Risk Factors in Total Joint Arthroplasty: Comparison of Infection Rates in Patients With Different Socioeconomic Backgrounds

Abstract

Infection after total joint arthroplasty is a serious complication. Several risk factors have been shown to increase the risk of total joint infections. The purpose of this study was to evaluate whether socioeconomic background was a risk factor for infection in primary total joint arthroplasty. A retrospective chart review was conducted over a 4-year period on a single surgeon’s split practice between private patients with mostly private insurance and Medicare and county based patients with predominately indigent county health coverage and Medicaid. An infection rate was calculated for each population in both primary total knee and hip arthroplasty. The two populations were statistically analyzed for differences in age, preoperative diagnoses, and socioeconomic background. To our knowledge, this is the first study showing an increased risk of infection in total joint arthroplasty based on socioeconomic background.

Infection after total joint arthroplasty can be a devastating complication to a usually highly successful surgery. Infection rates after primary total knee arthroplasty reported in the literature range from 0.39% to 2.5%.^1-4 Total hip infection rates are approximately 0.2% to 2.2% for primary procedures.^2,5,6

Several risk factors for infection after total joint arthroplasty have been identified. The risks are increased in patients with rheumatoid arthritis, immunosuppressive therapy including steroids, malignancy, diabetes mellitus, prior native joint infection, malnutrition, infection elsewhere in the body, human immunodeficiency virus (HIV), prior joint surgery, obesity, skin ulcerations, and hypokalemia.^3,6-12 The most common infecting organisms are Staphylococcus aureus and coagulase-negative staphylococci.^6,10,13 Once an infection is found, it must be eradicated with a combination of irrigation and debridement with possible resection and revision and a course of antibiotics.

Approximately 12,000 prosthetic joint infections occur annually in the United States, with an estimated cost of $600 million dollars a year.^6,14-16 Because of the morbidity and cost to society, identification of risk factors to help prevent infection is critical. No studies evaluating a possible increased risk of infection in total joint arthroplasty based on socioeconomic background have been reported previously in the literature.

This retrospective study evaluated whether socioeconomic background was a risk factor for primary total joint infections by comparing total joint replacement performed at our institution in 2 different patient populations: an indigent county population and a private insured population.

Materials and Methods

Study Population

The senior author (R.A.W.) has a split practice between private patients with predominantly private insurance and Medicare at a private hospital and county-based patients with predominantly indigent county health coverage and Medicaid at a county hospital. After receiving internal review board approval, this retrospective study examined and compared charts from January 1, 2002 through December 31, 2005 with a mean follow-up of 2.3 years for the 2 populations.

We used Current Procedural Terminology (CPT) and International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes for primary total knee arthroplasty (TKA), primary total hip arthroplasty (THA), removal of knee prosthesis, removal of hip prosthesis, and infection and inflammatory reaction due to internal joint prosthesis. A cross-reference also was conducted with the infection control nursing staff. A total of 765 patients were included in the study.

Prosthetic joint infection was diagnosed in cases in which cultures of joint aspirates or intraoperative specimens yielded organisms, gross purulence was noted during debridement or removal of a prosthesis, pathologic specimens showed acute inflammatory cells of ≥5 neutrophils per high-power field, or a communicating sinus tract existed. We also included those cases for review in which a high index of clinical suspicion due to increased pain or persistent drainage of the surgical wound for >10 days in hips and >4 days in knees led to a surgical debridement and a 6-week course of antibiotic therapy, even if cultures were negative.

Charts were reviewed on all total joint replacements in both populations. Data including gender, age, preoperative diagnosis, and financial classification were extracted for all patients. The indigent county insurance is a tax-supported medical program offered to eligible county residents. Eligibility is based on household size and gross monthly income according to the current Federal Poverty Income Levels. The program is available for patients who do not have Medicaid or Medicare with prescription benefits or any medical coverage that covers all or part of medical services and prescription pharmaceutical costs. Additional data including comorbidities, body mass index, potassium, albumin, total lymphocyte count, and infecting organism were extracted for the infected population. Infection rates were calculated for both procedures in each population.

