JBJS | Instructional Course Lectures, The American Academy of Orthopaedic Surgeons - Diagnosis of Infection following Total Hip Arthroplasty*†

The Journal of Bone & Joint Surgery, Volume 79, Issue 10

Instructional Course Lecture | October 01, 1997

Instructional Course Lectures, The American Academy of Orthopaedic Surgeons - Diagnosis of Infection following Total Hip Arthroplasty*†

M. J. SPANGEHL, M.D., F.R.C.S.(C)‡; A. S. E. YOUNGER, M.B., M.SC., F.R.C.S.(C)‡; B. A. MASRI, M.D., F.R.C.S.(C)‡; C. P. DUNCAN, M.B., M.SC., F.R.C.S.(C)‡, VANCOUVER, BRITISH COLUMBIA, CANADA

View Disclosures and Other Information

An Instructional Course Lecture, The American Academy of Orthopaedic Surgeons

The Journal of Bone & Joint Surgery. 1997; 79:1578-88

5 Recommendations (Recommend) | 3 Comments | Saved by 3 Users Save Case

text A A A

Introduction

High rates of infection complicated the early experience with total hip arthroplasty and, although the rates have decreased substantially over the last few decades, infection still is a source of considerable morbidity. In the 1960s, Charnley reported a rate of infection of 9.5 per cent (nineteen infections after 199 total hip arthroplasties)⁵. More recently, authors have reported that infection causes failure after 1 per cent (seventy-one of 5081 and twenty-seven of 2084) to 2 per cent (ninety-four of 5500 and thirty-six of 1798) of primary total hip arthroplasties^{6,13,20,22,37,38,42,58,61}, and the rate is higher after revision procedures. Although these percentages are small, the large number of hip arthroplasties performed each year results in a major burden on the health-care system⁶². Infection following total hip arthroplasty is costly to treat because of the subsequent need for reoperation and the prolonged hospitalization often required to eradicate the infection. In the United States, the cost per year to treat the 3500 to 4000 infections following total hip arthroplasty is between 150 and 200 million dollars⁶². Because of an aging population that will need an increasing number of arthroplasties, methods to prevent, diagnose, and treat infection must be perfected in order to reduce the cost of total hip arthroplasty to society.

Infection following total hip arthroplasty can present a diagnostic challenge. No test is 100 per cent sensitive and 100 per cent specific; thus, the diagnosis of infection relies on the surgeon's judgment of the clinical presentation, the findings on physical examination, and the interpretation of the results of previous investigations. The consequences of misdiagnosis are considerable. Reimplantation of a prosthesis into an infected tissue bed, without appropriate débridement, is likely to result in persistent infection⁴¹. Numerous investigations are available for the workup and diagnosis of failed total hip replacements. These investigations, as well as an algorithm to rule out the presence of infection as a cause of failure, will be presented.

*Printed with permission of The American Academy of Orthopaedic Surgeons. This article will appear in Instructional Course Lectures, Volume 47, The American Academy of Orthopaedic Surgeons, Rosemont, Illinois, March 1998.

†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. No funds were received in support of this study.

‡Department of Orthopaedics, Vancouver Hospital and Health Sciences Centre, Third Floor, 910 West 10th Avenue, Vancouver, British Columbia V5Z 4E3, Canada.

View Larger

TABLE I SENSITIVITY AND SPECIFICITY OF THE ERYTHROCYTE SEDIMENTATION RATE AND THE C-REACTIVE PROTEIN LEVEL

Laboratory Screening Test	No. of Patients	No. Who Had an Infection	Value for Erythrocyte Sedimentation Rate (mm/hr.)	Value for C-Reactive Protein Level (mg/L)	Sensitivity	Specificity
Erythrocyte sedimentation rate
Sanzen and Carlsson⁵⁹	56	23	>30		0.61	1.00
Thoren and Wigren⁶⁸	79	51	>35		0.88	0.96
Roberts et al.⁵⁶	69	14	>30		0.84	0.79
Spangehl et al.⁶⁵	171	34	>30		0.82	0.85

C-reactive protein level
Sanzen and Carlsson⁵⁹	56	23		>10	0.91	0.88
Spangehl et al.⁶⁵	142	26		>10	0.96	0.92

Add to My JBJS

TABLE I SENSITIVITY AND SPECIFICITY OF THE ERYTHROCYTE SEDIMENTATION RATE AND THE C-REACTIVE PROTEIN LEVEL

Laboratory Screening Test	No. of Patients	No. Who Had an Infection	Value for Erythrocyte Sedimentation Rate (mm/hr.)	Value for C-Reactive Protein Level (mg/L)	Sensitivity	Specificity
Erythrocyte sedimentation rate
Sanzen and Carlsson⁵⁹	56	23	>30		0.61	1.00
Thoren and Wigren⁶⁸	79	51	>35		0.88	0.96
Roberts et al.⁵⁶	69	14	>30		0.84	0.79
Spangehl et al.⁶⁵	171	34	>30		0.82	0.85

C-reactive protein level
Sanzen and Carlsson⁵⁹	56	23		>10	0.91	0.88
Spangehl et al.⁶⁵	142	26		>10	0.96	0.92

View Larger

TABLE II SENSITIVITY AND SPECIFICITY OF NUCLEAR IMAGING

*The study included various types of infections; not all were related to the prosthesis.

Imaging Modality	No. of Sites Tested	No. of Proved Infections	Sensitivity	Specificity
Technetium/gallium scanning
Kraemer et al.³⁴	43	13	0.38 (5/13)	1.00 (30/30)
Merkel et al.⁴³*	42	24	0.50 (12/24)	0.78 (14/18)
Indium-111-labeled whiteblood-cell scanning
Merkel et al.⁴³*	42	24	0.83 (20/24)	0.94 (17/18)
Glithero et al.²³	25	8	0.38 (5/13)	1.00 (12/12)
Wukich et al.⁷⁵	24	7	1.00 (7/7)	0.41 (7/17)
Johnson et al.³³	29	9	1.00 (9/9)	0.50 (10/20)
Technetium/indium-111-labeled white-blood-cell scanning
Johnson et al.³³	29	9	0.89 (8/9)	0.95 (19/20)
Palestro et al.⁵²	50	10	1.00 (10/10)	0.98 (39/40)
Indium-111-labeled immunoglobulin-G scanning
Oyen et al.¹⁵*
Entire study	120	72	0.97 (70/72)	0.85 (41/48)
Failed arthroplasties only	37	12	0.92 (11/12)	0.88 (22/25)
Wegener et al.^*74	15	11	0.91 (10/11)	1.00 (4/4)

Add to My JBJS

TABLE II SENSITIVITY AND SPECIFICITY OF NUCLEAR IMAGING

*The study included various types of infections; not all were related to the prosthesis.

Imaging Modality	No. of Sites Tested	No. of Proved Infections	Sensitivity	Specificity
Technetium/gallium scanning
Kraemer et al.³⁴	43	13	0.38 (5/13)	1.00 (30/30)
Merkel et al.⁴³*	42	24	0.50 (12/24)	0.78 (14/18)
Indium-111-labeled whiteblood-cell scanning
Merkel et al.⁴³*	42	24	0.83 (20/24)	0.94 (17/18)
Glithero et al.²³	25	8	0.38 (5/13)	1.00 (12/12)
Wukich et al.⁷⁵	24	7	1.00 (7/7)	0.41 (7/17)
Johnson et al.³³	29	9	1.00 (9/9)	0.50 (10/20)
Technetium/indium-111-labeled white-blood-cell scanning
Johnson et al.³³	29	9	0.89 (8/9)	0.95 (19/20)
Palestro et al.⁵²	50	10	1.00 (10/10)	0.98 (39/40)
Indium-111-labeled immunoglobulin-G scanning
Oyen et al.¹⁵*
Entire study	120	72	0.97 (70/72)	0.85 (41/48)
Failed arthroplasties only	37	12	0.92 (11/12)	0.88 (22/25)
Wegener et al.^*74	15	11	0.91 (10/11)	1.00 (4/4)

View Larger

TABLE III SENSITIVITY AND SPECIFICITY OF PREOPERATIVE ASPIRATION

Study	No. of Sites Tested	No. of Proved Infections	Sensitivity	Specificity
Phillips and Kattapuram⁵³	141	33	0.91 (30/33)	0.82 (89/108)
Tigges et al.⁷⁰	147	14	0.93 (13/14)	0.92 (122/133)
Kraemer et al.³⁴	45	14	0.57 (8/14)	0.97 (30/31)
Roberts et al.⁵⁶	78	15	0.87 (13/15)	0.95 (60/63)
Barrack and Harris³	260	4	0.50 (2/4)	0.88 (224/256)
Lachiewicz et al.³⁶	156	25	0.92 (23/25)	0.97 (127/131)
Mulcahy et al.⁴⁵	71	16	0.69 (11/16)	0.91 (50/55)
Fehring and Cohen¹⁵	166	6	0.50 (3/6)	0.88 (141/160)
Spangehl et al.⁶⁵	180	21	0.86 (18/21)	0.94 (149/159)

