Cost Effectiveness of Mesh Prophylaxis to Prevent Parastomal Hernia in Patients Undergoing Permanent Colostomy for Rectal Cancer

Transcription

1 Cost Effectiveness of Mesh Prophylaxis to Prevent Parastomal Hernia in Patients Undergoing Permanent Colostomy for Rectal Cancer Lawrence Lee, MD, MSc, Abdulaziz Saleem, MD, MSc, Tara Landry, MLIS, Eric Latimer, PhD, Prosanto Chaudhury, MD, MSc, FACS, Liane S Feldman, MD, FRCSC, FACS BACKGROUND: Parastomal hernia (PSH) is common after stoma formation. Studies have reported that mesh prophylaxis reduces PSH, but there are no cost-effectiveness data. Our objective was to determine the cost effectiveness of mesh prophylaxis vs no prophylaxis to prevent PSH in patients undergoing abdominoperineal resection with permanent colostomy for rectal cancer. STUDY DESIGN: Using a cohort Markov model, we modeled the costs and effectiveness of mesh prophylaxis vs no prophylaxis at the index operation in a cohort of 60-year-old patients undergoing abdominoperineal resection for rectal cancer during a time horizon of 5 years. Costs were expressed in 2012 Canadian dollars (CAD$) and effectiveness in quality-adjusted life years. Deterministic and probabilistic sensitivity analyses were performed. RESULTS: In patients with stage I to III rectal cancer, prophylactic mesh was dominant (less costly and more effective) compared with no mesh. In patients with stage IV disease, mesh prophylaxis was associated with higher cost (CAD$495 more) and minimally increased effectiveness (0.05 additional quality-adjusted life years), resulting in an incremental cost-effectiveness ratio of CAD$10,818 per quality-adjusted life year. On sensitivity analyses, the decision was sensitive to the probability of mesh infection and the cost of the mesh, and method of diagnosing PSH. CONCLUSIONS: In patients undergoing abdominoperineal resection with permanent colostomy for rectal cancer, mesh prophylaxis might be the less costly and more effective strategy compared with no mesh to prevent PSH in patients with stage I to III disease, and might be cost effective in patients with stage IV disease. (J Am Coll Surg 2014;218:82e91. Ó 2014 by the American College of Surgeons) Parastomal hernia (PSH) is a frequent complication after stoma creation. 1 Common symptoms of PSH include pain, bulging, poor cosmesis, and problems with stoma appliance fit, causing leakage and skin irritation. Disclosure Information: The Steinberg-Bernstein Centre for Minimally Invasive Surgery and Innovation is supported by an unrestricted educational grant from Covidien. Dr Feldman receives research and fellowship funding and is a paid consultant for Covidien and research support from Ethicon. All other authors have nothing to disclose. Dr Lee is supported by a scholarship from the Quebec Health Science Research Fund (FRQS) and the McGill Surgeon Scientist Program. Received July 7, 2013; Revised August 18, 2013; Accepted September 24, From the Steinberg-Bernstein Centre for Minimally-Invasive Surgery and Innovation (Lee, Feldman), Department of Surgery (Lee, Saleem, Chaudhury, Feldman), Montreal General Hospital Library (Landry), McGill University Health Centre, and Department of Epidemiology, Biostatistics and Occupational Health, McGill University (Latimer), Montreal, Quebec, Canada. Correspondence address: Liane S Feldman, MD, FRCSC, FACS, Steinberg-Bernstein Centre for Minimally-Invasive Surgery and Innovation, 1650 Cedar Ave, L9-303, Montreal, Quebec, Canada H3G 1A4. liane.feldman@mcgill.ca Although many patients with PSH can be treated conservatively, symptoms secondary to PSH can negatively affect quality of life of stoma patients. 2 Surgical repair might be required in patients with symptomatic PSH, but current surgical techniques are associated with a high recurrence rate, even if mesh is used. 3 Given the high incidence of PSH and difficulties in its treatment, several randomized trials have been performed to investigate prophylactic placement of mesh at the index operation to prevent PSH. 4-7 All studies report a substantial decrease in the incidence of PSH in patients that received prophylactic mesh placement, with no cases of mesh infection, intolerance, or fistula formation requiring mesh removal. These data have led some to suggest that prophylactic mesh placement should be performed routinely in all patients undergoing stoma formation. 8 However, before routine use, the cost effectiveness of this treatment strategy should be investigated. 9 There are several reasons why prophylactic mesh might not be cost effective, as the mesh used for PSH prophylaxis ª 2014 by the American College of Surgeons ISSN /13/$36.00 Published by Elsevier Inc. 82

