperative costs in patients without major complications. Recognizing groups of increased cost may lead to specific process analyses aimed at optimising their pathways of care and ultimately saving money. Moreover, these findings may help administrators to tailor future individualized lung resection reimbursement tariffs based on patient characteristics. Keywords: Lung resection Lung cancer Costs Risk modelling INTRODUCTION Cost variability after lung resection has been recently investigated in the attempt to identify groups of patients in which to improve their pathways of care and ultimately save money [1 3]. It is well known that the occurrence of postoperative complications with the attending increase in resource utilization and prolonged hospital stay is the principle factor affecting costs following surgery [4, 5]. However, only 20 30% of all patients undergoing to lung resection will develop complications. The majority will have an uneventful postoperative course. The advent of video-assisted thoracoscopic surgery (VATS) for performing major anatomic lung resections has reduced the incidence of complications even more [6, 7]. If every effort should be done to minimize the occurrence of complications by improving standards of care and patient selection, we should not forget that the majority of patients will never experience major postoperative complications. They represent the larger volume of patients and as such the main hospital cost driver. The identification of cost variability of patients not experiencing major complications will help to streamline care in selected groups with certain characteristics with the aim to improve their care and reduce costs. Therefore, the objective of this study was to verify postoperative cost variability following anatomic lung resections without major complications and to identify factors associated with increased costs in this group of patients. Presented at the 24th European Conference on General Thoracic Surgery, Naples, Italy, 29 May-1 June VC The Author Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved.

2 A. Brunelli et al. / European Journal of Cardio-Thoracic Surgery 231 PATIENTS AND METHODS This is a retrospective analysis on prospectively collected data of 250 consecutive patients submitted to anatomic lung resections (224 lobectomies, 15 pneumonectomies and 11 segmentectomies) performed in a single-centre during 1 fiscal year (1 April March 2015). One hundred and eighty five resections were performed by VATS. The study was reviewed by the Research and Innovation Department of the hospital and classified as service evaluation not requiring review by a NHS Research Ethics Committee. All patients were discussed at a tumour multidisciplinary meeting and inoperability criteria were in accordance with current guidelines [8]. All patients were operated on by qualified general thoracic surgeons either through a muscle sparing thoracotomy or a 2 3 port VATS. In general, patients were extubated in the operating room and transferred to a High Dependency Unit (HDU) for the first 24 h. They were then transferred to a dedicated thoracic ward bed when their conditions allowed. All patients were managed according to standardized pathways of care including pain control (paravertebral and intravenous patient controlled analgesia), early mobilization, chest physiotherapy and rehabilitation and venous thromboembolic prophylaxis. Fixed, semi-fixed and variable postoperative costs were obtained for each patient from the hospital Patient Level Information and Costing System (PLICS). The system categorises the Trust s costs as either direct, indirect or as an overhead to patient care in accordance with National Health Service (NHS) Clinical Costing Standards. Then by using patient activity taken from clinical or other data sources, it allocates these costs to each patient in the most appropriate and granular method. This generates a patient bill showing the resources consumed and an actual or calculated cost for the patient episode. Costs were collected in GBP and expressed in Euro (Eur:GBP exchange rate 1.4, as of January 2016). Fixed costs, such as senior management pay, building overheads and cleaning are allocated in one of two ways. The first method allocates costs to patients in a stepped process, firstly by assigning costs to an area or speciality within the hospital and then to the patients that utilize that area. For example, the cost of a ward stay will include heat and lighting costs for the ward and will be weighted based on the patient s length of stay. Alternatively, fixed costs can be allocated based on the actual usage of a service. For example, actual cleaning hours will be used to allocate a proportion of cleaning costs to individual patients based on their levels of usage. However, it is the validation and analyses of semi-fixed and variable costs that can inform the differences between pathways of care. These costs may include: radiology and pathology testing, theatre usage, medications, medical staff costs and other specific postoperative therapeutic procedures such as cardioversion, bronchoscopy and blood transfusions. The methods for allocating these costs to