Add to My JBJS

TABLE III SENSITIVITY AND SPECIFICITY OF PREOPERATIVE ASPIRATION

Study	No. of Sites Tested	No. of Proved Infections	Sensitivity	Specificity
Phillips and Kattapuram⁵³	141	33	0.91 (30/33)	0.82 (89/108)
Tigges et al.⁷⁰	147	14	0.93 (13/14)	0.92 (122/133)
Kraemer et al.³⁴	45	14	0.57 (8/14)	0.97 (30/31)
Roberts et al.⁵⁶	78	15	0.87 (13/15)	0.95 (60/63)
Barrack and Harris³	260	4	0.50 (2/4)	0.88 (224/256)
Lachiewicz et al.³⁶	156	25	0.92 (23/25)	0.97 (127/131)
Mulcahy et al.⁴⁵	71	16	0.69 (11/16)	0.91 (50/55)
Fehring and Cohen¹⁵	166	6	0.50 (3/6)	0.88 (141/160)
Spangehl et al.⁶⁵	180	21	0.86 (18/21)	0.94 (149/159)

View Larger

TABLE IV SENSITIVITY AND SPECIFICITY OF INTRAOPERATIVE FROZEN SECTIONS

Study	No. of Sites Tested	No. of Proved Infections	Sensitivity	Specificity
Fehring and McAlister¹⁶	97	11	0.18 (2/11)	0.90 (77/86)
Feldman et al.¹⁸	33	9	1.00 (9/9)	0.96 (23/24)
Athanasou et al.²	106	22	0.91 (20/22)	0.96 (81/84)
Lonner et al.³⁹	175	19	0.84 (16/19)	0.99 (154/156)
Fehring and Cohen¹⁵	130	4	0.50 (2/4)	0.90 (114/126)
Spangehl et al.⁶⁵	202	35	0.80 (28/35)	0.94 (157/167)

Add to My JBJS

TABLE IV SENSITIVITY AND SPECIFICITY OF INTRAOPERATIVE FROZEN SECTIONS

Study	No. of Sites Tested	No. of Proved Infections	Sensitivity	Specificity
Fehring and McAlister¹⁶	97	11	0.18 (2/11)	0.90 (77/86)
Feldman et al.¹⁸	33	9	1.00 (9/9)	0.96 (23/24)
Athanasou et al.²	106	22	0.91 (20/22)	0.96 (81/84)
Lonner et al.³⁹	175	19	0.84 (16/19)	0.99 (154/156)
Fehring and Cohen¹⁵	130	4	0.50 (2/4)	0.90 (114/126)
Spangehl et al.⁶⁵	202	35	0.80 (28/35)	0.94 (157/167)

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 1 Protocol for the diagnosis of infection following total hip arthroplasty. The asterisk indicates that repeat aspiration should be performed before the procedure if clinical suspicion of an infection is high or if the erythrocyte sedimentation rate (ESR) or the C-reactive protein (CRP) level, or both, are elevated for another reason. The dagger indicates that the surgeon should consider performing sequential technetium-99m and indium-labeled white blood-cell scanning. The box indicates that frozen sections should be obtained if the erythrocyte sedimentation rate and the C-reactive protein level are not elevated.

Clinical Presentation

A thorough history and physical examination are of paramount importance in the diagnosis of infection. Although a conclusive diagnosis can be made in many instances, there are no data, to our knowledge, with regard to the efficacy of clinical assessment alone. Even when a conclusive diagnosis cannot be established, a careful history and physical examination can help to guide the appropriate investigations.

Coventry⁹, and later Fitzgerald et al.²¹, described what is perhaps the most common system for the classification of infection after total hip arthroplasty. This three-stage classification system is based on the mode or timing of the presentation of infection.

Type-I infections occur in the immediate postoperative period. The patient usually is seen during the first postoperative month, and the diagnosis is evident on the basis of the medical history and the physical examination. Systemic signs of infection, such as fever, chills, and sweating, may be present. Pain is usually continuous. On examination, the wound may be erythematous, swollen, fluctuant, and tender. Wound drainage, if present, is usually purulent. Type-I infections are caused by infected hematomas or superficial wound infections spreading contiguously to the periprosthetic space. The diagnostic challenge is to determine whether or not a superficial infection has penetrated deep to the fascia.

Type-II infections also are believed to originate at the time of the operation, but because of a small inoculum or the low virulence of the organism the onset of symptoms is delayed. The patient usually is seen between six and twenty-four months after the index procedure. The hallmark of this type of infection is a gradual deterioration in function and an increase in pain. Pain is often present from the time of the original procedure; it may be activity-related or it may occur at night and during rest. Often, the only clue to infection is early loosening of the components. Systemic symptoms are not part of the presentation; however, there may be a history of prolonged wound drainage at the time of the index procedure. Specific questions about a delay in the patient's discharge from the hospital, a prolonged course of antibiotics, or ongoing wound drainage should be asked while the history is being obtained. The findings on examination of the hip in a patient who has a type-II infection usually are non-specific and are similar to those associated with aseptic loosening. Increased warmth or a draining sinus may be present. Careful examination for an old closed sinus or evidence of poor wound-healing may suggest the presence of deep infection. A type-II infection represents a diagnostic challenge.

Type-III infections are the least common and are caused by a hematogenous spread to a previously asymptomatic hip, usually two years or more after the arthroplasty. Generally, there is an acute febrile episode accompanied by sudden, rapid deterioration in the function of the hip. The acute onset of such an infection in a hip with a previously well functioning prosthesis is of far more importance than the temporal relationship between the onset of the infection and the insertion of the prosthesis. The diagnosis usually can be made on the basis of the history and the physical examination. The patient may recall a systemic or febrile illness that was followed by symptoms in the hip. Seeding can occur at the site of a loose prosthesis, a solidly osseointegrated prosthesis, or a solidly fixed cemented prosthesis. A type-III infection is likely to occur in patients who are immunosuppressed, such as those who have had a renal transplant or who have been managed with immunosuppressive medications for inflammatory arthropathy; those who have recurrent episodes of bacteremia, such as intravenous drug abusers; and those who need repeat urinary catheterization^28,30. Other factors that may be associated with type-III infection are dental manipulation^28,29,57,66, respiratory infection²⁹, remote periprosthetic infection²⁹, open skin lesions¹¹, endoscopy⁷², and contamination of the operative site²⁸. Early diagnosis may allow salvage of the joint by means of thorough débridement, whereas a delay in the diagnosis may necessitate a one or two-stage exchange procedure in order to eradicate the infection.

Recently, Estrada et al. expanded the classification to include patients for whom intraoperative cultures are positive despite a presumed preoperative diagnosis of aseptic loosening¹⁴. Tsukayama et al. reported the results of treatment of 106 infections⁷¹. Thirty-one patients who initially were thought to have aseptic failure were diagnosed as having an infection on the basis of positive intraoperative cultures. In that study, a minimum of two of five cultures had to be positive in order for the joint to be considered infected. Sixteen of these patients also had an elevated erythrocyte sedimentation rate (more than thirty millimeters per hour), whereas just one of the twenty-five patients who had had a histological examination had acute inflammation. Only the white blood-cell count and the erythrocyte sedimentation rate were used routinely in the preoperative workup. These results lead to the question of whether some of the patients had a pre-existing chronic infection or false-positive intraoperative cultures. It is debatable whether positive intraoperative cultures represent true infection if there is no other evidence of that diagnosis.

Preoperative Investigations

White Blood-Cell Count

The white blood-cell count is rarely abnormal in patients who have an infection following a total hip arthroplasty, and it is not helpful for ruling infection in or out as a cause of failure of the procedure. Canner et al. found that, of fifty-two patients who had an infection following a joint arthroplasty, only eight (15 per cent) had leukocytosis⁴. It has been our experience⁶⁵, as well as that of others¹², that the white blood-cell count is normal in most patients who have an infection following a total hip arthroplasty. When a patient does have an abnormal white blood-cell count, the systemic infection is usually clinically obvious and is either type I or type III.

Erythrocyte Sedimentation Rate and C-Reactive Protein Level

The erythrocyte sedimentation rate and the C-reactive protein level are the most useful laboratory screening investigations for the diagnosis of a potential infection following total hip arthroplasty. The erythrocyte sedimentation rate is a non-specific hematological test that measures the distance, in millimeters, that a column of red blood cells settles in one hour. This is a reflection of the formation of erythrocyte rouleaux. The formation of rouleaux generally is discouraged because of the normal negative charge on the erythrocytes; however, if there is an excess of positively charged macromolecules in the serum, the erythrocyte negative charge becomes diluted. Dilution decreases the repulsion between erythrocytes, thereby allowing the formation of rouleaux, which increases both the mass of settling red blood-cell units and the sedimentation rate¹⁰.

