KNEE
The economics of using prophylactic antibiotic-loaded bone cement in total knee replacement C. J. Gutowski, B. M. Zmistowski, C. T. Clyde, J. Parvizi From Thomas Jefferson University Hospital, Philadelphia, Pennsylvania, United States
The rate of peri-prosthetic infection following total joint replacement continues to rise, and attempts to curb this trend have included the use of antibiotic-loaded bone cement at the time of primary surgery. We have investigated the clinical- and cost-effectiveness of the use of antibiotic-loaded cement for primary total knee replacement (TKR) by comparing the rate of infection in 3048 TKRs performed without loaded cement over a three-year period versus the incidence of infection after 4830 TKRs performed with tobramycin-loaded cement over a later period of time of a similar duration. In order to adjust for confounding factors, the rate of infection in 3347 and 4702 uncemented total hip replacements (THR) performed during the same time periods, respectively, was also examined. There were no significant differences in the characteristics of the patients in the different cohorts. The absolute rate of infection increased when antibiotic-loaded cement was used in TKR. However, this rate of increase was less than the rate of increase in infection following uncemented THR during the same period. If the rise in the rate of infection observed in THR were extrapolated to the TKR cohort, 18 additional cases of infection would have been expected to occur in the cohort receiving antibiotic-loaded cement, compared with the number observed. Depending on the type of antibiotic-loaded cement that is used, its cost in all primary TKRs ranges between USD $2112.72 and USD $112 606.67 per case of infection that is prevented. Cite this article: Bone Joint J 2014;96-B:65–9.
C. J. Gutowski, MD, MPH, Orthopaedic Surgery Resident Thomas Jefferson University, Department of Orthopaedic Surgery, 1025 Walnut St, 516 College Building, Philadelphia, Pennsylvania 19107, USA. B. M. Zmistowski, BA, Medical Student C. T. Clyde, BA, Medical Student Thomas Jefferson University, 1025 Walnut St, 516 College Building, Philadelphia, Pennsylvania 19107, USA. J. Parvizi, MD, PhD, Professor of Orthopaedic Surgery Rothman Institute at Thomas Jefferson University, Sheridan Building 10th Floor, 125 South 9th St., Philadelphia, Pennsylvania 19017, USA. Correspondence should be sent to Dr C. J. Gutowski; e-mail: [email protected] ©2014 The British Editorial Society of Bone & Joint Surgery doi:10.1302/0301-620X.96B1. 31428 $2.00 Bone Joint J 2014;96-B:65–9. Received 7 December 2012; Accepted after revision 28 August 2013
Peri-prosthetic infection following total knee replacement (TKR) is a serious complication that leads to high morbidity for patients, technically demanding revision surgery with a complicated peri-operative course and high costs. Rates of revision TKR are projected to rise significantly in the future decades, 1 with one study estimating a 66% increase between 2005 and 2030, with hospital costs for these procedures expected to exceed US$2 billion.2 Revision procedures performed because of infection are associated with higher costs than those performed for aseptic failure. 2 As result of this burden, several strategies to combat this trend are under scrutiny. Intra-articular biomaterials are risk factors for bacterial contamination and subsequent infection.3 Polymethylmethacrylate (PMMA) cement carries a particularly high risk for colonisation by bacteria compared with other materials such as stainless steel and polyethylene.4 Given this, the antibiotic impregnation of cement is a logical aspect of prophylaxis and has proved effective in treating an established infection at revision surgery.3
VOL. 96-B, No. 1, JANUARY 2014
Low-dose antibiotic-loaded bone cement, generally accepted to be 1 g of antibiotic per 40 g of PMMA,5 has been approved by the United States Food and Drug Administration (FDA) for use during the second stage of a revision joint replacement after a previously established infection.3 Low-dose antibiotic loaded PMMA is not currently approved by the FDA for use in primary joint replacement; however, orthopaedic surgeons around the world routinely use it during these operations.6 Several premixed formulations combining different concentrations of antibiotics with PMMA are commercially available, but in the United States only single-antibiotic preparations using gentamicin or tobramycin are licensed. Although evidence in support of low-dose antibiotic-loaded PMMA as prophylaxis against infection is well established,7-9 many concerns remain regarding its widespread use. These include a risk of increased antibiotic toxicity and allergic reactions, and the possibility that the mechanical strength of the loaded PMMA may be compromised compared with that of unadulterated cement.10 However, these potential pitfalls have either been refuted 65
66
C. J. GUTOWSKI, B. M. ZMISTOWSKI, C. T. CLYDE, J. PARVIZI
or not proven.11,12 Although the potential risks are minor, there are two other major concerns that have yet to be fully addressed. Some fear the development of antimicrobial resistance, although this concept remains unproven. They reason that exposure of intra-articular bacteria to low levels of antibiotic could encourage resistance through mutation of the bacterial genotype.13 Also, in the current cost-conscious healthcare environment the second major concern relates to the scrutiny that surrounds the financial implications of treatment. The current cost of premixed antibiotic-loaded PMMA is between $200 and $500 per 40 g pack, with two packets generally being used per TKR.14 Jiranek et al3 estimated an increase of $117 million to the annual national healthcare costs in the United States if half of the primary total joint replacements included antibioticloaded cement. This does not take into account the increase in case load discussed earlier. Given these projections, surgeons are under pressure to limit the costs, and many are wary of the additional expense associated with including antibiotic-loaded PMMA at the time of TKR. The additional initial cost of incorporating antibioticloaded PMMA into a primary TKR protocol must be weighed against the cost savings that would be subsequently realised as a result of lowering the rate of infection, which is in turn contingent upon the rate of prevention of infection being directly attributable to the use of antibioticloaded PMMA. This study examined the possible contribution of antibiotic-loaded PMMA to the reduction in the rate of infection after TKR, and explored the financial cost–benefit analysis of its routine use.