patients vary depending on the resource type. Testing costs for example will be directly matched to patients whereas consultant costs will be allocated based on a proportion of their time spent with patients. A breakdown of the unitary costs used for this analysis is shown in Table 1. Table 1: Items In most cases, the largest postoperative costs incurred by each patient were the costs of the ward bed they occupied. As these costs vary depending on the patient s length of stay, they can be used as an indication of acuity for each case. Three types of ward have been utilized across the range of cases, a general thoracic ward, a HDU and an Intensive Care Unit (ICU). The predominant ward cost is staffing but also includes running costs such as consumables, drug stock and maintenance costs. Consequently, the average fully absorbed cost for 1 day stay in a general thoracic ward, thoracic HDU bed and ICU is calculated as 418e, 491e and 3060e, respectively. Thoracic Morbidity and Mortality (TMM) system was used to grade the severity of complications [9]. Regardless the type of complication, the TMM system grades each post-surgical complication in a I to V scale according to the complexity of its management. Grades I and II include minor complications requiring no therapy or pharmacologic intervention only. Grades III and IV are major complications that require surgical intervention or life support. Grade V complications result in patient death as illustrated in Table 2. We focused our analysis on 210 patients who did not develop major complications (TMM < 3). Statistical analysis Breakdown of unitary costs used for the analysis Cost (Euro) Ward stay (per day) High dependency unit stay (per day) Intensive care unit stay (per day) 3060 Chest X-ray Chest computed tomography with contrast The individual line-item costs were summed to obtain the total postoperative cost associated with each patient. The following baseline and surgical variables were tested for a possible association with postoperative costs in patients without major complications by using multivariable regression analysis validated by bootstrap resampling technique: age, sex, body mass index, extent of resection, forced expiratory volume in 1 s (FEV1%), carbon monoxide lung diffusion capacity (DLCO%), moderate to severe chronic obstructive pulmonary disease (COPD) status (FEV1 < 80% and FEV1/FVC ratio < 70%), coronary artery disease (CAD), cerebrovascular disease, diabetes, American Society of Anaesthesiologist (ASA) score, Eastern Cooperative Oncology Group performance score and induction chemotherapy. For the purpose of developing a user-friendly model, numeric variables were categorized by searching the best threshold value using a c-index analysis. Multiple regression analysis was then performed by including these categorized variables. Variables with P < 0.05 were retained in the final model and their reliability tested by bootstrap analysis with 1000 samples [10 12]. In the bootstrap procedure, repeated samples of the same number of observations as the original database were selected with replacement from the original set THORACIC

3 232 A. Brunelli et al. / European Journal of Cardio-Thoracic Surgery Table 2: Thoracic mortality and morbidity classification system Minor complications Grade I Adverse event which alters the standard postoperative course without requiring a specific treatment Grade II Pharmacologic treatment or minor intervention required Major Grade IIIa Surgical, radiologic, endoscopic treatment, or multitherapy required without general anesthesia complications Grade IIIb Surgical, radiologic, endoscopic treatment, or multitherapy required with general anesthesia Grade IVa Intensive care unit treatment for single organ dysfunction required Grade IVb Intensive care unit treatment for multiple organ dysfunction required Mortality Grade V Adverse event which leads to death Table 3: Characteristics of the patients included in the study (210 patients without major complications) Variables Age (10.7) Sex male (n, %) 78 (37.1%) BMI (kg/m 2 ) (5.3) FEV1% (20.4) DLCO% (16.6) CAD (n, %) 18 (8.6%) ASA score 2.44 (0.6) ECOG score >1 (n, %) 18 (8.6%) Neoadjuvant chemotherapy 11 (5.2%) Pneumonectomy (n, %) 6 (2.9%) Results are expressed as means and standard deviations unless otherwise specified. BMI: body mass index; FEV1: forced expiratory volume in 1 s; DLCO: carbon monoxide lung diffusion capacity; CAD: coronary artery disease; ASA: American Society of Anaesthesiology score; ECOG: Eastern Cooperative Oncology group performance score. observations. For each sample, multivariable regression was repeated. Variables occurring in the majority of the bootstrap models (>50%) were judged to be stable and retained in the final model. Finally, a model estimating the postoperative cost based on patient characteristics and type of surgery was developed. The analysis was performed by using the STATA 12.0 (Stata Corp., College Station, TX) statistical software. RESULTS Of the total 250 patients submitted to major anatomic resections during the study time, 40 developed major complications (TMM > 2) and were excluded from the analysis. One hundred and seventeen patients