Acute-phase reactants are one type of the positively charged macromolecules just mentioned. These macromolecules are manufactured in the liver in response to a number of inflammatory, infectious, and neoplastic processes. An elevated erythrocyte sedimentation rate is an indirect indicator of an abundance of acute-phase reactants¹⁰. Because acute-phase reactants are produced under a variety of conditions, the specificity of an elevated erythrocyte sedimentation rate is decreased; the sensitivity, however, remains high. Patients who have a chronic infection at the site of a total joint prosthesis generally have an increased erythrocyte sedimentation rate without systemic illness or an increased white blood-cell count. Values of more than thirty or thirty-five millimeters per hour generally are considered to be abnormal and indicative of infection unless proved otherwise. Raising or lowering the chosen value for the erythrocyte sedimentation rate in order to differentiate between septic and aseptic failure will inversely affect the sensitivity and directly affect the specificity of the test for a given sample population. As the usefulness of a test is determined by its ability to both rule in and rule out the presence of infection, the value chosen to differentiate between septic and aseptic conditions should result in sensitivities and specificities that approach each other. This value generally has been reported to be between thirty and thirty-five millimeters per hour (Table I)^56,59,65,68.

C-reactive protein is an acute-phase reactant that is synthesized in the liver and is found in only trace amounts under normal conditions⁵⁹. As is the case for the other acute-phase reactants, the C-reactive protein level increases in a non-specific manner as a result of infectious, inflammatory, or neoplastic disorders. The C-reactive protein level increases from trace amounts (the normal state) to reach maximum values within forty-eight hours after an operation and then returns to trace amounts in approximately two to three weeks^1,47,64. The erythrocyte sedimentation rate may remain elevated for months after an uncomplicated total hip replacement⁶⁴. Therefore, the ability of the C-reactive protein level to return to normal much faster than the erythrocyte sedimentation rate enables it to be a more sensitive indicator of infection, particularly in the early postoperative period.

In a study of seventy-nine patients who had had a revision hip replacement and had no known factors that would have elevated the erythrocyte sedimentation rate, twenty-seven of the twenty-eight patients who did not have an infection had an erythrocyte sedimentation rate that was thirty-five millimeters per hour or less⁶⁸. The fifty-one patients who had an infection had a mean erythrocyte sedimentation rate of fifty-nine millimeters per hour, and six of these patients had an erythrocyte sedimentation rate of thirty-five millimeters per hour or less. With use of an erythrocyte sedimentation rate of more than thirty-five millimeters per hour as an indication of infection, the sensitivity was 0.88 and the specificity was 0.96. In a similar study of fifty-six patients who had had a revision (twenty-three of whom had an infection and thirty-three of whom did not)⁵⁹, the use of an erythrocyte sedimentation rate of more than thirty millimeters per hour as an indication of infection yielded a sensitivity of 0.61 and a specificity of 1.00. In that study, the C-reactive protein level also was analyzed. With use of a C-reactive protein level of more than ten milligrams per liter as an indication of infection, the sensitivity and the specificity were 0.91 and 0.88, respectively. In our own series, in which an erythrocyte sedimentation rate of more than thirty millimeters per hour and a C-reactive protein level of more than ten milligrams per liter were considered to be indicative of infection, we found a sensitivity and specificity of 0.82 and 0.85 for the erythrocyte sedimentation rate and of 0.96 and 0.92 for the C-reactive protein level⁶⁵ (Table I). Combining the tests should improve the accuracy of diagnosis.

Some care must be taken in interpreting the erythrocyte sedimentation rate or the C-reactive protein level before a revision hip arthroplasty. The physician must determine whether any other factors, such as rheumatoid arthritis, a recent operation, neoplasia, collagen vascular disease, infection, or an inflammatory condition, are present. If no such conditions are applicable, an erythrocyte sedimentation rate of more than thirty or thirty-five millimeters per hour and a C-reactive protein level of more than ten milligrams per liter should be considered abnormal and should warrant additional investigation to rule out infection.

Plain Radiography

Plain radiographs should be made for all patients who have a failed arthroplasty, even though radiographs are of limited value as an investigative tool for the diagnosis of infection. Many radiographic findings, such as loosening, osteolysis, and endosteal scalloping, are common to both septic and aseptic failure. Occasionally, the patient will have diagnostic changes, such as periostitis, rapidly progressive and diffuse osteolysis, or endosteal scalloping^3,19,27,40. Periosteal new-bone formation, with or without loosening of a component, has been considered by some to be pathognomonic of deep infection¹⁹.

Endosteal scalloping has been shown to be suggestive of infection⁴⁰. In a retrospective review of fifty revision total hip replacements, twenty-nine (91 per cent) of thirty-two hips in which the prosthesis failed because of infection had evidence of endosteal scalloping on radiographs⁴⁰. Other investigators have reported endosteal scalloping in as many as sixty-two (24 per cent) of 260 hips that were not infected^27,31.

Early loosening and rapidly progressive radiolucent lines also are suggestive of infection^3,27. This may be particularly true if obvious causes of osteolysis, such as severe polyethylene wear, are not present. In recent years, the radiographic diagnosis of loosening has evolved^32,35. Previously, possible loosening was defined as a radiolucent line, occupying 50 to 100 per cent of the bone-cement interface, that had not been present on radiographs made immediately postoperatively; probable loosening, as a continuous radiolucent line surrounding the entire mantle at the bone-cement interface; and definite loosening, as a radiolucent line at the stem-cement interface, fracture of the cement mantle or the stem, or migration of the prosthesis²⁶. The clinical relevance of a radiolucent line at the bone-cement interface recently has been brought into question^32,35. Retrieval studies of asymptomatic patients have shown circumferential remodeling of trabecular bone adjacent to cement and resultant endosteal cortical-bone resorption causing what appears as a radiolucent line on radiographs^32,35. On the acetabular side, definite loosening is indicated by migration of the socket or the cement mantle, protrusio acetabuli, or acetabular fracture⁴⁰.

The addition of arthrography can improve the accuracy of radiographs in the diagnosis of loosening^25,48. Arthrography may show penetration of the contrast medium between the bone and the cement. Arthrography does not have notable advantages compared with plain radiography when used for the diagnosis of loosening of the femoral component, but it is of benefit when used for the assessment of loosening of a cemented socket⁴⁸. Loosening is often subtle, and the diagnosis is easier to make if old radiographs are reviewed in order to document fracture of the cement, migration of a component, or progression of radiolucent lines at the component-cement interface.

Although some patients initially are seen with septic loosening of a total hip replacement, loosening is not necessarily a feature of infection following total hip arthroplasty. Most patients who have a postoperative or an acute hematogenous infection have solidly fixed components. In our series of eighty-four patients who had an infection following total hip arthroplasty, thirty-four (40 per cent) had solid fixation of the femoral component at the time of revision. Occasionally, plain radiographs provide clues to infection; however, they are neither sensitive nor specific for its detection.

Radionuclide Imaging

Scintigraphy continues to receive attention with regard to its potential for the diagnosis of infection following joint replacement. Scintigraphy is limited by the cost of the scans; the time required for the patient to have the procedure; and, in some situations, the inability of the scans to yield consistently acceptable levels of sensitivity and specificity. Often, scans are no more accurate than serological investigations, which are much less expensive.

Technetium-99m bone scans, the first scans that were used for the diagnosis of infection following hip arthroplasty, are sensitive but not specific. Some investigators have found that a negative bone scan rules out infection⁶⁷; however, others have reported that a technetium-99m scan occasionally can be negative in a patient who has an infection if there is an inadequate blood supply to the bone⁷³. The causes of photopenic defects include subperiosteal pus, soft-tissue swelling, and vasospasm. Difficulties associated with technetium-99m bone scans include the fact that multiple conditions, such as fractures, tumors, heterotopic ossification, and inflammatory disorders, can result in increased uptake in the periprosthetic tissue; the fact that the scans can remain positive for as long as one year after an uncomplicated hip replacement and for more than two years after insertion of a prosthesis without cement⁴⁹; and, most importantly, the fact that the scans cannot be used to differentiate between infection and aseptic loosening.

Gallium-67 citrate is a radioisotope that accumulates in areas of inflammation. Like technetium, it is non-specific, as any process resulting in reactive bone formation may cause increased uptake⁷³. Although more accurate than individual scans, sequential technetium-gallium scans still lack sufficient accuracy to be clinically useful for the diagnosis of a potential infection following hip arthroplasty^34,43. Merkel et al., in a prospective study comparing sequential technetium-gallium scans with indium-labeled-leukocyte scans for the diagnosis of a variety of low-grade musculoskeletal infections, found a sensitivity of only 0.50 (twelve of twenty-four) and a specificity of 0.78 (fourteen of eighteen) for the sequential technetium-gallium scans⁴³.