2 Vol. 218, No. 1, January 2014 Lee et al Preventing Parastomal Hernia with Mesh 83 Abbreviations and Acronyms CAD$ ¼ Canadian dollars PSH ¼ parastomal hernia QALY ¼ quality-adjusted life years WTP ¼ willing to pay can be expensive, 10 and although PSH might be common, most patients are asymptomatic or can be treated conservatively. 11 As with any mesh, infection and erosion can occur and lead to considerable morbidity, 12 and this complication might not be apparent in randomized trials with short follow-up. In addition, many permanent colostomies (ie, as part of an abdominoperineal resection) are performed for malignancy, which can sufficiently decrease life expectancy to the point that PSH might not have time to develop. Therefore, the objective of this study was to determine the cost effectiveness of prophylactic mesh vs no prophylactic mesh placement to prevent PSH in patients undergoing abdominoperineal resection with permanent colostomy for rectal cancer. METHODS Model We performed a cost-effectiveness analysis using a cohort Markov model to simulate a cohort of 60-year-old patients with rectal cancer undergoing abdominoperineal resection and permanent colostomy (Fig. 1) during a time horizon of 5 years with 1-year cycles. We investigated 2 competing strategies that can impact the incidence of PSH: prophylactic mesh placement vs no mesh placement at the index operation. All patients undergoing elective abdominoperineal resection for rectal cancer enter the model (Fig. 1A), where patients might experience a postoperative death or survive and enter the cohort Markov model (Fig. 1B). All patients start in the no PSH state. With time, a PSH can develop and they would transition into the symptomatic PSH or asymptomatic PSH states. Because the incidence of PSH appears highest in the first year after stoma creation, 13 the probability of PSH developing was different between the first and subsequent years. Because length of follow-up in 2 of the 3 trials was only 12 months (and the majority of PSH occurred within the first 2 years in the third), the pooled results from the randomized trials were used for to determine the incidence of PSH for the first year only. After the first year, the yearly probability of PSH was derived from a 5-year follow-up study of a randomized trial by Janes and colleagues. 14 Patients in the asymptomatic PSH state might have symptoms develop and they transition into the symptomatic PSH state or remain asymptomatic. Patients in the symptomatic PSH state might experience mesh infection, obstruction secondary to PSH, undergo surgery to repair PSH, or remain in the symptomatic PSH state. Therefore, not all patients who have a PSH develop require operative repair, and only those with symptomatic PSH can undergo repair. The probability of requiring PSH repair was obtained from the pooled incidences reported in the randomized trials. 4,5,14 Although none of the 3 randomized trials reported mesh infection after prophylactic mesh placement, we incorporated a 2.3% probability of mesh infection after prophylactic placement or for PSH repair because of the relatively small sample sizes and short follow-up of the randomized trials. 3 We varied the range up to 22.5%, which was described to be the incidence of mesh infection in patients undergoing ventral hernia repair with concomitant bowel resection, for sensitivity analyses. 15 Mesh infections could be treated conservatively or by mesh removal, probabilities of which were derived from a study by Stremitzer and colleagues. 16 If the infected mesh required removal, it was assumed that suture repair would be performed, which was assigned a 69% probability of PSH recurrence. 3 All obstructions secondary to PSH could be managed operatively or conservatively (assumed to occur once yearly at maximum if managed conservatively). The probability of requiring operative management of an obstructive episode was obtained from pooled analysis from a review by Miller and colleagues specifically in colorectal patients. 17 We also assumed that all PSH undergoing repair were repaired electively using mesh (any approach or technique). After PSH repair, patients transition into the no PSH after repair state. For all states, patients might remain in their current state or die. Because our study population consisted of patients with rectal cancer, the analysis was stratified according to cancer stage. The yearly probability of death from cancer was derived from a large population-based rectal cancer registry. 18 The analysis was performed from the Canadian health care system (single payer) perspective. The discount rate used in base case analysis was 3%. 19 All costs were adjusted to 2012 Canadian dollars (CAD$) using the real health care inflation rate, specific for health care services and adjusted for population growth 20 (1 CAD$ ¼ 0.82 US$ in 2012, using purchasing power parity 21 ). The model was built and analyzed using TreeAge Pro 12 (TreeAge Software Inc.) or STATA 12 (StataCorp). Probabilities A systematic review was performed to identify studies comparing prophylactic mesh placement vs no mesh to prevent PSH in patients undergoing surgery in which a stoma is created. Studies were kept only if they were English-language randomized controlled trials comparing