did not develop any complications and 93 patients experienced minor complications (TMM grade 1 or 2). The most frequent complications classed as minor were the followings: 16 prolonged air leaks, 16 atrial fibrillation, 25 pneumonia, 11 sputum retentions treated with physiotherapy, 4 postoperative delirium, 5 acute kidney failures, 5 ileum, 5 subcutaneous emphysema, 4 urinary tract infections, 4 urinary retentions, 1 wound infection and 6 electrolyte imbalances. The characteristics of the 210 patients without major complications included in the study are shown in Table 3. Patients with minor or no complications had an average postoperative cost of 4040e (range e), 9000e lower than the cost observed in patients with major complications (13 156e, P < ). Table 4 shows the breakdown of postoperative costs by TMM class. Patients with complications graded as TMM 2 (pharmacologic treatment only) were 1480e more expensive than those without any complications (P < 0.001). The higher costs of the patients with minor complications were mainly associated with their longer hospital stays. These patients had an average postoperative stay 6.5 days longer than those without complications (10.7 vs 4.1 days, P < ). The results of the stepwise multivariable regression analysis (dependent variable: postoperative cost) are shown in Table 5. The presence of CAD (P = 0.004), low DLCO (P = 0.001) and performance of thoracotomy (P = 0.01) instead of VATS were independently and reliably associated with postoperative costs. The corresponding regression equation to estimate postoperative costs is the following: XCAD (presence of CAD coded as 1, absence of CAD coded as 0) XThoracotomy (presence of thoracotomy coded as 1, absence of thoracotomy coded as 0) -21.6XDLCO. Variables such as sex, age, FEV1, ECOG score, ASA score, Diabetes, cerebrovascular disease, COPD status, neoadjuvant chemotherapy and extent of resection were not retained in the final model as their P value was lower than DLCO resulted the only significant numeric variable. The best cut-off value associated with postoperative cost resulted 60%. Multivariable regression was repeated using DLCO < 60% as categorical variable (Table 6). According to the regression model, being submitted to thoracotomy would increase the postoperative cost by 648e, having DLCO < 60% by 935e and having a history of CAD by 1043e. If all three factors were present, an increase of postoperative costs from 3592e (cost of a patient without major complications and without any of the three factors) to 6219e would be expected. The higher postoperative costs observed in patients with one or more of the risk factors can be explained in part by their longer hospital stay compared to those without any of the risk factors (6.1 vs 4.7 days, P = ). In addition, the proportion of patients with a postoperative stay longer than 7 days was higher in those with at least one risk factor (24%) compared to those without any (11%) (P = 0.01). Patients with no major complications and with a history of CAD had a 1.5 days longer postoperative stay compared to those without CAD (7.6 vs 5.1 days, P = 0.009). Patients with no major complications with a preoperative DLCO < 60% had a 1.7 days longer postoperative stay compared to those with a higher DLCO value (6.7 vs 5.0 days, P = ).

4 A. Brunelli et al. / European Journal of Cardio-Thoracic Surgery 233 Table 4: Breakdown of postoperative costs by TMM class TMM class Number of patients Postoperative costs (e) Table 5: Results of the stepwise multivariable regression analysis (dependent variable: postoperative costs) Predictors Coefficients SE P-value Bootstrap Intercept 5344 CAD % Thoracotomy % DLCO % Parsimonious model is shown. CAD: coronary artery disease; DLCO: carbon monoxide lung diffusion capacity; bootstrap: percentage of significancy in 1000 bootstrap samples. Table 6: Results of the stepwise multivariable regression analysis with DLCO expressed as categorical variable (DLCO < 60%) (dependent variable: postoperative costs) Predictors Coefficients SE P-value Bootstrap Intercept CAD % Thoracotomy % DLCO< % Parsimonious model is shown. CAD: coronary artery disease; DLCO: carbon monoxide lung diffusion capacity; bootstrap: percentage of significancy in 1000 bootstrap samples. Finally, patients with no major complications submitted to thoracotomy instead of VATS had a 0.9 days longer postoperative stay (6.0 vs 5.1 days, P = 0.09). Interestingly, when all patients with and without major complications were combined, DLCO and CAD lost their association with postoperative costs at regression analysis. Multiple regression analysis performed on all 250 patients showed that presence of diabetes (P = 0.002) and thoracotomy (P = 0.002) remained significantly associated with postoperative costs. The corresponding following cost estimation model was derived for all patients: XDiabetes (presence of diabetes coded as 1, absence of diabetes coded as 0)+ 3186XThoracotomy (presence of thoracotomy coded as 1, absence of thoracotomy coded as 0). DISCUSSION Rationale and objectives Identification of cost variability amongst patients submitted to lung resection has been a recent focus of investigation. Previous