We describe the findings at six years in an ongoing prospective clinicoradiological and metal ion study in a cohort of 26 consecutive male patients with unilateral Birmingham Hip Resurfacing arthroplasties with one of two femoral head sizes (50 mm and 54 mm). Their mean age was 52.9 years (29 to 67). We have previously shown an early increase in the 24-hour urinary excretion of metal ions, reaching a peak at six months (cobalt) and one year (chromium) after operation. Subsequently there is a decreasing trend in excretion of both cobalt and chromium. The levels of cobalt and chromium in whole blood also show a significant increase at one year, followed by a decreasing trend until the sixth year.

Introduction: Hip Resurfacing has always been an attractive concept for the treatment of hip arthritis in young patients. Excellent early and medium-term results have been reported with the Birmingham Hip Resurfacing (BHR) device in single and multi-surgeon all-diagnoses and OA series. In the present report we present the results of BHR in inflammatory arthritis.

Methods: This is a single-surgeon consecutive series. There were 15 consecutive hips (12 patients) including 2 women (2 hips) with ankylosing spondylitis (AS) operated at a mean age of 41.7 years (range 29.5 to 54.3 years). Fortytwo hips (31 patients) with seronegative or rheumatoid (RA) arthritis treated with a BHR at a mean age of 40 (13 to 64) years and a follow-up of 2 to 9 (mean 5.9) years were also studied. One patient died 5 years later. Revision for any reason was the end-point and unrevised patients were assessed with Oxford hip scores and reviewed clinico-radiologically with AP and lateral radiographs.

Results: In the RA group there was one failure from femoral neck fracture two months after operation giving a failure rate of 2.4%. There were no failures in this cohort at a follow-up of 1.8 to 8.8 (mean 4.9) years. As a combined group the failure rate of BHRs in inflammatory arthritis is 1.75% and the cumulative survivorship at 9 years is 98.2% (figure).

Discussion: The good results of Birmingham Hip Resurfacing in inflammatory arthritis in this relatively young cohort of patients make this a viable treatment option for these patients. Selection of patients with a reason-able bone quality and adherence to precise operative technique are vital to the success of this procedure.

Introduction: The potential adverse effects of metal ion elevation in patients with metal-metal bearings continue to be assessed. We reported earlier that metal ions cross the placenta. The present report is a comparison of the rate of transfer in 14 study patients (with MM devices) and 24 control subjects (with no metal devices).

Methods: Whole blood from concurrent specimens of maternal and umbilical cord blood obtained at the time of delivery were analysed with high resolution inductively coupled plasma mass spectrometry.

Results: Cobalt and chromium were detectable in all specimens in the control subjects and study patients. The mean difference between maternal and cord blood cobalt concentrations was 0.56 μg/l (p < 0.001) in the study group and 0.03 μg/l (p > 0.5) in the control group respectively (figure). The mean difference between maternal and cord blood chromium concentrations was 0.96 chromium (p < 0.0005) in the study group and 0.002 chromium (p > 0.5) in the control group respectively. The mean cord cobalt in the study patients was significantly higher than that in the control subjects (difference 0.38μg/l, p < 0.01) but the difference in the cord levels of chromium between study and controls (difference 0.13μg/l, p> 0.05) was not significant.

Discussion: There was almost no difference between the maternal and cord blood levels in the control group implying that the placenta offers almost no resistance to their passage in subjects without a metal device. In the study patients the mean cord cobalt level was 59% of the maternal level and the mean cord chromium level was 26% of the maternal level suggesting that the placenta exerts a modulatory effect on the rate of metal transfer when the maternal levels are higher

Introduction: The Birmingham Hip Resurfacing (BHR) device was introduced in 1997 as a conservative hip arthroplasty option for young patients with severe arthritis. Primary osteoarthritis is the most common etiology of hip arthritis in the West. Excellent early and medium-term results have been reported with the BHR. This is a 9–year review of the first 100 consecutive BHRs performed by one surgeon (DJWM) for primary OA.

Methods: The first 100 BHRs (91 patients) performed for OA are now at a mean follow-up of 9.1 (range 8.9 to 9.3) years. Four patients (5 hips) died 5.2 to 6.9 years later due to unrelated causes. Patients with unrevised hips were reviewed clinico-radiologically and with questionnaire assessment.

