There are concerns about the risk of iatrogenic infection when employing local anaesthetic techniques with post-operative intra-articular infusions in total knee arthroplasty. This study aimed to determine the efficacy of intact epidural filters in preventing transit of bacteria and to develop a technique of administration which would prevent membrane rupture.Filter efficacy was assessed using a standardised test suspension of Pseudomonas aeruginosa. Twenty millilitres of suspension was injected through isolated epidural filters (n=10) or filters with 40cm of catheter tubing attached (n=30). For each filter, injections were carried out at 0, 8 and 24h. Filtrates were collected, incubated, sub-cultured onto Columbia horse blood agar and examined for bacterial growth.Three delivery techniques were tested: manually controlled syringe with 5ml of water at 20ml/min, forced administration syringe with 5ml of water at >240ml/min and an automated syringe driver delivering 40ml of water at 6.7ml/min. For the two techniques using syringes, three syringe sizes, 5ml, 10ml and 20ml, were tested. Each test condition was carried out on 10 filters (total n=70). Filters were examined for rupture.Intact epidural filters prevented bacterial transit in all cases. Manual controlled and automated syringe driver administration generated no filter ruptures. Manual forced administration generated 93% filter rupture. Ruptures occurred at peak pressures of approximately 620kPa.Epidural filters can be used to prevent bacterial transit. These results suggest automated devices remove the risk of filter rupture. This study is relevant to all specialties that utilise these filters during infiltration such as epidurals or other regional anaesthetic techniques. Clinical relevanceThis study identified that filters are prone to rupture with high infusion rates and that manual techniques are particularly vulnerable. From these results, it is recommended that pumps are used to minimise risk of filter rupture.