Surgical Technique

The same perioperative protocol was followed for the 2 groups. Orthopedic residents were involved to an equal degree in the care of both sets of patients. Intravenous antibiotics consisting of 1 g of vancomycin and 1 g of cefazolin or 900 mg of clindamycin in patients with severe penicillin allergy were administered preoperatively within 1 hour of skin incision. Patients were prepped with betadine solution or DuraPrep Surgical solution (3M, St Paul, Minnesota) or chlorhexidine gluconate solution if they had a history of an allergic reaction to iodine. Ioban (3M) also was placed on the wound.

Total knee arthroplasties were performed under tourniquet throughout the case; tourniquets were released after the final bandages were placed. Total knee arthroplasties were performed using a medial parapatellar approach, and THAs were performed using a modified Hardinge’s lateral approach described by Frndak et al.¹⁷

Palacos cement (Zimmer Inc, Warsaw, Indiana) impregnated with gentamicin was used for TKAs, and uncemented techniques were performed for THAs. No drains were implanted in either procedure unless excessive bleeding occurred, which was rare.

Postoperatively, 1 g of cefazolin (or 900 mg of clindamycin for allergic patients) was administered intravenously every 8 hours for 24 hours. Deep venous thrombosis prophylaxis was begun the morning after surgery and consisted of enoxaparin 40 mg subcutaneously for a total of 7 to 10 days. Patients who underwent TKA received continuous passive motion for 8 hours/day for the first 3 weeks. In both procedures, physical therapy commenced on postoperative day 1 and continued for at least 3 weeks, and then was extended on an individual basis.

Statistical Analysis

Statistical analysis was performed for differences in age, preoperative diagnoses, and socioeconomic and financial classification. The Fisher exact test was used to compare categorical data, and t tests were used to compare continuous data. A P value >.05 was considered significant.

Results

Total Knee Arthroplasty

A total of 483 primary TKAs were performed; 220 were performed in the private population and 263 were performed in the county population. There were 82 men (37.27%) and 138 women (62.73%) in the private population versus 65 men (24.71%) and 198 women (75.29%) in the county population. The county-based population was significantly younger than the private population, with mean ages of 59 and 64.2 years, respectively (P<.0001) (Table 1).

Based on insurance/financial classification, in the private group, 95 patients (43.18%) had private insurance, 110 patients (50%) had Medicare, and 15 patients (6.82%) had Medicaid. In the county-based group, 19 patients (7.22%) had private insurance, 80 patients (30.42%) had Medicare (39 of 80 patients were <65 years and received Medicare for disability versus 15 of 110 patients in the private group), 23 patients (8.75%) had Medicaid, and 141 patients (53.61%) had hospital indigent care. Therefore, for TKAs, the county population had statistically more indigent patients (Medicaid plus hospital indigent insurance) and the private population had statistically more insured (Medicare plus private insurance) (P<.0001) (Table 1, Figure 1).

For preoperative diagnoses, there were no statistically significant differences between the private and county-based populations. In the private population, preoperative diagnoses were degenerative joint disease in 212 patients (96.36%), rheumatoid arthritis in 2 patients (0.91%), posttraumatic in 4 patients (1.82%), and avascular necrosis in 2 patients (0.91%). In the county population, preoperative diagnoses were degenerative joint disease in 247 patients (93.92%), rheumatoid arthritis in 9 patients (3.42%), posttraumatic in 3 patients (1.14%), and avascular necrosis in 4 patients (1.52%). The county population approached having statistically more rheumatoid arthritis patients (P=.0739) (Table 1, Figure 2).

There were 4 infections in 220 patients (1.89%) in the private population versus 8 infections in 263 patients (3.04%) in the county population. This difference was not statistically significant (P=.5595) (Table 1).

Figure 1: Graph showing financial classification in the private and county patient populations for total knee arthroplasty (values are expressed as percentages of each population). Figure 2: Graph showing preoperative diagnosis in the private and county patient populations for total knee arthroplasty (values are expressed as number of diagnosis in each population). (Abbreviations: AVN, avascular necrosis; DJD, degenerative joint disease; and RA, rheumatoid arthritis.)