Results: With revision for any reason as the end-point there were four failures (two were post-operative AVN leading to femoral head collapse, 8 months and 7 years after their respective operations; and two were deep infections both 5 years after operation). Kaplan-Meier survival analysis showed a 96% cumulative survival at 9 years (figure). There were no failures from osteolysis or aseptic loosening and no patient is awaiting a revision for any reason.

Discussion: The performance of the BHR continues to be good at 9 years. Arthroplasty devices are known to manifest two phases of failure, one in the early years and another in the later years. Early failure with this device has been very low. The interim years are continuing to be promising and we are yet to find out when the late failures, if any, are likely to occur.

Introduction: source of elevated metal levels in patients with MM bearings continues to be debated. Under the controlled conditions of hip simulators most wear occurs within the first million cycles and subsequent motion produces negligible wear. However, in metal ion studies in patients, although there is a peak in metal levels in the early months after implantation, they never return to normal levels thereafter. In order to explain these conflicting observations it has been suggested that in vivo metal wear also occurs only during the early months and that continued corrosion of metal particles released during that period is responsible for metal level elevation later on.

If run-in wear is the only source of sustained metal release, then replacing the bearing with a non-MM bearing should not make a difference to metal release in patients and elevated levels should continue to persist even after such a revision. In order to verify this we studied metal release in patients who underwent revision of a MM bearing to a non-MM bearing after revision.

Methods. Using high resolution mass spectrometry, whole blood concentrations and daily output of cobalt and chromium were studied in four patients prospectively, whose hip resurfacings were revised to metal– polythylene THRs. None of the patients had other metal devices or compromised renal function.

Results. Preoperative levels in these patients were highly elevated as expected from a failing device. Thereafter there is a clear and progressively rapid trend of reducing metal levels in whole blood and urine.

Discussion The progressive and steep reduction of metal release following MM bearing removal suggests that corrosion from previously worn particles alone cannot account for the persistent elevation of systemic metal levels in patients with MM bearings and that bearing wear continues to occur after the initial run-in period

Introduction: The results of Birmingham Hip Resurfacing (BHR) device in several series reveal that the predominant mode of failure is femoral neck fracture or femoral head collapse and that careful patient selection and precise operative technique are vital to the success of this procedure. In this report we consider the results of BHR in patients with severe arthritis secondary to femoral head AVN.

Methods: This is a single-surgeon consecutive series of BHRs with a minimum follow-up of 5 years. Fifty six patients with Ficat-Arlet grade III or IV femoral head AVN and treated with BHRs at a mean age of 44.2 years (range 19 to 67.7 years) were followed-up for 5 to 9.2 years (mean 6.8 years). No patient died and none was lost to follow-up. Revision for any reason was the end-point and unrevised patients were assessed with Oxford hip scores. They were also reviewed clinically and with AP and lateral radiographs.

Results: There were four failures in this cohort giving a failure rate of 7% and a cumulative survivorship of 92.9% at 9 years (figure). In one further patient the femoral component has tilted into varus from further collapse of the femoral head. He is asymptomatic but knows that he will need a revision if he develops symptoms. No other patient shows clinical or radiological adverse signs.

Discussion: Several studies have shown that the results of arthroplasty are generally worse in AVN as compared to those in osteoarthritis. Reviewing the above results it appears to us that the relatively poorer cumulative survival observed in patients with a diagnosis of AVN (92.9%) compared to those with other diagnoses make AVN a relative contraindication to this procedure.

Introduction: Cementless cup designs in metal-on-metal (MoM) hip resurfacing devices generally depend on a good primary press-fit fixation which stabilises the components in the early post-operative period. Pressfitting the cup into the acetabulum generates non-uniform compressive stresses on the cup and consequently causes non-uniform cup deformation. That in turn may result in equatorial contact, high frictional torque and femoral head seizure. It has been reported that high frictional torque has the potential to generate micromotion between the implant and its surrounding bone and as a result adversely affect the longevity of the implant. The aim of this study was to investigate the effects of cup deformation on friction between the articulating surfaces in MoM bearings with various clearances.

Materials and methods: Six Birmingham Hip Resurfacing (BHR) devices with various clearances (80 to 306 μm) were tested in a hip friction simulator to determine the friction between the bearing surfaces. The components were tested in clotted blood which is the primary lubricant during the early post-operative period. The joints were friction tested initially in their pristine conditions and subsequently the cups were deflected by 25– 35 μm using two points pinching action before further friction tests were carried out.

