 |
 |
 |
|
| home | CME | help | feedback | subscriptions | archive |
|
© 1997 The Journal of Bone and Joint Surgery, Inc. Wear of Polyethylene Acetabular Components in Total Hip Arthroplasty. An Analysis of One Hundred and Twenty-eight Components Retrieved at Autopsy or Revision Operations*  , BOSTON, MASSACHUSETTSInvestigation performed at the Orthopaedic Biomechanics Laboratory and the Hip and Implant Unit, Department of Orthopaedic Surgery, Massachusetts General Hospital and Harvard Medical School, Boston
We evaluated the rates of volumetric wear and the patterns of wear of 128 acetabular components retrieved during an autopsy or a revision operation between one and twenty-one years after total hip arthroplasty. Twenty-two all-polyethylene components were retrieved at autopsy from hips that had been functioning well at the time of death (Group A). The remaining 106 components—eighty-four all-polyethylene components (Group B) and twenty-two metal-backed components (Group C)—were retrieved during revision operations. All 128 components had been inserted with cement. The mean rate of volumetric wear, determined directly with a fluid-displacement method, was thirty-five cubic millimeters per year (range, eight to 116 cubic millimeters per year) for Group A, sixty-two cubic millimeters per year (range, eight to 256 cubic millimeters per year) for Group B, and ninety-four cubic millimeters per year (range, twelve to 284 cubic millimeters per year) for Group C. Multivariate regression analysis showed a significant relationship (p < 0.05) between the size of the femoral head and the calculated mean annual rate of volumetric wear. The rate of volumetric wear was highest in association with thirty-two-millimeter femoral heads and lowest in association with twenty-two-millimeter heads; according to linear regression analysis, this represented a 7.5 per cent increase (Group A) or a 10 per cent increase (Group B) in the rate of wear for every one-millimeter increase in the size of the head. Linear regression analysis also showed a significant relationship between the duration that the implant had been in situ and the rate of wear (p < 0.05), with the rate being highest initially after the operation and decreasing with an increasing duration in situ. With the numbers available, the patient's age and gender and the side of the arthroplasty did not have a significant relationship to the annual rate of volumetric wear. Increased thickness of the polyethylene was related to a decreased rate of wear (p < 0.05) in the group of metal-backed components, which had a 25 per cent increase in the rate of wear for every one-millimeter decrease in thickness, but not in the other groups. The estimated median annual rates of wear, after adjustment of confounding variables to a hypothetical constant set of median values for the parameters (duration in situ, 132 months; diameter of the femoral head, twenty-six millimeters; and thickness of the polyethylene, eight millimeters), were significantly different among the three groups of components (p < 0.05). Histological evaluation of the worn surfaces showed the predominant mechanisms of wear to be abrasion and adhesion rather than fatigue-cracking or delamination. The highly worn areas were polished to a glassy finish on gross examination, but scanning electron microscopy showed numerous multidirectional scratches along with fine, drawn-out fibrils with a diameter of one micrometer or less oriented parallel to each other. These fibrils are the most likely source of submicrometer wear particles. Thus, wear appeared to occur mostly at the surface of the components and to be due to large-strain plastic deformation and orientation of the surface layers into fibrils that subsequently ruptured during multidirectional motion.
Total joint replacement with components made of metal alloys and synthetic polymers has revolutionized the treatment of end-stage osteoarthrosis. Because problems with fixation and infection have been reduced over the last several decades, wear of the polyethylene against the metal head at the articulation and periprosthetic osteolysis due to biological reactions to the wear debris are now emerging as the most important unsolved problems6,7. Radiographic measurement of the rate of wear at metal-to-polyethylene articulations has revealed a mean linear rate of penetration of the femoral head into the socket of approximately 0.1 cubic millimeter per year3. However, there has been a wide range, with rates as high as 0.6 cubic millimeter per year3. Variables that have been implicated in the acceleration of polyethylene wear include the patient's age, level of activity, and body weight; the thickness of the liner; the size of the femoral head; and oxidative degradation in vivo1,3,9. However, mechanisms of wear that most commonly occur only in vivo and the effects of the various clinical variables on the rates of wear remain incompletely understood. Examination of components after they have been retrieved from the body provides a means with which to investigate the rates and mechanisms of wear in vivo. However, some previous studies5,8 have been hampered by the unavailability of a large number of components of different designs, incomplete clinical and radiographic data, difficulties in examining the worn surfaces under high magnification, inclusion of knee components (which have different loading characteristics and therefore would be expected to have different mechanisms of wear) along with hip components, and difficulties inherent in interpreting the features of wear. The current study was undertaken to evaluate the amounts and patterns of wear, and the clinical variables that may have affected wear, in 128 polyethylene acetabular components that were retrieved at autopsy or during a revision operation. The volumetric wear was measured directly, as accurate measurement is critical for the evaluation of factors that can influence the rate of wear. Furthermore, the inclusion of specimens retrieved at autopsy from patients who had had a successful total hip replacement allowed evaluation of the rate of wear before loosening of the implant.