Patients and Methods Following institutional review board approval, we undertook a retrospective cohort study comparing the rate of peri-prosthetic infection identified during two periods, first from 1 January 2000 to 31 December 2002, during which no primary joint replacement received antibiotic-loaded PMMA at our institution; and secondly after 2003, when loaded PMMA (Simplex P with tobramycin; Stryker Orthopaedics, Kalamazoo, Michigan) was introduced. Primary total hip replacements (THRs) continued to be performed without PMMA. By the end of 2003, all surgeons at our institution were using PMMA with added tobramycin during routine primary TKR. The second period of the study therefore ran from 1 January 2004 to 31 December 2007. We recorded the number of joint replacements that were undertaken during these periods, as well as the patients’ age, gender, and the Deyo modification15 of the Charlson comorbidity index (CCI),16 which was used to assess the patient’s overall health status. All patients presenting with infection following joint replacement were identified using Musculoskeletal Infection Society (MSIS) methodology.17 This defines infection as the growth of identical organisms in two separate periprosthetic tissue or fluid cultures; a draining sinus tract; or four of the following six criteria: 1) elevated erythrocyte
sedimentation rate (ESR) and C-reactive protein (CRP) levels; 2) elevated white blood cell (WBC) count in the synovial fluid; 3) elevated neutrophil percentage in the synovial fluid; 4) positive histological analysis; 5) intra-articular purulence; and 6) a single positive peri-prosthetic tissue or fluid culture. As histological analysis was not routinely performed at this institution, this study required three out of five criteria to define an infection. After identifying patients who potentially had an infection, case notes were reviewed to ensure that they met the definition. The timing of the presentation of the infection was also noted. The number of patients who underwent TKR in whom infection was prevented was determined by comparing the observed incidence to the expected incidence. It has been shown that over the past two decades the rate of infection after TKR has risen at a rate closely matching that after THR.18 As this was a retrospective study, the expected rate of infection after TKR was calculated by extrapolating the observed rise in infection following THR over the same time period. We propose that this value represents the increasing background rate of infection due to any other confounding factors. The cost of preventing a single case of infection by the addition of antibiotic-loaded PMMA was calculated using the cost of using it for the entire cohort and subtracting the baseline cost of using unadulterated PMMA in all patients, divided by the number of infections prevented. We investigated four notional combinations of PMMA and antibiotics. Model 1 consisted of this institution’s current operative protocol, using two packets of pre-mixed Simplex P with tobramycin ($420 in additional cost per joint, since each of the two commercially mixed bags of Simplex P cost $250/ bag, as opposed to $40/bag if unadulterated PMMA had been used). Model 2 consisted of hand-mixing two packets of PMMA with two 1 g vials of vancomycin ($7.88 in additional cost per joint, as each gram of vancomycin costs $3.94). Model 3 consisted of hand-mixing two packets of PMMA with two 1 g vials of tobramycin ($138.66 per joint, as 1 g of tobramycin costs $69.33). Model 4 consisted of hand-mixing two packets of PMMA with 1 g (one vial) of tobramycin and 1 g (one vial) of vancomycin ($73.27 in additional cost). We justify this because these two antibiotics have been shown to act synergistically by improving elution from the cement and prolonging the antimicrobial activity.19 Statistical analysis. A Cochran–Mantel–Haenszel test was used to compare the two cohorts (before and after the use of antibiotic-loaded PMMA) while adjusting for changes in the rate of infection due to confounders represented by the incidence of infection after THR. For other measures, continuous variables were analysed with a Student’s t-test and categorical variables were measured with chi-squared analysis. Statistical analysis was performed with SPSS 16.0 (SPSS Inc., Chicago, Illinois). A p-value of < 0.05 was considered statistically significant. THE BONE & JOINT JOURNAL
THE ECONOMICS OF USING PROPHYLACTIC ANTIBIOTIC-LOADED BONE CEMENT IN TOTAL KNEE REPLACEMENT
67
Table I. Demographic information for study sample (PMMA, polymethylmethacrylate)
Admissions (n) Joint replacements (n) Mean age (yrs) (range) Female gender (n, %) Median Charlson index (range)
Before use of antibiotic-loaded PMMA (1 January 2000 to 31 December 2002)
After introduction of antibiotic-loaded PMMA (1 January 2004 to 31 December 2007)
Knee
Hip
Knee
Hip
2390 3048 65.9 (14 to 95) 1573 (66) 2 (0 to 8)
3054 3347 62.4 (13 to 99) 1605 (53) 2 (0 to 9)
4060 4830 65.7 (28 to 97) 2611 (64) 2 (0 to 8)
4433 4702 62.1 (11 to 98) 2239 (51) 2 (0 to 8)
Table II. Observed risk of developing peri-prosthetic joint infection (PJI) in total hip (THR) and total knee replacement (TKR) at various post-operative time points, with all TKRs (no THRs) receiving antibiotic-loaded cement (ALBC) after 1 January 2004 Developed PJI (n, %)
THR Prior to ALBC (1 January 2000 to 31 December 2002) After ALBC in TKR (1 January 2004 to 31 December 2007) TKR Prior to ALBC (1 January 2000 to 31 December 2002) After ALBC in TKR (1 January 2004 to 31 December 2007)
Total joints
Three months
Six months
Two years
3347 4701
6 (0.18) 18 (0.38)
7 (0.21) 19 (0.40)
11 (0.33) 25 (0.53)
3048 4826
6 (0.20) 17 (0.35)
10 (0.33) 19 (0.39)
23 (0.75) 40 (0.83)
Table III. Profile of infecting organisms in confirmed cases of peri-prosthetic joint infection (PJI), by joint and time period (ALBC, antibiotic-loaded bone cement; MR-/MSSA, methicillin-resistant/-sensitive Staphylococcus aureus) Prior to the use of ALBC (1 January 2000 to After the introduction of ALBC (1 January 31 December 2002) 2004 to 31 December 2007) Hip (n = 11)
Knee (n = 40)
Hip (n = 25)
Gram-positive 15 (65) Staphylococcus 13 (57) Staphylococcus aureus 10 (43) MRSA 7 (30) MSSA 3 (13) Coagulase-negative Staphylococcus 3 (13) Streptococcus 3 (13) Enterococcus 0
Knee (n = 23)
10 (91) 8 (73) 7 (64) 5 (45) 2 (18) 1 (9) 1 (9) 1 (9)
32 (80) 26 (65) 15 (38) 6 (15) 9 (23) 11 (28) 5 (13) 2 (5)
23 (92) 22 (88) 19 (76) 13 (52) 6 (24) 4 (16) 2 (8) 1 (4)
Gram-negative Escherichia coli Proteus mirabilis
3 (13) 0 0
1 (9) 0 0
6 (15) 3 (8) 1 (3)
3 (12) 2 (8) 1 (4)
Polymicrobial
1 (4.3)
0
2 (5)
5 (20)
Culture negative
5 (22)
0
3 (8)
0
