In this investigator-initiated randomized controlled trial it could be demonstrated that the use of a self-coiling catheter compared with a regular straight catheter for continuous popliteal sciatic block was more effective in terms of pain management and dislocation rate within the tissue. The latter appeared without visible external changes measured by leakage and external dislodgement of the catheter at the insertion site. The higher stability of the self-coiling catheter in its original positioning may result in lower pain levels on the first and second POD and a reduced opioid consumption compared to patients in the RSC group. To our knowledge this is the first trial investigating impact of different catheter characteristics on secondary dislocations and advantages of a self-coiling catheter for peripheral nerve block in clinical practice.
Pain intensity
Dislocation rate on the day of operation had no impact on pain levels because initial bolus of ropivacaine 0.5 % caused a long-lasting blockage of pain perception in both groups. Long acting ropivacaine provides nerve block duration of up to 18 hours, depending on dose and proximity of applied LA to the nerve. Christiansen et al. described a mean duration of action for distal sciatic blocks of more than 13 hours with a lower dose of 60 mg ropivacaine [22]. In this study 100 mg were administered. Various other studies have reported similar results. Several investigators compared a single popliteal block of the sciatic nerve with a continuous application of local anesthetic via catheter and determined no significant differences in pain intensity on the day of surgery for both groups [23–25]. The effect of the initially performed nerve block was effective in both groups and thus did not result in NRS-differences.
The significance of the lower pain level of the SCC group on the first and second POD showed a measurable difference of 0.7 points on NRS to the control group. Even if this difference seems to be small, it can become relevant for the individual patient. The reason for the higher pain scores in the RSC group on POD 1 and 2 may be related to a higher dislocation rate outward of the fascial space of the sciatic nerve. Thus, a lower effective dose of continuous infused ropivacaine could act on the sciatic nerve. In contrast, the lack of difference on POD 3 might be explained by spontaneous pain relief in the late postoperative period.
With the diminishing effect of the initial bolus of ropivacaine 0.5 % on the first postoperative day, the average pain level of all participants was highest compared to the other measure points. This might be caused by dislocations of indwelling catheters, insufficiency of lower continuous ropivacaine dose and pain perception in the regions not supplied by sciatic nerve when single shot of supplemental regional anesthesia had worn off. The latter issue is relevant if medial parts of the foot were involved in surgery and saphenous or femoral nerve block was established [26]. Pain perception in sensory territory of the saphenous nerve might have contributed to levelling pain intensity scores and scattering intergroup differences regarding perineural sciatic nerve catheter performance. Using an additional catheter for the saphenous nerve might have decreased pain scores and additional opioid use. However, reports about impact of the continuous or prolonged saphenous nerve block on pain after ankle surgery are conflicting. Whereas Fisker et al could not find advantage of continuous saphenous nerve block compared to single shot [27]. Jarell et al. showed lower pain scores and less opioid need with continuous block of saphenous nerve after ankle surgery [28]. Another investigator reported improved postoperative pain management with prolonged saphenous nerve block by additional perineural dexamethasone [29].
Surgery of the distal lower extremity or the foot has shown to be one of the most painful surgical interventions. In almost 71,000 patients with 179 different operations in all areas of the body, the calcaneus operation was revealed to be the most painful operation with a mean postoperative NRS of 6.68 on the first postoperative day. Other operations on the foot, forefoot or ankle were among the most painful fifteen operations with a mean NRS of at least 6 on the first postoperative day [2]. This underlines the importance of continuous regional anesthesia of the ankle and foot.
In our study almost, all patients were free of pain on POD 0. In the further course, all mean NRS values in both cohorts remained at all examination times below 3.5 but only patients with self-coiling catheters performed with mean NRS 2.7 even lower than the commonly accepted intervention threshold of NRS 3.
The absence of ultrasound-guidance and confirmation of proper catheter location, differently used local anesthetics and additional analgetic drugs, the heterogeneity of surgical procedures and the lack of information about mobilization strategies, complicate comparability with other studies. However, most studies show maximum pain intensity on POD 1 [23–25, 30–32]. Similar to our results, usually pain level decreases day by day in the further course. The pain intensity of patients with continuous sciatic nerve block varied between 0 and 4 on POD 1. Our results are in this range. Many of the studies mentioned above had an outpatient setting which may be another reason for the fluctuation range in reported pain intensity. Patients were discharged with the peripheral nerve catheters on the day of surgery and received defined amounts of analgetics as well as instructions for mobilisation. Previous surgery, the duration and length of stay in hospital varied greatly or were not specified. The indicated operation times were very heterogeneous and ranged from 32 min [24] to 110 min [33]. That may have an impact on surgical trauma and consequently on pain intensity. The evaluation of pain and other criteria in the following days was conducted by telephone. As a result, it could not be clearly elucidated whether patients' statements are reliable and how mobilisation was carried out as additional pain stimulus.
