Cost Analysis of 109 Microsurgical Reconstructions and Flap Monitoring with Microdialysis

Transcription

1 Cost Analysis of 109 Microsurgical Reconstructions and Flap Monitoring with Microdialysis Leena Setälä, M.D., Ph.D., 1 Heini Koskenvuori, M.Sc., 2 Daiva Gudaviciene, M.D., Ph.D., 3 Leena Berg, M.D., Ph.D., 1 and Paula Mustonen, M.D., Ph.D. 1 ABSTRACT Few studies have examined the cost-effectiveness of microsurgery, and little is known about the cost-effectiveness of flap monitoring. We studied the costs related to microsurgery during 2004 to 2006 in Kuopio University Hospital. A total of 99 patients were reconstructed with 109 flaps. Primary success was achieved in 64% of cases. Reoperation for anastomosis was conducted in 25% and for other surgical complications in 27%. The intended result was achieved in 94% of cases. The mean total cost of hospital care was 20,000 s in head and neck cancer surgery, 15,500 s in defects of the lower extremities, and 9200 s in breast reconstruction. The costs were greatly influenced by surgical complications (i.e., if the primary reconstruction failed, then the secondary microvascular flap almost doubled the expense involved; mean expenses per case 27,900 s). Microdialysis was used in flap monitoring with an additional cost of 535s per patient. We found that microdialysis provided an early diagnosis of perfusion failure and helped to save the flap. It was estimated that if one or two flaps per year are saved due to more effective monitoring, then the extra costs of using microdialysis are covered. KEYWORDS: Microsurgery, monitoring, microdialysis, cost analysis Since the introduction of microsurgery into plastic and reconstructive surgery, this technique has been considered as demanding, laborious, costly, and sometimes unreliable. The quality of microsurgery and medical treatment overall has increased, resulting in high success rates especially in hospitals performing a large number of microsurgical operations. Large volumes of microsurgery allow for more experience, better training, and the development of routine treatment protocols. However, there are also small-volume microsurgical units where the lack of routine must be substituted by other means to control the quality of the reconstructions. Today, it is not known which quality control measures are cost-effective. It is obvious that taking care of the safety of the flap transfer increases the expenses but at the same time it assists in decreasing the risk of costly complications. Low complication rates can be achieved by careful patient selection, high level of education, and expertise of the surgical and anesthesia team and effective patient monitoring. In addition, flap monitoring Departments of 1 Plastic Surgery and 2 Accounting and Finance, Kuopio University Hospital, Kuopio, Finland; 3 Department of Breast Surgery and Oncology, Vilnius University Institute of Oncology, Vilnius, Lithuania. Address for correspondence and reprint requests: Leena Setälä, M.D., Ph.D., Department of Surgery, Kuopio University Hospital, PO Box 1777, Kuopio, Finland ( leena.setala@kuh.fi). J Reconstr Microsurg. Copyright # by Thieme Medical Publishers, Inc., 333 Seventh Avenue, New York, NY 10001, USA. Tel: +1(212) Received: March 15, Accepted after revision: June 22, DOI /s ISSN X.