papers have tried to develop financial models adjusted on clinical characteristics and have been able to identify groups of patients or specific comorbidities associated with increased cost after lung resection [1 3]. A recent study has shown for instance that although intraoperative costs may be higher than postoperative costs, the latter displayed a much greater variability [3]. Knowing that certain groups of patients are at higher risk of being more expensive than others can trigger in-depth analyses to audit their pathways of care with the final aim to improve their postoperative course, outcome and ultimately save money. Although previous studies have shown that postoperative costs are mainly related to complications, the majority of patients after lung surgery will not develop major complications. They certainly represent the main financial driver of this operation. For this reason, we wanted to focus our attention on those patients, who did not develop major complications and try to identify factors associated with postoperative costs. In order to classify the severity of complications we used a previously validated system, the TMM classification system [9]. This grading system classifies the severity of the complications based on the complexity of their treatment. Complications managed with observation only or by pharmacologic intervention were classed as minor. The final objective was to develop a financial model able to estimate the postoperative costs in this specific group of patients. Main findings One of the main finding of this study was that patients with minor or no complications following lung resection had an average postoperative cost of 4040e, with a large variability. The cheapest patient cost 2205e whereas the most expensive e, a value 7-fold higher. This result confirmed previous findings. In the paper of Medbery et al. [2] postoperative costs showed the greatest variability ranging from 3000$ to $ after VATS lobectomy. Similarly, in a previous paper from our group [3], the variability of the postoperative costs was much larger than the intraoperative costs (from 530e to e). It appears clear that there are margins of saving by improving the care of those patients with the highest cost. The use of multivariable regression analysis showed that some clinical factors were significantly associated with postoperative cost. A history of CAD, a low DLCO and performance of thoracotomy instead of VATS increased the risk of having an expensive postoperative course. Patients with one or more of these risk factors had longer postoperative stays and this explains in part their increased costs. The presence of CAD and a low DLCO were found associated with postoperative costs in previous studies [2, 3]. Their association with cost appears logical since they are factors reputedly linked with increased risk of complications. The novel finding in this study was that these risk factors were also associated with increased costs in those patients with minor complications only or without complications. THORACIC

5 234 A. Brunelli et al. / European Journal of Cardio-Thoracic Surgery A possible explanation of this finding can be that the presence of CAD or low DLCO may increase the risk of complications such as AF, pneumonia or prolonged air leak, which are mostly managed without any major procedure or with pharmacologic intervention only (TMM class 1 or 2). Despite being classed as minor complications they can contribute to prolong hospital stay and increase cost. For the same reason, patients operated on by VATS experienced a shorter hospital stay, probably for a lower incidence of minor complications or less postoperative pain. This translated into a lower cost compared to patients operated on by thoracotomy. This finding confirmed previous studies showing the postoperative financial benefits of VATS over thoracotomy [13 16]. We were able to show this financial benefit also in patients without major postoperative complications. According to our analysis, a patient with an uneventful course submitted to thoracotomy would have an expected postoperative cost 648e higher than if operated on by VATS. Limitations This study may have several potential limitations. The risk adjusting cost analysis refers only to patients submitted to anatomic lung resection and who did not experience major complications. The model is not valid for those patients experiencing major complications, who are expected to have higher costs and for other types of operations (as patients will likely have different case mixes). Dichotomization of numeric factors in the aggregate model to estimate costs may have inherent downsides such as imprecision in cost estimation. Nevertheless, we performed this categorization only after having run the regression analysis keeping all numeric variables as continuous. DLCO resulted significant and reliable when used as continuous variable and was categorized only post hoc to produce a more user-friendly model. The financial data used in this study has been derived from the Trust s new PLICS system, which was implemented from March 2015 onwards. The main methodologies used to allocate costs in PLICS have