1. Aetna considers infusion pumps for intralesional administration of narcotic analgesics and anesthetics experimental and investigational because the effectiveness of these pumps has not been demonstrated in well-designed clinical studies in the peer-reviewed published medical literature.

2. Aetna considers infusion pumps for intraarticular administration of narcotic analgesics and anesthetics experimental and investigational because they have not been proven to improve post-operative pain control.

Note: This policy does not apply to continuous peripheral nerve blocks (e.g., brachial plexus blocks, inter-costal blocks, femoral nerve blocks).

See also CPB 010 - Continuous Passive Motion (CPM) Machines, and CPB 161 - Infusion Pumps.

Pain relief after surgery is often provided by patient-controlled systemic analgesia, which uses an intravenous infusion pump and a patient-activated switch to administer narcotic analgesics.

In order to avoid the complications associated with systemically administered narcotic analgesia, infusion pumps have been developed to administer narcotic analgesics and anesthetics directly into the lesion.  The On-Q Pain Management System, the Pain Buster Pain Management System, the Don Joy Pain Pump, and the Stryker Pain Pumps are brand names of devices designed to provide pain relief at the operative site for patients recovering at home from day surgery.  These devices have been used most frequently for patients who have undergone orthopedic or "sports medicine" surgery to repair knee and shoulder problems. 

Attached to the catheter is a small plastic pump that automatically directs a local anesthesia to the source of the pain. The pumps have been used to dull the pain and eliminate the need for systemic narcotic and non-narcotic analgesics. Narcotics have also been infused directly into inflamed tissue. The device is secured to the body until the narcotic medication or anesthetic is depleted, and the patient can remove it him/herself.  The manufacturers of these devices claim that patients treated in this way are able to move around sooner following surgery and participate in rehabilitation with greater ease, and require fewer drugs to aid in recovery.

Studies in the medical literature, however, have not shown better patient outcomes (in terms of enhanced pain relief, reductions in disability, improvements in function or faster recovery) when these devices are used in place of or in addition to standard (systemic) administration of narcotics. 

Well-designed randomized controlled clinical studies evaluating both subjective endpoints of reduction in pain and objectively measured functional endpoints (reductions in disability and improvement in function) are especially important in evaluating pain interventions because of the susceptibility of pain to placebo effects.  The study by Alford, et al. (2003) found reductions in pain and narcotic use in subjects both subjects receiving intra-articular anesthetic and subjects receiving intra-articular saline compared to a comparison group receiving no catheter, suggesting an important placebo effect from intra-articular infusion pumps.  These findings were consistent with a study by Rosseland, et al. (2004), which found significant effects of intraarticular infusion of saline.

A study by Alford, et al. (2003) is significant in that it reported on functional outcomes (reductions in disability, improvements in function) in addition to subjective pain scores and narcotic consumption.  The investigators found no significant differences in functional outcomes (range of motion, straight leg raises) between the group receiving intra-articular anesthetic and the group receiving intra-articular saline.  Other randomized controlled clinical studies of intraarticular and intralesional anesthetic pumps are of weaker design than this study if they report only on pain scores and supplemental analgesic use, and not on functional outcomes.

Available studies do not consistently demonstrate clinically significant reductions in narcotic consumption in subjects receiving intra-articular or intralesional anesthetic. In some studies, there was no significant reduction in narcotic usage in subjects assigned to intra-articular or intralesional anesthetic infusion compared to subjects assigned to intra-articular or intralesional saline infusions. 

In available studies, the reported reductions in pain scores in groups receiving intra-articular or intralesional anesthetics were generally modest and inconsistent, with some studies reporting significant reductions in some types of pain with intra-articular or intralesional anesthetic but not others. 

A number of studies have failed to find any significant effect of intra-articular or intralesional infusion of anesthetics (see, e.g., Boss, et al., 2004; Drosos, et al., 2002; Aasbo, et al., 1996; Henderson, et al., 1990; Joshi, et al., 1993; Klasen et al, 1999; Schwarz et al, 1999; Rautoma et al, 2000; DeWeese et al, 2001).