Results and Discussions: It has been reported that reduced clearance results in reduced friction. However, none of the previous studies have taken cup deflection into consideration nor have they used physiologically relevant lubricant. The results presented in this study show that for the reduced clearance components, friction was significantly increased when the cups were deflected by only 30 μm. However, for the components with higher clearance the friction did not change before and after deflection. It is postulated that the larger clearances can accommodate for the amount of distortion introduced to the cups in this study.

Introduction: The increasing use of metal-metal bearings in hip arthroplasty raises concerns of elevated systemic metal levels. Lubrication theory and in-vitro studies suggest that larger diameter bearings generate less wear. Does this hold true in real life?

Methods: This is a cross-sectional study of 181 patients with either a Birmingham Hip Resurfacing or a 28mm Metasul THR. Whole blood levels (at 1year) and urinary metal ion output in the early (1 to 3) years and medium term (4 to 6years) in patients with these bearings were compared. None of the patients had other metal devices or compromised renal function. Both devices used a high-carbon cobalt-chrome bearing with no post-casting heat treatment. Cementless porous-ingrowth titanium cups and cemented polished tapered stainless steel stems or cementless porous-ingrowth titanium alloy stems were used in THRs.

Results: Whole blood metal levels at 1-year and daily output of metal ions in the early years and medium term (figure 1) in both bearings were in the same range and without a significant difference.

Discussion: Metal ion monitoring is the best way to estimate bearing wear in vivo and the best measure of device safety in the long term. Total wear over a given time period is best estimated from timed metal ion excretion rates. Blood levels represent a balance between the release of metal from the device and its renal clearance.

An earlier study (Clarke et al JBJS(Br) 2003) suggests that smaller bearings generate less wear. In that study bearings with different metallurgy and wear properties were grouped together, a potential confounding factor. The present study does not suffer from that error and our findings do not support the view that a larger bearing diameter leads to either an increase or decrease in metal ion generation.

Introduction. Painful post-collapse femoral head osteonecrosis (AVN) continues to be a therapeutic challenge. Joint preserving surgery does not produce satisfactory results after femoral head collapse, making an arthroplasty almost inevitable. Does metal-metal resurfacing offer a conservative option that matches the consistent results of a stemmed THA in these patients?

Methods. 104 consecutive resurfacings (94 patients) performed for Ficat stage III/IV AVN were reviewed clinically, radiologically and with Oxford hip scores. Mean age at operation was 43.9 years. Aetiology included trauma (20%), steroids/chemotherapy (25%), alcohol abuse (8%), AVN secondary to Perthes’/SUFE (4%) and idiopathic (43%). Two types of devices were used a) McMinn Resurfacing Arthroplasty, HA-coated smooth uncemented cup and cemented femoral component and b) Birmingham Hip Resurfacing, porous HA cup and cemented femoral component.

Results. At 2 to 11.5 years (mean 7.7 years), there were seven (6.7%) failures: four from further femoral head collapse, two infections and one osteolysis, aseptic loosening. The cumulative survival at 11years is 90% (figure 1). Mean Oxford score at follow-up is 15.4. 55% participate in moderate to heavy work or participate in regular sporting activity.

Discussion. The results of any treatment modality in AVN are not as good as those in osteoarthritis. One reason is that the etiopathological factors that caused non-traumatic AVN (steroids etc) have the potential to cause continued femoral head damage. Post-traumatic patients need complex reconstructions and risk a higher failure rate. Those on immunosuppressants are prone to infections. The results of hip resurfacing in AVN in this series are no different from THRs in many published series. Metal-metal resurfacing is a good conservative option for post-collapse femoral head AVN and gives acceptable results if strict patient selection criteria are followed.

Introduction: Dislocation rates with THA vary from 3% to 15%. One specialist centre reported a 6.4% early dislocation rate with a 28mm ceramic on polyethylene THA in young patients (mean age 56 years) in a single surgeon series. Although young patients have the advantage of better soft tissues, their greater mobility demands increase dislocation risk.

Dislocation rates in large headed metal-on-metal resurfacings are extremely low. However, many patients are unsuitable for resurfacing and need a replacement. In such cases, it is attractive to transfer the large-headed metal-metal bearing advantage to replacement arthroplasty in order to reduce wear and dislocation rates. Does large diameter metal-metal total hip replacement really reduce the early dislocation rate?

Methods: 206 consecutive primary metal-metal THRs (189 patients) were included. The device consists of an uncemented cup, a matching modular cobalt chrome head (head diameter ranged 38 – 58mm) fixed on a stem through a 12/14 cone. Cemented stems were used in 107 procedures and 99 were proximal-porous uncemented stems.