Results No significant differences were found in patient demographics that could affect the rate of infection between the four cohorts (THR vs TKR, pre-antibiotic-loaded PMMA vs antibiotic-loaded PMMA in TKR) (Table I). Between the two study periods, the rate of infection within two years of THR increased by 60%, from 0.33% (11 of 3347) to 0.53% (25 of 4701). The rate of infection within two years of TKR increased by 10.7%, from 0.75% (23 of 3048) to 0.83% (40 of 4826) after the introduction of antibiotic-loaded PMMA. VOL. 96-B, No. 1, JANUARY 2014
Similar findings were noted three and six months after surgery (Table II). The expected rate of infection in TKRs at two years post-operatively was 1.2% (calculated by applying the observed 60% increase in THRs), which corresponds to 58 cases of infection in the antibiotic-loaded cohort. Compared with the observed value of 40 patients who had infection (p = 0.40), the addition of antibioticloaded PMMA was associated with 18 prevented cases of infection. This is also seen in the data at three months (p = 0.06) and six months (p = 0.23) post-operatively.
68
C. J. GUTOWSKI, B. M. ZMISTOWSKI, C. T. CLYDE, J. PARVIZI
Table IV. Cost analysis of four different forms of antibiotic-loaded polymethylmethacrylate (PMMA) in total knee replacement (TKR)
Added cost per TKR ($) Total cost of cohort prophylaxis ($) (4826 knees) Cost per case of infection prevented ($)
Model 1: Simplex P
Model 2: Hand-mixed Model 3: Hand-mixed Model 4: Hand-mixed vancomycin/ vancomycin tobramycin tobramycin
420 2 026 920.00 112 606.67
7.88 38 028.88 2112.72
The infecting organisms were also investigated (Table III), but the numbers were too small to allow any statistically supported conclusion to be reached. The number of patients in whom infection after TKR was prevented by the use of antibiotic PMMA at two years (n = 18) was then used to calculate the cost per case of preventing infection, based on the additional costs associated with each of the models of antibiotic-loaded PMMA (Table IV). Simplex P with tobramycin provided the most expensive model, with a cost to prevent infection of > $100 000 per case. This is significantly higher than the other three models, with hand-mixed vancomycin-loaded PMMA having the lowest cost at $2112.72 associated with the prevention of one case.
Discussion There are reports that support the use of low-dose antibioticloaded PMMA as prophylaxis against infection following joint replacement.7 Chiu et al20 showed a statistically significant reduction in the rate of deep infection after TKR when using antibiotic-loaded PMMA compared with the use of unadulterated cement (0% vs 3.1%, respectively, p = 0.0238). Josefsson et al8 showed a significant difference (p < 0.05) in post-operative infection rate between patients undergoing primary THR using gentamicin-loaded bone cement (0.8% infection rate) compared with those using systemic antibiotics only (1.9% infection rate). Engesaeter et al,9 in a paper from The Norwegian Arthroplasty Registry, surveyed primary THRs performed between 1987 and 2001 and found that systemic antibiotic prophylaxis was associated with a 1.8-fold higher risk of infection than systemic antibiotics in conjunction with antibiotic-loaded PMMA. They also found a similar benefit from the use of loaded PMMA in primary TKR.9 The costs incurred by a hospital treating infection after joint replacement are difficult to quantify. Our data are patient and institution specific, and will vary significantly depending on the form of surgery which is undertaken. The mean charge for revision TKR secondary to infection was found by Lavernia et al2 to be $109 805. Revision arthroplasty because of infection has been reported to be $60 000 more expensive than revision for aseptic loosening or mechanical failure.21,22 Other studies have examined the cost of treating infected joint replacements and focused on economic differences that exist between resistance patterns and treatment methodologies. Parvizi et al21 reported that
138.66 669 173.16 37 176.29
73.27 353 601.02 19 644.50
the mean cost of treating infection caused by methicillinresistant organisms was $97 651, compared with $65 470 for infections caused by methicillin-sensitive organisms. The cost of preventing infection after joint replacement must be weighed against the cost of treating it. We found that the cost of preventing a single case of infection was between $2112.72 and $112 606.67, depending on the cement/antibiotic mix. Although we did not directly evaluate the antimicrobial efficacy of each model, these data might lead one to believe that commercially available antibiotic-loaded PMMA would have to be 50 times more effective than manually mixed PMMA and vancomycin to be financially justifiable. Our study has limitations. First, owing to the relative low rate of infection, the sample size studied may be susceptible to a type II statistical error. Secondly, also because of this smaller sample size and the retrospective nature of the study, it was not possible to account for all potential confounding variables that might increase the risk of infection. These may include demographic differences between the groups not accounted for by the characteristics examined in Table I, such as MRSA colonisation in the nares or presence of autoimmune disease. Additionally, due to the retrospective nature of the study, we were obliged to estimate the expected rate of infection following TKR between 2004 and 2007 using extrapolation from THR data. In a prospectively-designed study, a control group of TKR patients could have been created during the second time period (2004 to 2007) who did not receive ALBC, in order to more accurately estimate the background rate of infection in the TKR population receiving unadulterated bone cement. Thirdly, when analysing the cost of the four models, we assumed their efficacy to be equivalent, an assumption for which we have no supporting evidence. Perhaps most importantly, we took no account for the criticisms of using hand-mixed cement with antibiotics, such as the potentially decreased fatigue or tensile strength, or mixing inconsistency of such a product.23 The resulting cost–benefit analysis of using antibioticloaded PMMA to prevent infection after joint replacement was variable, depending on the type of preparation created. According to our findings, the current market prices of premixed products may not be justified. However, less expensive methods may be extremely cost-effective. Surgeons are urged to compare the data presented here with the cost of their own treatment protocols for treating THE BONE & JOINT JOURNAL
THE ECONOMICS OF USING PROPHYLACTIC ANTIBIOTIC-LOADED BONE CEMENT IN TOTAL KNEE REPLACEMENT
patients with an infected joint replacement. Future research is needed to investigate the efficacy of manually mixed and compiled antibiotic-loaded PMMA formulas, which will further clarify whether they are cost-effective. No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. This article was primary edited by D. Rowley and first-proof edited by J. Scott.