Further investigation showed of the 8 infections in the county population, 4 patients had indigent county insurance and 4 patients had Medicare. Three of the 4 Medicare patients were <65 years old and received Medicare for disability; these 3 patients would have qualified for the indigent insurance based on income. In the private population, 3 patients had private insurance and 1 patient had Medicare (Table 2).

The diagnoses in patients with infection were as follows: degenerative joint disease in 7 patients and rheumatoid arthritis in 1 patient in the county population, and degenerative joint disease in 2 patients, posttraumatic in 1 patient, and rheumatoid arthritis in 1 patient in the private population (Table 2). Therefore, by diagnosis, the infection rate in the county and private populations, respectively, was 2.8% and 0.94% for degenerative joint disease, 11.1% and 50% for rheumatoid arthritis, 0% and 25% for posttraumatic.

	Table 2: Click to view larger table. (PDF)

Patient 12 in the private total population underwent open reduction and internal fixation of tibial plateau and tubercle fracture 1 year prior to TKA. This patient developed a coagulase-negative staphylococci and Escherichia coli infection after removal of the internal fixation hardware and implantation of the prosthetic joint.

Other factors investigated included body mass index (BMI), potassium and albumin levels, and total lymphocyte count. Seven of the 8 patients with infections in the county population and 3 of the 4 patients with infections in the private population were considered obese with a BMI >30 kg/m². Hypokalemia was not a strong predictor of infection rate as the potassium level was <3.5 mEq/L in only 1 of the 12 patients with infection. Only 1 patient with an infection had a low total lymphocyte count and albumin levels, defined as a lymphocyte count <1500 cells/mm³ and an albumin level <3.5 g/dL, respectively (Table 2).

Total Hip Arthroplasty

A total of 282 primary THAs were performed, with 150 performed in the private population and 132 performed in the county population. There were 57 men (38%) and 93 women (62%) in the private population versus 69 men (52.27%) and 63 women (47.73%) in the county population. The county-based population was significantly younger than the private population, with mean ages of 50.9 and 60.8 years, respectively (P<.0001) (Table 1).

Based on insurance/financial classification, in the private group, 64 patients (42.67%) had private insurance, 77 patients (51.33%) had Medicare, and 9 patients (6%) had Medicaid. In the county-based group, 12 patients (9.09%) had private insurance, 18 patients (13.64%) had Medicare (14 of 18 patients were <65 years and received Medicare for disability versus 19 of 77patients in the private group), 33 patients (25%) had Medicaid, and 69 patients (52.27%) had hospital indigent care. Therefore, for THAs, the county population had significantly more indigent patients (Medicaid plus hospital insurance) and the private population had significantly more insured (Medicare plus private insurance) (P<.0001) (Table 1 and Figure 3).

There were several significant differences between the private and county-based populations when comparing preoperative diagnoses. In the private population, preoperative diagnoses were degenerative joint disease in 123 patients (82%), rheumatoid arthritis in 4 patients (2.67%), posttraumatic in 2 patients (1.33%), and avascular necrosis in 21patients (14%). In the county population, preoperative diagnoses were degenerative joint disease in 65 patients (49.24%), rheumatoid arthritis in 1 patient (0.76%), posttraumatic in 13 patients (9.85%), and avascular necrosis in 47 patients (35.61%). The county population had significantly more patients with a preoperative diagnosis of posttraumatic (P=.0022) and avascular necrosis (P<.0001), whereas the private population had more patients with a preoperative diagnosis of degenerative joint disease (P<.0001) (Table 1 and Figure 4).

Figure 3: Graph showing financial classification in the private and county patient populations for total hip arthroplasty (values are expressed as percentages of each population). Figure 4: Graph showing preoperative diagnosis in the private and county patient populations for total hip arthroplasty (values are expressed as number of diagnosis in each population). (Abbreviations: AVN, avascular necrosis; DJD, degenerative joint disease; and RA, rheumatoid arthritis.)