Oxycodone consumption
Patients in the self-coiling catheter group had a lower need for oxycodone in the first three days after surgery. While the lower consumption of oxycodone in the SCC group differed by 1.1 mg from the RSC group on the day of surgery, already on POD 1 the consumption was 3.9 mg lower. This difference in lower consumption in the SCC group reached significant levels at the day of surgery, POD 2 (4.9 mg) and POD 3 (3 mg), with an overall decrease in the need for additional opioids already from the second postoperative day on. Interestingly, the significant elevated pain intensity in the RSC group did not reach the level to create a difference for opioid consumption on POD 1.
Two facts might have influence accuracy of discrimination. First, we used a single prefixed dose of oxycodone 10 mg as rescue pain medication obeying our institutional multimodal pain concept. Therefore, discriminatory power regarding additional opioid need was low. Assumably, with lower opioid bolus provided by patient controlled intravenous analgesia we could have traced the opioid requirement more precisely. Second, over 50 % of patients in both groups had surgery that involved also innervation territory of saphenous nerve. Despite reliable sciatic nerve block, the fading effect of the single shot saphenous nerve block might have caused pain at the medial aspect to the ankle and foot and consequently might have led to elevated opioid consumption on POD 1. Thus, we could not distinguish, if opioid request was referred to poor catheter performance or terminated saphenous nerve block. However, the number of saphenous nerve blocks applied was not different between both catheter groups. Overall, the amount of opioid consumption in the present study is consistent with previous studies investigating continuous sciatic nerve block for foot and ankle surgery [25, 30–32].
Additional anesthesia procedures
A further aspect of the discussion is the possible influence of additional anesthetic procedures on postoperative pain intensity. Most patients (80.7 %) underwent adjunct general anesthesia, whereas 11.4 % received additional spinal anesthesia to the distal sciatic block. Sedation or additional peripheral regional anesthesia was given to the remaining 7.9 %. The choice of procedure was individually adapted to the comorbidities and patients request. Due to randomisation and group size there were no significant differences regarding the distribution between of additional anesthetic procedures between both groups, hence a possible influence might be negligible in this investigation. The question of whether spinal anesthesia has an influence on the postoperative pain level has not been conclusively clarified until now. YaDeau et al. compared general and spinal anesthesia, each in combination with PNB for operations on the ankle and foot in a recent randomised controlled trial. A significant difference of pain scores in favour of spinal anesthesia was found only one hour after the end of the operation [34]. In contrast to long acting morphine we used fentanyl as intrathecal supplemental opioid for spinal anesthesia. Thus, any effect on pain intensity beyond first 12 hours seems unlikely.
Catheter orifices
We compared in the present study a self-coiling catheter with a closed tip and six lateral microholes with a regular straight catheter that has only a single orifice at the end. One might question if this could have influenced our results. Fredrickson et al. investigated the outflow of injected fluid on catheters with a different number of orifices. They showed a dependency of the fluid spread pattern on the fluid flow rate. Below 80 ml per hour, fluid left multi-orifice catheters only on the most proximal orifice [35]. Only an injection rate over 100 ml per hour delivered all microholes. Considering our study´s flowrate of local anesthetics of 6- 10 ml per hour, the self-coiling catheter likely functioned rather as a single orifice catheter. Thus, we do not expect any relevant advantage of the multiple catheter orifice configuration in the present study. Moreover, considering that the SCC had been positioned only 2.4 cm into the target space within the perineural facial sheath, with continuous infusion the local anesthetic may have left the most proximal orifice only missing the target space partially if the catheter gets retracted by muscle movements. Clinical data regarding the influence of catheter orifice design on quality of pain management is conflicting, despite LA bolus application was used. Consequently, it can be concluded that catheters with multiple openings, such as the SCC, also function like conventional end-hole catheters at clinically relevant infusion rates. In this study there was a maximum continuous flow rate of 10 ml per hour. Only bolus applications by the acute pain service were probably applicated at speeds above 100 ml per hour. Since this injection was performed manually from a 10 ml syringe, no more precise statements can be made here about the application speed [36, 37]
Dislocation rate
Despite the widespread use of continuous regional anesthesia the topic of perineural catheter dislocation is not well elucidated, neither in studies nor in clinical practice. Thus, the results of our study provide new insights regarding factors contributing to catheter dislocation. To our best knowledge, there is no clinical study investigating dislocation rate of self-coiling catheters compared to regular catheters so far. Luyet et al had shown a decreased initial misplacement rate for self-coiling catheters in human cadavers. However, the cadaver study design did not address dislocation rates in the further course [19].
