Objectives

The diagnosis of surgical site infection following endoprosthetic reconstruction for bone tumours is frequently a subjective diagnosis. Large clinical trials use blinded Central Adjudication Committees (CACs) to minimise the variability and bias associated with assessing a clinical outcome. The aim of this study was to determine the level of inter-rater and intra-rater agreement in the diagnosis of surgical site infection in the context of a clinical trial.

Introduction: Compromised by pre-existing medical co-morbidities, weakened by the second hit of surgery, elderly patients with hip fractures are amongst the most challenging cases to manage appropriately in the acute hospital setting.

Aim: To document the frequency and outcome of post operative medical complications in elderly patients following surgery for hip and proximal femoral fractures.

Methods: Retrospective analysis of acute admissions to a University teaching hospital with hip and pertrochanteric fractures over an 18 month period. Medical records and radiographs were reviewed for details relating to location of fracture, in-patient morbidity, mortality and length of stay. In-patient fracture referrals and patients under 65 years of age were excluded.

Results: From a total of 438 fracture fixation procedures or prosthetic replacements, 368 patients were eligible for analysis. The mean age of patients was 83.6 years. The mean length of stay was 14.2 days.

142 patients experienced significant postoperative morbidity consisting of 24 myocardial infarctions, 46 respiratory tract infections, 33 urinary tract infections, 3 cerebral vascular accidents and 36 exacerbations of congestive cardiac failure. There was no correlation between morbidity and location or type of fracture. The mean age (86.1 yrs) and length of stay (26 days) was greater in the morbidity group (p< 0.05). The overall post operative in-patient mortality rate was 9%, rising to 50% in those who suffered a myocardial infarction and 33% in those with exacerbations of congestive cardiac failure.

Conclusion: This study reinforces the poor prognosis conferred by a hip fracture on the elderly patient both in terms of morbidity and mortality. As the population ages and life expectancy increases, health services will come under ever more pressure to cater for the expansion in numbers of hip fracture patients. Greater resources and dedicated professional medical staff are required in order that the best possible care and outcome may be achieved for these patients

Aim: An epidemiological analysis of spontaneous community acquired septic arthritis cases in an elderly population at a university teaching hospital

Method: We studied confirmed cases of spontaneous septic arthritis in the over 65 population. Patients with prosthetic joint infections were excluded from the study. We analysed data relating to initial presenting complaint and clinical examination, haematological and microbio-logical tests along with ultimate patient outcome.

Results: There were 7 confirmed cases of spontaneous septic arthritis in over 65 population in the last 6 months (2 hips, 5 knees). The mean age was 72.14 (range 65–82) with a mean length of stay of 49 days. Those with septic arthritis of the knee presented with swelling, pain and immobility. Hip cases presented with pain and immobility. All patients were systemically well at time of presentation, with no other foci of infection detected after septic screening. All patients had aspirate and arthroscopic/arthrotomy confirmed infection. Staphylococcus Aureus was isolated from 6 joint aspirations and Pseaudomonas Auruginosa from one patient. Complications of treatment included acute renal failure, cardio/respiratory failure, disseminated infection and death (1 case).

Conclusion: Septic arthritis must be considered as a differential diagnosis in all patients with joint pain, swelling and immobility. This diagnosis is not confined to the paediatric population. A backround of degenerative disease and the occult presentation in the elderly may delay diagnosis. Sepsis must be considered in the elderly with joint pathology, with treatment initiated in a prompt and aggressive manner to prevent the sequelae that ensues.

Background: Infection remains the single most devastating complication of joint arthroplasty. In cases of established prosthetic infection, where implant retention is not feasible, there is limited consensus on an optimum management protocol.

Aim: To assess the outcome of revision for infected hip prostheses using a novel treatment regimen.

Materials and Methods: Retrospective study of a consecutive case series of 40 patients with late chronic hip joint prosthetic infection treated by a single surgeon over a 4 year period. The mean interval between index arthroplasty and revision for infection was 40 months, with patients having prior symptoms of infection for a mean of 22 months.

The treatment protocol consisted of a two stage exchange with removal of infected components via a posterior approach incorporating an extended trochanteric osteotomy, insertion of an interim antibiotic eluting cement spacer and re-implantation of an extensively coated uncemented prosthesis on the femoral side. Systemic antibiotic treatment following each stage consisted of an abridged course of 5 days post operative intravenous administration followed by complete cessation of anti-microbial therapy. The mean interval between implant removal and re-implantation was 111 days.

Results: At a mean follow up of 29.6 months (minimum 12 months), there were 2 cases of recurrent prosthetic infection. Dislocation following the second stage occurred in 7 patients. There was one mortality and one case of post operative sciatic nerve palsy. The Harris hip score increased from a pre-operative mean of 43.8 to a post operative mean of 83.9. At follow up, no patient had required revision for aseptic loosening or mechanical instability on the femoral side

Conclusion: The combination of effective staged surgical joint debridement, a shortened post operative course of systemic antibiotic treatment and an adequate latent period before re-implantation has led to encouraging early results in this series of revised chronic hip joint prosthetic infections.

The in-cement technique for revision hip arthroplasty involves retaining the original cement-bone interface. This has been proven to be a biomechanically stronger method than recementing after complete removal of the original cement mantle.

This study reviewed a series of 54 consecutive revision hip arthroplasty procedures, using the in-cement technique, between November 1999 and November 2003. Clinical and radiological follow-up included functional assessment.

There were 54 procedures performed in 51 patients, whose mean age at surgery was 70.3 years (45 to 85). A total of 42 were available at a mean follow-up of 29.2 months (6 to 51). There was no radiological evidence of loosening. Functional assessments were available for 40 patients who had a mean Harris hip score of 85.2 (51.9 to 98.5), a mean Oxford hip score of 19.6 (12 to 41), a mean UCLA activity profile score of 5.9 (3 to 8) and a mean SF-36 score of 78.0 (31.6 to 100).

The in-cement technique provides consistent, high functional outcomes and should be considered in appropriately selected cases.

Revision of the femoral component during revision hip arthroplasty may pose significant technical challenges, most notably femoral fracture and bone perforation. The in-cementing technique allows use of the original bone-cement interface that has been proven to be biomechanically stronger than recementing after complete removal of the original cement mantle.

This historical prospective study reviews a series of 54 consecutive revision hip arthroplasty procedures carried out by the senior author using the in-cementing technique from November 1999 to March 2003. Patients were followed up clinically and radiologically with serial outpatient reviews and their functional outcome was assessed using the Harris hip scoring system, the Oxford hip scoring system and the University of California at Los Angeles (UCLA) activity profile. Their physical and mental well being was also assessed using the SF-36 self-questionnaire.

Fifty-four procedures were performed on 51 patients. There were 31 males and 20 females. The average age was 70.3+/−8.1 years (range: 45 – 83 years). The average time to revision from the original procedure was 132.8+/−59.0 months (range: 26 – 286 months). The average length of follow up was 29.2+/−13.4 months (range: 6 – 51 months) post revision arthroplasty. Two patients suffered dislocations, one of which was recurrent and was revised with a Girdlestone’s procedure. No patient displayed any evidence of radiographical loosening. The average Harris hip score of the study group was 85.2+/−11.6 (range: 51.9 – 98.5). The average Oxford hip score recorded was 19.6+/−7.7 (range: 12 – 41) and the average UCLA activity profile score was 5.9+/−1.6 (range: 3 – 8). The SF-36 questionnaire had an average value of 78.0+/−18.3 (range: 31.6 – 100) with an average physical score of 73.3+/−22.2 (range: 20.5 – 100).

In conclusion, we feel the results of this study show excellent clinical and radiological results of the in-cementing technique with high patient satisfaction in terms of functional outcome. This technique merits consideration where possible in revision hip arthroplasty.

We assessed the outcome of patients with Vancouver type B2 and B3 periprosthetic fractures treated with femoral revision using an uncemented extensively porous-coated implant. A retrospective clinical and radiographic assessment of 22 patients with a mean follow-up of 33.7 months was performed. The mean time from the index procedure to fracture was 10.8 years. There were 17 patients with a satisfactory result. Complications in four patients included subsidence in two, deep sepsis in one, and delayed union in one. Concomitant acetabular revision was required in 19 patients. Uncemented extensively porous-coated femoral stems incorporate distally allowing stable fixation. We found good early survival rates and a low incidence of nonunion using this implant.

Background: CT scanning is an essential part of the preoperative planning process prior to surgical fixation of acetabular fractures. Considerable disparity exists between the clinical and radiological outcome following open reduction and internal fixation of these fractures. It is suggested that this disparity is due to poor assessment of the quality of reduction using plain radiographs alone.

Aim: To investigate the role of post-operative CT scanning following ORIF of acetabular fractures.

Methods: Prospective study commenced in January 2000 of all patients in our institution undergoing internal fixation of acetabular fractures. Post operative axial CT scans were compared with plain radiographs (AP pelvis and 45 degree oblique Judet views) with regard to the sensitivity to detect articular fracture reduction in terms of gap displacement and step deformity or offset. A simplified binary measurement of radiological outcome was used stratifying radiological result into anatomical and non-anatomical. Three observers independently reviewed the plain radiographs and CT scans at two separate time points and categorized the radiographic outcome as described. The interobserver reproducibility and intraobserver reliability of these measurements was expressed as a kappa statistic. In addition in those patients greater than 18 months following surgery we attempted to correlate the radiographic with the clinical outcome using the Harris hip score and the SF-36 score.

Results: 20 patients were recruited. Plain films were equieffective in detecting post-operative articular fragment displacement (p=0.24). The interobserver and intraobserver agreement between the radiological outcome measurements were good with respective kappa values of 0.61 and 0.65. There was a weak association between clinical and radiographic outcome as ascribed by post operative CT scans.

Conclusion: While there may be an argument for the use of post operative CT scanning acetabular fractures in selective cases, we did not find any significant benefit of CT scans over plain radiographs in the assessment of reduction or radiological outcome following these injuries. Hence we do not routinely advocate their use in the post operative setting.

Background: Periprosthetic fractures are now recognised as the second most frequent reason for revision following total hip replacement, less so than aseptic loosening but more so than dislocation and infection. The post-operative incidence of peri-prosthetic fracture is approximately 1% after primary arthroplasty and 4% after revision surgery. 75% of postoperative femoral fractures are associated with prosthetic loosening. The Vancouver system, a reliable and valid classification scheme, sub-classifies fractures around the stem of the prosthesis into three types. B2 is where there is a loose implant but adequate bone stock and B3 is where the implant is loose and bone stock is deficient.

Aim: To assess the outcome of patients with Vancouver type B2 and B3 fractures treated with femoral revision using an uncemented extensively porous coated implant (Solution ® stem).

Methods: A retrospective chart review was performed. Patients in addition attended for a clinical and radiographic assessment. Engh’s criteria for osseointegration of cementless components and the length of time to fracture union were the radiographic endpoints. The Harris hip score was used for clinical assessment with a score of above 80 indicating a satisfactory result.

Results: From July 1999 to present, we identified 22 such patients treated with this method. The mean duration of follow-up was 33.7 months with a minimum of 12 months. The mean age of patients was 78.7 years (range 67–88). The mean time from the index procedure to fracture was 10.8 years (range 7–20 years). The index procedure was a primary total hip replacement in all but 2 cases (revisions). 18 patients had a good result. Four patients had a poor result. The mean Harris Hip Score in the 18 patients was with good results was 82.7 and 69 in those with poor results. The mean duration to fracture union was 5.2 months (range 4–8 months). 17 patients needed concomitant acetabular cup revision. No patient showed any evidence of loosening or subsidence. 2 patients presented with deep seated infection (1 early, 1 late).

Conclusion: Periprosthetic fractures in the setting of a loose prosthesis present a difficult reconstructive challenge. Uncemented extensively porous coated femoral implants incorporate distal with intramedullary fixation of most fractures, permitting fracture healing as well as achieving osseointegration. We report good early survival rates, stable fixation and a low incidence of non union using this implant.

The aim of this study is to assess the success of posterior lumbar interbody fusion in the treatment of degenerative spinal instability. Methods: Historical prospective study containing sixty-five consecutive patients who underwent posterior lumbar interbody fusion (PLIF) using carbon cages and pedicle fixation between 1993 and 2000.

Clinical outcome was assessed by the postoperative symptomatic relief, complications rate and the fusion rate. The fusion rate was assessed using plain radiographs and the Brantigan and Steffee scoring system. Functional outcome was measured by the improvement in the Oswestry disability index, PROLO score, return to work and satisfaction with the surgical outcome. The determinants of functional relief were analysed against the improvement in disability using multiple regression analysis. Results: Overall fusion rate was ninety eight percent. There was a significant improvement in Oswestry disability index P< 0.001. There was 85% satisfaction with the surgical procedure and 58% return to pre-disease activity level. We found preoperative level of disability to be best the determinant of functional recovery irrespective of age or the degree of psychological morbidity (p< 0.0001). Conclusion: The combination of posterior lumbar interbody fusion (PLIF) and posterior instrumented fusion is a safe and effective method of achieving segmental fusion with sustained functional relief and high satisfaction rate. Direct relationship between preoperative level of disability and functional recovery suggests that spinal fusion should be performed to alleviate disability caused by degenerative spine.

Introduction: Hip arthroplasty is growing in importance and relevance in the detection and treatment of various pathologies affecting the hip joint. A growing number of indications exist including the diagnosis and treatment of synovial-based disease, extraction of loose bodies within the joint, and treating chondral flaps and labral tears. Hip arthroplasty represents a minimally invasive, joint preserving, effective and reproducible tool to achieve these. The literature cites few complications arising from the procedure.

Study: Retrospective study of the outcomes of 20 patients who underwent hip arthroplasty between March 1999 and October 2002.

Methods: All the patients were first clinically assessed and then underwent Magnetic Resonance Imaging of their hips prior to undergoing arthroplasty. The indications for the MR arthrography included hip or groin pain, joint stiffness with associated impaired exercise tolerance. Arthroscopy was performed as a day case under general anaesthetic. The patients were placed supine on a fracture table and traction applied, under image control 30 and 70 degree arthoscopes were placed through an anterior and peritrochanteric portals. The patients followed up for an average of 20 months (6–44 months) and assessed for pain, mechanical symptoms, activity level, work status, sports ability and performance of activities of daily living. Data collected was retrospectively reviewed.

Results: The correlation between MR imaging and arthroscopy was approximately 80%. MR imaging detected all the labral tears present in 10 patients. This sensitivity was however reduced for the detection of small osteophytes (105), synovitis (5%), and minor labral radial fibrillation (5%). Four patients had loose bodies, which were successfully extracted. The patients whose symptoms were the result of mechanical pathology achieved complete resolution of their symptoms post operatively. No complications were seen.

Conclusion: Hip arthroplasty can be performed for a variety of conditions with reasonable expectations of success and minimal complication rate.

Introduction: Conservative treatment of humeral diaphyseal fractures yields reliable union rates with satisfactory clinical outcome. The high incidence of shoulder complications following intramedullary fixation of humeral fractures has led clinicians to consider other less invasive treatment modalities. There is a growing body of evidence supporting the role of functional bracing in the conservative management of these fractures.

Aims: To compare the outcome of patients with humeral fractures treated non-operatively using functional bracing and those treated by internal fixation.

Methods: Retrospective clinical and radiographic study of two groups of patients with humeral diaphyseal fractures matched for age, sex and mechanism of injury. The study group consisted of 46 patients who had presented to our institution between January 1999 and July 2002 with closed diaphyseal humeral fractures. Mean follow-up was 21 months 96–42 months). Patients were assessed for pain, range of motion, return to work and sporting activities. Group 1 comprised of patients who underwent functional bracing whereas Group 2 underwent internal fixation. There were 12 males and 10 females in Group 1, with a mean age of 42 (16–75 years). Group 2 comprised 13 males and 11 females, with a mean age of 37 years (20–80 years). Fourteen patients were treated with antegrade locked intramedullary nailing, whereas 10 patients underwent plate osteosynthesis. All functional braces were specifically moulded and customised for each patient by the senior upper limb occupational therapist. Three patients were lost to follow-up (Group 2).

Results: All patients treated with functional bracing went on to union at an average of 13 weeks (10–18 weeks). There were no major complications in this group. Two patients had residual loss of shoulder range of motion. Two patients had radial nerve palsies at presentation, which recovered fully. Of the twenty-one patients treated operatively, four required further procedures. Three patients had removal of IM nails with ORIF and bone grafting. One patient went on to non-union following ORIF, and had subsequent bone grafting. Eight patients developed restricted shoulder function, four requiring manipulations under anaesthesia.

Conclusion: Functional bracing of humeral diaphyseal fractures is a safe, non-invasive treatment providing reliable outcome and high rates of union. In this retrospective study, we found the results following functional bracing to be superior to that following internal fixation. Close supervision and patient compliance with treatment and rehabilitation is empiric to a satisfactory outcome in these patients.

Introduction: The majority of patients with osteoarthritis of the knee suffer from femorotibial pain with a smaller proportion suffering predominantly patello-femoral symptoms. No clear consensus exists as to the need for patellar resurfacing when performing total knee replacement for patients with symptomatic femorotibial osteoarthritis but without prominent patello-femoral symptomatic and radiographic disease.

Aims: To identify the advantages and disadvantages of both resurfacing and non-resurfacing of the patella during cemented total knee replacement performed for osteoarthritis predominantly of the femorotibial joint. To objectively clarify the rationale for the use of either procedure in clinical practice.

Methods: Prospective randomized double blinded clinical trail. Patients with osteoarthritis of the knee and principally femorotibial symptoms were included. Patients with rheumatoid arthritis, gross deformity of the knee and gross radiological or clinical patello-femoral arthritis were excluded. The implant used was a cemented posterior stabilized AMK (Depuy, Leeds UK) prosthesis. Pre-operative American Knee Society Score, SF-36 questionnaire and WOMAC scores were calculated for each patient. These instruments were repeated and combined with clinical and radiological follow up at 3 months, 6 months and one year.

Results: 58 patients were recruited into the study, 53 of whom completed follow-up and were included in the analysis. Baseline characteristics were similar in each group. Operating room time was less in the non-resurfaced group (p< 0.05). At one year, no patient in either group had needed to undergo a revision procedure. There was no difference between the resurfaced and non-resurfaced groups in terms of global functional outcome as measured by SF36 and WOMAC scores at one-year post operatively. The American Knee Society score showed no difference between the two groups (p=0.86) at one-year post surgery.

Conclusion: There is no significant difference in clinical outcome at one year following surgery vis-à-vis those who did and did not have patellar resurfacing performed during knee replacement for predominantly femorotibial symptomatic osteoarthritis. Patellar resurfacing as a procedure is not without complications. In patients with osteoarthritis of the knee and predominantly femorotibial disease based on clinical and radiographic findings, we do not advocate the routine use of patellar resurfacing.

The management of long bone infected non-unions; posttraumatic chronic osteomyelitis and primary segmental bone defects constitute some of the most difficult reconstructive challenges encountered by orthopaedic surgeons. Measures employed to treat these conditions are tantamount to limb salvage with amputation a likely outcome if reconstructive endeavors prove unsuccessful. The Ilizarov method of distraction osteogenesis and bone transport, following radical debridement in the case of infection, is one potential management option in such cases.

Aim: To assess outcome in a cohort of patients with long bone defects treated with this technique.

Methods: Clinical review coupled with retrospective chart study and radiographic assessment.

Results: 24 patients (20 adults, 4 children) underwent treatment over a 6-year period. 21 had defects of a primary traumatic aetiology (18 tibial and 3 femoral fractures). The mean interval between injury and commencing bone transport was 41.6 months. The mean defect dimension requiring transport was 9cm (range 4.5 to 28cm). The mean external fixation time was 14.6 months. Union occurred in 21 cases. Autologous bone grafting of the docking site was performed in 6 cases in order to stimulate union. Following removal, frames were reapplied in 4 cases due to refracture or development of angular deformity. Two patients proceeded to below knee amputations. According to ASAMI* criteria, the bone result was excellent in 12 patients, good in 5 patients, fair in one patient and poor in 6 patients. Functional outcome was excellent in 7 patients, good in 12, fair in 2 and poor in 3.

Conclusion: The Ilizarov method of bone transport is a reliable procedure providing consistent results in complex cases when frequently alternative treatment options have been exhausted. Outcome compares favorably with other treatment modalities such as vascularised free tissue transfer or Papineau type grafting procedures. The treatment period is lengthy and both major and minor complications are common but limb salvage is successful in the main part. More aggressive treatment and appropriate fracture management in the early stages may have a role to play in improving the prognosis of these patients.

*Association for the Study and Application of the Methods of Ilizarov

Posterior lumbar interbody fusion is a well described procedure for the treatment of back pain associated with degenerative disc disease and segmental instability. It allows decompression of the spinal canal and circumferential fusion through a single posterior incision.

Sixty-five consecutive patients who underwent posterior lumbar interbody fusion (PLIF) using carbon cages and pedicle fixation between 1993 and 2000 were recruited and contacted with a postal survey.

Clinical outcome was assessed by the assessment of postoperative clinical findings and complications and the fusion rate, which was assessed using standard X-rays with the scoring system described by Brantigan and Steffee. Functional outcome was measured by using improvement in the Oswestry Disability Index, PROLO score, return to work and satisfaction with the surgical outcome. The determinants of functional relief were analysed against the improvement in disability using multiple regression analysis.

The mean postoperative duration at the time of the study was 4.4 years. The response rate to the survey was 84%. Overall radiological fusion rate was ninety eight percent. There was a significant improvement in Oswestry Disability Index P< 0.001. There was 85% satisfaction with the surgical procedure and 58% return to pre-disease activity level and full employment. In the presence of near total union rate we found preoperative level of disability to be best the determinant of functional recovery irrespective of age or the degree of psychological morbidity (p< 0.0001).

The combination of posterior lumbar interbody fusion (PLIF) and posterior instrumented fusion is a safe and effective method of achieving circumferential segmental fusion. This procedure gives sustained long-term improvement in functional outcome and high satisfaction rate. Direct relationship between preoperative level of disability and functional recovery suggests that spinal fusion should be performed to alleviate disability caused by degenerative spine.

Assessment and referral of spinal disease in a primary care setting is a challenge for the general practitioner. This has led to establishment of spinal assessment clinic to insure prompt access to the patient who requires treatment by a spinal surgeon. These clinics are run by a trained physiotherapist who liaises with a member of the spinal team and decides the need for referral to the spinal clinic on the bases of the patient’s history and clinical examination. In our clinic each patient is also assessed with Oswestry disability index, Short form-36, visual analogue score and hospital anxiety score (HADS), although these scores do not contribute to the clinical decision-making. The aim of this study is to assess the screening value of Oswestry disability score, Short form-36 scores in diagnosing acute spinal pathology.

Sixty-nine patients who were referred to the spine clinic from the assessment clinic between March and December 2001 were recruited. Sixty-nine age and sex-matched patients were randomly chosen from five hundred and twelve patients who were seen in the spinal assessment clinic and did not need referral to the specialised spine clinic. The Oswestry disability score, Short form-36 scores and pain visual analogue scores between the two groups were statistically compared. The correlation between the level of psychological morbidity, length of symptoms and presence of past history of symptoms against the level of disability was statistically assessed.

Although there was a significant increase in the level of disability in the referred group with each score (Oswestry Disability Score P< 0.001, SF-36 physical component score P=0.014, Visual analogue pain score P< 0.001). The variation in the scores makes the scoring system unspecific for use as a screening tool. We also found strong relationship between psychological disability and length of symptoms indicating the need for prompt treatment for back pain.

Two hundred patients with Adhesive Capsulitis according to Codman’s criteria were treated with manipulation under anaesthetic and hydraulic distension by the senior author. The procedure and subsequent rehabilitation was uniform for all patients. The average age at time of procedure was sixty years (range 36–91 years). Follow-up was performed using a self-assessment booklet which we devised to examine outcome in the following areas: Pain Visual Analogue Score, Ten Activities of Daily Living, Ability to Sleep & Lie on Affected Shoulder, Range of Motion and overall satisfaction. The average length of follow-up was sixty-two months (range 12–125 months). One hundred and forty-five patients were available for follow-up and completed the assessment correctly.

Results: There were no operative complications in this group. Pain was significantly decreased from a mean pre-operative pain visual analogue score of 7.9 to 1.4 post-operatively. Shoulder pain causing difficulty sleeping was reduced from 85% of patients pre-operatively to 15% post-operatively. Range of motion was assessed in comparison to pre-operative values of the affected side and current values of the unaffected side. Regarding patient satisfaction 90% of patients were improved post-operatively, 7.5% unchanged and 1.5% felt that their symptoms were worse following the procedure. The procedure was well tolerated and 97% of patients would have the procedure again. Patients who had the procedure within nine months following onset of symptoms had better long-term range of motion and functional outcome than those who had a greater delay in treatment. Our results indicate manipulation and hydraulic distension is a safe effective treatment for adhesive capsulitis and that a more favourable outcome is achieved if it is performed at an early stage.

Neonatal septic arthritis is a true orthopaedic emergency posing significant threat to life and limb.

Objective: To examine the clinical presentation, diagnosis, treatment and outcome of children presenting with septic arthritis in the neonatal period.

Design: Retrospective review of clinical notes and radiographs of children presenting over a 20 year period (1977–97).

Subjects: 34 patients with septic arthritis in a total of 36 joints.

Outcome measures: Clinical outcome was classified as satisfactory or unsatisfactory as per Morrey et al. Radiological outcome was graded I–IV as per Choi et al. Joint instability, destruction, limb length discrepancy and angular deformity were assessed.

Results: The hip joint was affected in 24 of the 34 cases. Pseudoparalysis was the most reliable clinical finding occurring in 29 out of 34 cases. Staph Aureus was isolated as the infecting pathogen in 22/34 patients. Sequelae occurred in 16 hips and 1 knee. Poor prognostic indicators were delayed diagnosis (p< 0.05) and the hip as site of infection (p< 0.01). Clinical outcome was unsatisfactory in 15 patients and satisfactory in 17 patients. Radiological outcome was Choi I or IIA (good) in 12 hips and Choi II to IV in 13 hips (poor). Multiple further reconstructive procedures were required in 15 cases.

Conclusions: Despite optimum treatment, neonatal septic arthritis results in significant long-term morbidity for a high proportion of cases.