References 1. Kurtz S, Ong K, Lau E, Mowat F, Halpern M. Projections of primary and revision hip and knee arthroplasty in the United States from 2005 to 2030. J Bone Joint Surg [Am] 2007;89-A:780–785. 2. Lavernia C, Lee DJ, Hernandez VH. The increasing financial burden of knee revision surgery in the United States. Clin Orthop Relat Res 2006;446:221–226. 3. Jiranek WA, Hanssen AD, Greenwald AS. Antibiotic-loaded bone cement for infection prophylaxis in total joint replacement. J Bone Joint Surg [Am] 2006;88A:2487–2500. 4. Oga M, Sugioka Y, Hobgood CD, Gristina AG, Myrvik QN. Surgical biomaterials and differential colonization by Staphylococcus epidermidis. Biomaterials 1988;9:285–289. 5. Hanssen AD, Spangehl MJ. Practical applications of antibiotic-loaded bone cement for treatment of infected joint replacements. Clin Orthop Relat Res 2004;427:79–85. 6. Bourne RB. Prophylactic use of antibiotic bone cement: an emerging standard: in the affirmative. J Arthroplasty 2004;19(Suppl):69–72. 7. Parvizi J, Saleh KJ, Ragland PS, Pour AE, Mont MA. Efficacy of antibioticimpregnated cement in total hip replacement. Acta Orthop 2008;79:335–341. 8. Josefsson G, Gudmundsson G, Kolmert L, Wijkström S. Prophylaxis with systemic antibiotics versus gentamicin bone cement in total hip arthroplasty: a five-year survey of 1688 hips. Clin Orthop Relat Res 1990;253:173–178. 9. Engesaeter LB, Lie SA, Espehaug B, et al. Antibiotic prophylaxis in total hip arthroplasty: effects of antibiotic prophylaxis systemically and in bone cement on the revision rate of 22,170 primary hip replacements followed 0-14 years in the Norwegian Arthroplasty Register. Acta Orthop Scand 2003;74:644–651.
VOL. 96-B, No. 1, JANUARY 2014
69
10. Hanssen AD. Prophylactic use of antibiotic bone cement: an emerging standard: in opposition. J Arthroplasty 2004;19(Suppl):73–77. 11. Randelli P, Evola FR, Cabitza P, et al. Prophylactic use of antibiotic-loaded bone cement in primary total knee replacement. Knee Surg Sports Traumatol Arthrosc 2010;18:181–186. 12. Lewis G. Effect of two variables on the fatigue performance of acrylic bone cement: mixing method and viscosity. Biomed Mater Eng 1999;9:197–207. 13. Arciola CR, Campoccia D, Montanaro L. Effects on antibiotic resistance of Staphylococcus epidermidis following adhesion to polymethylmethacrylate and to silicone surfaces. Biomaterials 2002;23:1495–1502. 14. Clyburn TA, Cui Q. Antibiotic laden cement: current state of the art. AAOS Now 2007. http://www.aaos.org/news/bulletin/may07/clinical7.asp (date last accessed 4 November 2013). 15. Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol 1992;45:613–619. 16. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987;40:373–383. 17. Parvizi J, Zmistowski B, Berbari EF, et al. New definition for periprosthetic joint infection: from the Workgroup of the Musculoskeletal Infection Society. Clin Orthop Relat Res 2011;469:2992–2994. 18. Kurtz SM, Lau E, Schmier J, et al. Infection burden for hip and knee arthroplasty in the United States. J Arthroplasty 2008;23:984–991. 19. Penner MJ, Masri BA, Duncan CP. Elution characteristics of vancomycin and tobramycin combined in acrylic bone-cement. J Arthroplasty 1996;11:939–944. 20. Chiu FY, Chen CM, Lin J, Lo WH. Cefuroxime-impregnated cement in primary total knee arthroplasty: a prospective, randomized study of three hundred and forty knees. J Bone Joint Surg [Am] 2002;84-A:759–762. 21. Parvizi J, Pawasarat IM, Azzam KA, et al. Periprosthetic joint infection: the economic impact of methicillin-resistant infections. J Arthroplasty 2010;25(Suppl):103– 107. 22. Bozic KJ, Ries MD. The impact of infection after total hip arthroplasty on hospital and surgeon resource utilization. J Bone Joint Surg [Am] 2005;87-A:1746–1751. 23. Klekamp J, Dawson JM, Haas DW, DeBoer D, Christie M. The use of vancomycin and tobramycin in acrylic bone cement: biomechanical effects and elution kinetics for use in joint arthroplasty. J Arthroplasty 1999;14:339–346.
