Increased femoral head size may reduce dislocation rates following total hip replacement. The National Joint Registry for England and Wales has highlighted a statistically significant increase in the use of femoral heads ≥ 36 mm in diameter from 5% in 2005 to 26% in 2009, together with an increase in the use of the posterior approach. The aim of this study was to determine whether rates of dislocation have fallen over the same period. National data for England for 247 546 procedures were analysed in order to determine trends in the rate of dislocation at three, six, 12 and 18 months after operation during this time. The 18-month revision rates were also examined.
Between 2005 and 2009 there were significant decreases in cumulative dislocations at three months (1.12% to 0.86%), six months (1.25% to 0.96%) and 12 months (1.42% to 1.11%) (all p < 0.001), and at 18 months (1.56% to 1.31%) for the period 2005 to 2008 (p < 0.001). The 18-month revision rates did not significantly change during the study period (1.26% to 1.39%, odds ratio 1.10 (95% confidence interval 0.98 to 1.24), p = 0.118). There was no evidence of changes in the coding of dislocations during this time.
These data have revealed a significant reduction in dislocations associated with the use of large femoral head sizes, with no change in the 18-month revision rate.
Dislocation remains a major complication of total hip replacement (THR) with revision procedures required in 13% to 42% of patients who recurrently dislocate.1,2 Dislocation rates following THR vary between 0.5% and 5%,3–6 depending in part on the initial indication for replacement.3,7,8 More than three-quarters of all such dislocations occur within the first postoperative year; 30% to 50% take place within the first three months.3,8 The theoretical advantages of using components with a larger diameter of femoral head include a greater stable range of movement by reducing the likelihood of impingement and increasing the distance required to lift the femoral head from within the acetabular component. Additionally, careful capsular and abductor repair enhances soft-tissue resistance to dislocation.9–13 However, there is little evidence regarding the long-term effects of using large-diameter modular femoral heads and their longevity in modern bearings, and rates of volumetric wear, loosening and osteolysis remain unknown.
The latest annual report from the National Joint Registry for England and Wales has highlighted a marked increase in the use of large-diameter modular heads over the last five years.14 In 2005, 5% of THRs used a modular femoral head of ≥ 36 mm diameter. This proportion increased to 26% by 2009. During 2009, 1069 patients (18%) undergoing single-stage revision had recurrent dislocation recorded as the indication for surgery, but the interval between primary and revision procedure remains unknown.14 The cost of dislocation is high in terms of morbidity and healthcare resources. Each closed reduction represents around 20% of the hospital cost of an uncomplicated THR, and revision surgery to achieve stability can cost an additional 150%.15 The aim of this study was to determine whether national dislocation rates have fallen during this five-year period.
Patients and Methods
Data for English NHS patients, who underwent THR between January 2005 and December 2009 in whatever setting their treatment was provided were collected from the administrative hospital episode statistics database. Those undergoing THR following fracture of the hip were not included. This database covers all admissions to English hospitals providing care for NHS patients, and includes 15 diagnostic fields coded using the tenth revision of the International Statistical Classification of Diseases and Related Health Problems (ICD-10). In addition, 15 surgical procedure fields are coded using the fourth revision of the Office of Population, Censuses and Surveys Classification of Surgical Operations and Procedures (OPCS-4). Records of the same patient (defined using a combination of NHS number, date of birth, gender and postcode) were linked, and the number of days between the index operation and any subsequent orthopaedic re-admission to any NHS hospital were collected. Patients with missing dates of operation were excluded. By using the appropriate ICD-10 or OPCS-4 codes, complication rates were established. Data linkage was carried out at the NHS Information Centre.
In order to identify THR dislocations, OPCS-4 codes for closed reduction in theatre were used. This assumes that most dislocations are reduced closed in theatre and not in the emergency department, and rarely require open reduction. Dislocations were recorded at three, six, 12 and 18 months after the primary procedure. Laterality was examined in the coding string to ensure that the redislocation had occurred on the same side as the primary procedure. If this was not possible, patients were excluded. Further dislocations were not recorded to ensure that patients with recurrent dislocations did not create erroneous results. In order to make certain that any changes in dislocation rates over time were not the result of changes in coding practice, 18-month dislocation rates following hemiarthroplasty for intracapsular fractures of the femoral neck were also analysed (2005 to 2008) using the same criteria.
Secondary outcomes were 30-day re-admission to an orthopaedic ward and the 18-month revision THR rate. In addition, we also examined rates of wound complications, including haematomas, superficial wound infections and deep infections requiring a return to the operating theatre, within 30 days of the procedure, as this might affect early dislocation rates. Data were examined by year of operation (2005 to 2009) and compared. Age, gender and Charlson score,16 which predicts the one-year mortality for a patient, were recorded for each patient.
Trend analyses were performed on data from 2005 to 2009 using the chi-squared method,17,18 with p < 0.05 considered to show a statistically significant linear trend. Odds ratios (OR) and 95% confidence intervals (CI)19 were calculated for 2005 and 2008 data. The null hypothesis of no difference in dislocation rate (OR = 1) was tested using the chi-squared test (α = 0.05) for a 2 × 2 contingency table.