Indium-111-labeled white blood cells are useful for the diagnosis of conditions of increased vascularity and white blood-cell uptake; however, their usefulness for the diagnosis of infection following hip arthroplasty continues to be debated^23,43,54,75. Glithero et al. reported a poor sensitivity (0.38; three of eight) but a high specificity (1.00; seventeen of seventeen) in an analysis of twenty-five failed arthroplasties, eight of which were believed to be associated with an infection²³. Merkel et al., in their previously mentioned study of mixed infections, found a sensitivity of 0.83 (twenty of twenty-four) and a specificity of 0.94 (seventeen of eighteen) with use of indium-labeled leukocytes⁴³. In an attempt to increase its ability to aid in the diagnosis of infection, indium-111-labeled white blood-cell scanning was combined, into a sequential protocol, with scanning with various preparations of technetium^33,52. When a zonal analysis of the prosthesis was performed, Palestro et al. found greater accuracy (greater uptake) in the region of the prosthetic femoral head in infected hips⁵². The combined scans generally had higher sensitivities and specificities (Table II); however, cost and time constraints still allow these scans only a limited role.

Other radiolabeled markers have been investigated in an attempt to improve the accuracy of nuclear imaging. Radiolabeled immunoglobulin-G has been used for the investigation of musculoskeletal infections^50,51,63,74. Radiolabeled immunoglobulin-G scans are similar to indium-111-labeled white blood-cell scans in that the radiopharmaceutical agent is labeled to a carrier that targets areas of acute inflammation. However, the advantage of immunoglobulin-G-labeling is that the patient does not need to have a phlebotomy before the scan is made and the lengthy laboratory preparation and subsequent reinjection of white blood cells can be avoided. Oyen et al. prospectively compared the results of indium-111-labeled white blood-cell scans with those of indium-111-labeled immunoglobulin-G scans⁵⁰. They reported superior results with the indium-111-labeled immunoglobulin-G scans (a sensitivity and a specificity of 0.80 [twenty of twenty-five] and 1.00 [twenty-four of twenty-four] compared with 0.56 [fourteen of twenty-five] and 0.79 [nineteen of twenty-four] for the indium-111-labeled white blood-cell scans)⁵⁰. Unfortunately, the study population included patients who had a variety of bone and soft-tissue infections, including infections outside of the musculoskeletal system. When patients who had musculoskeletal infections were separated from the rest of the series, the sensitivity and specificity for the indium-111-labeled immunoglobulin-G scans were 1.00 (fifteen of fifteen and eight of eight, respectively) whereas those for the indium-111-labeled white blood-cell scans were 0.73 (eleven of fifteen) and 0.88 (seven of eight). In a subsequent, non-comparative study from the same institution, the sensitivity of the indium-111-labeled immunoglobulin-G scans was 0.92 (eleven of twelve) and the specificity was 0.88 (twenty-two of twenty-five) when only scans made after failed arthroplasties were analyzed⁵¹ (Table II). Until additional comparative studies specifically addressing failed total joint arthroplasties are available, the routine use of indium-111-labeled immunoglobulin-G scans cannot be recommended.

Although costly and time-consuming, radionuclide scans can be of benefit in equivocal situations in which the results of screening serological investigations may be falsely elevated and cultures of specimens aspirated from the joint may be unreliable because of the administration of antibiotics. The use of sequential technetium-99m and indium-111-labeled white blood-cell scans currently is recommended; however, for the convenience of the patient, the use of radiolabeled immunoglobulin-G scans may supersede the use of sequential scans, provided that they are proved to be equivalent or superior to sequential scans for the diagnosis of infection following hip arthroplasty.

Other Imaging Modalities

Magnetic resonance imaging can be of value after an infection has been diagnosed in a patient in whom a total hip arthroplasty was performed with use of radiolucent cement. Magnetic resonance imaging can be used to determine the extent of the cement mantle within the femur and the pelvis so that the revision procedure can be planned appropriately¹⁷.

Ultrasound has a limited role in the diagnosis of infection. It can be used to measure the thickness of the joint capsule, with a thick capsule being indicative of infection²⁴. Soft-tissue abscesses also may be evaluated with ultrasound.

Although these imaging modalities may assist in preoperative planning, they are not currently recommended as a means of excluding infection in patients who have a failed or painful total hip replacement.

Aspiration of the Hip Joint

Aspiration of the hip is perhaps the most useful investigative tool for definitive confirmation of the presence or absence of infection. Recently, however, the role of aspiration has generated considerable debate. In the past, aspiration was advocated for all patients who had a failed hip replacement^21,48,56. Currently, most authors favor a more limited role^{3,15,36,45,70}, with aspiration being used to confirm a clinical suspicion of infection or to support or negate the findings of other preoperative investigations, such as determination of the erythrocyte sedimentation rate and the C-reactive protein level, which may be falsely elevated because of connective-tissue disease. An additional benefit of aspiration in instances of suspected infection is the ability to identify the organism and its antibiotic-sensitivity profile, which may influence preoperative planning as well as the type of antibiotic that is chosen if use of an antibiotic depot is planned.

As in other investigations, the usefulness of aspiration is determined by its accuracy (the sum of the numbers of true-positive and true-negative results divided by the total number of tests). The reported rates of sensitivity and specificity have varied widely, with the sensitivity of preoperative aspiration ranging from 0.50 to 0.93 and the specificity ranging from 0.82 to 0.97^{3,15,34,36,45,53,56,65,70} (Table III). These results suggest that aspiration is better for ruling infection in than for ruling it out.

The wide variation in the number of positive results that are observed when aspiration is used to ascertain the presence of infection is partly a function of the different methods used to calculate the results. One such variation in methodology involves the type of sample that is obtained at the time of aspiration. If only joint fluid is aspirated, a so-called dry tap may be interpreted as a negative result. However, it is possible to obtain synovial tissue by means of a needle biopsy, and cultures of this tissue may result in a diagnosis of infection despite the dry tap. Other variations include the number of samples obtained at the time of aspiration; the performance of repeat aspirations after the initial aspiration; the lack of a so-called gold standard for comparison (in many studies, the results of aspiration are compared with those of intraoperative cultures alone, which also are associated with a certain prevalence of false results); and the unrecognized or unreported use of antibiotics before aspiration, which is an inherent problem in many retrospective studies. In a prospective review of the results of diagnostic procedures performed for the evaluation of failed total hip replacements, we found that thirteen of 193 patients had been receiving antibiotics before aspiration⁶⁵. Twelve of the thirteen patients had an infection, but only six had a positive result on aspiration.

Additionally, the wide variation in positive results may be due in part to the technique of aspiration. To reduce the number of false-positive results and thereby improve accuracy, a standard protocol should be established either by the radiology department or by the surgeon performing the aspiration. Strict aseptic technique must be observed not only to reduce the rate of false-positive results but also to prevent the unlikely but definite possibility of joint contamination and subsequent infection in a painful yet uninfected hip. To reduce the rate of false-negative results, all antibiotics must be discontinued for two to three weeks before the aspiration. The intracapsular position of the needle should be confirmed with arthrography. Local anesthetics should be used only in the skin and not in the joint as they are bacteriostatic⁶⁰. Provided that sufficient fluid is obtained, the specimen should be separated into three samples for culture; however, if insufficient fluid is obtained, the joint can be irrigated with non-bacteriostatic saline solution and then reaspirated. A needle biopsy of synovial tissue also should be performed at the time of the aspiration.

In our aspiration protocol, a diagnosis of infection is made if all three specimens are positive for the same organism and if this result coincides with the clinical profile. If only one sample is positive, the aspiration is repeated. If the result on any of the repeat aspirations is positive for the same organism and the antibiotic-sensitivity profile also is identical, then the presence of infection is confirmed. If two of the three initial samples are positive, the result is interpreted in accordance with those of other investigations. For example, if hematological parameters are elevated and there is no other apparent cause apart from the suspected infection, then infection is likely and the aspiration need not be repeated. However, if the results of the laboratory investigations are normal or the parameters are elevated for other reasons, then the aspiration should be repeated.

Intraoperative Investigations

Intraoperative Frozen Sections

Mirra et al. reported on the value of intraoperative frozen sections in 1976, stating that more than five polymorphonuclear leukocytes per high-power field is evidence of probable infection and that a lack of polymorphonuclear leukocytes is evidence of an absence of infection⁴⁴.

Intraoperative frozen sections have become a valuable tool for the diagnosis of infection. They are most useful in equivocal situations, when preoperative investigations are confounded by false elevations in the erythrocyte sedimentation rate or the C-reactive protein level, or both, or when the intraoperative appearance of the joint raises a suspicion of infection. Most investigators have reported favorable results (a sensitivity of 0.80 or more and a specificity of 0.90 or more)^{2,15,16,18,39,65}, although the authors of two studies^15,16 reported very poor sensitivities (Table IV). As with aspiration, some of the variation in results may be due in part to the low prevalence of infection in some series as well as the various criteria used to define a positive result. Fehring and McAlister used the over-all histological picture rather than a specific number of polymorphonuclear cells per high-power field as the criterion¹⁶. Lonner et al. recommended the use of ten polymorphonuclear cells per high-power field to improve specificity without reducing sensitivity³⁹. In their prospective analysis of the results of 175 revision arthroplasties, specificity improved from 0.96 to 0.99 when ten instead of five polymorphonuclear leukocytes per high-power field was used as an index of infection. The sensitivity remained the same, at 0.84, for both indices. An important observation from that study, as well as from our own analysis of frozen sections, is that tissue should be obtained from areas that appear to be most inflamed. In addition, the pathologist should be experienced in the preparation and interpretation of specimens; we have found substantial interobserver variations among pathologists who were less experienced in the interpretation of tissue obtained from patients who had a failed total hip replacement.