2 JOURNAL OF RECONSTRUCTIVE MICROSURGERY using clinical supervision or technical devices is considered mandatory by many authors. 1 3 The primary aim of the study was to investigate the costs of microsurgical reconstructions in relation to its outcome. We also wanted to make an economic analysis of microdialysis (MD) as a monitoring method and to evaluate its impact on the outcome. METHODS The study population consisted of 99 consecutive patients who underwent microsurgical reconstructions in 2004 to The 50 male and 49 female patients were operated at Kuopio University Hospital. All reconstructions were made by senior plastic surgeons usually assisted by registrars of the plastic surgery training program. Patient information was collected by retrospective chart review, and financial data were extracted from the hospital database. In these patients, 109 microsurgical flaps were used, chosen according to the needs of the defect and varying from a simple standard flap to multicomponent osteomusculocutaneous flaps. The most usual indication for the reconstruction was resection of oropharyngeal cancer, followed by lower-leg trauma and breast cancer. Postoperative MD has been used in our hospital as a routine monitoring system to support the standard clinical follow-up of each microvascular flap since In this series, it was used in 106 flaps, and it was the only monitoring method for 22 buried or otherwise poorly visualized flaps. The method has been described previously. 4 Shortly, a semipermeable catheter was implanted into the flap intraoperatively and perfused by a battery-operated pump. The perfusate was collected hourly and analyzed at bedside to assess the glucose and lactate concentrations, which reflect the state of flap perfusion and oxygenation. In our hospital, glucose concentration over 1.0 mmol/l and lactate concentration less than 7.0 mmol/l are considered as signs of optimal flap perfusion. The medical and financial data were analyzed using a statistical analysis program (SPSS for Windows, 11.2, SPSS Inc., Chicago, IL). The patient data were analyzed using descriptive statistics. The costs between different groups were analyzed using analysis of variance (ANOVA) and expressed as means and standard deviations when appropriate; p values less than 0.05 were considered statistically significant. RESULTS The intended result was ultimately achieved in 102 of the 109 patients (94%). Primary success without a reoperation was achieved in 64% (Table 1). Attempts to make a reconstruction early after a trauma or in a primary cancer operation were more prone to complications. The rate of anastomosis revisions was 25%, including seven single arterial reanastomoses for thrombosis, five venous and five multiple reanastomoses, and an exploration in eight flaps. One arterial reanastomosis was conducted for bleeding and another for pedicle rupture. Seven flaps were lost; the correction of perfusion failure was successful in 14 of 19 attempts. Twenty-four patients needed additional treatment because of surgical complications (total flap loss in seven cases, partial flap necrosis in seven, skin graft loss in two, wound necrosis in six, wound infection in two). The treatments included 11 new flaps (six microvascular, five local), four skin grafts, four cases of prolonged conservative wound treatment, and one lower leg amputation. Costs The mean total cost for the period of hospital care of these microsurgical patients was 16,325 s. Thisconsisted of costs of the operating theater (3722 s, 23%), anesthesia costs (2121 s, 13%), intensive care unit (2011 s, 12%), and the hospital room day fees (6457 s, 40%). The latter includes all treatment at the ward: nutrition, medication, supplies, basic diagnostics, the attending doctors fees, and overhead costs. The remaining mean 2013 s per patient is made up of physiotherapy, major radiological procedures, and the fees for consulting physicians and surgeons. The indication of the surgery affected significantly the total costs of the treatment (Table 2). The costs were clearly increased if there were surgical complications (Table 3). The mean total cost of an uncomplicated case was 14,597 s; a case with surgical complications (other than flap loss), 17,367 s; and a case with flap loss and a new reconstruction, Table 1 Success of the Surgery Subdivided According to the Indication Cancer Primary Cancer Secondary Trauma Primary* Trauma Secondary Other Total Result as planned 34 (64%) 8 (89%) 12 (52%) 11 (73%) 6 (67%) 71 (64%) Planned result with complications 16 (30%) 0 (0%) 9 (39%) 4 (27%) 2 (22%) 31 (39%) Result not achieved 3 (6%) 1 (11%) 2 (9%) 0 (0%) 1 (11%) 7 (6%) Total *Within 30 days after the trauma.