One hundred and twenty-eight polyethylene acetabular components of various designs were retrieved from 117 patients or cadavera twelve to 252 months after total hip arthroplasty. Clinical and radiographic data were available for all patients. The patient's age at the time of the operation, gender, primary diagnosis, and weight; the type of components; the status of the fixation of the components; the reason for removal; and the duration of the operation were recorded (Table I). Twenty-two all-polyethylene components were retrieved at autopsy from hips that had been functioning well at the time of death; none of these implants had been loose (Group A). The remaining 106 components were retrieved during a revision operation for prosthetic loosening, periprosthetic osteolysis, dislocation, or infection. These included eighty-four all-polyethylene components (Group B) and twenty-two metal-backed components (Group C). The mean duration that the components had been in situ was 149 months (range, twelve to 252 months) for Group A, 127 months (range, twenty-four to 237 months) for Group B, and 131 months (range, thirty-two to 228 months) for Group C.
All of the acetabular and femoral components had been inserted with cement. The polyethylene in almost all of the acetabular components (Harris Design-1, Harris Design-2, Müller, Aufranc-Turner, Charnley, Harris metal-backed, and Harris mini; Howmedica, Rutherford, New Jersey) was machined from RCH-1000 bar stock (Hoechst, Frankfurt, Germany). All of these components were sterilized with use of gamma irradiation in air. The polyethylene in the Ti-Bac components (Zimmer, Warsaw, Indiana) was machined from Hifax bar stock (Himont, Wilmington, Delaware) and was sterilized with use of gamma irradiation. The Trapezoidal-28 components (Zimmer) were direct-compression-molded from Hifax bar stock and were sterilized with use of ethylene oxide. Information on the few remaining components was not available. Four femoral components (Bechtol; Richards Medical, Memphis, Tennessee) were made of stainless-steel alloy. One Trapezoidal-28 component had a titanium-alloy femoral head. All of the remaining femoral components were made from cobalt-chromium-molybdenum alloy. (The Harris Design-1, Müller, Aufranc-Turner, CAD, Charnley, Harris metal-backed, and Harris CDH components were made from cast alloys, and the Harris Design-2 and Trapezoidal-28 components were made from forged alloys.) Titanium alloy was used for the metal backing of seven acetabular components (Omnifit [Osteonics, Allendale, New Jersey] and Ti-Bac). All liners of the metal-backed components were securely fixed in the shell. The Harris metal-backed components were preassembled at the factory. All components were manufactured before 1985, with the exception of the Ti-Bac, Omnifit, and anatomic medullary locking (AML; DePuy, Warsaw, Indiana) components. After retrieval, all components were inspected visually for the presence of damage due to wear. All components were evaluated for the presence of a high-wear area and a low-wear area at the articulation and for the formation of a ridge between the two, as described by Dowling et al. The components were then examined under a dissection microscope at a magnification of ten times. The features indicating wear damage were recorded for both the high and the low-wear area. These features included wear-polishing (a smooth, shiny area on the articulating surface); fine scratching (scratches fifty micrometers or less in width); flaking (delamination of the subsurface); pitting; cracking; coarse, abrasive wear (roughening of the surface with scratches 100 micrometers or more in width); embedded particles of metal or cement; and discoloration. Each feature was graded in a semiquantitative fashion on a scale ranging from grade 0 to grade 3 (with grade 3 being the most severe). The amount of wear of the polyethylene at the articulating surface was quantitated with a volumetric method. The wear of the polyethylene was almost always eccentric to the apex of the liner; the worn surface could usually be distinguished by the presence of a distinct contour that was different from the original articulating contour and was separated by a transitional ridge. The volume of wear was determined by inserting a femoral head of the appropriate size into the original contour and measuring the volume of fluid (oil) that was required to fill the remaining contour. Part of the original contour, with a distinct radius of curvature into which the femoral head could be placed, had always been preserved in these components. The reproducibility of this method for the determination of volumetric wear was assessed by obtaining multiple measurements in forty of the worn and ten of the new components. The interobserver variability among three different technicians was within 10 per cent. The intraobserver variability among multiple measurements made by one technician was also within 10 per cent. The accuracy of this method was assessed by removing a known amount of material from the articulation of new components with a ball end mill and measuring the volume with use of the fluid-displacement method. The technique was found to be accurate (95 per cent confidence interval) to within 10 per cent. Using multivariate regression analysis, we analyzed the influence of the patient's age, gender, and weight; the side of involvement; the design of the femoral component; the type of retrieval (autopsy or revision operation); the duration that the implant had been in situ; the size of the femoral head; and the thickness of the polyethylene on the rate of wear. Because of possible differences in the demographic characteristics among the three groups, each group was also analyzed separately with use of multivariate analysis. Representative components were then divided in half with a band-saw, with separation of the anterior half from the posterior half parallel to the direction of the wear. Each half was coated with gold and was examined under a scanning electron microscope (Cambridge Instruments, Cambridge, United Kingdom) with use of secondary electrons at ten kilovolts. Fine details of the wear morphology were evaluated with the scanning electron microscope.