The economics of using prophylactic antibiotic-loaded bone cement in total knee replacement C. J. Gutowski, B. M. Zmistowski, C. T. Clyde, J. Parvizi From Thomas Jefferson University Hospital, Philadelphia, Pennsylvania, United States
The rate of peri-prosthetic infection following total joint replacement continues to rise, and attempts to curb this trend have included the use of antibiotic-loaded bone cement at the time of primary surgery. We have investigated the clinical- and cost-effectiveness of the use of antibiotic-loaded cement for primary total knee replacement (TKR) by comparing the rate of infection in 3048 TKRs performed without loaded cement over a three-year period versus the incidence of infection after 4830 TKRs performed with tobramycin-loaded cement over a later period of time of a similar duration. In order to adjust for confounding factors, the rate of infection in 3347 and 4702 uncemented total hip replacements (THR) performed during the same time periods, respectively, was also examined. There were no significant differences in the characteristics of the patients in the different cohorts. The absolute rate of infection increased when antibiotic-loaded cement was used in TKR. However, this rate of increase was less than the rate of increase in infection following uncemented THR during the same period. If the rise in the rate of infection observed in THR were extrapolated to the TKR cohort, 18 additional cases of infection would have been expected to occur in the cohort receiving antibiotic-loaded cement, compared with the number observed. Depending on the type of antibiotic-loaded cement that is used, its cost in all primary TKRs ranges between USD $2112.72 and USD $112 606.67 per case of infection that is prevented. Cite this article: Bone Joint J 2014;96-B:65–9.
C. J. Gutowski, MD, MPH, Orthopaedic Surgery Resident Thomas Jefferson University, Department of Orthopaedic Surgery, 1025 Walnut St, 516 College Building, Philadelphia, Pennsylvania 19107, USA. B. M. Zmistowski, BA, Medical Student C. T. Clyde, BA, Medical Student Thomas Jefferson University, 1025 Walnut St, 516 College Building, Philadelphia, Pennsylvania 19107, USA. J. Parvizi, MD, PhD, Professor of Orthopaedic Surgery Rothman Institute at Thomas Jefferson University, Sheridan Building 10th Floor, 125 South 9th St., Philadelphia, Pennsylvania 19017, USA. Correspondence should be sent to Dr C. J. Gutowski; e-mail: [email protected] ©2014 The British Editorial Society of Bone & Joint Surgery doi:10.1302/0301-620X.96B1. 31428 $2.00 Bone Joint J 2014;96-B:65–9. Received 7 December 2012; Accepted after revision 28 August 2013
Peri-prosthetic infection following total knee replacement (TKR) is a serious complication that leads to high morbidity for patients, technically demanding revision surgery with a complicated peri-operative course and high costs. Rates of revision TKR are projected to rise significantly in the future decades, 1 with one study estimating a 66% increase between 2005 and 2030, with hospital costs for these procedures expected to exceed US$2 billion.2 Revision procedures performed because of infection are associated with higher costs than those performed for aseptic failure. 2 As result of this burden, several strategies to combat this trend are under scrutiny. Intra-articular biomaterials are risk factors for bacterial contamination and subsequent infection.3 Polymethylmethacrylate (PMMA) cement carries a particularly high risk for colonisation by bacteria compared with other materials such as stainless steel and polyethylene.4 Given this, the antibiotic impregnation of cement is a logical aspect of prophylaxis and has proved effective in treating an established infection at revision surgery.3
VOL. 96-B, No. 1, JANUARY 2014
Low-dose antibiotic-loaded bone cement, generally accepted to be 1 g of antibiotic per 40 g of PMMA,5 has been approved by the United States Food and Drug Administration (FDA) for use during the second stage of a revision joint replacement after a previously established infection.3 Low-dose antibiotic loaded PMMA is not currently approved by the FDA for use in primary joint replacement; however, orthopaedic surgeons around the world routinely use it during these operations.6 Several premixed formulations combining different concentrations of antibiotics with PMMA are commercially available, but in the United States only single-antibiotic preparations using gentamicin or tobramycin are licensed. Although evidence in support of low-dose antibiotic-loaded PMMA as prophylaxis against infection is well established,7-9 many concerns remain regarding its widespread use. These include a risk of increased antibiotic toxicity and allergic reactions, and the possibility that the mechanical strength of the loaded PMMA may be compromised compared with that of unadulterated cement.10 However, these potential pitfalls have either been refuted 65
66
C. J. GUTOWSKI, B. M. ZMISTOWSKI, C. T. CLYDE, J. PARVIZI
or not proven.11,12 Although the potential risks are minor, there are two other major concerns that have yet to be fully addressed. Some fear the development of antimicrobial resistance, although this concept remains unproven. They reason that exposure of intra-articular bacteria to low levels of antibiotic could encourage resistance through mutation of the bacterial genotype.13 Also, in the current cost-conscious healthcare environment the second major concern relates to the scrutiny that surrounds the financial implications of treatment. The current cost of premixed antibiotic-loaded PMMA is between $200 and $500 per 40 g pack, with two packets generally being used per TKR.14 Jiranek et al3 estimated an increase of $117 million to the annual national healthcare costs in the United States if half of the primary total joint replacements included antibioticloaded cement. This does not take into account the increase in case load discussed earlier. Given these projections, surgeons are under pressure to limit the costs, and many are wary of the additional expense associated with including antibiotic-loaded PMMA at the time of TKR. The additional initial cost of incorporating antibioticloaded PMMA into a primary TKR protocol must be weighed against the cost savings that would be subsequently realised as a result of lowering the rate of infection, which is in turn contingent upon the rate of prevention of infection being directly attributable to the use of antibioticloaded PMMA. This study examined the possible contribution of antibiotic-loaded PMMA to the reduction in the rate of infection after TKR, and explored the financial cost–benefit analysis of its routine use.