Gram Stain

The gram stain is used as a simple intraoperative investigation to confirm the presence or absence of bacteria and thereby the presence or absence of an infection of the joint. In equivocal situations, when intraoperative investigations are the most useful, an investigation must be both sensitive (able to detect infection) and specific (able to rule out infection). While the gram stain may be specific, it lacks any acceptable level of sensitivity (range, 0 to 0.23^7,18,34,65). The gram stain is not reliable for determining the presence of infection in patients who are having a revision hip arthroplasty, and it should not be used as a basis for determining treatment. In equivocal situations that call for intraoperative investigation, we have abandoned its use in favor of the frozen section.

Opinion of the Surgeon

Despite normal findings on preoperative investigation, some hips may appear infected to the surgeon during the procedure. Frank pus and the formation of an abscess are the most obvious signs of infection. Features such as diffuse synovitis, turbid joint fluid, and formation of slime are all suspicious features.

The appearance of the joint is occasionally so overwhelmingly suspicious for infection that the surgeon cannot proceed with a single-stage exchange procedure. However, false impressions do occur, and we have delayed definitive management, because an infection initially was suspected, for a number of patients who did not have an infection. In two instances, the definitive operation was delayed because purulent material was found in the joint despite normal findings on preoperative aspiration. Intraoperative cultures subsequently were negative, and severe erosion of the polyethylene liner and wear debris were thought to be responsible for the tissue reaction. Severe accumulations of wear debris can incite a reaction that is similar in appearance to that of an infection.

Unfortunately, the true value of the surgeon's intraoperative opinion may never be known because the preoperative findings inevitably will influence that opinion. Feldman et al. studied the association between the surgeon's intraoperative opinion and the pathological diagnosis and found a sensitivity of 0.70 (seven of ten) and a specificity of 0.87 (twenty of twenty-three)¹⁸. There are a number of treatment options for patients in whom the macroscopic appearance of the joint is suspicious, the cultures of specimens obtained from preoperative aspiration are negative, the erythrocyte sedimentation rate or the C-reactive protein level is equivocal, and an intraoperative frozen section is not available. One option is to leave the prosthesis in place, obtain multiple specimens for culture, and reoperate at a later date after the final results of culture are available. Another option is to perform a single-stage exchange with a thorough débridement in a patient who may not tolerate repeated procedures. Finally, the surgeon may elect to perform a Girdlestone arthroplasty on the basis of the macroscopic appearance of the joint. This procedure is preferable to implantation of a new prosthesis into an infected tissue bed without performing a full débridement or a two-stage exchange. Despite the potential for a false interpretation of the appearance of the joint, this appearance should not be disregarded because the consequences of misdiagnosis can be devastating, and it should be kept in mind that other preoperative and intraoperative investigations are not 100 per cent accurate.

Intraoperative Cultures

Intraoperative cultures often are used as the standard for the diagnosis of infection following an arthroplasty; however, the results of these cultures are not 100 per cent accurate. Fitzgerald et al.²⁰ noted positive cultures in association with 111 (25 per cent) of 437 hips that had not been operated on previously, and others^46,69 have reported rates of false-positive cultures of as high as 29 per cent (188 of 658) in association with total hip arthroplasty. In more recent studies, the rate of intraoperative false-positive cultures was between 6 per cent (ten of 159) and 13 per cent (thirty-two of 254) for patients who had had a revision hip replacement^3,15,65.

As with joint aspiration, careful technique must be followed in order to avoid false results. Preoperative antibiotics should be withheld until specimens have been obtained. Clean instruments that have not been used on the skin should be employed to obtain the specimens. The specimens should be taken from an area that has not been previously cauterized, before any irrigation fluid is used and immediately after the pseudocapsule has been opened, to decrease the chance of colonization and to allow for subsequent administration of antibiotics. Samples should be obtained from close to the surface of the prosthesis and, if applicable, from inflamed tissue.

A minimum of three tissue specimens should be sent fresh to the laboratory for immediate processing. Cultures should not be considered negative until the final results of the broth subcultures are available. Although late growth of bacteria or growth in liquid medium alone often is considered a contaminant, the final results of culture should be interpreted in the context of all of the preoperative and intraoperative findings. We have detected late growth of bacteria on solid medium or growth in liquid medium alone in some cases of infection, although this finding is not common in our experience.

Molecular Analysis

Molecular technology may be used to diagnose the presence of bacterial DNA and RNA. Specific bacteria can be identified by their DNA structure. Polymerase chain reaction enables the production of large amounts of specific sequences of target DNA from small quantities of starting material. In essence, large volumes of identical copies of the original starting material are produced. The specimen then is heat-cycled in an amino-acid broth to allow exposure and polymerization of the DNA chains. After twenty-five to forty cycles, small quantities of bacterial DNA are amplified to create sufficient volumes of DNA for analysis⁵⁵. The DNA then can be identified by sequencing, blot analysis, or enzyme digestion. Bacterial RNA, which is produced in large volumes by a single DNA gene segment, also can be sequenced with use of reverse transcriptase enzyme^8,22,55. These techniques are currently under development and may hold promise for improved diagnostic accuracy. However, the inherent advantage of polymerase chain reaction—namely, the ability to identify bacteria from small quantities of DNA—makes these techniques susceptible to contamination⁵⁵. Whether polymerase chain reaction is too sensitive to any bacterial particles that are encountered, thereby leading to a high rate of false-positive results, remains to be determined.

Protocol for the Diagnosis of Infection following Arthroplasty

To exclude infection as a cause of failure of an arthroplasty, certain key investigations should be performed. This should help to keep the over-all number of investigations to a minimum, thereby being cost-effective yet still confirming the diagnosis of infection.

After a careful history is obtained and a physical examination is performed, with special attention being paid to the details of the index procedure and the chronology of the pain, both the erythrocyte sedimentation rate and the C-reactive protein level should be determined for every patient who is to have a revision. If both results are normal—that is, the erythrocyte sedimentation rate is less than thirty millimeters per hour and the C-reactive protein level is ten milligrams per liter or less—and there is no suggestion of infection on clinical presentation, no additional investigations are needed. If the erythrocyte sedimentation rate or the C-reactive protein level is elevated for any reason or if there is a clinical suspicion of infection, then an aspiration of the hip joint should be performed. A diagnosis of infection is made if the clinical suspicion is high; the erythrocyte sedimentation rate or the C-reactive protein level, or both, are elevated for no other known reason; and the cultures of the aspirated fluid are positive. If the erythrocyte sedimentation rate or the C-reactive protein level, or both, are falsely elevated, an intraoperative frozen section may be used to confirm the diagnosis, particularly if the cultures of the aspirated fluid are negative and the clinical suspicion remains high or if the clinical suspicion is low but the cultures of the aspirated fluid are positive. However, in these situations, the determination of the erythrocyte sedimentation rate and the C-reactive protein level, as well as the aspiration, should be repeated before the procedure. A sequential indium bone scan also may be used preoperatively to assist the physician in making the diagnosis if it is expected that an intraoperative frozen section will not be available or that it may be unreliable because of lack of local expertise (Fig. 1).

The single most important factor in determining the treatment options for a patient in whom a total hip arthroplasty has failed is the exclusion of a diagnosis of infection. In most patients, the diagnosis can be successfully confirmed preoperatively on the basis of the clinical presentation, the results of serological tests, or positive cultures of aspirated joint fluid. In other patients, the diagnosis can be established according to intraoperative findings at the time of the revision procedure and confirmed with frozen section. The diagnosis may be confirmed in the postoperative period by the final results of the intraoperative cultures.