3 84 Lee et al Preventing Parastomal Hernia with Mesh J Am Coll Surg Figure 1. Simplified schematic representing the decision between prophylactic mesh vs no mesh placement at the index operation. (A) Patients entering the model undergo abdominoperineal resection for rectal cancer. (B) If they survive, they enter into the cohort Markov model. APR, abdominoperineal resection; OR, operating room; PSH, parastomal hernia. prophylactic permanent mesh vs no mesh to prevent PSH in patients undergoing permanent colostomy. Three studies met the inclusion criteria (Table 1). Main model parameters were derived from these 3 studies (Table 2). Remaining model parameters were obtained from specific literature searches or by expert opinion. If more than one study reported a relevant probability parameter, the probabilities were pooled using a weighted randomeffects model. In cases of probability expressed over n-years (eg, 95.1% 5-year overall survival for stage I rectal cancer 18 ), yearly transition probabilities were calculated. 34 Perioperative mortality estimates for elective abdominoperineal resection for malignancy (CPT codes and 45395) and PSH repair (CPT code 44346) were obtained from the NSQIP database during 2005 to Costs All 3 randomized trials used a variation of a lightweight, partly absorbable mesh for prophylactic mesh placement. We imputed the cost of a cm lightweight, polypropylene mesh with an oxidized regenerated cellulose barrier (which was obtained from the manufacturer) as the cost of the mesh for both the prophylactic mesh placement and any subsequent PSH repairs (Table 2). The costs of abdominoperineal resection, PSH repair, and mesh removal were obtained from the Canadian Institute for Health Information case-mix groups, 25 which are similar to diagnosis-related groups. These costs represent the mean overall cost of the procedure, including the operation and subsequent hospitalization, not inclusive of the mesh. The costs of the mesh and physician

4 Vol. 218, No. 1, January 2014 Lee et al Preventing Parastomal Hernia with Mesh 85 Table 1. Characteristics of the Studies Used in Pooled Analysis to Estimate Probabilities of Outcomes First author Sample size, mesh/no mesh, n Mesh Mesh placement Surgical approach Parastomal hernia assessment Follow-up Janes 4 27/ cm large-pore, lightweight polypropylene and polyglactin (Vypro; Ethicon) Sublay* Open Physical examination 12 mo Lopez-Cano 5 19/ cm large-pore, lightweight polypropylene encapsulated with polydioxanone (Proceed; Ethicon) Serra-Aracil 6 27/ cm large-pore lightweight polypropylene and polyglecaprone (Ultrapro; Ethicon) Modified keyhole Laparoscopic CT 12 mo (intraperitoneal) 22 Sublay* Open Physical examination, y CT Median 29 mo All included studies were randomized trials comparing prophylactic mesh insertion with no prophylactic mesh insertion in patients undergoing permanent end colostomy. *Sublay was defined as between the rectus and the posterior fascia. y Hernias detected by physical examination were used to calculate pooled proportions. fees 26,27 were added to procedural costs, as necessary. No differences in operative time or materials were directly imputed because none of the randomized trials reported a difference in operation duration as a result of the mesh placement. 6 Rather, these costs were considered as part of the sensitivity analysis by varying the added cost of the mesh across a wide range of possible values (to account for possible costs related to additional operative time or equipment due to the mesh). The cost of conservative treatment of mesh infection was calculated based on the assumption that treatment would require 2 weeks of broad-spectrum antibiotics in addition to vacuumassisted closure (daily cost ¼ CAD$ ) for a median of 81 days, based on a study describing the treatment of mesh infections after ventral hernia repair by Stremitzer and colleagues. 16 It was assumed that infected meshes would be explanted only after failure of conservative therapy, therefore, the cost of infected mesh removal was calculated to be equivalent to the procedural cost in addition to conservative therapy. Utilities Quality-adjusted life years (QALY) are calculated by multiplying the utility of a health state by the time spent in that state. Utilities, also known as health-related quality-of-life weights, are weighted from 0.0 (death) to 1.0 (perfect health). The utility of a permanent colostomy was obtained from a study evaluating patient preferences from Smith and colleagues. 