The annual rates of wear ranged widely, from eight to 284 cubic millimeters, among the patients (Table I). The mean annual rate of wear was thirty-five cubic millimeters (range, eight to 116 cubic millimeters) for the all-polyethylene components retrieved at autopsy (Group A), sixty-two cubic millimeters (range, eight to 256 cubic millimeters) for the all-polyethylene components retrieved during revision (Group B), and ninety-four cubic millimeters (range, twelve to 284 cubic millimeters) for the metal-backed polyethylene components retrieved during revision (Group C). The annual rate of wear was between 100 and 200 cubic millimeters for one specimen (5 per cent) in Group A, eleven specimens (13 per cent) in Group B, and seven specimens (32 per cent) in Group C. The annual rate was more than 200 cubic millimeters for none of the specimens in Group A, one specimen (1 per cent) in Group B, and two specimens (9 per cent) in Group C. Multivariate analysis of all 128 components showed significant relationships (p < 0.05) between the annual rate of volumetric wear and the size of the femoral head (the rate was highest in association with the thirty-two-millimeter heads, intermediate in association with the twenty-eight and twenty-six-millimeter heads, and lowest in association with the twenty-two-millimeter heads), the duration that the implant had been in situ (the rate decreased in association with increased durations in situ), the type of retrieval (the rate was lower in association with specimens retrieved at autopsy as compared with those retrieved during revision), and the presence of a metal backing (the rate was lower in association with the all-polyethylene components as compared with the metal-backed polyethylene components). None of the other variables, including the patient's gender or age, the side of involvement, the design of the femoral component, or the thickness of the polyethylene, had a significant relationship to volumetric wear. The patient's age, gender, and weight and the side of the arthroplasty did not have significant relationships to the rate of wear in any group. The size of the femoral head had a significant relationship to the rate of wear (p < 0.05) in Groups A and B; Group C did not have enough variability in the size of the head to allow valid statistical statements. The number of acetabular components with a twenty-two-millimeter inner diameter was too small in Groups A and C for any meaningful data to be derived. After exclusion of the twenty-two-millimeter heads, linear regression analysis demonstrated that for every one-millimeter increase in the diameter of the head there was a 7.5 per cent increase in the annual rate of volumetric wear in Group A. With inclusion of the twenty-two-millimeter heads, it was found that for every one-millimeter increase in the diameter of the head there was a 10 per cent increase in the annual rate of volumetric wear in Group B. The duration that the implant had been in situ also had a significant relationship (p < 0.05) to the rate of wear in each group. For every yearly increase in the duration there was a 9 per cent decrease in the rate of wear in Group A, a 7.7 per cent decrease in Group B, and an 11 per cent decrease in Group C. A finding of special interest was that the thickness of the polyethylene did not have a significant relationship to the annual rate of wear in either group of all-polyethylene components (Groups A and B). However, a significant relationship was found between the thickness of the polyethylene and the rate of wear (p < 0.05) of the metal-backed components (Group C). In this group, for every one-millimeter decrease in the thickness of the polyethylene the rate of wear increased by 25 per cent, according to linear regression analysis. The estimated median rates of wear were compared among the three groups after adjustment of the important confounding variables to a hypothetical constant set of median values for the parameters (duration in situ, 132 months; diameter of the femoral head, twenty-six millimeters; and thickness of the polyethylene, eight millimeters). The estimated median rates of wear were significantly different (p < 0.05) between Group A and Group B and between Group B and Group C. The wear at the articulating surface was characterized by a highly worn polished area superiorly and a less worn area inferiorly, separated by a transitional ridge, in all components (Fig. 1). In general, the transition from the high-wear area to the low-wear area was less prominent in components with a thirty-two-millimeter inner diameter than in those with a twenty-two, twenty-six, or twenty-eight-millimeter inner diameter, and it was also less prominent in specimens that had been in situ for a shorter duration than in those that had been in situ for a longer duration.