There are a paucity of studies that have directly compared the effectiveness and safety of intra-articular or intralesional infusions with established methods of postoperative analgesia.  Several such studies have been published, showing intra-articular or intralesional infusion to offer inferior postoperative pain relief (see., e.g., Dauri, et al., 2003; Iskandar, et al., 2003).

Available studies are small and not sufficiently powered to evaluate uncommon but clinically significant adverse effects of intralesional catheters.  The maintenance of a catheter in the wound may have an effect on infection and wound healing.  In addition, anesthetics used in continuous wound perfusion have vasoconstrictive properties that may adversely affect wound healing by decreasing blood flow to injured tissues.  Systemic absorption of large doses of anesthetic may be toxic.

In sum, available studies suggest that pain relief from intralesional and intra-articular anesthetics, if any, is modest and it remains unclear whether any analgesia produced by intra-articular and intralesional anesthetics is clinically useful. 

Alford, et al. (2003) reported on the effectiveness of postoperative intra-lesional anesthetic infusion after anterior cruciate ligament reconstruction.  This study is significant in that it is a blinded, randomized, controlled clinical study that examined not only subjective pain endpoints and narcotic consumption but also objective endpoints of physical function.  In this study, 49 patients were randomly assigned to one of three groups: no catheter, an infusion catheter filled with saline, and an infusion catheter filled with anesthetic.   The only statistically shown benefit of intra-lesional anesthetic infusion over saline infusion was in maximum pain ratings.  Median pain ratings were significantly lower in both catheter groups compared with the group receiving no catheter; however, there were no significant differences in median pain ratings between the catheter groups.   Only the saline catheter group had significantly less narcotic consumption than the no catheter group.  Narcotic consumption of the anesthetic catheter group was intermediate between the saline catheter group and the no catheter group, and not statistically significantly different than the no catheter group.   Physical therapy data showed no significant differences in range of motion on postoperative day 4 among groups.  Significantly more patients were able to perform straight leg raises during the first physical therapy session in both the saline catheter group (70%) and the anesthetic catheter group (72%) than the control group (50%).  This study suggests a strong placebo effect from the use of a saline catheter.  There were no consistent differences in outcomes between the saline catheter and anesthetic catheter groups.

Gupta, et al. (2002) reported on a prospective, double-blind, randomized controlled clinical study of 40 subjects undergoing laparoscopic cholecystectomy.  This study was of stronger design than many other randomized controlled clinical studies of intralesional anesthetic pumps in that it includes as outcome measures both subjective assessments of pain and objective assessments of supplemental narcotic pain medication consumption, reductions in disability and improvements in function.  This study found a modest benefit to intralesional anesthetic pumps that was limited to only the first few hours after surgery.  Statistically significant differences in pain intensity (VAS scores) between patients receiving intralesional anesthesia versus intralesional saline infusion were limited to deep pain and pain during coughing during the early post-operative period (within 4 hours following surgery), with no differences in pain at the shoulder or incisional sites.  There were no significant differences in VAS scores between groups more than 4 hours after surgery.  However, these investigators noted that, in general, the pain intensity was mild, even in the placebo group.  There were no significant differences between groups in the amount of supplemental narcotic analgesic medication used, in the number of patients requiring no supplemental narcotic analgesic medication, or in the number of patients requiring higher doses of narcotic medication.  There were also no differences between groups in objective measures of post-operative recovery: time to transfer from Phase 1 to Phase 2 recovery, time to sit up in bed, time to stand and walk without support, time to drink and eat, time to void, and time to discharge home.  The most common post-operative complication was nausea, which was significantly more common in subjects receiving intralesional anesthesia.  No differences were seen between the groups during the first week.  The median times to start eating regularly, walking normally, defecating, driving the car, and return to normal activities of daily living were also similar between groups. 