been validated with the input of clinical, financial and managerial staff. However, the validation of outputs is inevitably an on-going process that will require further clinical review and discussion to improve the accuracy, robustness and confidence in the data produced. Cost accuracy is also dependent upon the level of costs that can be allocated using electronic feeds that capture the resources actually consumed by identifiable patients, which otherwise would be allocated using an appropriate apportionment methodology. Even where electronic sources are available, poor data quality may limit their usefulness within PLICS (for example, if patient identifiers are not correctly recorded). The process of validation should therefore focus attention on the way information is recorded and coded across the organization, helping to improve the quality and variety of data that is inputted into the system. This will reduce the need to use manufactured methodologies to allocate costs to patients and add layers of granularity to the information that PLICS produces. Notably this includes advances in initiative aimed at developing a patient barcoding system, which will allow instant tracking of consumables to the patients that have used them. In addition, refinement in hospital stay cost, which would take into consideration a cost variability proportional with the duration of stay assuming a progressively lower intensity of care may help in the future to improve the cost model. Our hospital has a high patient volume, allowing negotiable high discounts of consumables, which may not be reproducible in other settings. This study only looked at the hospital postoperative costs associated with lung resections and did not capture the entire episode of care, including readmissions, emergency room visits, postoperative clinic visits and home care. Such financial information was not available at the time of this study. Therefore, our financial figures represent only a portion of the entire postoperative care episode. Future analyses will be needed to evaluate and risk adjust the costs associated with the other elements, since they will be utilized in future bundled payment arrangements. Clinical and research implications The knowledge that certain patient characteristics or certain types of operation, independently by the occurrence of major complications, are associated with increased postoperative costs may have important clinical and managerial implications. This information may be used to assess the financial performance of specific surgical services and to define more precisely the required investments and resources utilization during the annual budget discussion. The use of a cost model such as the one derived from this analysis fits in the conceptual framework of the accountable care organizations. Providers have both financial and clinical accountabilities. They are responsible to deliver high quality outcomes and reduce total cost of care for specific population of patients. Furthermore, the model proposed in this study represents a further methodological template, which can be used by health administrators and regulators to redefine reimbursement tariffs for specific groups of patients with certain characteristics. Finally, the identification of CAD, low DLCO and thoracotomy as factors associated with postoperative costs should prompt root cause analyses to identify the cause of the increased costs in these groups of patients with the aim to improve their pathways of care, reduce cost variability and ultimately save money. Future research is needed to evaluate whether the institution of specific measures or processes of care would lead to a reduction in cost variability in these classes of patients. Conflict of interest: none declared. REFERENCES [1] Brunelli A, Salati M, Refai M, Xiumé F, Rocco G, Sabbatini A. Risk-adjusted econometric model to estimate postoperative costs: an additional instrument for monitoring performance after major lung resection. J Thorac Cardiovasc Surg 2007;134: [2] Medbery RL, Perez SD, Force SD, Gillespie TW, Pickens A, Miller DL et al. Video-assisted thoracic surgery lobectomy cost variability: implications for a bundled payment era. Ann Thorac Surg 2014;97: [3] Brunelli A, Tentzeris V, Sandri A, McKenna A, Liew SL, Milton R et al. A risk-adjusted financial model to estimate the cost of a video-assisted thoracoscopic surgery lobectomy programme. Eur J Cardiothorac Surg 2016;49: [4] Birkmeyer JD, Gust C, Dimick JB, Birkmeyer NJ, Skinner JS. Hospital quality and the cost of inpatient surgery in the United States. Ann Surg 2012;255:1 5.