29 To our knowledge, there are no studies that report utilities specifically for PSH. Therefore, we used utilities generated from the difference in quality of life of patients with and without incisional hernia, 30 which were weighted on a scale of 0 to 1 using an algorithm. 35 We calculated the disutility of a mesh infection based on the difference in quality of life in patients who experienced a surgical site infection. 31 The utility of the hospitalization for obstruction secondary to PSH was obtained from utilities of nonsurgically managed hospitalized patients. 33 Sensitivity analyses We performed deterministic and probabilistic sensitivity analyses to account for parameter uncertainty. One-way deterministic analyses were performed by varying variables one at a time over a specified range of values. Probabilistic sensitivity analysis was performed by simulating the model 10,000 times (Monte Carlo simulation). In each simulation, results are recalculated using values for each model parameter that are chosen at random from the distribution fitted around each variable (Table 2). The 95% CIs were obtained from the 2.5 th and 97.5 th percentiles of the 10,000 simulations. Ranges were derived from the literature (directly or from pooled analysis) using the 95% CIs or from expert opinion. Beta distributions were fitted for probabilities and utilities, gamma distributions for costs. Given that one of the randomized trials (that of Lopez-Cano and colleagues 5 ) diagnosed PSH by means different from the other 2 (radiologically vs clinically) and, therefore, might have skewed the pooled proportions of PSH incidence, an analysis was performed excluding this study. RESULTS Results of the cost-effectiveness analysis for all rectal cancer stages are reported in Table 3. In patients with stages I to III rectal cancer, mesh prophylaxis was associated with substantially lower costs and more QALYs over 5 years compared with no mesh placement. The additional QALYs gained in patients with stages I, II, and III were equivalent to 117, 103, and 84 additional days

5 86 Lee et al Preventing Parastomal Hernia with Mesh J Am Coll Surg Table 2. Markov Model Inputs and Associated Distributions Used for Deterministic and Probabilistic Sensitivity Analyses Parameter Base case (range*) Distribution y References Probabilities Probability of PSH after prophylactic mesh placement, 1 st year z ( ) 0.08 x ( ) Beta: a 3, b 12 Beta: a 2, b 24 Probability of PSH after prophylactic mesh placement, 2 nd year 14 and beyond 0.01 ( ) Beta: a 1, b z (0.29e0.90) Probability of PSH if no mesh placed, 1 st year 0.44 x ( ) Beta: a 6, b 4 Beta: a 24, b Probability of PSH if no mesh placed, 2 nd year and beyond 0.09 ( ) Beta: a 7, b Probability of symptoms from PSH 0.59 ( ) Beta: a 5, b 3 23,24 Probability of obstruction secondary to PSH 0.18 ( ) Beta: a 1, b 4 24 Probability of obstruction requiring operation 0.22 ( ) Beta: a 33, b Probability of mesh infection ( ) Beta: a 8, b Probability of mesh explantation if mesh becomes infected 0.45 ( ) Beta: a 14, b Proportion of PSH undergoing repair after prophylactic mesh placement 0.11 ( ) Beta: a 1, b 8 5,6,14 Proportion of PSH undergoing repair if no mesh placed 0.23 ( ) Beta: a 10, b 33 5,6 Probability of recurrence after PSH repair 0.14 ( ) Beta: a 46, b Operative mortality after APR 0.02 ( ) Beta: a 17, b 1021 NSQIP Operative mortality after PSH repair ( ) Beta: a 1, b 333 NSQIP Yearly probability of death from stage I rectal cancer jj Yearly probability of death from stage II rectal cancer jj Yearly probability of death from stage III rectal cancer jj Yearly probability of death from stage IV rectal cancer jj Costs, CAD$ Lightweight, polypropylene mesh with cellulose barrier (15 15 cm) 850 (300 10,000) Gamma: a 1, b 850 Manufacturer Abdominoperineal resection 13,940 (5,000 30,000) Gamma: a 1, b ,26 Operation to repair PSH 4,683 (2,000 15,000) Gamma: a 1, b ,26 Operation to remove infected mesh 6,804 (3,000 20,000) Gamma: a 1, b ,27 Obstruction secondary to PSH 3,041 (1,500 10,000) Gamma: a 1, b Operation for obstruction 6,126 (3,000 20,000) Gamma: a 1, b ,26 Conservative treatment of mesh infection 19,252 (5,700 51,000) Gamma: a 1, b ,28 Utilities Permanent colostomy 0.84 ( ) Beta: a 185, b Disutility of a symptomatic PSH 0.08 ( ) Beta: a 4, b Disutility of a mesh infection 0.05 ( ) Beta: a 5, b 97 31,32 During hospitalization 0.42 ( ) Beta: a 4, b 6 33 *Used in deterministic (1- and 2-way) sensitivity analyses. y Used in probabilistic sensitivity analyses. z Pooled proportion using all 3 studies listed in Table 1. x Pooled proportion excluding study of Lopez-Cano and colleagues. 