3 COST ANALYSIS OF MICROSURGERY AND FLAP MONITORING/SETÄLÄ ET AL Table 2 Indication Costs of the Patient Care and Their Components Subdivided According to the Indication Total Costs, Operation Costs, Anesthesia Costs, Costs of the ICU, Costs of the Ward, Cancer primary (n ¼ 50) 15,340 (7914) 4011 (2493) 1886 (1087) 1815 (3822) 5899 (3629) Cancer secondary (n ¼ 9) 9869 (5367) 2359 (1767) 2354 (2099) 0 (0) 3620 (1645) Trauma primary* (n ¼ 17) 27,208 (22,871) 5109 (4724) 2959 (1798) 5362 (8070) 10,209 (7042) Trauma secondary (n ¼ 14) 12,338 (4934) 3123 (3030) 2028 (511) 0 (0) 6057 (2652) Other (n ¼ 9) 13,897 (7063) 1787 (1042) 1759 (767) 1908 (5188) 5935 (3011) Total (n ¼ 99) 16,325 (12,412) 3722 (3042) 2122 (1315) 2011 (4801) 6457 (2268) p value y *Within 30 days after the trauma. y Analysis of variance. All costs shown in euros. ICU, intensive care unit; SD, standard deviation. 27,954 s (p ¼ 0.021, ANOVA). The length of stay (LOS) was similarly associated with the success of the reconstruction. If no complications occurred, the LOS was 12 days (standard deviation [SD] 8), with complications it rose to 14 days (SD 7), and with flap loss it was further extended 18 days (SD 11). The need of intensive care unit also affected the costs but was also related to coexistent medical conditions such as multiple trauma. Microdialysis The mean direct cost for monitoring one flap with MD was 535 s. The costs were composed of two independent parts: the costs of the investment in the monitoring system and those associated with the single-use supplies per patient. The former was counted by dividing the yearly sum of amortization of the purchase price of the equipment and its maintenance costs by the number of patients each year. The latter was the cost of the catheter, the sample tubes, and the reagents needed per patient. No labor costs were included in the direct costs of MD. The impact of MD on the clinical course of the patient was estimated individually (Fig. 1). Eight flaps were recognized to show deteriorating MD values despite normal clinical appearance. Among these, one true venous thrombosis was treated early with reanastomosis. In three flaps, a potentially compressing hematoma was removed and two were explored for a sudden transitory drop of glucose values. No clot was found, and the situation was interpreted as an arterial spasm, because the patients had expressed feeling pain or cold before the drop. In one case, the poor MD values were related to poor location of the catheter, thus this pedicle exploration was unnecessary. In contrast, MD values were good in six pale-looking flaps, which were followed successfully with no intervention. In 11 flaps, both clinical signs and MD pointed to failing perfusion. In two such cases, the surgeon failed to recognize the true nature of the intervention needed, and the flaps were lost. Of the nine remaining cases, six were salvaged and three were lost due to repeated pedicle thrombosis. Of the 22 buried flaps, eight were reoperated for a perfusion failure diagnosed solely by MD. One flap was lost because the ischemic period was extended through a soon-recurring thrombosis. In these 106 flaps, the specificity of MD in diagnosing disturbed perfusion was 99% and the sensitivity was 100%. In our series, the pedicles of 20 flaps were thrombosed, in four of them repeatedly and in two unrecognized by the surgeon. The remaining 14 flaps seem to have benefited from early diagnosis, resulting in a reasonably good salvage rate (74% for the whole series). Historically, the salvage rate has been 44% in our hands, so it is estimated that due to better monitoring, five flaps were saved during the 3-year period. Two additional flaps could possibly have been saved, if the surgeon would have reacted to the results of the analysis. In three cases, MD caused extra cost through an unnecessary reoperation. As the total costs of the treatment increased by a mean of s for each patient with a lost Table 3 Costs Subdivided According to the Success of the Reconstruction Total Costs, Operation Costs, Anesthesia Costs, Costs of the ICU, Mean (SD) Costs of the Ward, Result as planned (n ¼ 64) 14,597 (12,895) 3471 (3187) 1852 (1154) 1873 (4929) 5847 (4348) Planned result with complication (n ¼ 28) 17,367 (7838) 3765 (2059) 2301 (869) 1450 (2992) 7262 (4451) Result not achieved (n ¼ 7) 27,954 (17,293) 5836 (4419) 3866 (2530) 5517 (8070) 8821 (4991) Total 16,325 (12,412) 3722 (3042) 2122 (1315) 2011 (4801) 6457 (4468) p value* *Analysis of variance. All costs shown in euros. ICU, intensive care unit; SD, standard deviation.