There were no infections in the private population versus 5 infections in 132 patients (3.78%) in the county population. This difference was statistically significant (P=.0216) (Table 1).

Breakdown of insurance type of the 5 infections in the county population showed 2 patients had indigent county insurance, 1 patient had Medicaid, 1 patient had Medicare, and 1 patient had private insurance. The patient on Medicare was <65 years and received Medicare for disability and would have qualified for the indigent insurance based on income (Table 2).

The diagnoses in patients with infection were posttraumatic in 2 patients and avascular necrosis in 3 patients, with infection rates of 15.4% and 6.4%, respectively (Table 2). Three of the 5 county-based infections occurred in patients who underwent previous surgeries.

The first patient (patient no. 13), a 38-year-old obese man, fell 5 weeks after undergoing primary THA for posttraumatic arthritis and experienced a periprosthetic fracture distal to the femoral stem. The fracture was repaired with a cable plate without complications. Five weeks later, he developed a draining sinus and was found to have a methicillin-resistant Staphylococcus aureus (MRSA) prosthetic joint and femur cable plate infection.

The second patient (patient no. 14), a 46-year-old man with a history of diabetes, former drug abuse, and heavy alcohol abuse, underwent primary THA for avascular necrosis 7 years after unsuccessful core decompression. He underwent irrigation and debridement 1 week postoperatively for unrelieved radiating hip pain and persistent wound drainage. All intraoperative cultures were negative but the patient had been given intravenous antibiotics prior to obtaining the cultures, and given a high index of clinical suspicion, he was treated with the infection protocol.

The third patient (patient no. 17), a 52-year-old man with a history of HIV, hepatitis C, diabetes, intravenous drug abuse, and heavy alcohol use, underwent percutaneous pinning of a femoral neck fracture 9 months before THA. He developed an infected nonunion that required serial debridement and intravenous antibiotics before eradication and implantation of the THA prosthesis. There was no sign of infection in laboratory or hip aspirations prior to THA; however, the patient developed a postoperative MRSA infection in the hip requiring serial debridements, resection arthroplasty, intravenous antibiotics, and reimplantation.

Four of the 5 county patients with infections were considered obese with a BMI >30 kg/m². Hypokalemia was not a strong predictor of infection rate as only 1 of the 5 patients with infection had a potassium level <3.5 mEq/L. None of the patients with an infection had a low total lymphocyte count or albumin level, defined as a lymphocyte count <1500 cells/mm³ and an albumin level <3.5 g/dL, respectively (Table 2).

Discussion

Using retrospective data, we compared the risk of infection after primary total joint arthroplasty between 2 patient populations with different socioeconomic backgrounds in a single surgeon’s practice during a 4-year period using the same perioperative protocol. The overall infection rate in the mainly indigent county population was significantly higher for THA (3.78%) than in the mainly private insured population (0%). Although not statistically significant, there was a difference in the infection rate for TKA between the county (3.04%) and the private (1.89%) populations. The overall infection rates were similar to those reported in the literature, which range from 0.2% to 2.5%.^1-6

We used a strict definition of infection in our study. Although other studies^6,18-20 have used similar criteria, the addition of counting those infections with a high index of clinical suspicion but without positive culture, pathology, or purulence in our study likely increased our overall infection rate. If patients treated for infection who had negative cultures were excluded from our study, the infection rates would have been 0.91% in the private TKA population, 2.66% in the county TKA population, and 3.03% in the county THA population.

In their study, Ridgeway et al¹⁰ used the surgical site infection protocol used by the Centers for Disease Control and Prevention. The problem with this protocol is that it lacked standard methods for postdischarge surveillance and that it only identified infections developed during patients’ hospitalization.

Our study is the first to analyze the effect of socioeconomic and insurance background on total joint infections. Ibrahim et al²¹ evaluated the racial and ethnic differences in surgical outcomes in veterans following knee and hip arthroplasty. They concluded black and Hispanic patients had significantly higher rates of infection-related complications following TKA compared with white patients. They did not observe the same differences in THA. These infection-related rates included pneumonia and urinary tract infections, along with superficial and deep surgical wound infections. There were no statistically significant differences by race or ethnicity when these were compared individually, except black and Hispanic patients had higher rates of urinary tract infections. Therefore, there was no difference in superficial or deep infections of total joint replacements depending on race.