5 jj Not varied in deterministic or probabilistic sensitivity analyses. APR, abdominoperineal resection; CAD$, Canadian dollars; PSH, parastomal hernia. of perfect health, respectively. In these patients, prophylactic mesh placement was dominant (less costly and more effective). However, in patients with stage IV rectal cancer, mesh prophylaxis was associated with higher costs and more QALYs. This strategy cost CAD$495 more over 5 years than no mesh, and the incremental QALYs in patients with stage IV disease were minimal (þ0.05), equivalent to only 18 additional days of perfect health, resulting in an incremental cost-effectiveness ratio of CAD$10,818/QALY. On deterministic sensitivity analysis, the decision was sensitive to the cost of the mesh used for prophylaxis and the probability of mesh infection. None of the remaining variables changed the incremental costs or

6 Vol. 218, No. 1, January 2014 Lee et al Preventing Parastomal Hernia with Mesh 87 Table 3. Cost-Effectiveness Base Case Results Comparing Prophylactic Mesh Insertion vs No Prophylactic Mesh Insertion Incremental cost, CAD$ (95% CI) Incremental QALY (95% CI) ICER, CAD$/QALY Cost, CAD$ (95% CI) QALY (95% CI) Stage I Prophylactic mesh 15,869 (7,648 27,879) 3.78 ( ) 4,922 ( 18,234 to 93) þ0.32 ( ) Dominant No mesh 20,791 (9,840 39,068) 3.46 ( ) Stage II Prophylactic mesh 15,910 (7,522 27,532) 3.50 ( ) 4,182 ( 15,681 to 187) þ0.28 ( ) Dominant No mesh 20,092 (9,450 36,431) 3.22 ( ) Stage III Prophylactic mesh 15,811 (7,630 27,809) 3.14 ( ) 3,084 ( 11,999 to 384) þ0.23 ( ) Dominant No mesh 18,895 (9,032 33,566) 2.81 ( ) Stage IV Prophylactic mesh 15,282 (7,098 26,469) 1.55 ( ) þ495 ( 722 to 1,453) þ0.05 ( ) 10,818 No mesh 14,787 (6,619 26,043) 1.50 ( ) CAD$, Canadian dollars; ICER, incremental cost-effectiveness ratio; QALY, quality-adjusted life years. effectiveness substantially, including the cost of mesh infection. Threshold analysis reported that the mesh used for prophylaxis needed to cost less than CAD$4,556, CAD$3,584, CAD$3,123, and CAD$601 for patients with stages I, II, III, and IV cancer, respectively, to remain the less costly option over 5 years. The interaction between the cost of the mesh and the probability of mesh infection demonstrated that only in cases where both the mesh cost and the probability of mesh infection are high, does the prophylactic mesh strategy become more costly for patients with stage I cancer (Fig. 2). On probabilistic sensitivity analysis, 96%, 96%, and 93% of all simulations reported that prophylactic mesh was less costly and more effective in patients with stage I, II, III cancer, respectively (Fig. 3). The cost-effectiveness acceptability curve, which reports the probability of cost Figure 2. Results of 2-way sensitivity analysis demonstrating the effect of co-varying the cost of the prophylactic mesh and probability of mesh infection in patients with stage I rectal cancer. $CAN, Canadian dollars. effectiveness at varying willing to pay (WTP) thresholds (ie, the amount a decision maker is willing to pay to gain an additional unit of effectiveness), demonstrates a 99% probability of cost effectiveness for patients with stages I to III cancer, and a 91% probability for patients with stage IV cancer at a WTP of CAD$50,000/QALY, which is a conservative WTP estimate (Fig. 4A). The exclusion of the study by Lopez-Cano and colleagues did not change the incremental cost-effectiveness ratio for stages I to III, but increased the incremental cost-effectiveness ratio in stage IV to CAD$15,636/QALY. However, the uncertainty around the results increased, as the probability of cost effectiveness at a WTP of CAD$50,000/QALY was reduced to 77% for stage I to 65% for stage IV (Fig. 4B). DISCUSSION In an era of increasing financial constraints for health care, it is important to establish the cost effectiveness of medical and surgical technologies and interventions before widespread adoption. We estimated the cost effectiveness of mesh prophylaxis at the index operation compared with no mesh in patients undergoing abdominoperineal resection with permanent colostomy for rectal cancer using a cohort Markov model. The model suggested that prophylactic mesh can be associated with lower costs and increased QALYs compared with no mesh for patients with stages I to III cancer, and is potentially cost effective in patients with stage IV disease. There is increasing evidence to support the use of mesh prophylaxis to prevent PSH, as multiple randomized trials have demonstrated to date. Meta-analyses of these trials have reported substantial reductions in PSH in patients who underwent prophylactic mesh placement at the index operation without important mesh-related