The morphological appearance of the wear surfaces was similar in all groups of components (including those retrieved at autopsy), but the frequency and severity of these features and the rates of wear varied widely (Table II). The morphological studies indicated that most of the wear was at the superior portion of the articulating surface and that the predominant mechanisms of wear of the polyethylene were abrasion and adhesion rather than fatigue-cracking or delamination of the subsurface. Wear was characterized by multidirectional scratching on a highly reflective surface in 117 (91 per cent) of the 128 components. There were numerous fine scratches running in multiple directions in the highly worn area, indicating three-body abrasive wear. Pitting, cracking, and delamination were extremely uncommon in the highly worn area but were present in the less worn area or the transitional ridge. The less worn area was yellowish in ninety-four components (73 per cent), but discoloration of the highly worn area was uncommon. Impingement of the femoral neck and subsequent deformation were noted in fifty-seven components (45 per cent).
Scanning electron microscopy showed several additional fine details of the worn surfaces (Fig. 2). The multidirectional fine scratches ranged from less than one micrometer to twenty micrometers. There were periodic parallel striations perpendicular to the direction of many of the scratches (Fig. 3). At higher magnification, these striations were visualized as incomplete tears in the material that had occurred during abrasive wear. In areas where the scratches crossed, material that appeared to have been displaced by one scratch seemed to have been reorganized and displaced further by a second scratch that had occurred later. This indicated a repetitive process of wear-polishing of the material by constant formation and obliteration of the scratch marks running in multiple directions.
At higher magnification, some of the worn surfaces had a fine fibrillar appearance, suggesting that the material at the surface had been pulled and drawn into fine fibrils oriented parallel to each other during abrasive or adhesive wear (Fig. 4). These fibrils were one micrometer or less in diameter and were loosely connected to each other. Most often, they were tangential to the surface and were organized parallel to one another along the direction of the adjacent scratches. Within the scratches themselves, the fibrils were oriented at a 0 to 45-degree angle to the direction of the scratches. In other areas, the fibrils were shorter, perpendicular to the surface, and oriented randomly, but they were less numerous than the highly oriented fibrils. These findings suggest that, with repeated abrasion and adhesion, the polyethylene at the articulation had been elongated and drawn into highly oriented fine fibrils. Some of these features of wear-polishing and fine scratching were found relatively early after the operation (even in some specimens that were retrieved at less than two years), whereas the formation of fine fibrils was usually observed in specimens that had been in place for more than five years. The components with the most severe wear showed massive disorganization of the surfaces, sometimes making it difficult for us to determine the actual mechanism of wear. However, the appearance of the fine fibrils in many of the specimens suggests that considerable deformation of the surface occurred during the wear process and that rearrangement of the molecular chain, from folded to extended chains, may have been taking place. The submicrometer size of the fibrils is also consistent with the submicrometer wear particles previously reported in periprosthetic tissues7.
These observations of multidirectional scratches, wear-polishing, and formation of fine fibrils in highly worn areas, combined with the relative paucity of cracks and delamination, suggest that the dominant mechanisms involved in the wear of the acetabular components were abrasion and adhesion at the surface. Stress-related cracking or delamination and corrosion did not appear to have been a major cause of wear at the articulation of the polyethylene acetabular components.