In their study, SooHoo et al⁴ reviewed 222,684 primary TKAs performed in California. They found Hispanic ethnicity was associated with a higher odds ratio for primary TKA infections. In our study, we did not specifically evaluate race.

When comparing our 2 arthroplasty populations based on insurance/financial classification, the county population had statistically more indigent patients (Medicaid plus hospital indigent insurance) and the private population had statistically more insured patients (Medicare plus private insurance). A large percentage of the county’s Medicare population (53 of 98) was <65 years and received Medicare for disability versus a small portion (34 of 187) in the private group. This subgroup of patients would have qualified for county indigent care based on income guidelines and would have made the percentage of patients with indigent county health care much higher than the percentage we reported. Similar results were seen when looking at THAs. The county population had statistically more indigent patients (Medicaid plus hospital insurance) and the private population had statistically more insured (Medicare plus private insurance), even counting the patients with disability in the Medicare population.

In their large population of primary TKAs, SooHoo et al⁴ found Medicaid insurance was associated with a higher odds ratio for infections, whereas private insurance patients had lower infection rates. Their study did not specifically include indigent county patients.

Although the infection rate was higher, there was no statistically significant increase in infection rates between the 2 populations when looking at TKAs. Our county population was younger than our private population, but unlike in the THA population, there were no statistical differences in preoperative diagnoses between the 2 populations.

When searching for the reason the county population had a higher rate of THA infections, one must look at the risk factors in our patient population. Overall, our county population was a younger group with more risky lifestyles including a higher level of substance abuse and dangerous activities. The 5 infections all occurred in men <60 years. Four of these 5 patients had a history of substance abuse including intravenous drug abuse and alcohol abuse. One of the infections occurred after the patient was discharged from the hospital. The patient reportedly was on cocaine when he fell and opened his incision. Although no studies have shown avascular necrosis to be an independent risk factor for increased infection, the diagnosis can suggest a higher level of alcoholism in our patient population. Age <65 years increased the risk of infection in the baseline probability analysis in the study by SooHoo et al⁴ who showed an individual increase of infection of 46% compared to the base case of a white woman >65 years who had no comorbidities and was insured by Medicare.

Our county-based population also had a higher percentage of posttraumatic hip disease than the private population, which usually presents at a younger age. Many of these patients underwent previous surgery for pelvis or hip fractures. Peersman et al³ and Wilson et al,¹² in large series of 6489 and 4171 TKAs, respectively, showed previous same-site surgery to be a risk factor for future prosthetic joint infection. Three of the 5 hip infections in our county-based population occurred in patients who had undergone previous surgeries.

Although Berbari et al⁶ showed prior native joint infection to be a risk factor for prosthetic joint infection, no study has shown previous infections in contralateral total joint arthroplasties to be a risk factor for a contralateral total joint infection. One patient in each of our 2 groups developed an infection of their knee after being successfully treated for an infected TKA on the contralateral side.

Rheumatoid arthritis is a documented risk factor for prosthetic joint infection.^6,8,12,22 In our TKA county and private populations, an infection occurred in a patient with rheumatoid arthritis in each group. In the private population, infection occurred in 1 of 2 patients with rheumatoid arthritis, and in the county population, infection occurred in 1 of 9 patients with rheumatoid arthritis. Wilson et al¹² theorized the risk is increased because of chronic immunosuppression from steroid use. Luessenhop et al⁸ believed the risk might be increased in patients with rheumatoid arthritis because of the altered or diminished host immune function.