7 88 Lee et al Preventing Parastomal Hernia with Mesh J Am Coll Surg Figure 3. Plot of incremental costs and quality-adjusted life years (QALYs) of mesh prophylaxis vs no mesh. Each point represents the result of one simulation trial. $CAN, Canadian dollars. adverse events. 36,37 However, none of these studies have addressed the cost effectiveness of this strategy. 8,9 Figel and colleagues 38 performed a value analysis of mesh prophylaxis using biologic mesh and reported that this strategy would be less costly only if a large number of PSH repairs are prevented, or if the cost of the mesh was low. However, their study was a small case series with 18 patients with biologic mesh, for which evidence is poor. This study is the first to perform a formal cost-effectiveness analysis of mesh prophylaxis using permanent mesh to prevent PSH in patients undergoing permanent colostomy. Similar to the previous study, we also found that the cost of the mesh was one of the most important variables affecting the cost-effectiveness results. We limited our analysis to a time horizon of 5 years because there were no long-term studies of mesh prophylaxis with follow-up longer than 5 years. The incidence of PSH is highest in the first few years, but can occur up to 20 years after stoma creation. 1 It is unknown whether mesh prophylaxis alters the long-term rate of PSH development, as only Janes and colleagues have reported 5-year data of their randomized trial. 14 We performed the analysis from the perspective of the Canadian health care system, which is a single-payer system. Differences in the health care systems and costs between Canada and other countries 39 might limit the generalizability. However, sensitivity analyses demonstrated that the decision was insensitive to large variations in costs. Sensitivity analysis also demonstrated that the decision was subject to uncertainty due to the variability in the method to detect PSH. Radiologic diagnosis using CT is more sensitive than physical examination, which can lead to increased diagnoses of clinically insignificant PSH. When we excluded the study that diagnosed PSH solely by CT, 5 there was an important increase in the uncertainty around the results (Fig. 4B), suggesting that a more conservative interpretation of the results is warranted. Although certainly other postoperative complications exist, we only considered mesh infection in the model. Although randomized trials did not report any differences Figure 4. Cost-effectiveness acceptability curve of prophylactic mesh vs no mesh placement to prevent parastomal hernia, stratified by rectal cancer stage. (A) Including all studies listed in Table 1. (B) Excluding study by Lopez-Cano and colleagues. 5 QALY, quality-adjusted life years.

8 Vol. 218, No. 1, January 2014 Lee et al Preventing Parastomal Hernia with Mesh 89 in stoma-related complications between the 2 groups, 4-6 we chose to impute the pooled probability of mesh infection of primary PSH repair derived from a systematic review of PSH mesh repair techniques (2.3%). 3 Small trials might be underpowered to detect rare adverse events, and concern persists about potential mesh infection in the presence of bowel surgery. The upper limit of the tested range (22.5%) included the reported incidences of mesh infection in patients with high-risk comorbidities (16% 40 ) and infection-prone mesh material (14% for PTFE meshes 41 ). Despite the risk and high costs associated with mesh infection, mesh prophylaxis remained a cost-effective option. Mesh prophylaxis became more costly only at extreme values of risk of mesh infection or mesh cost. Several studies have also investigated the use of biologic mesh for PSH prophylaxis 7 or repair 10 ; however, the effectiveness of biologic materials remains unknown, as most studies lack a control group or did not detect a difference. Even if equal effectiveness of biologic mesh compared with synthetic mesh (along with minimal mesh infections) was assumed, then it remains likely that biologic mesh would still be the more costly option, given the relationship between these 2 variables (Fig. 2) and the fact that biologics would likely exceed the threshold costs of the mesh. Multiple risk factors for PSH have been identified, including patient and operative factors. 8 One of the limitations of cohort Markov models is the inability to account for individual patient risk factors in the analysis. 42 A meta-analysis of mesh prophylaxis in patients at high risk of incisional hernia after laparotomy reported a significantly reduced incidence of incisional hernia without mesh-specific complications. 