Femoral nerve catheters are now used more frequently following TKA due to their improved efficacy in relation to lower pain scores, and an associated decrease in narcotic use post operatively. The benefits afforded by improved pain management related to the use of CFNCs, may also extend to benefits in short term functional recovery [7–9, 23] and this study has sought to contribute further knowledge in this area by investigating the relationship between the use of a CFNC and patient recovery in terms of their pain management, and functional mobility prior to discharge from hospital. Both the CFNC and the PCA only groups compared in this study had similar characteristics in regard to age, body mass index, operation site, length of hospital stay and whether a gait aid was used pre-operatively. There were also no significant differences in either the type of gait aid used postoperatively or in flexion attained when performing CPM exercises, whilst in hospital. Intravenous narcotic PCA was available to patients in both groups for a similar duration and both groups appeared to have a similar understanding of how to use PCA, as evidenced by comparable percentages of successful PCA dose attempts delivered. However there was a higher representation of women in the CFNC group than in the PCA only group.
This study's results have supported previous findings that patients with a CFNC use less supplemental narcotic analgesia during the postoperative period [6, 24, 25]. Having a CFNC in situ had a significantly large effect on the total fentanyl used with lower doses used, not only on the day of operation, but also on the first and second days postoperatively. The amount of fentanyl used also remained significantly lower when BMI was taken into account (Table 2). In general, patients in both groups appeared to manage their pain appropriately with an average pain score below 4 for both groups, however the highest pain score reported by patients was significantly higher (medium effect size) in the PCA only group on both the day of, and the day after, the operation. Although nausea scores were higher in the PCA only group, the difference did not reach significance in this sample and this is comparable with findings from previous studies [6, 24, 26].
Short term functional mobility was measured using a TUG test and AROM on Day 4 following surgery, which was the day prior to assessment for hospital discharge on the clinical pathway. In a study that examined physical performance measures after TKA  the TUG test was found to be useful in the early recovery period between one and nine to ten weeks postoperatively, by which time the test had reached a ceiling effect where most patients had met the 10 second criteria for being functionally independent . Preoperative TUG results in our preliminary study indicate a degree of disability prior to surgery with the sample averaging a time of 12.4 seconds. This was greater for women, who had slower TUG times and supports previous research that found gender differences exist, with women reporting greater disability at the time of arthroplasty and lower self-reported function [18, 27, 28]. This may be explained by a study that found that women undergo arthroplasty at a more advanced disease state than men and also, as a result, had reduced muscle activation and increased atrophy preoperatively . This level of disability was also reflected for women in our study's post operative TUG scores with women again having significantly slower times than men. It should also be noted that the researchers subjectively observed the men to be very competitive when completing their TUG tests, often asking the researchers how they compared to others and then "swapping results" on return to their shared rooms. Despite the gender influence, and the higher proportion of women in the FNC group, median TUG times were quicker for the FNC group (45 secs) than the PCA only group (58 secs), although the difference was not significant for this sample.
In comparison, the AROM was significantly lower in the CFNC group than in the PCA only group. Support for this finding is mixed. These results are not supported by previous findings from studies by Kadic et al.  and DeRuyter et al.  who found increased flexion on Days 3 to 6, and at Day 1 and discharge, respectively, in patients with a CFNC in situ for 48 hours. However a study by Carli et al.  using only a slightly lower dose regimen of 8 ml/h, compared to the10 ml/h of 0.2% ropivacaine in this study, found no significant difference in knee flexion on Days 1 and 2. The results in our study were not explained by the significant difference in the time analgesia was administered prior to testing, or by gender differences in each group. Most likely they can be explained by studies that found that quadriceps strength is the strongest predictor of functional performance [29, 30]. This is a matter for concern in regard to short term functional mobility considering the recent published caution that femoral nerve blockade may result in prolonged quadriceps weakness and an increased risk of falls . The variance in our results compared to the short term improvements seen in the study by Kadic et al.  may be related to a reduced dose of ropivacaine received by participants in their study. They described a dose of between 5 and 10 ml/h of 0.2% ropivacaine for the first 48 h (as opposed to an average dose of 10 ml/h in this study) although there were no details of the mean hourly or total dose received by participants making comparisons between the studies difficult. However recent data on the minimum local anaesthetic concentration (MLAC) showed that the minimum concentration at which patients did not require rescue analgesia using levobupivicaine was 0.024% for the femoral nerve and 0.014% for the sciatic nerve . Even with this ultra-low concentration, which equates to a reduction in the commercial preparation's concentration of four to six fold, there was mild motor block manifested by an inability to dorsiflex, which prevented early mobilisation. This may indicate that the doses of ropivicane used in this study are related to the significantly lower AROM observed. However this may not be significant in terms of long term recovery as Kadic et al. also observed that short term improvements in knee flexion did not correlate with increased knee flexion and improved functional outcomes at three months.
The preoperative decline of quadriceps' strength and function has been shown to impact negatively on functional recovery [33, 34] and may also explain our findings of a strong positive correlation between slower preoperative TUG times and poorer postoperative TUG performance. However the correlations between slower TUG times and an increased time between analgesia and exercise, and a higher pain score, also reinforces the importance of timely effective analgesia prior to undertaking exercise in the post operative period prior to discharge. Our findings suggest that oral analgesia prior to exercise, once the femoral nerve catheter has been removed, is influential in improved functional mobility although this is at odds with recent findings from Denmark that pain has limited impact on functional recovery past the first postoperative day .
Positive impact on functional recovery may also be achieved through earlier surgical intervention to prevent severe disability prior to surgery [34–36]. However, TKA is often delayed by either wait lists or the requirement to decrease the need for future revision arthroplasty. In these cases improved preoperative function through the use of physical therapies has been shown to be effective in improving postoperative function .
Study limitations and considerations for the follow-up study
As a preliminary study this design was limited by the lack of probability sampling and small sample number. Subsequently there was a difference in gender representation with the increased proportion of females in the CFNC group being more representative of the general population, than in the PCA only group. This emphasises the importance of conducting a randomised design with an a priori power analysis in the follow-up study. Despite the absence of these design features in this preliminary study, the comparison groups were similar in their characteristics and the study was conducted prospectively with differences between groups identified as having medium to large effect sizes, so enabling important information for the conduct of the follow up study.