Age at operation ranged from 37 to 83 years. Thirty patients were 55 years or under, eighty one were 56 – 65 years and ninety five were over 65 years. There were 122 females and 67 males. Posterior approach was used in all.

Results: There were no dislocations in these 206 consecutive procedures.

Discussion: Metal-metal hips have lower dislocation rates than hips containing polyethylene (0.9% against 6.4% in a matched series). This is attributed to the suction-fit effect of metal-metal bearings. Large diameter bearings have the additional benefit of having to translate a greater jump distance before a dislocation. This dual advantage leading to extremely low dislocation rates was first noted in metal-metal resurfacings. In large headed metal-metal THRs, the head-neck ratio is even more favourable and these devices appear to eliminate early dislocation as a major complication.

Introduction: Hip resurfacing is a bone conserving option that offers a better revision prospect for young and active patients. Encouraging results from several centres prove that they function well in the early years. Their long-term survival will be known from continued monitoring of early resurfacings.

Methods: This is a retrospective study of two cohorts of young (< 55 years) patients of osteoarthritis treated with hybrid-fixed metal-metal resurfacings. The cohorts are a) consecutive patients treated by the senior author in 1994 and 95 with a hydroxyapatite-coated smooth uncemented cup and a cemented femoral component and b) consecutive patients treated with hydroxyapatite-coated porous uncemented cup and a cemented femoral component since 1997 with a minimum follow-up of 5 years. 420 resurfacings (360 patients, 287 males and 73 females) were reviewed with Oxford hip scores and activity level monitoring (UCLA scale). Mean age at operation was 48.3 years.

Results: Ten patients (11 hips) died from unrelated causes. Out of the remaining 409 hips (350 patients) at a follow-up of 5 to 11.5 years (mean 7.1 years), there was one failure (cumulative failure rate 0.25% at 11 years) from avascular necrosis of the femoral head. The mean Oxford score of the 350 patients is 13.4. 87% had a UCLA score of 7 and above. 55% participated in impact sports or were involved in heavy occupational work.

Discussion: In the present study, with no loss to follow-up, excellent hip survival (99.75%) and activity level were seen. Young patients regard return to activities as one of their highest priorities. None of these patients were advised to change their activities at work or leisure.

The extremely low failure rate in the medium term proves the suitability of resurfacing in young active patients. However, caution needs to be exercised until long term results are available.

Introduction. The usage of metal-metal bearings in young patients has rekindled the debate about the potential adverse effects of mutagenecity on offspring born to them. This question could be answered in part if it was known whether metal ions are transferred to the developing foetus. One recent study seems to suggest that such transfer does not occur [Brodner et al, J Arthroplasty2004; 19 Suppl(3) p102–107]. Unfortunately the instrument used there was not sensitive enough (5 out of 6 analyses were below the limit of detection), leaving the question of transplacental metal ion transfer unanswered. The present study uses a more powerful analytic technique.

Methods. After informed consent, whole blood specimens were obtained at the time of delivery from five patients who had undergone a Birmingham Hip Resurfacing and from their babies’ umbilical cords. High resolution plasma mass spectrometry (HRICPMS) was used for analysis.

Results. Cobalt and chromium ions have been detected in all the specimens obtained so far. The cord blood cobalt levels were lower than the mothers blood levels in all the specimens. A similar relationship was found in all but one individual chromium measurement. The mean (±95% CI) of the two groups are shown in figure 1.

Discussion and Conclusion. The present study shows that with the use of whole blood specimens rather than serum; and the use of a better analytic method than previously used in the only other study on the topic, metal ions can be detected in all specimens of patients with metal-metal devices and in the cord blood of babies born to them.

There is sufficient evidence in this study to prove that metal ions do cross the placenta. There is therefore a continuing need for vigilance on the possible effects on the offspring born to patients with metal-metal devices.

Introduction: modern cementless joints depend on bony ingrowth for durable long term fixation. Increased friction and micromotion in the early weeks can prevent ingrowth and affect long-term success.

Most friction studies are conducted in a bovine serum- carboxymethylcellulose (bs-cmc) medium. Following implantation however, the joint is bathed in blood which contains macromolecules and cells. The effect of these on friction is not fully understood.

A progressive radiolucent line (fig 1) observed in some low clearance resurfacings raises the concern that increased friction may be affecting component fixation. The purpose of this investigation was to study the effect of clearance on friction for a given bearing diameter in the presence of blood as lubricant.