References

Aalto, K.; Österman, K.; Peltola, H.; and Räsänen, J.: Changes in erythrocyte sedimentation rate and C-reactive protein after total hip arthroplasty. Clin. Orthop.,184: 118-120, 1984.184118 1984 [PubMed]

Athanasou, N. A.; Pandey, R.; De Steiger, R.; Crook, D.; and McLardy Smith, P.: Diagnosis of infection by frozen section during revision arthroplasty. J. Bone and Joint Surg.,77-B(1): 28-33, 1995.77-B(1)28 1995

Barrack, R. L., and Harris, W. H.: The value of aspiration of the hip joint before revision total hip arthroplasty. J. Bone and Joint Surg.,75-A: 66-76, Jan. 1993.75-A66 1993

Canner, G. C.; Steinberg, M. E.; Heppenstall, R. B.; and Balderston, R.: The infected hip after total hip arthroplasty. J. Bone and Joint Surg.,66-A: 1393-1399, Dec. 1984.66-A1393 1984

Charnley, J.: A clean-air operating enclosure. British J. Surg.,51: 202-205, 1964.51202 1964

Charnley, J.: Postoperative infection after total hip replacement with special reference to air contamination in the operating room. Clin. Orthop.,87: 167-187, 1972.87167 1972 [PubMed]

Chimento, G. F.; Finger, S.; and Barrack, R. L.: Gram stain detection of infection during revision arthroplasty. J. Bone and Joint Surg.,78-B(5): 838-839, 1996.78-B(5)838 1996

Clarke, A. M.; Mapstone, N. P.; and Quirke, P.: Molecular biology made easy. The polymerase chain reaction. Histochem. J.,24: 913-926, 1992.24913 1992 [PubMed]

Coventry, M. B.: Treatment of infections occurring in total hip surgery. Orthop. Clin. North America,6: 991-1003, 1975.6991 1975

Covey, D. C., and Albright, J. A.: Current concepts review. Clinical significance of the erythrocyte sedimentation rate in orthopaedic surgery. J. Bone and Joint Surg.,69-A: 148-151, Jan. 1987.69-A148 1987

Deacon, J. M.; Pagliaro, A. J.; Zelicof, S. B.; and Horowitz, H. W.: Current concepts review. Prophylactic use of antibiotics for procedures after total joint replacement. J. Bone and Joint Surg.,78-A: 1755-1770, Nov. 1996.78-A1755 1996

Eftekhar, N. S.: Diagnosis of infection in joint replacement surgery. In Infection in Joint Replacement Surgery: Prevention and Management, pp. 115-130. Edited by N. S. Eftekhar. St. Louis, C. V. Mosby, 1984.

Eftekhar, N. S., and Tzitzikalakis, G. I.: Failures and reoperations following low-friction arthroplasty of the hip. A five- to fifteen-year follow-up study. Clin. Orthop.,211: 65-78, 1986.21165 1986 [PubMed]

Estrada, R.; Tsukayama, D.; and Gustilo, R.: Management of THA infections. A prospective study of 108 cases. Orthop. Trans.,17: 1114-1115, 1993.171114 1993

Fehring, T. K., and Cohen, B.: Aspiration as a guide to sepsis in revision total hip arthroplasty. J. Arthroplasty,11: 543-547, 1996.11543 1996 [PubMed]

Fehring, T. K., and McAlister, J. A., Jr.: Frozen histologic section as a guide to sepsis in revision joint arthroplasty. Clin. Orthop.,304: 229-237, 1994.304229 1994 [PubMed]

Fehrman, D. A.; McBeath, A. A.; DeSmet, A. A.; and Tuite, M. J.: Imaging barium-free bone cement. Am. J. Orthop.,25: 172-174, 1996.25172 1996 [PubMed]

Feldman, D. S.; Lonner, J. H.; Desai, P.; and Zuckerman, J. D.: The role of intraoperative frozen sections in revision total joint arthroplasty. J. Bone and Joint Surg.,77-A: 1807-1813, Dec. 1995.77-A1807 1995

Fitzgerald, R. H., Jr.: Total hip arthroplasty sepsis. Prevention and diagnosis. Orthop. Clin. North America,23: 259-264, 1992.23259 1992

Fitzgerald, R. H., Jr.; Peterson, L. F. A.; Washington, J. A., II; Van Scoy, R. E.; and Coventry, M. B.: Bacterial colonization of wounds and sepsis in total hip arthroplasty. J. Bone and Joint Surg.,55-A: 1242-1250, Sept. 1973.55-A1242 1973

Fitzgerald, R. H., Jr.; Nolan, D. R.; Ilstrup, D. M.; Van Scoy, R. E.; Washington, J. A., II; and Coventry, M. B.: Deep wound sepsis following total hip arthroplasty. J. Bone and Joint Surg.,59-A: 847-855, Oct. 1977.59-A847 1977

Garvin, K. L., and Hanssen, A. D.: Current concepts review. Infection after total hip arthroplasty. Past, present, and future. J. Bone and Joint Surg.,77-A: 1576-1588, Oct. 1995.77-A1576 1995

Glithero, P. R.; Grigoris, P.; Harding, L. K.; Hesslewood, S. R.; and McMinn, D. W.: White cell scans and infected joint replacements. Failure to detect chronic infection. J. Bone and Joint Surg.,75-B(3): 371-374, 1993.75-B(3)371 1993

Graif, M.; Schwartz, E.; Strauss, S.; Mouallem, M.; Schecter, M.; and Morag, B.: Occult infection of hip prosthesis: sonographic evaluation. J. Am. Geriat. Soc.,39: 203-204, 1991.39203 1991 [PubMed]

Harris, W. H., and McGann, W. A.: Loosening of the femoral component after use of the medullary-plug cementing technique. Follow-up note with a minimum five-year follow-up. J. Bone and Joint Surg.,68-A: 1064-1066, Sept. 1986.68-A1064 1986

Harris, W. H.; McCarthy, J. C., Jr.; and O'Neill, D. A.: Femoral component loosening using contemporary techniques of femoral cement fixation. J. Bone and Joint Surg.,64-A: 1063-1067, Sept. 1982.64-A1063 1982

Huddleston, H. D.: Femoral lysis after cemented hip arthroplasty. J. Arthroplasty,3: 285-297, 1988.3285 1988 [PubMed]

Hughes, P. W.; Salvati, E. A.; Wilson, P. D., Jr.; and Blumenfeld, E. L.: Treatment of subacute sepsis of the hip by antibiotics and joint replacement. Criteria for diagnosis with evaluation of twenty-six cases. Clin. Orthop.,141: 143-157, 1979.141143 1979 [PubMed]

Hunter, G., and Dandy, D.: The natural history of the patient with an infected total hip replacement. J. Bone and Joint Surg.,59-B(3): 293-297, 1977.59-B(3)293 1977

Irvine, R.; Johnson, B. L., Jr.; and Amstutz, H. C.: The relationship of genitourinary tract procedures and deep sepsis after total hip replacements. Surg., Gynec. and Obstet.,139: 701-706, 1974.139701 1974

Jasty, M. J.; Floyd, W. E., III; Schiller, A. L.; Goldring, S. R.; and Harris, W. H.: Localized osteolysis in stable, non-septic total hip replacements. J. Bone and Joint Surg.,68-A: 912-919, July 1986.68-A912 1986

Jasty, M.; Maloney, W. J.; Bragdon, C. R.; O'Connor, D. O.; Haire, T.; and Harris, W. H.: The initiation of failure in cemented femoral components of hip arthroplasties. J. Bone and Joint Surg.,73-B(4): 551-558, 1991.73-B(4)551 1991

Johnson, J. A.; Christie, M. J.; Sandler, M. P.; Parks, P. F., Jr.; Homra, L.; and Kaye, J. J.: Detection of occult infection following total joint arthroplasty using sequential technetium-99m HDP bone scintigraphy and indium-111-WBC imaging. J. Nucl. Med.,29: 1347-1353, 1988.291347 1988 [PubMed]

Kraemer, W. J.; Saplys, R.; Waddell, J. P.; and Morton, J.: Bone scan, gallium scan, and hip aspiration in the diagnosis of infected total hip arthroplasty. J. Arthroplasty,8: 611-616, 1993.8611 1993 [PubMed]

Kwong, L. M.; Jasty, M.; Mulroy, R. D.; Maloney, W. J.; Bragdon, C.; and Harris, W. H.: The histology of the radiolucent line. J. Bone and Joint Surg.,74-B(1): 67-73, 1992.74-B(1)67 1992

Lachiewicz, P. F.; Rogers, G. D.; and Thomason, H. C.: Aspiration of the hip joint before revision total hip arthroplasty. Clinical and laboratory factors influencing attainment of a positive culture. J. Bone and Joint Surg.,78-A: 749-754, May 1996.78-A749 1996

Lidwell, O. M.: Clean air at operation and subsequent sepsis in the joint. Clin. Orthop.,211: 91-102, 1986.21191 1986 [PubMed]

Lidwell, O. M.; Elson, R. A.; Lowbury, E. J.; Whyte, W.; Blowers, R.; Stanley, S. J.; and Lowe, D.: Ultraclean air and antibiotics for prevention of postoperative infection. A multicenter study of 8,052 joint replacement operations. Acta Orthop. Scandinavica,58: 4-13, 1987.584 1987

Lonner, J. H.; Desai, P.; DiCesare, P. E.; Steiner, G.; and Zuckerman, J. D.: The reliability of analysis of intraoperative frozen sections for identifying active infection during revision hip or knee arthroplasty. J. Bone and Joint Surg.,78-A: 1553-1558, Oct. 1996.78-A1553 1996

Lyons, C. W.; Berquist, T. H.; Lyons, J. C.; Rand, J. A.; and Brown, M. L.: Evaluation of radiographic findings in painful hip arthroplasties. Clin. Orthop.,195: 239-251, 1985.195239 1985 [PubMed]