Significantly fewer self-coiling catheters (14 %) slipped out of the subparaneural target space during the study period than regular catheters with straight ends (60 %). This significant difference might be caused by different insertion distances within the perineural fascia sheath. Ilfeld et al. [38] and Steffel et al. [16] described a higher dislodgement rate for lower insertion distance at the nerve. We used a regular straight catheter with an indwelling metal wire in the control group. Such firm catheters often pass the target structure and protrude out of the perineural fascia sheath during initial insertion of around 3 cm beyond the needle tip, especially if catheters are advanced in-plane perpendicular to the SAX imaged nerve. Thus, we had to adjust the catheter by retraction until LA injection was distributed well within the perineural fascia sheath under sonographic view. The self-coiling catheter provide a more reliable initial placement without passing the target space if the insertion distance of 3 cm beyond the needle tip is not exceeded using an in-plane approach [19]. Thus self-coiling catheters require less likely a withdrawal due to initial misplacement and have a longer catheter segment remaining around the target structure.
For postoperative evaluation of catheter position and initial confirmation of correct catheter placement we used sonographic imaging of saline bolus via the catheter. Although the efficacy of continuous regional anesthesia is depending on proper catheter location, assessment of catheter position is still not a common procedure neither in studies nor in clinical routine [39]. Postoperative break through pain and need of additional systemic analgesic medication are often considered surrogate markers for insufficient catheter performance due to any reason. However, this concept is misleading to prove an incorrect catheter position. It has to be considered that the visualization of catheter position is often compromised by sterile dressing and swollen tissue in the affected area.
Several methods for visualization of catheter position have been described in the literature, a.e. imaging of injected fluid (either saline, LA, or contrast medium) or air spreading out of catheter orifice or direct visualization of the catheter [14, 16, 40–42]. Imaging techniques include high resolution ultrasound (HRUS), computer tomography (CT), or magnetic resonance imaging (MRI) [19, 40, 41, 43]. There are pros and cons of each technique that we would like to discuss briefly. We decided for the injection of saline to prove catheter tip position because it is a safe and straightforward method that is part of our daily clinical routine. Moreover, it avoids unnecessary local anesthetic doses and harm to vulnerable structures nearby. The use of air or agitated fluid with microbubbles may enhance contrast and visualization of the injectate. However, spreading air within the tissue decreases markedly sonographic imaging quality of the target structure and surrounding tissue by scattered ultrasound waves. In contrast, visualization of saline via the catheter is similar to common procedure of observing local anesthetic injection via the cannula. Finally, direct visualization of the catheter is less favourable. On the one hand, despite the improved echogenity of catheters they hardly alignm with the ultrasound beam plane like a firm needle. This issue is aggravated by the use of self-coiling catheters. On the other hand, according similar to the injection over needles, control of spread of LA around the target structure is more important for a reliable block success than the catheter tip position itself. Compared to CT and MRI high resolution ultrasound imaging is commonly available. HRUS is a real point of care technique that can be applied as often as desired by the anesthesiologist avoiding unnecessary patient transports and staff expenses. Furthermore, CT examination exposes patients to radiation burden.
In our study the most dislocations were discovered after arrival at the PACU. It can be assumed that passive and/or active movement of the thigh musculature occurs when the patient is positioned prior to surgery or during patient transfer to bed. This causes mechanical traction to the catheter within the biceps femoris muscle pulling the catheter back outward of the target space within the nerve-surrounding fascia.
The dislocation rate within tissue was 14 % for the self-coiling catheter or 60 % for the regular straight catheter. Marhofer et al. examined the dislocation rates within the tissue in healthy volunteers. They determined a dislocation rate after movement of 25 % and 5 % for perineural femoral and interscalene catheter [15]. In contrast to our investigation, catheters were only examined for six hours after insertion. In addition, both catheters were placed using the out-of-plane technique, which is more robust against dislocation [43].