Patients and Methods Following institutional review board approval, we undertook a retrospective cohort study comparing the rate of peri-prosthetic infection identified during two periods, first from 1 January 2000 to 31 December 2002, during which no primary joint replacement received antibiotic-loaded PMMA at our institution; and secondly after 2003, when loaded PMMA (Simplex P with tobramycin; Stryker Orthopaedics, Kalamazoo, Michigan) was introduced. Primary total hip replacements (THRs) continued to be performed without PMMA. By the end of 2003, all surgeons at our institution were using PMMA with added tobramycin during routine primary TKR. The second period of the study therefore ran from 1 January 2004 to 31 December 2007. We recorded the number of joint replacements that were undertaken during these periods, as well as the patients’ age, gender, and the Deyo modification15 of the Charlson comorbidity index (CCI),16 which was used to assess the patient’s overall health status. All patients presenting with infection following joint replacement were identified using Musculoskeletal Infection Society (MSIS) methodology.17 This defines infection as the growth of identical organisms in two separate periprosthetic tissue or fluid cultures; a draining sinus tract; or four of the following six criteria: 1) elevated erythrocyte
sedimentation rate (ESR) and C-reactive protein (CRP) levels; 2) elevated white blood cell (WBC) count in the synovial fluid; 3) elevated neutrophil percentage in the synovial fluid; 4) positive histological analysis; 5) intra-articular purulence; and 6) a single positive peri-prosthetic tissue or fluid culture. As histological analysis was not routinely performed at this institution, this study required three out of five criteria to define an infection. After identifying patients who potentially had an infection, case notes were reviewed to ensure that they met the definition. The timing of the presentation of the infection was also noted. The number of patients who underwent TKR in whom infection was prevented was determined by comparing the observed incidence to the expected incidence. It has been shown that over the past two decades the rate of infection after TKR has risen at a rate closely matching that after THR.18 As this was a retrospective study, the expected rate of infection after TKR was calculated by extrapolating the observed rise in infection following THR over the same time period. We propose that this value represents the increasing background rate of infection due to any other confounding factors. The cost of preventing a single case of infection by the addition of antibiotic-loaded PMMA was calculated using the cost of using it for the entire cohort and subtracting the baseline cost of using unadulterated PMMA in all patients, divided by the number of infections prevented. We investigated four notional combinations of PMMA and antibiotics. Model 1 consisted of this institution’s current operative protocol, using two packets of pre-mixed Simplex P with tobramycin ($420 in additional cost per joint, since each of the two commercially mixed bags of Simplex P cost $250/ bag, as opposed to $40/bag if unadulterated PMMA had been used). Model 2 consisted of hand-mixing two packets of PMMA with two 1 g vials of vancomycin ($7.88 in additional cost per joint, as each gram of vancomycin costs $3.94). Model 3 consisted of hand-mixing two packets of PMMA with two 1 g vials of tobramycin ($138.66 per joint, as 1 g of tobramycin costs $69.33). Model 4 consisted of hand-mixing two packets of PMMA with 1 g (one vial) of tobramycin and 1 g (one vial) of vancomycin ($73.27 in additional cost). We justify this because these two antibiotics have been shown to act synergistically by improving elution from the cement and prolonging the antimicrobial activity.19 Statistical analysis. A Cochran–Mantel–Haenszel test was used to compare the two cohorts (before and after the use of antibiotic-loaded PMMA) while adjusting for changes in the rate of infection due to confounders represented by the incidence of infection after THR. For other measures, continuous variables were analysed with a Student’s t-test and categorical variables were measured with chi-squared analysis. Statistical analysis was performed with SPSS 16.0 (SPSS Inc., Chicago, Illinois). A p-value of < 0.05 was considered statistically significant. THE BONE & JOINT JOURNAL
THE ECONOMICS OF USING PROPHYLACTIC ANTIBIOTIC-LOADED BONE CEMENT IN TOTAL KNEE REPLACEMENT
67
Table I. Demographic information for study sample (PMMA, polymethylmethacrylate)
Admissions (n) Joint replacements (n) Mean age (yrs) (range) Female gender (n, %) Median Charlson index (range)
Before use of antibiotic-loaded PMMA (1 January 2000 to 31 December 2002)
After introduction of antibiotic-loaded PMMA (1 January 2004 to 31 December 2007)
Knee
Hip
Knee
Hip
2390 3048 65.9 (14 to 95) 1573 (66) 2 (0 to 8)
3054 3347 62.4 (13 to 99) 1605 (53) 2 (0 to 9)
4060 4830 65.7 (28 to 97) 2611 (64) 2 (0 to 8)
4433 4702 62.1 (11 to 98) 2239 (51) 2 (0 to 8)
Table II. Observed risk of developing peri-prosthetic joint infection (PJI) in total hip (THR) and total knee replacement (TKR) at various post-operative time points, with all TKRs (no THRs) receiving antibiotic-loaded cement (ALBC) after 1 January 2004 Developed PJI (n, %)
THR Prior to ALBC (1 January 2000 to 31 December 2002) After ALBC in TKR (1 January 2004 to 31 December 2007) TKR Prior to ALBC (1 January 2000 to 31 December 2002) After ALBC in TKR (1 January 2004 to 31 December 2007)
Total joints
Three months
Six months
Two years
3347 4701
6 (0.18) 18 (0.38)
7 (0.21) 19 (0.40)
11 (0.33) 25 (0.53)
3048 4826
6 (0.20) 17 (0.35)
10 (0.33) 19 (0.39)
23 (0.75) 40 (0.83)
Table III. Profile of infecting organisms in confirmed cases of peri-prosthetic joint infection (PJI), by joint and time period (ALBC, antibiotic-loaded bone cement; MR-/MSSA, methicillin-resistant/-sensitive Staphylococcus aureus) Prior to the use of ALBC (1 January 2000 to After the introduction of ALBC (1 January 31 December 2002) 2004 to 31 December 2007) Hip (n = 11)
Knee (n = 40)
Hip (n = 25)
Gram-positive 15 (65) Staphylococcus 13 (57) Staphylococcus aureus 10 (43) MRSA 7 (30) MSSA 3 (13) Coagulase-negative Staphylococcus 3 (13) Streptococcus 3 (13) Enterococcus 0
Knee (n = 23)
10 (91) 8 (73) 7 (64) 5 (45) 2 (18) 1 (9) 1 (9) 1 (9)
32 (80) 26 (65) 15 (38) 6 (15) 9 (23) 11 (28) 5 (13) 2 (5)
23 (92) 22 (88) 19 (76) 13 (52) 6 (24) 4 (16) 2 (8) 1 (4)
Gram-negative Escherichia coli Proteus mirabilis
3 (13) 0 0
1 (9) 0 0
6 (15) 3 (8) 1 (3)
3 (12) 2 (8) 1 (4)
Polymicrobial
1 (4.3)
0
2 (5)
5 (20)
Culture negative
5 (22)
0
3 (8)
0