4 JOURNAL OF RECONSTRUCTIVE MICROSURGERY Normal looking flap Normal MD values (n = 59) (n = 56) (n = 3) Normal looking flap Abnormal MD values Reoperation (n = 8) (n = 7) Abnormal looking flap Normal MD values (n = 6) (n = 6) Abnormal looking flap Reoperation (n = 10) (n = 4) (n = 2) Flap lost (n = 4) Abnormal MD values Flap lost Buried flap Normal MD values (n = 14) (n = 14) Buried flap Abnormal MD values Reoperation (n = 8) (n = 5) (n = 2 ) Flap lost Figure 1 Postoperative clinical courses in 106 flaps monitored by microdialysis. 1 One venous thrombosis, three compressing hematomas. Two cases of suspected transient spasm, one false-positive with misplaced catheter. 2 Two cases with repeated venous thrombosis, one triple thrombosis with thrombophilia, one missed venous thrombosis. 3 Repeated arterial thrombosis, ischemia-reperfusion injury. flap, and the yearly cost of MD monitoring system was 18,900 s, it seemed that one or two saved flaps per year would cover the extra costs of MD. DISCUSSION Plastic surgeons, those working in public and those in the private health care sector, need to be cost conscious. When operative resources are allocated, plastic surgeons need to be able to present reliable data about treatment costs, surgical effectiveness, and patient outcome to compete with other surgical specialities. It was recently suggested that cost-effectiveness analysis should be incorporated into clinical studies in plastic surgery. 5 In an attempt to increase the success in microvascular reconstruction, we wish to include improved safety at all stages of the patient management. If we are to accomplish this aim, we must be able to demonstrate the efficiency of the safety measures and justify their costs. We chose to analyze the expenses of reconstructive microsurgery in our hospital because this type of surgery is believed to be costly and there is a need to justify its use of resources. 6,7 Our hospital is publicly funded and functions as a nonprofit organization. The costs of the treatment are paid by the patient s home municipality and to a lesser extent by the patient him- or herself or by an insurance company. Each unit of our hospital charges the other units the true costs of their services (e.g., diagnostic analyses, anesthesia, time in the intensive care unit). In this way, the total cost of the treatment period, even one consisting of several activities, can be traced back to their origins. The prices of all internal services are set annually using the cost-price principle. Therefore, we believe that the financial data presented in this study reflect the true cost of each performance with reasonable accuracy. The costs incurred in a microvascular reconstruction are related to the type and site of the reconstruction, the patient s comorbidity and complications, and also the experience of the physician. 6,8 10 In our series, the average cost (20,000 s) of each microvascular reconstruction in the head and neck region was comparable to previously published data (from 23,000 USD to 35,000 USD). 6,9,10 In breast reconstruction, our costs (9200 s) were somewhat higher than those published Preminger et al 11 but comparable to those described by Thoma et al (15,238 Canadian dollars, 9500s). 12 In 1999, Heinz and his coworkers published a comparable analysis of 107 microsurgical cases with carefully detailed cost allocations. 8 In their series, two factors, the length of hospital stay and

5 COST ANALYSIS OF MICROSURGERY AND FLAP MONITORING/SETÄLÄ ET AL the success of microsurgery, were identified as the most important factors affecting costs. The average cost for a successful reconstruction was 37,500 to 39,000 USD, and a successful reoperation for a postoperative thrombosis increased the average costs by 20,000 USD. If a flap was lost and a new one had to be created, the average total costs increased to more than 90,000 USD per patient. These figures are significantly higher than those presented here and may reflect differences in the patient material (trauma cases with long hospital stay, need for intensive care) and differences in the treatment protocols. However, these results and those of others 6 clearly demonstrate the financial impact of surgical safety and success. Another study, with a remarkably low frequency in surgical flap complications, 10 focused on frequent medical