Methods: Six Birmingham Hip Resurfacing devices with a nominal diameter of 50mm each and a range of diametral clearances (80, 135, 175, 200, 243 and 306μm) were used. Frictional measurements were carried out on a Prosim Hip Friction Simulator (Simsol Simulation Solutions, Stockport, UK). The test was conducted sequentially with whole blood (viscosity 0.009Pas) and a BS-CMC mixture as the lubricants (viscosity 0.01Pas).

Results: Low clearance devices (80–175μm) generated higher friction with blood than with BS-CMC (fig 2). With blood as the lubricant, low clearance devices generated much higher friction than higher clearance devices (200–306μm).

Discussion: Ongoing research into the in vitro performance of bearings is performed in hip simulators with lubricants that are believed to simulate joint fluid in terms of viscosity. However these lubricants are unable to simulate the friction effects of macromolecules.

The results of this study suggest that reduced clearance bearings have the potential to generate higher friction when blood is the lubricant. this higher friction in the low clearance bearings may produce micromotion in the early postoperative period and hamper bony ingrowth resulting in impaired fixation with long-term implications for survival.

Introduction: Hip resurfacing is a good conservative option for young patients with arthritis. Resurfacings risk two unique failure mechanisms that do not occur in THA, i.e. femoral neck fracture and femoral head collapse.

Old age, osteopaenia, alcohol abuse, and large cysts are risk factors for fractures. It has been suggested that performing a bilateral resurfacing puts the first side at risk of fracture from the force used in implanting the second resurfacing. Is this a true risk or a sampling error?

Methods: Out of 2576 consecutive resurfacings performed by the senior author (July 1997 – May 2005), 191 patients (382 hips, 14.8% of all resurfacings) presented with bilateral arthritis and had both hips operated in the same hospital admission. 133 patients had the two operations a week apart and 58 had both the same day. A posterior approach was used in all cases with the patient in the lateral position on the contralateral side.

Results: Of the 382 resurfacings, only two failed from a femoral neck fracture. Both had the second operation a week after the first. A 35-year lady (rheumatoid arthritis) sustained a femoral neck fracture of the first hip following a fall nine weeks after the operation. A 57-year man (osteoarthritis) fractured his femoral neck at 3.5 months. He fractured the side operated second.

Discussion: The incidence of femoral neck fracture in the author’s series of 2576 resurfacings is 0.4%. Patients who present with bilateral severe arthritis are more likely to have non-primary OA such as inflammatory arthritis. It is difficult to conclude if such bilateral cases are more predisposed to a fracture by virtue of the pathology itself.

The low incidence of fractures (2/382, 0.5%) in this bilateral resurfacing series does not support the view that there is an increased risk of fracture from a bilateral procedure.

Introduction: Metal ions generated from metal-metal joints are transported in plasma and within cells. Whole blood (WB) analysis is a good measure of systemic metal exposure because it includes both intracellular and extracellular compartments. Analysis of whole blood is technically more challenging than serum. But can serum levels be considered equivalent to whole blood levels in the assessment of systemic metal exposure?

Methods: 262 concurrent WB and serum specimens from patients after metal-metal hip arthroplasty were analyzed for metal levels by high-resolution mass-spectrometry (reporting limits 0.06μg/l for serum and 0.1μg/l for WB). The relationship between serum and WB levels were studied using four methods of agreement testing: paired t-test, Pearson correlation, Bland-Altman agreement and variability on normalized-scatter.

Results: Mean WB and serum chromium concentrations were 2.2 and 4.2μg/l (p< 0.001). Corresponding values for cobalt were 2.4 and 3.2 μg/l (p< 0.001). The normalized scatter (fig 1B) shows poor congruence between whole blood and serum levels especially at lower concentrations. Pearson correlation confirms poor agreement at lower concentrations even after excluding outliers.

Bland-Altman analysis (Figure 1A) shows the limits of agreement between serum and WB are unacceptably wide (1.7 to -5.1 for chromium) suggesting poor agreement.

Discussion: The proportion of metal ion levels in serum and blood cells shows great variability. With advances made inmass-spectrometry, poor instrument sensitivity is a non-issue. Is there then a case for continued use of serum as a measure of systemic metal ion exposure? It can be justified only if serum levels show good agreement with WB levels throughout the range of measurements. Concurrent analysis of serum and whole blood metal ion concentrations in the present study, do not show such an agreement with four methods of testing, suggesting that serum metal concentrations cannot be reliably used as an instrument to measure systemic metal ion exposure.

