Terminated prior to its planned completion as anticipated by the protocol).

The anaesthetic management of patients undergoing total knee replacement is still not standardised. Epidural analgesia is common, but is associated with bilateral lower limb motor block and limited mobilisation. Spinal anaesthesia with intrathecal morphine provides good pain relief, but only for 12 to 16 hours, and is often associated with nausea and vomiting. Combined single injection femoral / sciatic blocks give good pain relief, but for a variable length of time (between 12 and 24 hours).

In contrast, continuous femoral perineural infusion of local anaesthetic provides very good pain relief for several days. Pain relief is maintained by a constant infusion of local anaesthetic using an elastomeric ball. The investigators overriding aim is to develop a local anaesthetic regimen which offers complete pain relief for at least 48 hours, yet allows full mobilisation within 24 hours Given that ropivacaine may offer a more advantageous pharmacological profile (less lipid solubility) compared to levobupivacaine, the investigators feel it is pertinent to investigate the capacity of ropivacaine to prevent pain relief after surgery. Thus, the investigators aim in this study is to compare the median effective dose of levobupivacaine with the of ropivacaine for preventive pain relief after total knee replacement. Further, calculation of the equipotent median effective dose's of each local anaesthetic allied to objective measurement of quadriceps motor block using an electromyogram will allow us to determine the sensory - motor split of each local anaesthetic.

Primary objective The primary aim of this study is to measure the molar of levobupivacaine and the molar median effective dose of ropivacaine when administered as a femoral perineural infusion to prevent pain 30 hours after total knee replacement Secondary objectives

To assess the:

Degree of postoperative pain, need for rescue analgesia,side effect profile of both groups, self-efficacy, expectancy, quality of life / health economics Type of study: Prospective single blind randomised controlled trial Planned number of subjects: 48 Study duration: 24 months Randomisation procedures Following informed consent, participants will be randomised into two groups. All preparations for study will be coded so that participants have an equal chance of receiving one of the two local anesthetics. Randomisation will be carried out by computer program, and codes kept in pharmacy in a sealed envelope.

Protocol for patient management Preoperative - Anaesthetic room Antibiotics: augmentin 1.2 grams Sedation with Target Controlled Infusion of propofol at plasma level of 0.5 micrograms per millilitre (μg.ml-1) Spinal anaesthesia with 3.2 ml hypobaric "plain" bupivacaine 0.5% and intrathecal morphine 0.1 mg All regional nerve blocks performed under ultrasound control Femoral block: 15 ml 0.3% levobupivacaine or ropivacaineFemoral perineural catheter inserted by Anaesthetist 5cm in a cranial direction Sciatic block: single injection 12 ml 0.5% levobupivacaine or ropivacaine Obturator block: single injection 6 ml 0.5% levobupivacaine or ropivacaine

Postoperative 8 hour anaesthetic observation

Determine median effective dose of levobupivacaine and ropivacaine for anaesthesia 8 hours after insertion of spinal needle, resolution of spinal anaesthesia confirmed and verbal rating pain score of knee bending recorded. Clinical measurement and subsequent patient dosing is as follows:

1. "Anaesthetic success": verbal rating pain score = 0: Next patient receives 10 microMolar (μM) less of levobupivacaine or 10μM ropivacaine.
2. "Anaesthetic failure": verbal rating pain score ≥ 1. Resolved by rescue analgesia of 15ml, 150μM levobupivacaine (5.8ml, 0.75%) or 15ml, 150μM ropivacaine (6.2ml, 0.75%).Next patient receives 10μM more of levobupivacaine or 10μM ropivacaine.
3. Technical failure: verbal rating pain score ≥ 1 Not resolved by rescue analgesia of 15ml, 150μM levobupivacaine (5.8ml, 0.75%) or 15ml, 150μM ropivacaine (6.2ml, 0.75%.Next patient receives same mass levobupivacaine or ropivacaine.

After the 8 hour assessment, all patients (excluding technical failures but including "anaesthetic failures") will start a femoral perineural infusion with levobupivacaine or ropivacaine at a rate of 10ml h-1 and concentration of 1(microMolar per millilitre) μM.ml-1 (400μM in 400ml elastomeric ball) for 48 hours.Patients classed as "technical failure" will have a femoral perineural catheter reinserted.

Postoperative 30 hour (±2hour) anaesthetic observation

The postoperative 30 hour anaesthetic observation will only be performed on the "anaesthetic successes" from the 8 hour observation. The median effective dose of levobupivacaine and ropivacaine for analgesia will be determined by choosing an end point of verbal rating pain score 0 at 30h after the start of spinal anaesthesia. Patients will divide into three groups:

1. "Successful pain relief" verbal rating pain score = 0 Next patient receives 10μM less of levobupivacaine or 10μM ropivacaine.
2. "Unsuccessful pain relief" verbal rating pain score ≥ 1 resolved by rescue analgesia of 15ml, 50μM levobupivacaine (5.8ml, 0.25% + saline) or 15ml, 50μM ropivacaine (7.7ml, 0.20% + saline).Next patient receives 10μM more of levobupivacaine or 10μM ropivacaine.
3. Technical failure: verbal rating pain score ≥ 1. Not resolved by rescue analgesia of 15ml, 50μM levobupivacaine (5.8ml, 0.25% + saline) or ml, 50μM ropivacaine (7.7ml, 0.20% + saline).

Rescue analgesia Rescue will be standardised: 15ml, 50μM levobupivacaine (5.8ml, 0.25% + saline) or 15ml, 50μM ropivacaine (7.7ml, 0.20% + saline). The number of rescue administrations will be counted.

.

Inclusion Criteria:

• All patients >18 years of age and non pregnant presenting for elective knee replacement

Exclusion Criteria:

• Signs of cardiac failure (3rd heart sound, lung crepitations)
• Type I and type II diabetes
• Abnormal cardiac arrhythmias
• Hypovolaemia
• Presence of seizures,
• Dementia,
• Depression
• Encephalopathy,
• Terminal illness with a life expectancy < 3 months
• Age < 18 years
• Pregnant
• Coagulopathy

30h