A study by Schurr, et al. (2004) of intralesional anesthetic infusion in 80 patients undergoing inguinal herniorrhaphy is also a prospective double-blind, randomized, controlled clinical trial, that assessed both pain and objective functional outcomes (activity, return of bowel function).  These investigators reported a “mild reduction” in worst pain in patients receiving intralesional anesthesia (mean 6.7) than patients receiving saline (mean 5.0).  There was no reduction in the total amount of time spent in moderate pain between groups.  On day 1, least pain ratings were also lower, and patients ambulated more frequently than those who received placebo.  The investigators reported no differences between groups from post-operative day 2 to 5.  In addition, the investigators reported no differences between groups in hydrocodone consumption.  The investigators concluded that intralesional anesthetic infusion provided modest improvements in pain scores and functional outcomes when compared with placebo.  The investigators noted, however, that these effects were limited to the first post-operative day only.   The investigators considered that the same effect may be achieved by administering a pre-operative dose of an extended-release oral opioid or a NSAID without anti-platelet effects to control background pain in the immediate post-operative period and for the first 24 hours.  The investigators reported 5 % of the infusion pumps failed immediately, and 19.4 % of subjects who completed the study reported leakage of the infusion fluid from around the catheter infusion site.  The investigators noted a 4 % infection rate among study subjects, which is 10 times the historical rate of infections associated with this procedure for the investigators’ institution.  The investigators stated that this study was too small to evaluate infection risk, and that a larger prospective study comparing intralesional anesthetic infusion versus no infusion is needed to completely define this risk.   The investigators concluded that “[a]lthough continuous infusion of bupivacaine after inguinal herniorrhaphy provides multimodal postoperative pain therapy, the pain-related outcomes are modestly improved at best and are limited to the first postoperative day.  The high incidence of leakage from the skin site and suggestion of increased infection risk alter the risk-to-benefit ratio of this technique”.   The investigators concluded that the additional costs associated with intralesional anesthesia may limit its widespread use in clinical practice.

A study by Sanchez et al (2004) of 45 patients undergoing inguinal hernia repair is also of weaker design than previously described studies.   Although this is a randomized, blinded study, only patients’ perception of pain and analgesic use were assessed, and objective measures of post-operative recovery were not assessed.   Although the investigators reported significant differences in pain scores in patients assigned to intralesional anesthesia versus placebo on post-operative days 2 through 5, there were no significant differences between groups in the amount of narcotic analgesics that were used.

A study by LeBlanc et al (2005) of 52 patients undergoing open inguinal herniorrhaphy is also of weaker study design because outcomes were limited to pain scores and analgesic use, and post-operative recovery was not assessed.   Pain (VAS) scores were not significantly different between groups.  Narcotic use was significantly higher in placebo subjects, but narcotic use decreased significantly in both groups beyond the first post-operative day.  There was no difference in duration of hospital stay between groups.

Noting that "the effectiveness of continuous intra-bursal infusion of analgesics for prolonged pain is yet unproven," Park, et al. (2002) undertook a prospective, randomized, double-blind, controlled clinical study of intra-bursal infusion of anesthesia versus saline in 60 patients following subacromial arthroscopy procedures.  All subjects received a postoperative intra-bursal bolus of anesthetics.  One group also received a continuous infusion of anesthetic into the subacromial space, and the control group received a continuous infusion of saline into the subacromial space.  The anesthetic group reported significantly less rest pain, but there was no difference in pain caused by movement.  In the anesthetic group, lesser amounts of supplemental analgesics were used in the first two days postoperatively, and there was no significant difference in supplemental analgesics on the third day postoperatively.  This study is of weaker design than the previously described study by Alford in that it only assessed postoperative pain and medication use, and did not assess objective functional measures. 

Noting that "at present, there is no clinical evidence of real effectiveness and safety of continuous wound perfusion after spinal surgery," Bianconi, et al. (2004) reported on a study of 37 patients undergoing posterior lumbar arthrodesis who were randomized into two groups: one group received an intravenous analgesic infusion following surgery, and the other group received an infusion of local anesthetic directly into the surgical area. Pain scores, use of rescue medication, and duration of hospital stay were less in the group receiving a local anesthetic infusion. However, the intravenous analgesic infusion was discontinued after 24 hours following surgery, while the continuous wound perfusion was maintained for 55 hours.