Results No significant differences were found in patient demographics that could affect the rate of infection between the four cohorts (THR vs TKR, pre-antibiotic-loaded PMMA vs antibiotic-loaded PMMA in TKR) (Table I). Between the two study periods, the rate of infection within two years of THR increased by 60%, from 0.33% (11 of 3347) to 0.53% (25 of 4701). The rate of infection within two years of TKR increased by 10.7%, from 0.75% (23 of 3048) to 0.83% (40 of 4826) after the introduction of antibiotic-loaded PMMA. VOL. 96-B, No. 1, JANUARY 2014
Similar findings were noted three and six months after surgery (Table II). The expected rate of infection in TKRs at two years post-operatively was 1.2% (calculated by applying the observed 60% increase in THRs), which corresponds to 58 cases of infection in the antibiotic-loaded cohort. Compared with the observed value of 40 patients who had infection (p = 0.40), the addition of antibioticloaded PMMA was associated with 18 prevented cases of infection. This is also seen in the data at three months (p = 0.06) and six months (p = 0.23) post-operatively.
68
C. J. GUTOWSKI, B. M. ZMISTOWSKI, C. T. CLYDE, J. PARVIZI
Table IV. Cost analysis of four different forms of antibiotic-loaded polymethylmethacrylate (PMMA) in total knee replacement (TKR)
Added cost per TKR ($) Total cost of cohort prophylaxis ($) (4826 knees) Cost per case of infection prevented ($)
Model 1: Simplex P
Model 2: Hand-mixed Model 3: Hand-mixed Model 4: Hand-mixed vancomycin/ vancomycin tobramycin tobramycin
420 2 026 920.00 112 606.67
7.88 38 028.88 2112.72
The infecting organisms were also investigated (Table III), but the numbers were too small to allow any statistically supported conclusion to be reached. The number of patients in whom infection after TKR was prevented by the use of antibiotic PMMA at two years (n = 18) was then used to calculate the cost per case of preventing infection, based on the additional costs associated with each of the models of antibiotic-loaded PMMA (Table IV). Simplex P with tobramycin provided the most expensive model, with a cost to prevent infection of > $100 000 per case. This is significantly higher than the other three models, with hand-mixed vancomycin-loaded PMMA having the lowest cost at $2112.72 associated with the prevention of one case.
Discussion There are reports that support the use of low-dose antibioticloaded PMMA as prophylaxis against infection following joint replacement.7 Chiu et al20 showed a statistically significant reduction in the rate of deep infection after TKR when using antibiotic-loaded PMMA compared with the use of unadulterated cement (0% vs 3.1%, respectively, p = 0.0238). Josefsson et al8 showed a significant difference (p < 0.05) in post-operative infection rate between patients undergoing primary THR using gentamicin-loaded bone cement (0.8% infection rate) compared with those using systemic antibiotics only (1.9% infection rate). Engesaeter et al,9 in a paper from The Norwegian Arthroplasty Registry, surveyed primary THRs performed between 1987 and 2001 and found that systemic antibiotic prophylaxis was associated with a 1.8-fold higher risk of infection than systemic antibiotics in conjunction with antibiotic-loaded PMMA. They also found a similar benefit from the use of loaded PMMA in primary TKR.9 The costs incurred by a hospital treating infection after joint replacement are difficult to quantify. Our data are patient and institution specific, and will vary significantly depending on the form of surgery which is undertaken. The mean charge for revision TKR secondary to infection was found by Lavernia et al2 to be $109 805. Revision arthroplasty because of infection has been reported to be $60 000 more expensive than revision for aseptic loosening or mechanical failure.21,22 Other studies have examined the cost of treating infected joint replacements and focused on economic differences that exist between resistance patterns and treatment methodologies. Parvizi et al21 reported that
138.66 669 173.16 37 176.29
73.27 353 601.02 19 644.50
the mean cost of treating infection caused by methicillinresistant organisms was $97 651, compared with $65 470 for infections caused by methicillin-sensitive organisms. The cost of preventing infection after joint replacement must be weighed against the cost of treating it. We found that the cost of preventing a single case of infection was between $2112.72 and $112 606.67, depending on the cement/antibiotic mix. Although we did not directly evaluate the antimicrobial efficacy of each model, these data might lead one to believe that commercially available antibiotic-loaded PMMA would have to be 50 times more effective than manually mixed PMMA and vancomycin to be financially justifiable. Our study has limitations. First, owing to the relative low rate of infection, the sample size studied may be susceptible to a type II statistical error. Secondly, also because of this smaller sample size and the retrospective nature of the study, it was not possible to account for all potential confounding variables that might increase the risk of infection. These may include demographic differences between the groups not accounted for by the characteristics examined in Table I, such as MRSA colonisation in the nares or presence of autoimmune disease. Additionally, due to the retrospective nature of the study, we were obliged to estimate the expected rate of infection following TKR between 2004 and 2007 using extrapolation from THR data. In a prospectively-designed study, a control group of TKR patients could have been created during the second time period (2004 to 2007) who did not receive ALBC, in order to more accurately estimate the background rate of infection in the TKR population receiving unadulterated bone cement. Thirdly, when analysing the cost of the four models, we assumed their efficacy to be equivalent, an assumption for which we have no supporting evidence. Perhaps most importantly, we took no account for the criticisms of using hand-mixed cement with antibiotics, such as the potentially decreased fatigue or tensile strength, or mixing inconsistency of such a product.23 The resulting cost–benefit analysis of using antibioticloaded PMMA to prevent infection after joint replacement was variable, depending on the type of preparation created. According to our findings, the current market prices of premixed products may not be justified. However, less expensive methods may be extremely cost-effective. Surgeons are urged to compare the data presented here with the cost of their own treatment protocols for treating THE BONE & JOINT JOURNAL
THE ECONOMICS OF USING PROPHYLACTIC ANTIBIOTIC-LOADED BONE CEMENT IN TOTAL KNEE REPLACEMENT
patients with an infected joint replacement. Future research is needed to investigate the efficacy of manually mixed and compiled antibiotic-loaded PMMA formulas, which will further clarify whether they are cost-effective. No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. This article was primary edited by D. Rowley and first-proof edited by J. Scott.