There was a statistically significant increase in the use of femoral heads of ≥ 36 mm during this time (p < 0.001, Table I⇓). Coding for dislocation appears consistent across the study period, as there was no significant trend (p = 0.368, Table II⇓) in 18-month dislocation rates following hemiarthroplasty for fracture (Fig. 1⇓).
Data were then retrieved for 247 546 patients who underwent primary THR between January 2005 and December 2009 in the English NHS. Patients who underwent THR in 2005 were slightly older and had lower Charlson scores than those in 2008, and a significantly smaller proportion had rheumatoid arthritis (Table III⇓). There was a significant decrease in the 18-month dislocation rate (OR = 0.84 (95% CI 0.75 to 0.93), p = 0.002). When cumulative dislocation trends were analysed, there were significant decreases over the period of the study at three, six, 12 and 18 months (Table IV⇓). There were no significant changes in 30-day infection rates or 18-month revision rates.
In the past, it was thought that the diameter of the femoral head influenced the risk of dislocation following THR, but there were contradictory findings in the literature.3,20–22 This association has yet to be proven, as dislocation can be attributed to several causes. There are a number of studies that show a strong relationship between larger femoral heads and lower dislocation rates, especially with careful repair of the posterior soft tissues,23–30 but the high number of variables weakens the statistical power of the large historical studies.3,31
A review of 10 500 patients at the Mayo Clinic found that previous surgery to the hip, a posterior approach and trochanteric osteotomy were the factors most strongly associated with dislocation.3 The National Joint Registry data show that THRs undertaken through a posterior approach in England and Wales have increased from 34% in 200432 to 57% in 2009.14 One might therefore expect the data presented here to demonstrate an increase in dislocations, not a reduction. However, more than ten confounding factors were identified in the Mayo study, making it difficult to determine the true cause of dislocation.
Berry et al33 examined 21 000 primary THRs in which 22 mm, 28 mm, or 32 mm diameter femoral heads were used. Regardless of the approach, dislocation rates were reduced when a larger size was used. The posterior approach was consistently associated with a higher dislocation rate across all head sizes compared with other approaches. These findings were also corroborated by other groups, including the Norwegian Arthroplasty Registry, which found that rates were lower with 32 mm heads than with 28 mm heads.11,30,34 This registry also showed that 22 mm and 28 mm heads performed equally well.
There are other potential benefits of large femoral heads. Amstutz et al9 reported an improvement in the range of movement with a 32 mm diameter head compared to a 22 mm diameter design, owing to the greater head-neck ratio (1.98 vs 1.74). These findings were supported by Chandler et al,35 who reported delayed contact between the femoral neck and the acetabular component with larger femoral head sizes, thereby allowing a greater range of movement. However, Bartz et al23 reported only very small gains in movement before impingement when the head size increased from 22 mm to 28 mm, and no significant increase in between 28 mm and 32 mm beads.
Larger heads require thinner acetabular bearings. This may be achieved with alternative bearing surfaces such as ceramic, or highly cross-linked polyethylene. The long-term survival rates of highly cross-linked polyethylene are awaited, but short-term studies demonstrate a 45% to 99% reduction in wear at three to five years.36,37
A proportion of these large heads are likely to be part of a metal-on-metal bearing. Resurfacing devices are not included in this analysis, but resurfacing bearings on a stem would be coded as a THR. According to the latest National Joint Registry data, five-year revision rates are four times higher with this type of THR than with cemented THRs.14 Additionally, there are reports of early failures secondary to elevated levels of wear and localised soft-tissue reactions.38,39
During this study period the 18-month rate of revision following THR did not change. As recurrent dislocation is an infrequent indication for revision and approximately 70% of dislocations are single events, the numbers required to study the effect of dislocations on revision rate are beyond the power of this study.3,5,40–42 However, we can say that the revision rate has not increased through other causes, such as peri-prosthetic fracture or excessive wear, at the expense of lower dislocation rates. The longer term consequences of the use of larger femoral heads will require analysis in future registry reports.
We have demonstrated that the increasing use of large femoral heads in THR is associated with a reduction in the dislocation rate, despite an increase in the use of the posterior approach. We are confident that this is a real observation and not a coding anomaly, as dislocation for hemiarthroplasty remained constant throughout the period. We have no data regarding implant position or method of posterior capsular repair, which may influence dislocation.
This is a large dataset reflecting a national experience of hard metrics, including implant size, surgical approach and dislocation. The increasing use of large heads has shown a reduction in the early complication of dislocation, although the long-term effects remain unknown.
Tables detailing the International Statistical Classification of Disease and Related Health Problems (10th revision; ICD-10) and Office of Population, Censuses and Surveys Classification of Surgical Operations and Procedures (4th revision; OPCS-4) codes used in the analysis, and their relation to the Charlson comorbidity index, are available with the electronic version of this article on our website at www.jbjs.org.uk
C.H.K.S. has approval to re-use hospital episode statistics data with the permission of the Health and Social Care Information Centre.
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.
- Received January 6, 2011.
- Accepted March 7, 2011.
- © 2011 British Editorial Society of Bone and Joint Surgery