With the availability of a large number of acetabular components, it has been possible to evaluate the amounts and mechanisms of wear as well as the clinical variables that influence wear. The observations made in the current study provide several insights. The annual, volumetric, in vivo rate of wear of the well fixed all-polyethylene acetabular components retrieved at autopsy averaged thirty-five cubic millimeters. This rate was substantially higher in the failed components that were retrieved during revision operations, regardless of the type of design or fixation of the implant, suggesting either that the components failed because of a high rate of wear or that the high rate of wear occurred in the components that failed. The annual volumetric rates of wear were the highest initially after the operation and decreased with time in situ. Increases in the size of the femoral head and the presence of metal backing on the acetabular component appeared to lead to increased rates of wear, whereas the patient's age, weight, and gender and the side of involvement had no effect. Decreased thickness of the polyethylene had an adverse effect on the rate of wear of the metal-backed components but did not influence the rate of wear of the all-polyethylene components. The appearance of the worn surfaces suggests that polyethylene wear in the hip is predominantly due to abrasion or adhesion rather than to fatigue-cracking or delamination. Three-body abrasive wear-polishing was a dominant mode of wear in the current study and led to the production of fine (one-micrometer-diameter or smaller) particles of polyethylene wear debris. The relationship between decreased rates of wear and increased time in situ, which was also observed by Charnley and Halley, has several implications. It is possible that a so-called wearing-in phenomenon occurs early on, during which time the incongruities between the femoral head and the acetabular polyethylene are worn away at a higher rate. The articulation may then become more congruous, leading to a lower rate of wear. The data showing decreased rates of wear with time are not consistent with the theory that oxidative degradation occurs in vivo and is responsible for most wear. Such a phenomenon should lead to higher rates of wear in association with increased durations in situ, although extensive oxidative degradation in vivo may contribute to wear in some components. These data also are not consistent with the theory that subsurface fatigue is a dominant mechanism of wear, as such fatigue would be expected to lead to higher rates of catastrophic wear at late time-periods after the operation. The data are more consistent with abrasive and adhesive mechanisms of wear, which tend to produce higher rates of wear initially because of surface incongruities and lower rates after the initial wearing-in period has taken place. It was surprising that the patient's age, gender, and weight and the side of involvement did not have a significant relationship to the rate of wear. Charnley and Halley made similar observations on the basis of radiographic measurements. This finding is also consistent with abrasive and adhesive mechanisms of wear rather than a fatigue mechanism, as with the former mechanisms the distance traveled by the femoral head within the acetabular polyethylene (sliding distance) would be a more important determinant of wear than the loads imposed on the hip. It is generally assumed that in younger patients, heavier patients, and men the hips are subjected to a higher load and more repetitions of load and that they therefore are more likely to have higher rates of wear. However, these parameters may not be good indicators of levels of activity or of the number of cycles imposed on the hip each year. The number of cycles and the sliding distance may be more important determinants of wear than the joint load if abrasion and adhesion are the predominant mechanisms of wear. The relationship between the size of the femoral head and the rate of wear in two of the groups in this study is also consistent with abrasive and adhesive mechanisms of wear, with which a larger surface area of contact is expected to produce a higher rate of wear. Livermore et al., in a radiographic follow-up study, found that thirty-two-millimeter heads were associated with more wear than twenty-eight-millimeter heads were. However, these authors did not find a lower rate of volumetric wear in association with twenty-two-millimeter heads. Clarke evaluated the effect of the size of the femoral head on the rate of wear with use of an experimental wear simulator; he found the lowest rate of volumetric wear in association with twenty-two-millimeter heads, an intermediate rate in association with twenty-six-millimeter heads, and the highest rate in association with thirty-two-millimeter heads. These findings are consistent with the findings in the current study, although the number of twenty-two-millimeter heads was too small in two of our groups to allow any firm conclusions. It should be realized that, while twenty-two-millimeter heads may produce the lowest rates of wear, it may not be desirable to use such a head in most patients because of the higher potential for postoperative dislocation, impingement after wear has occurred, and wearing through due to the higher rates of linear penetration than is found with the twenty-six-millimeter heads9. Larger-diameter (particularly thirty-two-millimeter) heads are clearly associated with higher rates of volumetric wear and increased production of polyethylene debris and periprosthetic osteolysis, and their use therefore may be undesirable except in special circumstances. It is generally believed that thin-walled components are susceptible to high