References 1. Kurtz S, Ong K, Lau E, Mowat F, Halpern M. Projections of primary and revision hip and knee arthroplasty in the United States from 2005 to 2030. J Bone Joint Surg [Am] 2007;89-A:780–785. 2. Lavernia C, Lee DJ, Hernandez VH. The increasing financial burden of knee revision surgery in the United States. Clin Orthop Relat Res 2006;446:221–226. 3. Jiranek WA, Hanssen AD, Greenwald AS. Antibiotic-loaded bone cement for infection prophylaxis in total joint replacement. J Bone Joint Surg [Am] 2006;88A:2487–2500. 4. Oga M, Sugioka Y, Hobgood CD, Gristina AG, Myrvik QN. Surgical biomaterials and differential colonization by Staphylococcus epidermidis. Biomaterials 1988;9:285–289. 5. Hanssen AD, Spangehl MJ. Practical applications of antibiotic-loaded bone cement for treatment of infected joint replacements. Clin Orthop Relat Res 2004;427:79–85. 6. Bourne RB. Prophylactic use of antibiotic bone cement: an emerging standard: in the affirmative. J Arthroplasty 2004;19(Suppl):69–72. 7. Parvizi J, Saleh KJ, Ragland PS, Pour AE, Mont MA. Efficacy of antibioticimpregnated cement in total hip replacement. Acta Orthop 2008;79:335–341. 8. Josefsson G, Gudmundsson G, Kolmert L, Wijkström S. Prophylaxis with systemic antibiotics versus gentamicin bone cement in total hip arthroplasty: a five-year survey of 1688 hips. Clin Orthop Relat Res 1990;253:173–178. 9. Engesaeter LB, Lie SA, Espehaug B, et al. Antibiotic prophylaxis in total hip arthroplasty: effects of antibiotic prophylaxis systemically and in bone cement on the revision rate of 22,170 primary hip replacements followed 0-14 years in the Norwegian Arthroplasty Register. Acta Orthop Scand 2003;74:644–651.
VOL. 96-B, No. 1, JANUARY 2014
69
10. Hanssen AD. Prophylactic use of antibiotic bone cement: an emerging standard: in opposition. J Arthroplasty 2004;19(Suppl):73–77. 11. Randelli P, Evola FR, Cabitza P, et al. Prophylactic use of antibiotic-loaded bone cement in primary total knee replacement. Knee Surg Sports Traumatol Arthrosc 2010;18:181–186. 12. Lewis G. Effect of two variables on the fatigue performance of acrylic bone cement: mixing method and viscosity. Biomed Mater Eng 1999;9:197–207. 13. Arciola CR, Campoccia D, Montanaro L. Effects on antibiotic resistance of Staphylococcus epidermidis following adhesion to polymethylmethacrylate and to silicone surfaces. Biomaterials 2002;23:1495–1502. 14. Clyburn TA, Cui Q. Antibiotic laden cement: current state of the art. AAOS Now 2007. http://www.aaos.org/news/bulletin/may07/clinical7.asp (date last accessed 4 November 2013). 15. Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol 1992;45:613–619. 16. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987;40:373–383. 17. Parvizi J, Zmistowski B, Berbari EF, et al. New definition for periprosthetic joint infection: from the Workgroup of the Musculoskeletal Infection Society. Clin Orthop Relat Res 2011;469:2992–2994. 18. Kurtz SM, Lau E, Schmier J, et al. Infection burden for hip and knee arthroplasty in the United States. J Arthroplasty 2008;23:984–991. 19. Penner MJ, Masri BA, Duncan CP. Elution characteristics of vancomycin and tobramycin combined in acrylic bone-cement. J Arthroplasty 1996;11:939–944. 20. Chiu FY, Chen CM, Lin J, Lo WH. Cefuroxime-impregnated cement in primary total knee arthroplasty: a prospective, randomized study of three hundred and forty knees. J Bone Joint Surg [Am] 2002;84-A:759–762. 21. Parvizi J, Pawasarat IM, Azzam KA, et al. Periprosthetic joint infection: the economic impact of methicillin-resistant infections. J Arthroplasty 2010;25(Suppl):103– 107. 22. Bozic KJ, Ries MD. The impact of infection after total hip arthroplasty on hospital and surgeon resource utilization. J Bone Joint Surg [Am] 2005;87-A:1746–1751. 23. Klekamp J, Dawson JM, Haas DW, DeBoer D, Christie M. The use of vancomycin and tobramycin in acrylic bone cement: biomechanical effects and elution kinetics for use in joint arthroplasty. J Arthroplasty 1999;14:339–346.