McDonald, D. J.; Fitzgerald, R. H., Jr.; and Ilstrup, D. M.: Two-stage reconstruction of a total hip arthroplasty because of infection. J. Bone and Joint Surg.,71-A: 828-834, July 1989.71-A828 1989

Maderazo, E. J.; Judson, S.; and Pasternak, H.: Late infections of total joint prostheses. A review and recommendations for prevention. Clin. Orthop.,229: 131-142, 1988.229131 1988 [PubMed]

Merkel, K. D.; Brown, M. L.; Dewanjee, M. K.; and Fitzgerald, R. H., Jr.: Comparison of indium-labeled-leukocyte imaging with sequential technetium-gallium scanning in the diagnosis of low-grade musculoskeletal sepsis. A prospective study. J. Bone and Joint Surg.,67-A: 465-476, March 1985.67-A465 1985

Mirra, J. M.; Amstutz, H. C.; Matos, M.; and Gold, R.: The pathology of the joint tissues and its clinical relevance in prosthesis failure. Clin. Orthop.,117: 221-240, 1976.117221 1976 [PubMed]

Mulcahy, D. M.; Fenelon, G. C.; and McInerney, D. P.: Aspiration arthrography of the hip joint. Its uses and limitations in revision hip surgery. J. Arthroplasty,11: 64-68, 1996.1164 1996 [PubMed]

Murray, W. R.: Results in patients with total hip replacement arthroplasty. Clin. Orthop.,95: 80-90, 1973.9580 1973 [PubMed]

Niskansen, R. O.; Korkala, O.; and Pammo, H.: Serum C-reactive protein levels after total hip and knee arthroplasty. J. Bone and Joint Surg.,78-B(3): 431-433, 1996.78-B(3)431 1996

O'Neill, D. A., and Harris, W. H.: Failed total hip replacement: assessment by plain radiographs, arthrograms, and aspiration of the hip joint. J. Bone and Joint Surg.,66-A: 540-546, April 1984.66-A540 1984

Oswald, S. G.; Van Nostrand, D.; Savory, C. G.; and Callaghan, J. J.: Three-phase bone scan and indium white blood cell scintigraphy following porous coated hip arthroplasty: a prospective study of the prosthetic tip. J. Nucl. Med.,30: 1321-1331, 1989.301321 1989 [PubMed]

Oyen, W. J.; Claessens, R. A.; van der Meer, J. W.; and Corstens, F. H.: Detection of subacute infectious foci with indium-111-labeled autologous leukocytes and indium-111-labeled human nonspecific immunoglobulin G: a prospective comparative study. J. Nucl. Med.,32: 1854-1860, 1991.321854 1991 [PubMed]

Oyen, W. J.; van Horn, J. R.; Claessens, R. A.; Slooff, T. J.; van der Meer, J. W.; and Corstens, F. H.: Diagnosis of bone, joint, and joint prosthesis infections with In-111-labeled nonspecific human immunoglobulin G scintigraphy. Radiology,182: 195-199, 1992.182195 1992 [PubMed]

Palestro, C. J.; Kim, C. K.; Swyer, A. J.; Capozzi, J. D.; Solomon, R. W.; and Goldsmith, S. J.: Total-hip arthroplasty: periprosthetic indium-111-labeled leukocyte activity and complementary technetium-99m-sulfur colloid imaging in suspected infection. J. Nucl. Med.,31: 1950-1955, 1990.311950 1990 [PubMed]

Phillips, W. C., and Kattapuram, S. V.: Efficacy of preoperative hip aspiration performed in the radiology department. Clin. Orthop.,179: 141-146, 1983.179141 1983 [PubMed]

Pring, D. J.; Henderson, R. G.; Rivett, A. G.; Krausz, T.; Coombs, R. R. H.; and Lavender, J. P.: Autologous granulocyte scanning of painful prosthetic joints. J. Bone and Joint Surg.,68-B(4): 647-652, 1986.68-B(4)647 1986

Rapley, R.; Theophilus, B. D.; Bevan, I. S.; and Walker, M. R.: Fundamentals of the polymerase chain reaction: a future in clinical diagnostics?. Med. Lab. Sci.,49: 119-128, 1992.49119 1992 [PubMed]

Roberts, P.; Walters, A. J.; and McMinn, D. J. W.: Diagnosing infection in hip replacements. The use of fine-needle aspiration and radiometric culture. J. Bone and Joint Surg.,74-B(2): 265-271, 1992.74-B(2)265 1992

Rubin, R.; Salvati, E. A.; and Lewis, R.: Infected total hip replacement after dental procedures. Oral Surg., Oral Med., Oral Pathol.,41: 18-23, 1976.4118 1976

Salvati, E. A.; Robinson, R. P.; Zeno, S. M.; Koslin, B. L.; Brause, B. D.; and Wilson, P. D., Jr.: Infection rates after 3175 total hip and total knee replacements performed with and without a horizontal unidirectional filtered air-flow system. J. Bone and Joint Surg.,64-A: 525-535, April 1982.64-A525 1982

Sanzen, L., and Carlsson, A. S.: The diagnostic value of C-reactive protein in infected total hip arthroplasties. J. Bone and Joint Surg.,71-B(4): 638-641, 1989.71-B(4)638 1989

Schmidt, R. M., and Rosenkranz, H. S.: Antimicrobial activity of local anesthetics: lidocaine and procaine. J. Infect. Dis.,121: 597-607, 1970.121597 1970 [PubMed]

Schutzer, S. F., and Harris, W. H.: Deep-wound infection after total hip replacement under contemporary aseptic conditions. J. Bone and Joint Surg.,70-A: 724-727, June 1988.70-A724 1988

Sculco, T. P.: The economic impact of infected total joint arthroplasty. In Instructional Course Lectures, The American Academy of Orthopaedic Surgeons. Vol. 42, pp. 349-351. Rosemont, Illinois, The American Academy of Orthopaedic Surgeons, 1993.

Serafini, A. N.; Garty, I.; Vargas-Cuba, R.; Friedman, A.; Rauh, D. A.; Neptune, M.; Landress, L.; and Sfakianakis, G. N.: Clinical evaluation of a scintigraphic method for diagnosing inflammations/infections using indium-111-labeled nonspecific human IgG. J. Nucl. Med.,32: 2227-2232, 1991.322227 1991 [PubMed]

Shih, L.-Y.; Wu, J.-J.; and Yang, D.-J.: Erythrocyte sedimentation rate and C-reactive protein values in patients with total hip arthroplasty. Clin. Orthop.,225: 238-246, 1987.225238 1987 [PubMed]

Spangehl, M. J.; Duncan, C. P.; O'Connell, J. X.; and Masri, B. A.: Prospective analysis of preoperative and intraoperative studies for the diagnosis of infection in 210 consecutive revision total hip arthroplasties. Read at the Annual Meeting of The American Academy of Orthopaedic Surgeons, San Francisco, California, Feb. 15, 1997.

Stinchfield, F. E.; Bigliani, L. U.; Neu, H. C.; Goss, T. P.; and Foster, C. R.: Late hematogenous infection of total joint replacement. J. Bone and Joint Surg.,62-A: 1345-1350, Dec. 1980.62-A1345 1980

Tehranzadeh, J.; Gubernick, L.; and Blaha, D.: Prospective study of sequential technetium-99m phosphate and gallium imaging in painful hip prostheses (comparison of diagnostic modalities). Clin. Nucl. Med.,13: 229-236, 1988.13229 1988 [PubMed]

Thoren, B., and Wigren, A.: Erythrocyte sedimentation rate in infection of total hip replacements. Orthopedics,14: 495-497, 1991.14495 1991 [PubMed]

Tietjen, R.; Stinchfield, F. E.; and Michelsen, C. B.: The significance of intracapsular cultures in total hip operations. Surg., Gynec. and Obstet.,144: 699-702, 1977.144699 1977

Tigges, S.; Stiles, R. G.; Meli, R. J.; and Roberson, J. R.: Hip aspiration: a cost effective and accurate method of evaluating the potentially infected hip prosthesis. Radiology,189: 485-488, 1993.189485 1993 [PubMed]

Tsukayama, D. T.; Estrada, R.; and Gustilo, R. B.: Infection after total hip arthroplasty. A study of the treatment of one hundred and six infections. J. Bone and Joint Surg.,78-A: 512-523, April 1996.78-A512 1996

Vanderhooft, J. E., and Robinson, R. P.: Late infection of a bipolar prosthesis following endoscopy. A case report. J. Bone and Joint Surg.,76-A: 744-746, May 1994.76-A744 1994

Wegener, W. A., and Alavi, A.: Diagnostic imaging of musculoskeletal infection. Roentgenography; gallium, indium-labeled white blood cell, gammaglobulin, bone scintigraphy; and MRI. Orthop. Clin. North America,22: 401-418, 1991.22401 1991

Wegener, W. A.; Velchik, M. G.; Weiss, D.; Ter, S.; Byars, A.; Neptune, M.; and Alavi, A.: Infectious imaging with indium-111-labeled nonspecific polyclonal human immunoglobulin. J. Nucl. Med.,32: 2079-2085, 1991.322079 1991 [PubMed]