43 If these results can be extrapolated to PSH, then mesh prophylaxis in high-risk patients is likely to remain cost effective, although this remains to be investigated. The results of this study have to be interpreted in view of several limitations. As with any model-based economic evaluation, inputs for model parameters are based on data from other studies, and can, therefore, only be as reliable as the results from those studies. A large number of model parameters were derived from pooled data of 3 randomized trials that consisted of relatively small sample sizes and short follow-up, which might not detect rare outcomes such as mesh-related complications or longterm recurrences. Economic modeling represents the synthesis and analysis of the existing literature to inform the decision maker of the best treatment given what is currently known. Finally, there were no studies that reported utilities for PSH or mesh infection. The imputed disutility of mesh infection might underestimate the true effect, considering that it was obtained from a quality-of-life score in patients with surgical site infections rather than true mesh infections. However, results were insensitive to the wide range of values on sensitivity analyses, suggesting that this limitation had little impact. In addition, studies that have investigated quality of life in patients affected by PSH have used the Stoma-QoL instrument, 2,44 and there is no algorithm to transform this measure into the utility scale (0.0 to 1.0). Therefore, we extrapolated the impact of incisional hernia on quality of life 30 to PSH. The resulting disutility ( 0.08) from this method was similar to the negative quality-of-life impact due to peristomal bulging of a sigmoid colostomy on the Stoma-QoL reported by Kald and colleagues (3.2 points lower in patients with peristomal bulging compared with those without bulging, out of a total possible score of 100 points). 2 CONCLUSIONS Modeling suggests that mesh prophylaxis can be a costeffective strategy to prevent PSH in patients undergoing abdominoperineal resection and permanent colostomy for rectal cancer, as it is likely to be less costly and more effective than no mesh for stages I to III disease, and might be cost effective for stage IV disease, but this decision is subject to some uncertainty. However, mesh prophylaxis might not be cost effective if the cost of the mesh and the incidence of mesh infection are high. Author Contributions Study conception and design: Lee, Latimer, Chaudhury, Feldman Acquisition of data: Lee, Saleem, Landry, Chaudhury Analysis and interpretation of data: Lee, Saleem, Landry, Latimer, Chaudhury, Feldman Drafting of manuscript: Lee, Saleem, Landry Critical revision: Latimer, Chaudhury, Feldman REFERENCES 1. Carne PW, Robertson GM, Frizelle FA. Parastomal hernia. Br J Surg 2003;90:784e Kald A, Juul KN, Hjortsvang H, Sjodahl RI. Quality of life is impaired in patients with peristomal bulging of a sigmoid colostomy. Scand J Gastroenterol 2008;43:627e Hansson BM, Slater NJ, van der Velden AS, et al. Surgical techniques for parastomal hernia repair: a systematic review of the literature. Ann Surg 2012;255:685e Janes A, Cengiz Y, Israelsson LA. Randomized clinical trial of the use of a prosthetic mesh to prevent parastomal hernia. Br J Surg 2004;91:280e Lopez-Cano M, Lozoya-Trujillo R, Quiroga S, et al. Use of a prosthetic mesh to prevent parastomal hernia during laparoscopic abdominoperineal resection: a randomized controlled trial. Hernia 2012;16:661e667.

9 90 Lee et al Preventing Parastomal Hernia with Mesh J Am Coll Surg 6. Serra-Aracil X, Bombardo-Junca J, Moreno-Matias J, et al. Randomized, controlled, prospective trial of the use of a mesh to prevent parastomal hernia. Ann Surg 2009;249: 583e Hammond TM, Huang A, Prosser K, et al. Parastomal hernia prevention using a novel collagen implant: a randomised controlled phase 1 study. Hernia 2008;12:475e Hotouras A, Murphy J, Thaha M, Chan CL. The persistent challenge of parastomal herniation: a review of the literature and future developments. Colorectal Dis 2013;15: e202ee Temple LK, Buie WD, Ellis CN, et al. Canadian Association of General Surgeons, the American College of Surgeons, the Canadian Society of Colorectal Surgeons, and the American Society of Colon and Rectal Surgeons: evidence based reviews in surgery-colorectal surgery. Dis Colon Rectum 2011;54: 1202e Slater NJ, Hansson BM, Buyne OR, et al. Repair of parastomal hernias with biologic grafts: a systematic review. J Gastrointest Surg 2011;15:1252e Bafford AC, Irani JL. Management and complications of stomas. Surg Clin North Am 2013;93:145e Leber GE, Garb JL, Alexander AI, Reed WP. Long-term complications associated with prosthetic repair of incisional hernias. Arch Surg 1998;133:378e Londono-Schimmer EE, Leong AP, Phillips RK. Life table analysis of stomal complications following colostomy. Dis Colon Rectum 1994;37:916e Janes A, Cengiz Y, Israelsson LA. Preventing parastomal hernia with a prosthetic mesh: a 5-year follow-up of a randomized study. World J Surg 2009;33:118e121; discussion El-Gazzaz GH, Farag SH, El-Sayd MA, Mohamed HH. The use of synthetic mesh in patients undergoing ventral hernia repair