Diabetes and obesity are major problems in the United States. Several studies have shown diabetes increases the risk of infection of prosthetic joints.^3,6,22 In our study, 3 of the 5 hip infections occurred in patients with diabetes (patients 14, 16, and 17), and 2 of the 4 private knee infections occurred in patients with diabetes (patients 9 and 10). Although Wilson et al¹² did not find obesity to increase the risk for prosthetic joint infection, Peersman et al,³ Ridgeway et al,¹⁰ and Lentino²² discussed obesity as an individual risk factor. In our study, obesity appeared to be a large risk factor in both of our populations. Fourteen of the 17 infections occurred in patients who were considered obese with a BMI >30 kg/m². Five of the 14 patients were considered morbidly obese, with BMI >40 kg/m².

We further investigated the infections that occurred, looking for other common risks factors for infection. Hypokalemia³ was not a strong predictor of infection rate as only 2 of the 17 patients with infections had a potassium level <3.5 mEq/L. We expected to find albumin level and total lymphocyte count, which has been shown to be a predictor of general nutrition level,^3,6,22 would be low in patients with infections. Although we did not have levels for all patients, only 1 patient with an infection (patient 2) had a low total lymphocyte count and albumin level, defined as a lymphocyte count <1500 cells/mm³ and an albumin level <3.5 g/dL, respectively.

This study has some limitations. First, it is a retrospective study with an average follow-up of 2.3 years. Because infections can occur for many years after total joint arthroplasty, we may have underestimated our overall infection rate. However, Phillips et al⁵ showed the incidence of deep hip infections was highest in the period between hospital discharge and the end of the first postoperative month; the rate dropped significantly during successive follow-up periods. In addition, we did not have a perfect way of tracking our patients’ long-term status, and some may have sought medical attention at other medical facilities if problems developed.

Poor medical record documentation prevented the evaluation of other variables, such as surgery time and exact timing of preoperative antibiotics. However, the senior author used the same protocol for antibiotics and the same technical procedure for each case. Neither hospital had an adequate survey system to accurately identify the overall infection rate for operations performed at that hospital; consequently, we were unable to determine the effect of the facility and professional staff on the infection rate. There are other possible variables such as operating room or patient room cleanliness, infection cases in the same patient room, and single versus double occupancy rooms.

We believe the county hospital treated a larger proportion of patients who presented to the emergency room with an admitting diagnosis of infection somewhere in the body. Therefore, one could argue the hospital’s nosocomial infection rate would be higher at the county hospital, which could explain the difference in the infection rate between the 2 patient populations studied. The strengths of the study include the sample sizes of 2 different populations for a single surgeon with the same protocol over the same time period.

Currently, the idea of “pay for performance” is being considered to improve the overall outcomes of medical treatment. One way to implement pay for performance would be to reward surgeons with lower complication rates or penalize surgeons with higher complication rates. This study illustrated the fact that not all patients and not all patient populations are equal, and payment methods that reimburse surgeons based on patient outcome rather than adherence to established best practice protocols would unjustly penalize surgeons who care for patients with greater risk factors. This financial disincentive to care for these patients could lead to discrimination and must be avoided.

Conclusion

This large, retrospective study showed a statistically increased risk of infection in a county-based indigent population undergoing THA versus a privately insured population. Although higher, there was no statistically significant difference in the risk of infection when comparing TKA in the county-based and privately insured populations. Further prospective studies are needed to investigate the exact reasons for the differences.