Only two studies address internal dislocation rates for continuous popliteal sciatic nerve block. Steffel et al. compared a catheter-over-needle (CON) with a conventional catheter-through-needle (CTN) technique in human cadaver [16]. 27 % of the CON catheters dislocated from the fascial sheath of the sciatic nerve. In contrast, all conventional catheters remained perineural. Remarkably, Steffel et al. did not evaluate the spread of the local anesthetic, but only visualized sonographically the suspected end of the catheter in the sonography. However, LA spread around the target is the decisive determinant for a sufficient nerve block. Comparisons with our study are difficult, since only a tiny cohort of 30 persons was involved and the tissue characteristics of the body donors cannot be considered identical to those of living subjects [16]. In contrast to our patients, only passive flexion movements were performed on the body donors. However, we believe that an active contraction and relaxing of the muscle contributes considerably to the movements of the catheter within the tissue. Hauritz and colleagues compared two different approaches for popliteal sciatic blockade, regarding dislocation within the tissue. They confirmed catheter location 48 h after application by means of an MRI contrast bolus [10]. Whereas a dislocation rate of 10 % for the out-of-plane approach was reported, the use of an in-plane technique similar to our study protocol resulted in a much higher dislocation rate of 40 %. We found an even higher dislocation rate of 60 % for conventional catheters in our study. Important difference is the longer part of catheter remaining under the perineural fascia sheath. A catheter distance within the perineural fascia sheath of 1.6 cm in our study compared to 3.4 cm in the investigation of Hauritz et al. may increase catheter dislocations due to traction forces as already mentioned before. Interestingly, the application of self-coiling catheter using the in-plane approach decreased the dislocation rate to a level reported by Hauritz et al. for the out-of-plane technique. Whereas regular straight perineural catheters seem more reliable in point of dislocation rate using out-of-plane approach, whereas the self-coiling catheters may be placed with good results performing the in-plane technique.
Leakage and catheter shift at the insertion site
The maximum leakage rate was 31.4 % in both groups in our study. This is comparable with the leaking rate of 31 % for CTN technique in another study [40]. Lower leakage rates of 13.9 % for continuous distal sciatic catheters were reported [44]. Leakage problems commonly occur with catheter-through-needle approaches, because needle´s diameter is larger than the catheter. Thus, the tissue is not sealing the puncture track along the catheter and injected fluid as well as interstitial fluid or blood can flow retrogradely. Though catheter-over-needle technique decreases the occurrence of leaking, no superiority regarding dislocation rates has been shown in clinical trials so far [40, 45, 46]. Accordingly, we did not observe any relation between leakage and dislocations since both study groups showed the same leakage rate. Dislocation rates in terms of the catheter slipping out of the skin at the insertion site are reported with an incidence of 0.5 to 26 % [8, 9]. In our study we considered an outward slipping at the insertion site of 1 cm or more as a clinically relevant movement. This was evident in 10 % in SCC group and 12.6 % in RSC group. According to findings of Marhofer et al. we could not observe a significant impact of outwardly slipped catheters at insertion site on the dislocation rate at target area [15].
Complications
The overall complication rate was very low. No persistent neurologic deficit was observed. The most severe complication was a patient with a metallic taste several hours after catheter application considered as a mild sign of local anesthetic systemic toxicity. Ropivacaine infusion was stopped. Neither a negative aspiration test nor sonographic imaging of fluid spread via the catheter revealed a secondary intravascular dislocation. The catheter was removed immediately, followed by complete regression of symptoms. The overall mild infection rate of 5.7 % observed here ranges within the results of other studies [47, 48]. Mild infection was considered as any redness of the puncture site during daily visit. The catheters were removed immediately without any further sequels.
In this study, catheter occlusion occurred in one patient from each cohort (total 1.4 %). Ma et al. reported this rare incident with a rate of 1 % [44].
Limitations
Our results are limited to the use of ropivacaine for the initial bolus. The dislocated catheters probably would have been earlier discovered by using a short- or middle-long acting local anesthetic for the initial bolus. Additionally, for rescue pain management patients received opioids orally by nurses only on demand in a fixed dosage of 10 mg Oxycodone. A more finely tuned opioid application, e.g. by patient controlled intravenous analgesia could have led to a clearer reflection of the actual need for additional painkillers.
Another limitation is the single shot concept for saphenous nerve block with limited pain relief of 12- 18 hours. Thus, NRS scores may have been influenced by pain perception in saphenous nerve innervation area.
Furthermore, the design of our study was not double blinded since the ultrasound examiner occasionally could have drawn conclusions regarding catheter type by watching insertion depth specific catheter length graduations.
The results of this study apply only to our setting of short axis view of the nerve and in-plane needle approach. Beyond that, our results are not applicable for other catheter designs or alternative insertion sites.
In the case of sonographically confirmed catheter dislocation in situ, no further positional checks were performed. Whether it is possible that the misalignment could spontaneously convert to a renewed perineural position or not remains unclear.
Furthermore, extrafascial dislocation does not necessarily mean a complete loss of effectiveness. Our continuous ropivacaine dose may have been still sufficient to release pain by local anesthetic spreading toward the nerve along the residual puncture pathway or by diffusion through the connective tissue. Experience from the efficacy of interfascial plane blocks (e.g. the erector spinae block), we suspect some analgetic effects of even low amounts of LA by blocking small C-fibres, even if the application site is not close to the target nerve.