The dysplasia cup, which was devised as an adjunct to the Birmingham Hip Resurfacing system, has a hydroxyapatite-coated porous surface and two supplementary neutralisation screws to provide stable primary fixation, permit early weight-bearing, and allow incorporation of morcellised autograft without the need for structural bone grafting.

A total of 110 consecutive dysplasia resurfacing arthroplasties in 103 patients (55 men and 48 women) performed between 1997 and 2000 was reviewed with a minimum follow-up of six years. The mean age at operation was 47.2 years (21 to 62) and 104 hips (94%) were Crowe grade II or III.

During the mean follow-up of 7.8 years (6 to 9.6), three hips (2.7%) were converted to a total hip replacement at a mean of 3.9 years (2 months to 8.1 years), giving a cumulative survival of 95.2% at nine years (95% confidence interval 89 to 100). The revisions were due to a fracture of the femoral neck, a collapse of the femoral head and a deep infection. There was no aseptic loosening or osteolysis of the acetabular component associated with either of the revisions performed for failure of the femoral component. No patient is awaiting a revision.

The median Oxford hip score in 98 patients with surviving hips at the final review was 13 and the 10th and the 90th percentiles were 12 and 23, respectively.

We report a retrospective review of the incidence of venous thromboembolism in 463 consecutive patients who underwent primary total hip arthroplasty (487 procedures). Treatment included both total hip replacement and hip resurfacing, and the patients were managed without anticoagulants. The thromboprophylaxis regimen included an antiplatelet agent, generally aspirin, hypotensive epidural anaesthesia, elastic compression stockings and early mobilisation. In 258 of these procedures (244 patients) performed in 2005 (cohort A) mechanical compression devices were not used, whereas in 229 (219 patients) performed during 2006 (cohort B) bilateral intermittent pneumatic calf compression was used.

All operations were performed through a posterior mini-incision approach. Patients who required anticoagulation for pre-existing medical problems and those undergoing revision arthroplasty were excluded. Doppler ultrasonographic screening for deep-vein thrombosis was performed in all patients between the fourth and sixth post-operative days. All patients were reviewed at a follow-up clinic six to ten weeks after the operation. In addition, reponse to a questionnaire was obtained at the end of 12 weeks post-operatively.

No symptomatic calf or above-knee deep-vein thrombosis or pulmonary embolism occurred. In 25 patients in cohort A (10.2%) and in ten patients in cohort B (4.6%) asymptomatic calf deep-vein thromboses were detected ultrasonographically. This difference was statistically significant (p = 0.03). The regimen followed by cohort B offers the prospect of a low incidence of venous thromboembolism without subjecting patients to the higher risk of bleeding associated with anticoagulant use.

The early designs of hip resurfacing implants suffered high rates of early failure, making it impossible to obtain valuable mid-term radiostereophotogrammetric (RSA) results. The metal-on-metal Birmingham Hip Resurfacing arthroplasty has shown promising mid-term results and we present here the first mid-term RSA analysis of a hip resurfacing implant.

The analysis was performed in 19 hips at five years post-operatively. The mean acetabular component translation and rotation, and femoral component translation were compared with the previous RSA measurements at two and six months, and one and two years.

There was no statistical significance (t-test, p ≤ 0.05) between these consecutive movements, indicating the mid-term stability of the implant.

Metal ions generated from joint replacements are a cause for concern. There is no consensus on the best surrogate measure of metal ion exposure. This study investigates whether serum and whole blood concentrations can be used interchangeably to report results of cobalt and chromium ion concentrations.

Concentrations of serum and whole blood were analysed in 262 concurrent specimens using high resolution inductively-coupled plasma mass-spectrometry. The agreement was assessed with normalised scatterplots, mean difference and the Bland and Altman limits of agreement.

The wide variability seen in the normalised scatterplots, in the Bland and Altman plots and the statistically significant mean differences between serum and whole blood concentrations suggest that they cannot be used interchangeably. A bias was demonstrated for both ions in the Bland-Altman plots. Regression analysis provided a possible conversion factor of 0.71 for cobalt and 0.48 for chromium. However, even when the correction factors were applied, the limits of agreement were greater than ±67% for cobalt and greater than ±85% for chromium, suggesting that serum and whole blood cannot be used interconvertibly. This suggests that serum metal concentrations are not useful as a surrogate measure of systemic metal ion exposure.