6 A. Brunelli et al. / European Journal of Cardio-Thoracic Surgery 235 [5] Handy JR Jr, Denniston K, Grunkemeier GL, Wu YX. What is the inpatient cost of hospital complications or death after lobectomy or pneumonectomy? Ann Thorac Surg 2011;91: [6] Paul S, Altorki NK, Sheng S, Lee PC, Harpole DH, Onaitis MW et al. Thoracoscopic lobectomy is associated with lower morbidity than open lobectomy: a propensity-matched analysis from the STS database. J Thorac Cardiovasc Surg 2010;139: [7] Falcoz PE, Puyraveau M, Thomas PA, Decaluwe H, Hürtgen M, Petersen RH et al. ESTS Database Committee and ESTS Minimally Invasive Interest Group. Video-assisted thoracoscopic surgery versus open lobectomy for primary non-small-cell lung cancer: a propensity-matched analysis of outcome from the European Society of Thoracic Surgeon database. Eur J Cardiothorac Surg 2016;49: [8] Brunelli A, Kim AW, Berger KI, Addrizzo-Harris DJ. Physiologic evaluation of the patient with lung cancer being considered for resectional surgery: diagnosis and management of lung cancer, 3rd ed: American College of Chest Physicians evidence-based clinical practice guidelines. Chest 2013;143(5 Suppl):e166S 90S. [9] Seely AJ, Ivanovic J, Threader J et al. Systematic classification of morbidity and mortality after thoracic surgery. Ann Thorac Surg 2010;90: [10] Blackstone EH. Breaking down barriers: helpful breakthrough statistical methods you need to understand better. J Thorac Cardiovasc Surg 2001;122: [11] Grunkemeier GL, Wu YX. Bootstrap resampling method: something for nothing? Ann Thorac Surg 2004;77: [12] Brunelli A. A synopsis of resampling techniques. J Thorac Dis 2014;6: [13] Farjah F, Backhus LM, Varghese TK, Mulligan MS, M. Cheng A, Alfonso- Cristancho R et al. Ninety-day costs of video-assisted thoracic surgery versus open lobectomy for lung cancer. Ann Thorac Surg 2014;98: [14] Casali G, Walker WS. Video-assisted thoracic surgery lobectomy: can we afford it? Eur J Cardiothorac Surg 2009;35: [15] Swanson SJ, Meyers BF, Gunnarsson CL, Moore M, Howington JA, Maddaus MA et al. Video-assisted thoracoscopic lobectomy is less costly and morbid than open lobectomy: a retrospective multiinstitutional database analysis. Ann Thorac Surg 2012;93: [16] Deen SA, Wilson JL, Wilshire CL, Vallières E, Farivar AS, Aye RW et al. Defining the cost of care for lobectomy and segmentectomy: a comparison of open, video-assisted thoracoscopic, and robotic approaches. Ann Thorac Surg 2014;97: APPENDIX. CONFERENCE DISCUSSION Dr M. Okumura (Osaka, Japan): VATS involves more procedural costs compared to thoracotomy. Yet the net cost of the operative treatment is the same. Can you comment on that? Dr Drosos: We have focused on the postoperative cost. The operative cost of VATS is well known to be higher than open thoracotomy. But then the postoperative stay is shorter, which makes the whole postoperative cost less than the one for thoracotomy. So the overall cost is less for VATS than for open thoracotomy. Dr Drosos: We have focused on the postoperative cost. The operative cost of VATS is well known to be higher than open thoracotomy. But then the postoperative stay is shorter, which makes the whole postoperative cost less than the one for thoracotomy. So the overall cost is less for VATS than for open thoracotomy. Dr S. Shiono (Yamagata, Japan): You showed that thoracotomy cases are very expensive, but the patients who receive a thoracotomy are advanced cases. I think that advanced lung cancer is related to cost. What is your opinion? Dr Drosos: We have not analysed according to the stage of cancer. We have focused on the postoperative cost according to the grade of complications. As I showed, we put all anatomic lung resections - all the pneumonectomies, lobectomies and segmentectomies - in the same pool, regardless of the stage of the cancer. So we have not made a breakdown analysis on that point of yours, but it is an interesting one and could be investigated further. Dr H. Eid (Dubai, UAE): I think the postoperative hospital stay doesn t depend on thoracotomy or VATS. It depends on the drains, when you can remove the drains. If I remove the drains two days after thoracotomy, I send the patient home; the same for VATS. So it is not VATS or thoracotomy that provides the benefit factor in terms of the cost. However, I believe the consumables cost for VATS lobectomy in the OR is more than for open thoracotomy. Do you agree with me? Dr Drosos: Yes,theconsumablesofVATSaremoreexpensive,theonesthatwe use are more expensive than those for open thoracotomy in units that do not use staplers for all vessels. Dr Eid: But the approach, VATS or thoracotomy, is not the issue. What determines the postoperative stay is the drainage, when you remove the drain. Once you remove the drain, you can send the patient home. Dr Drosos: I would like maybe to highlight that this is what we have found here. This is an association between those two variables, thoracotomy and the postoperative cost. We cannot be exactly sure about the causal link between these factors and the results, because this is retrospective analysis. It would need further investigation to elucidate this. Dr I. Polyakov (Krasnodar, Russian Federation): This study is measuring surgical or therapeutic factors by cost. But might it be possible to measure the cost by units other than pounds or Euros in order to make it more universally applicable in measuring the effectiveness of the surgical procedure? Dr Drosos: The initial calculations were GB pounds, but we converted it to Euros, as this is a European conference. But this is something that we can look at inthefuturewithaviewtousingmoregloballyrelevantunits.wehavechosen Euros as one of the main currencies in the world and the main currency of Europe. THORACIC