References

1. Blom AW, Brown J, Taylor AH, Pattison G, Whitehouse S, Bannister GC. Infection after total knee arthroplasty. J Bone Joint Surg Br. 2004; 86(5):688-691.
2. Hanssen AD, Rand JA. Evaluation and treatment of infection at the site of a total hip or knee arthroplasty. Instr Course Lect. 1999; 48:111-122.
3. Peersman G, Laskin R, Davis J, Peterson M. Infection in total knee replacement: a retrospective review of 6489 total knee replacements. Clin Orthop Relat Res. 2001; (392):15-23.
4. SooHoo NF, Lieberman JR, Ko CY, Zingmond DS. Factors predicting complication rates following total knee replacement. J Bone Joint Surg Am. 2006; 88(3):480-485.
5. Phillips CB, Barrett JA, Losina E, et al. Incidence rates of dislocation, pulmonary embolism, and deep infection during the first six months after elective total hip replacement. J Bone Joint Surg Am. 2003; 85(1):20-26.
6. Berbari EF, Hanssen AD, Duffy MC, et al. Risk factors for prosthetic joint infection: case-control study. Clin Infect Dis. 1998; 27(5):1247-1254.
7. Fitzgerald RH Jr, Nolan DR, Ilstrup DM, Van Scoy RE, Washington JA II, Coventry MB. Deep wound sepsis following total hip arthroplasty. J Bone Joint Surg Am. 1977; 59(7):847-855.
8. Luessenhop CP, Higgins LD, Brause BD, Ranawat CS. Multiple prosthetic infections after total joint arthroplasty: risk factor analysis. J Arthroplasty. 1996; 11(7):862-868.
9. Parvizi J, Sullivan TA, Pagnano MW, Trousdale RT, Bolander ME. Total joint arthroplasty in human immunodeficiency virus-positive patients: an alarming rate of early failure. J Arthroplasty. 2003; 18(3):259-264.
10. Ridgeway S, Wilson J, Charlet A, Kafatos G, Pearson A, Coello R. Infection of the surgical site after arthroplasty of the hip. J Bone Joint Surg Br. 2005; 87(6):844-850.
11. Tsukayama DT, Goldberg VM, Kyle R. Diagnosis and management of infection after total knee arthroplasty. J Bone Joint Surg Am. 2003; 85(suppl 1):S75-S80.
12. Wilson MG, Kelley K, Thornhill TS. Infection as a complication of total knee-replacement arthroplasty: risk factors and treatment in sixty-seven cases. J Bone Joint Surg Am. 1990; 72(6):878-883.
13. Gaine WJ, Ramamohan NA, Hussein NA, Hullin MG, McCreath SW. Wound infection in hip and knee arthroplasty. J Bone Joint Surg Br. 2000; 82(4):561-565.
14. Darouiche RO. Treatment of infections associated with surgical implants. N Engl J Med. 2004; 350(14):1422-1429.
15. Lentino JR. Infections associated with prosthetic knee and prosthetic hip. Curr Infect Dis Rep. 2004; 6(5):388-392.
16. Sculco TP. The economic impact of infected joint arthroplasty. Orthopedics. 1995; 18(9):871-873.
17. Frndak PA, Mallory TH, Lombardi AV Jr. Translateral surgical approach to the hip: the abductor muscle “split.” Clin Orthop Relat Res. 1993; (295):135-141.
18. Bauer TW, Brooks PJ, Sakai H, Krebs V, Borden L. A diagnostic algorithm for detecting an infected hip arthroplasty. Orthopedics. 2003; 26(9):929-930.
19. Segawa H, Tsukayama DT, Kyle RF, Becker DA, Gustilo RB. Infection after total knee arthroplasty: a retrospective study of the treatment of eight-one infections. J Bone Joint Surg Am. 1999; 81(10):1434-1445.
20. Spangehl MJ, Masri BA, O’Connell JX, Duncan CP. Prospective analysis of preoperative and intraoperative investigations for the diagnosis of infection at the sites of two hundred and two revision total arthroplasties. J Bone Joint Surg Am. 1999; 81(5):672-683.
21. Ibrahim SA, Stone RA, Han X, et al. Racial/ethnic differences in surgical outcomes in veterans following knee or hip arthroplasty. Arthritis Rheum. 2005; 52(10):3143-3151.
22. Lentino JR. Prosthetic joint infections: bane of orthopedists, challenge for infectious disease specialists. Clin Infect Dis. 2003; 36(9):1157-1161.

Authors

Drs Webb, Lichtman, and Wagner are from the Department of Orthopedic Surgery, John Peter Smith Hospital, Fort Worth, Texas.

Drs Webb, Lichtman, and Wagner have no relevant financial interests to disclose.

Correspondence should be addressed to: Brian G. Webb, MD, JPS Department of Orthopedic Surgery, 1500 S Main St, Ft Worth, TX 76104.