complications, which significantly increased the LOS and treatment costs by a mean of 20,000 USD. The complications were related to patient comorbidity, especially to tobacco and alcohol abuse, which postoperatively lead to increased use of the intensive care unit. The high flap survival was explained by the use of most reliable flap types and reliable sources of recipient vessels. Patient age, comorbidity, smoking and drinking habits, and nutrition affect the safety of microsurgery but are not factors that can be controlled by the physicians. Surgeons sometimes have the possibility to choose between a simple or delayed reconstruction and a complicated or immediate one, between pedicled flaps and microsurgery, but if the choice is determined solely by surgical safety, this may lead to a less favorable result in terms of complication rate, function, or cosmesis. Any efforts to increase surgical safety must focus on controlling those risk factors that can be modified. The risk of flap loss can be decreased by microsurgical training of the surgeon, so they can select the correct method for each patient and also ensure they have a proper reconstructive design and technique. 13 In addition, it is essential to train all other members of the perioperative team to provide patient preparation, anesthesia, postoperative patient care, and flap monitoring of good quality. It is wise to centralize the microsurgical activity so that in each unit, different types of reconstructions are conducted on a frequent basis so as to become routine procedures 14 and the staff is trained and experienced. Under such circumstances, flap monitoring is a routine safety measure, and simple clinical flap surveillance can result in a high success rate. 3 In 2001, we were compelled to find a new objective monitoring method for our flaps, because many flaps were lost due to delayed diagnosis of pedicle thrombosis. When we analyzed this problem, we felt that it was attributable to the small annual number of microsurgical patients, lack of centralization of their postoperative care within our hospital, and thus limited experience among the staff. After starting flap monitoring with MD, we were able to obtain reliable information for any flap. The information that glucose and lactate values were normal could be reliably obtained from any of the attending nurses, even inexperienced nurses and by phone. If the values started to deteriorate toward the set critical levels, then the surgeon on call made a clinical evaluation. This usually resulted in a reoperation to determine the cause of the deterioration. In this way, we were able to undertake early reoperations, often finding only a short plug of thrombosis in the anastomosis and having a reasonable time to arrange a new anastomosis without the risk of permanent tissue damage. In our series, the rate of reoperations is rather high. This can be explained by the fact that we had not been able to centralize the reconstructions but all our senior surgeons, four to five in all, shared the annual number of operations. Naturally, our junior surgeons wanted to participate in this activity. Therefore, we cannot compete with the high success rate of 98 to 99% reported elsewhere. 3,10 However, we think that a reliable monitoring method provides an advantage in promoting the prompt repair of failures with a high salvage rate. Using MD, we achieved an acceptable success rate (94%) with reasonable costs. There are no studies published about the costeffectiveness of any flap-monitoring method. In clinical monitoring, costs consist of the staff taking care of the patient and simultaneously conducting the surveillance of the flap. With special devices, one has to consider not only the purchase price but the maintenance, spare parts, and additional labor costs. One weakness of our study is that we were not able to analyze the labor costs that could be attributed to the monitoring. However, we estimated that the insertion of the catheter does not extend the duration of the operation and the sample analysis takes only 3 to 5 minutes per hour. The laboratory technician requires about 1 hour to start up and calibrate the analyzer. In our series, the estimated cost savings due to better monitoring were equal or higher than the cost of MD. So we can state that MD has proved its worth in our hands