rates of wear, but we were not able to find such a trend in the non-metal-backed components in this study. This may have been due to the greater range of variability in the sizes of the heads articulating with the non-metal-backed components; the size of the head may have had a more profound influence on the rates of wear than did the thickness of the polyethylene. Conversely, the thickness of the polyethylene may be an important determinant of wear only when the components are metal-backed. These data should not be interpreted to mean that thin-walled acetabular components may be acceptable for insertion with cement; rather, it should be assumed that the thickness of the polyethylene is particularly important when the components are metal-backed. Many investigators have emphasized the influence of pits, delamination, and cracks in the polyethylene. Dowling et al. and Kurth et al. noted numerous pits, flakes, and cracks in polyethylene acetabular components that had been retrieved during revision operations. Although we found these features in a few components, they were rare and were essentially confined to the low-wear area and the transitional ridge. We believe that these features may be indicative of stress-related material deformation at the edges of loaded and unloaded areas of the articulation, similar to what occurs in the polyethylene tibial components of total knee prostheses because of high subsurface shear stresses. The contribution of these characteristics to the over-all wear of the polyethylene is very small in the hip, where the conformity at the articulation is high. The results of this study should not be interpreted to mean that other variables are not important in the wear of polyethylene. Sutula et al. showed that oxidative degradation can occur after irradiation in some components and that it is related to wear. We did not find evidence of substantial oxidative degradation or cracking in the components in our study. This may have been because of the types of resins used or because of differences in the methods of fabrication or the doses of radiation. Thus, these data are useful only for evaluating variables that influence wear in the absence of substantial oxidative degradation after irradiation. Similarly, many of the components were not metal-backed, and those that were had firm fixation of the polyethylene in the shell. They did not have modular connections and were not made of softer titanium or stainless-steel alloy. Such variables could also influence the rates of wear. The marked variability in the rates of wear in the current study, even in the well fixed components in Group A (range, eight to 116 cubic millimeters per year), suggests that other factors, such as the patient's level of activity, the design of the components, the offset of the femoral head, the surface finish, and differences in the polyethylene itself, may also be involved. The multidirectional scratches as well as the wear-polishing observed in the highly worn areas suggest that abrasion is an important mechanism of wear of polyethylene acetabular components. The repetitive abrasion of the articulating surface produced a fine fibrillar topography on the worn surfaces with time. Polymers such as polyethylene contain highly folded molecules arranged in lamellar form10. Abrasion of such materials in a single direction has been shown to lead to marked elongation and large-strain plastic deformation of the material at the surface, which in turn leads to the rearrangement and orientation of chain-folded molecules into chain-extended molecules. The surface layer of the material thus becomes anisotropic, oriented, and strain-hardened in the predominant direction of sliding10. This molecular rearrangement with repeated abrasion may be the basis for the formation of the fine fibrils that we observed at the articulating surfaces. The highly oriented fibrils are expected to be more brittle than the underlying bulk material and to be susceptible to fracture if they are abraded in directions perpendicular to the direction of orientation. Bragdon et al. found that ultra-high molecular weight polyethylene did not become measurably worn in a hip simulator when the femoral head was subjected to motions in a single direction (flexion-extension), but wear averaged thirty cubic millimeters when the femoral head was subjected to multidirectional motion (flexion-extension, internal-external rotation, and abduction-adduction). Their observation lends support to this theory. Thus, the wear of acetabular components appears to occur mostly at the surface and to be due to large deformation and orientation of the surface layers into fibrils and to subsequent rupture of these fibrils during multidirectional motion. These morphological features of the wear surfaces, combined with the data on the clinical variables that influenced the rates of wear, suggest that abrasion may be a dominant mechanism of wear of polyethylene acetabular components in vivo. The fine fibrils produced by these mechanisms are very small (one micrometer or less in diameter) and may represent the major source of submicrometer wear particles observed in previous histological studies7.
*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. Funds were received in total or partial support of the research or clinical study presented in this article. The funding source was the William H. Harris Foundation, Boston, Massachusetts.
  CiteULike  Connotea  Del.icio.us  Facebook  Technorati  Twitter What's this?
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Top
Abstract
Introduction
| home | CME | help | feedback | subscriptions | archive | site index |
|
Stanford University Libraries' HighWire Press (TM) assists in the publication of JBJS Online. |