during colorectal resection: risk of infection and recurrence. Asian J Surg 2012;35:149e Stremitzer S, Bachleitner-Hofmann T, Gradl B, et al. Mesh graft infection following abdominal hernia repair: risk factor evaluation and strategies of mesh graft preservation. A retrospective analysis of 476 operations. World J Surg 2010;34: 1702e Miller G, Boman J, Shrier I, Gordon PH. Natural history of patients with adhesive small bowel obstruction. Br J Surg 2000;87:1240e Pahlman L, Bohe M, Cedermark B, et al. The Swedish rectal cancer registry. Br J Surg 2007;94:1285e Weinstein MC, Siegel JE, Gold MR, et al. Recommendations of the panel on cost-effectiveness in health and medicine. JAMA 1996;276:1253e Canadian Institute of Actuaries. Health Care Trend Rate. Available at: /212031e.pdf. Accessed January 23, Organisation for Economic Co-Operation and Development PPP Benchmark Results. Available at: org/index.aspx?datasetcode¼cpl. Accessed January 6, Lopez-Cano M, Lozoya-Trujillo R, Espin-Basany E. Prosthetic mesh in parastomal hernia prevention. Laparoscopic approach. Dis Colon Rectum 2009;52:1006e Cheung MT. Complications of an abdominal stoma: an analysis of 322 stomas. ANZ J Surg 1995;65:808e Ripoche J, Basurko C, Fabbro-Perray P, Prudhomme M. Parastomal hernia. A study of the French federation of ostomy patients. J Visc Surg 2011;148:e435ee Canadian Institute for Health Information. Case mix groups (CMGþ). Available at: Accessed March 30, Régie de l assurance maladie Québec. Manuel de facturation. Available at: medecins-specialistes/manuels/pages/facturation.aspx. Accessed May 1, Canadian Institute for Health Information. National Physician Database, data release. Available at: ca/estore/productfamily.htm?pf¼pfc2032&lang¼en&media¼0. Accessed April 2, Canadian Coordinating Office for Health Technology Assessment. Vacuum assisted wound closure therapy. Available at: Accessed July 2, Smith DM, Sherriff RL, Damschroder L, et al. Misremembering colostomies? Former patients give lower utility ratings than do current patients. Health Psychol 2006;25: 688e van Ramshorst GH, Eker HH, Hop WC, et al. Impact of incisional hernia on health-related quality of life and body image: a prospective cohort study. Am J Surg 2012;204: 144e Perencevich EN, Sands KE, Cosgrove SE, et al. Health and economic impact of surgical site infections diagnosed after hospital discharge. Emerg Infect Dis 2003;9:196e Colavita PD, Tsirline VB, Belyansky I, et al. Prospective, long-term comparison of quality of life in laparoscopic versus open ventral hernia repair. Ann Surg 2012;256:714e722; discussion Hays RD, Eastwood JA, Kotlerman J, et al. Health-related quality of life and patient reports about care outcomes in a multidisciplinary hospital intervention. Ann Behav Med 2006;31:173e Fleurence RL, Hollenbeak CS. Rates and probabilities in economic modelling: transformation, translation and appropriate application. Pharmacoeconomics 2007;25:3e Ara R, Brazier J. Deriving an algorithm to convert the eight mean SF-36 dimension scores into a mean EQ-5D preference-based score from published studies (where patient level data are not available). Value Health 2008; 11:1131e Shabbir J, Chaudhary BN, Dawson R. A systematic review on the use of prophylactic mesh during primary stoma formation to prevent parastomal hernia formation. Colorectal Dis 2012; 14:931e Wijeyekoon SP, Gurusamy K, El-Gendy K, Chan CL. Prevention of parastomal herniation with biologic/composite prosthetic mesh: a systematic review and meta-analysis of randomized controlled trials. J Am Coll Surg 2010;211: 637e Figel NA, Rostas JW, Ellis CN. Outcomes using a bioprosthetic mesh at the time of permanent stoma creation in preventing a parastomal hernia: a value analysis. Am J Surg 2012;203:323e326; discussion Organisation for Economic Co-operation and Development. Health Expenditure and Financing. Available at: oecd.org/index.aspx?datasetcode¼health_stat. Accessed October 8, Krpata DM, Blatnik JA, Novitsky YW, Rosen MJ. Evaluation of high-risk, comorbid patients undergoing open ventral hernia repair with synthetic mesh. Surgery 2013;153:120e125.

10 Vol. 218, No. 1, January 2014 Lee et al Preventing Parastomal Hernia with Mesh Brown RH, Subramanian A, Hwang CS, et al. Comparison of infectious complications with synthetic mesh in ventral hernia repair. Am J Surg 2013;205: 182e Caro JJ. Pharmacoeconomic analyses using discrete event simulation. Pharmacoeconomics 2005;23:323e Nachiappan S, Markar S, Karthikesaligam A, et al. Prophylactic mesh placement in high-risk patients undergoing elective laparotomy: a systematic review. World J Surg 2013;37:1861e Scarpa M, Ruffolo C, Boetto R, et al. Diverting loop ileostomy after restorative proctocolectomy: predictors of poor outcome and poor quality of life. Colorectal Dis 2010;12:914e920.