in economic perspective. This may not be the case in institutes where pedicle thrombosis is less common and clinical monitoring is well executed. The cost-effectiveness of a monitoring method is related to the cost of the investment and to the number of patients needed to monitor until one encounters the first patient with compromised perfusion. A monitoring method can also cause additional costs if it is not reliable a false-negative observation is potentially disastrous if the flap is lost, and a false-positive observation may lead to an unnecessary reoperation. The three unnecessary reoperations, one for misplaced catheter and two for a spasm, could have been avoided by following our basic protocol: keep the patient warm and free of pain and insert the catheter to a wellperfused part of the flap. In microsurgical literature, not much is discussed regarding how to diagnose and treat an arterial spasm. We have found that only a sensitive

6 JOURNAL OF RECONSTRUCTIVE MICROSURGERY monitoring method can visualize a spasm. In our hospital, we have been able to prevent an imminent perfusion failure in some cases by conservative means using active warming, pain medication, and intravenous fluids. A monitoring devise may be essential to reveal an early perfusion failure, transient or imminent, but good nursing and sound clinical and surgical judgment is still necessary to handle the situation. In this series of 99 patients, we have shown that the costs of microsurgical reconstructions are related to the complexity of the reconstruction and to the success of the reconstruction. In our hands, MD has proved to be a reliable and cost-effective tool in controlling the success of the flap transfer. This kind of special monitoring method may be useful in flaps with limited visual access, in cases with an intraoperative pedicle thrombosis, and in institutes where inexperienced personnel are responsible for treating microsurgical patients. REFERENCES 1. Jones NF. Intraoperative and postoperative monitoring of microsurgical free tissue transfers. Clin Plast Surg 1992;19: Kamolz LP, Giovanoli P, Haslik W, Koller R, Frey M. Continuous free-flap monitoring with tissue-oxygen measurements: three-year experience. J Reconstr Microsurg 2002; 18: ; discussion Disa JJ, Cordeiro PG, Hidalgo DA. Efficacy of conventional monitoring techniques in free tissue transfer: an 11-year experience in 750 consecutive cases. Plast Reconstr Surg 1999; 104: Setälä L, Papp A, Romppanen EL, Mustonen P, Berg L, Härmä M. Microdialysis detects postoperative perfusion failure in microvascular flaps. J Reconstr Microsurg 2006; 22: Thoma A, Strumas N, Rockwell G, McKnight L. The use of cost-effectiveness analysis in plastic surgery clinical research. Clin Plast Surg 2008;35: Miller MJ, Swartz WM, Miller RH, Harvey JM. Cost analysis of microsurgical reconstruction in the head and neck. J Surg Oncol 1991;46: de Bree R, Reith R, Quak JJ, Uyl-de Groot CA, van Agthoven M, Leemans CR. Free radial forearm flap versus pectoralis major myocutaneous flap reconstruction of oral and oropharyngeal defects: a cost analysis. Clin Otolaryngol 2007; 32: Heinz TR, Cowper PA, Levin LS. Microsurgery costs and outcome. Plast Reconstr Surg 1999;104: Petruzzelli GJ, Brockenbrough JM, Vandevender D, Creech SD. The influence of reconstructive modality on cost of care in head and neck oncologic surgery. Arch Otolaryngol Head Neck Surg 2002;128: Jones NF, Jarrahy R, Song JI, Kaufman MR, Markowitz B. Postoperative medical complications not microsurgical complications negatively influence the morbidity, mortality, and true costs after microsurgical reconstruction for head and neck cancer. Plast Reconstr Surg 2007;119: Preminger BA, Pusic AL, McCarthy CM, Verma N, Worku A, Cordeiro PG. How should quality-of-life data be incorporated into a cost analysis of breast reconstruction? A consideration of implant versus free TRAM flap procedures Plast Reconstr Surg 2008;121: Thoma A, Khuthaila D, Rockwell G, Veltri K. Cost-utility analysis comparing free and pedicled TRAM flap for breast reconstruction. Microsurgery 2003;23: Khouri RK. Avoiding free flap failure. Clin Plast Surg 1992; 19: Lin CS, Lee HC, Lin CT, Lin HC. The association between surgeon case volume and hospitalization costs in free flap oral cancer reconstruction operations. Plast Reconstr Surg 2008; 122: