We reviewed the current screening protocol for MRSA detection in patients admitted for elective lower limb arthroplasty as we thought it may not be cost effective and there is no evidence base for validity of MRSA screening swabs.

The current institutional screening protocol includes MRSA swabs from nose and groin taken on 2 different occasions; the 1st swab taken at time patient was placed on the waiting list and the 2nd swab taken at the Orthopaedic pre-clerking clinic around 6–8 weeks later.

We found that during the study period, of 1791 patients who underwent MRSA screening swabs, only 16 patients had positive results. Of these, 9 patients had known MRSA (2 in care home) and no data was available in 2 patients. However in 5 patients, there was an initial negative swab which later was found to be positive. The time interval in these ranged from 0.6 to 85 weeks (Conversion rate = 0.3%).

The laboratory cost for these screening swabs was £9,027. Given the low prevalence rate of MRSA and low conversion rate at 6 weeks, there is a role for accepting a single-swab screening protocol. This would also result in significant cost savings of around £ 4,500.

We can therefore conclude that in the indigenous population, it is unlikely a patient will de-novo acquire MRSA in the community unless exposed to independent risk factors. We have adopted the new screening protocol where patients are screened routinely for detection and eradication only once prior to admission unless independent risk factors such as previous MRSA colonisation, hospital admission or if being admitted from residential/care home.

There is however no evidence in literature about the time period of validity of MRSA screening swabs.

Late wound infection is a recognised complication of instrumented spinal deformity surgery. In most cases it is a benign complication which usually resolves after implant removal. However, some of our patients with late infection developed a pseudoarthrosis.

To investigate this further we undertook a retrospective review of all patients undergoing implant removal for deep infection between 1991 and 2000.

Twenty-one patients were identified, representing a late infection rate of at least 6%. They showed no specific pre- or intra-operative risk factors. Nine had some problems with early post-operative wound healing, which settled with minimal treatment. Late infection presented as localised swelling or a discharging sinus between 4 and 84 months (average 31 months) post-surgery. Blood parameters were abnormal in 15 cases, frank infection demonstrated in 19 cases, loosening of the implant in four cases and positive bacteriology culture in 14 cases. Wounds healed within 2 to 17 weeks (average 5 weeks) following implant removal, wound debridement and antibiotic therapy lasting 2 to 20 weeks (average 6 weeks). This was delayed until one year post-surgery in the three cases presenting early. Follow-up of between 6 and 92 months (average 38 months) revealed no persistent infection. Pseudoarthrosis developed in seven patients (33%). Four of these patients had progressive deformity warranting refusion and three produced minimal symptoms. Patients developing a pseudoarthrosis had an excess of post-operative wound problems, presented much earlier and had more severe infections compared to those without sequelae.

Late infection is confirmed as a relatively common complication of scoliosis surgery. Implant removal, aggressive debridement and primary closure is confirmed as effective treatment to eradicate the infection. A high rate of pseudoarthrosis is the only sequelae. The excess of early infections in these cases may indicate interference with a critical stage of the fusion process. Preservation of the implants until one year post surgery was unsuccessful at preventing a pseudoarthrosis.