Wukich, D. K.; Abreu, S. H.; Callaghan, J. J.; Van Nostrand, D.; Savory, C. G.; Eggli, D. F.; Garcia, J. E.; and Berrey, B. H.: Diagnosis of infection by preoperative scintigraphy with indium-labeled white blood cells. J. Bone and Joint Surg.,69-A: 1353-1360, Dec. 1987.69-A1353 1987

Topics

View Large | Download Slide(.ppt)
Add to My JBJS

View Larger

TABLE I SENSITIVITY AND SPECIFICITY OF THE ERYTHROCYTE SEDIMENTATION RATE AND THE C-REACTIVE PROTEIN LEVEL

Laboratory Screening Test	No. of Patients	No. Who Had an Infection	Value for Erythrocyte Sedimentation Rate (mm/hr.)	Value for C-Reactive Protein Level (mg/L)	Sensitivity	Specificity
Erythrocyte sedimentation rate
Sanzen and Carlsson⁵⁹	56	23	>30		0.61	1.00
Thoren and Wigren⁶⁸	79	51	>35		0.88	0.96
Roberts et al.⁵⁶	69	14	>30		0.84	0.79
Spangehl et al.⁶⁵	171	34	>30		0.82	0.85

C-reactive protein level
Sanzen and Carlsson⁵⁹	56	23		>10	0.91	0.88
Spangehl et al.⁶⁵	142	26		>10	0.96	0.92

Add to My JBJS

TABLE I SENSITIVITY AND SPECIFICITY OF THE ERYTHROCYTE SEDIMENTATION RATE AND THE C-REACTIVE PROTEIN LEVEL

Laboratory Screening Test	No. of Patients	No. Who Had an Infection	Value for Erythrocyte Sedimentation Rate (mm/hr.)	Value for C-Reactive Protein Level (mg/L)	Sensitivity	Specificity
Erythrocyte sedimentation rate
Sanzen and Carlsson⁵⁹	56	23	>30		0.61	1.00
Thoren and Wigren⁶⁸	79	51	>35		0.88	0.96
Roberts et al.⁵⁶	69	14	>30		0.84	0.79
Spangehl et al.⁶⁵	171	34	>30		0.82	0.85

C-reactive protein level
Sanzen and Carlsson⁵⁹	56	23		>10	0.91	0.88
Spangehl et al.⁶⁵	142	26		>10	0.96	0.92

View Larger

TABLE II SENSITIVITY AND SPECIFICITY OF NUCLEAR IMAGING

*The study included various types of infections; not all were related to the prosthesis.

Imaging Modality	No. of Sites Tested	No. of Proved Infections	Sensitivity	Specificity
Technetium/gallium scanning
Kraemer et al.³⁴	43	13	0.38 (5/13)	1.00 (30/30)
Merkel et al.⁴³*	42	24	0.50 (12/24)	0.78 (14/18)
Indium-111-labeled whiteblood-cell scanning
Merkel et al.⁴³*	42	24	0.83 (20/24)	0.94 (17/18)
Glithero et al.²³	25	8	0.38 (5/13)	1.00 (12/12)
Wukich et al.⁷⁵	24	7	1.00 (7/7)	0.41 (7/17)
Johnson et al.³³	29	9	1.00 (9/9)	0.50 (10/20)
Technetium/indium-111-labeled white-blood-cell scanning
Johnson et al.³³	29	9	0.89 (8/9)	0.95 (19/20)
Palestro et al.⁵²	50	10	1.00 (10/10)	0.98 (39/40)
Indium-111-labeled immunoglobulin-G scanning
Oyen et al.¹⁵*
Entire study	120	72	0.97 (70/72)	0.85 (41/48)
Failed arthroplasties only	37	12	0.92 (11/12)	0.88 (22/25)
Wegener et al.^*74	15	11	0.91 (10/11)	1.00 (4/4)

Add to My JBJS

TABLE II SENSITIVITY AND SPECIFICITY OF NUCLEAR IMAGING

*The study included various types of infections; not all were related to the prosthesis.

Imaging Modality	No. of Sites Tested	No. of Proved Infections	Sensitivity	Specificity
Technetium/gallium scanning
Kraemer et al.³⁴	43	13	0.38 (5/13)	1.00 (30/30)
Merkel et al.⁴³*	42	24	0.50 (12/24)	0.78 (14/18)
Indium-111-labeled whiteblood-cell scanning
Merkel et al.⁴³*	42	24	0.83 (20/24)	0.94 (17/18)
Glithero et al.²³	25	8	0.38 (5/13)	1.00 (12/12)
Wukich et al.⁷⁵	24	7	1.00 (7/7)	0.41 (7/17)
Johnson et al.³³	29	9	1.00 (9/9)	0.50 (10/20)
Technetium/indium-111-labeled white-blood-cell scanning
Johnson et al.³³	29	9	0.89 (8/9)	0.95 (19/20)
Palestro et al.⁵²	50	10	1.00 (10/10)	0.98 (39/40)
Indium-111-labeled immunoglobulin-G scanning
Oyen et al.¹⁵*
Entire study	120	72	0.97 (70/72)	0.85 (41/48)
Failed arthroplasties only	37	12	0.92 (11/12)	0.88 (22/25)
Wegener et al.^*74	15	11	0.91 (10/11)	1.00 (4/4)

View Larger

TABLE III SENSITIVITY AND SPECIFICITY OF PREOPERATIVE ASPIRATION

Study	No. of Sites Tested	No. of Proved Infections	Sensitivity	Specificity
Phillips and Kattapuram⁵³	141	33	0.91 (30/33)	0.82 (89/108)
Tigges et al.⁷⁰	147	14	0.93 (13/14)	0.92 (122/133)
Kraemer et al.³⁴	45	14	0.57 (8/14)	0.97 (30/31)
Roberts et al.⁵⁶	78	15	0.87 (13/15)	0.95 (60/63)
Barrack and Harris³	260	4	0.50 (2/4)	0.88 (224/256)
Lachiewicz et al.³⁶	156	25	0.92 (23/25)	0.97 (127/131)
Mulcahy et al.⁴⁵	71	16	0.69 (11/16)	0.91 (50/55)
Fehring and Cohen¹⁵	166	6	0.50 (3/6)	0.88 (141/160)
Spangehl et al.⁶⁵	180	21	0.86 (18/21)	0.94 (149/159)

Add to My JBJS

TABLE III SENSITIVITY AND SPECIFICITY OF PREOPERATIVE ASPIRATION

Study	No. of Sites Tested	No. of Proved Infections	Sensitivity	Specificity
Phillips and Kattapuram⁵³	141	33	0.91 (30/33)	0.82 (89/108)
Tigges et al.⁷⁰	147	14	0.93 (13/14)	0.92 (122/133)
Kraemer et al.³⁴	45	14	0.57 (8/14)	0.97 (30/31)
Roberts et al.⁵⁶	78	15	0.87 (13/15)	0.95 (60/63)
Barrack and Harris³	260	4	0.50 (2/4)	0.88 (224/256)
Lachiewicz et al.³⁶	156	25	0.92 (23/25)	0.97 (127/131)
Mulcahy et al.⁴⁵	71	16	0.69 (11/16)	0.91 (50/55)
Fehring and Cohen¹⁵	166	6	0.50 (3/6)	0.88 (141/160)
Spangehl et al.⁶⁵	180	21	0.86 (18/21)	0.94 (149/159)

View Larger

TABLE IV SENSITIVITY AND SPECIFICITY OF INTRAOPERATIVE FROZEN SECTIONS

Study	No. of Sites Tested	No. of Proved Infections	Sensitivity	Specificity
Fehring and McAlister¹⁶	97	11	0.18 (2/11)	0.90 (77/86)
Feldman et al.¹⁸	33	9	1.00 (9/9)	0.96 (23/24)
Athanasou et al.²	106	22	0.91 (20/22)	0.96 (81/84)
Lonner et al.³⁹	175	19	0.84 (16/19)	0.99 (154/156)
Fehring and Cohen¹⁵	130	4	0.50 (2/4)	0.90 (114/126)
Spangehl et al.⁶⁵	202	35	0.80 (28/35)	0.94 (157/167)

Add to My JBJS

TABLE IV SENSITIVITY AND SPECIFICITY OF INTRAOPERATIVE FROZEN SECTIONS

Study	No. of Sites Tested	No. of Proved Infections	Sensitivity	Specificity
Fehring and McAlister¹⁶	97	11	0.18 (2/11)	0.90 (77/86)
Feldman et al.¹⁸	33	9	1.00 (9/9)	0.96 (23/24)
Athanasou et al.²	106	22	0.91 (20/22)	0.96 (81/84)
Lonner et al.³⁹	175	19	0.84 (16/19)	0.99 (154/156)
Fehring and Cohen¹⁵	130	4	0.50 (2/4)	0.90 (114/126)
Spangehl et al.⁶⁵	202	35	0.80 (28/35)	0.94 (157/167)