Dauri, et al. (2003) compared the effectiveness of epidural, continuous femoral block, and intraarticular analgesia in 60 patients undergoing anterior cruciate ligament reconstruction. Patients were randomly assigned to receive continuous epidural ropivicaine, continuous ropivicaine femoral block, or continuous intraarticular ropivicaine. The investigators reported that visual analog pain scores were significantly higher in the group receiving intraarticular anesthetic 24 hours following surgery, and that use of supplementary analgesics was significantly higher in the group receiving intraarticular anesthetic throughout the postoperative observation.  The investigators also reported that intraarticular analgesia was associated with a lower degree of patient satisfaction.  The investigators concluded that epidural or continuous femoral nerve block provide adequate pain relief in patients undergoing anterior cruciate ligament reconstruction, whereas intraarticular analgesia seems unable to cope satisfactorily with the analgesic requirements of this surgical procedure.   

Gupta, et al. (2004) reported on the results of a randomized controlled clinical trial comparing continuous intraperitoneal infusion of levobupivacaine versus normal saline placebo in 40 women undergoing elective abdominal hysterectomy. The investigators found a reduction in opioid consumption in the levobupivacaine group lasting from 4 to 24 hours after surgery which was associated with a reduced incidence of nausea. Despite a reduction in analgesic requirement during this period with levobupivacaine infusion, patients had moderate pain during coughing, which the investigators concluded was “unsatisfactory.” In addition, no differences were found between the groups in other endpoints, including vomiting, time to eating, drinking, mobilizing, or home discharge.

Boss, et al. (2004) examined the effectiveness of continuous subacromial bupivacaine infusion in 42 patients undergoing acromioplasty and rotator cuff repair. Patients were randomly assigned to subacromial continuous infusions of bupivacaine or saline (placebo). The investigators reported no significant differences in supplemental opioid consumption by intravenous patient controlled analgesia, in antiemetic use, or in subjective pain perception by visual analog scale between the groups. The investigators concluded that the continuous subacromial infiltration of bupivacaine anesthetic is ineffective in providing pain relief after rotator cuff repair and acromioplasty surgery.

Fredman, et al. (2001) reported on the analgesic efficacy of patient-controlled wound instillation of the analgesic bupivacaine in 50 patients undergoing major abdominal surgery. Subjects were randomly assigned to either bupivacaine or sterile water. The investigators found no significant differences between groups in the amount of rescue opioid requirements during the 24 hour study period. The investigators reported that visual analog scores for pain were similar between groups at rest, on coughing, and after leg raise. The investigators concluded that bupivacaine wound instillation via patient controlled analgesia pump does not decrease pain or postoperative opioid requirements after abdominal surgery.

In a randomized study, Zieren, et al. (1999) compared the effect of repeated intralesional boluses of local anesthetic to oral analgesic in 104 patients undergoing tension-free inguinal hernia repair. Patients were randomly assigned to postoperative repeated boluses of bupivacaine analgesic through a subcutaneous catheter or oral analgesic dipyrone administered 6, 12, and 24 hours after operation. The investigators reported no significant differences between groups in absolute pain scores, course of pain, and the effects of analgesics. There were no differences in duration of hospital stay between groups. The investigators concluded that repeated intralesional boluses of local anesthetic did not result in better pain control than oral analgesics after tension-free inguinal hernia repair.

Schurr, et al. (2004) evaluated postoperative continuous wound infusion of the local anesthetic bupivacaine to saline placebo in patients undergoing inguinal herniorrhaphy. The investigators reported that patients ho received bupivacaine had lower ratings for worst pain than patients who received saline. On day 1, least pain ratings were lower in patients receiving bupivacaine, and patients ambulated more frequently than those who received placebo. However, these differences did not persist beyond the first postoperative day, and there were no differences between groups between postoperative days 2 through 5. The investigators also reported no differences between groups in rescue narcotic consumption. The investigators concluded that continuous infusion of local anesthetic after inguinal herniorrhaphy provided “modest” improvements in pain scores and functional outcomes when compared with placebo. However, the investigators noted that these effects were limited to the first postoperative day only.

Bianconi, et al. (2003) reported on the results of a randomized trial comparing intravenous infusion of morphine plus ketorolac to continuous wound infusion of the anesthetic ropivicaine in 37 patients undergoing hip or knee joint replacement surgery.   The investigators reported that the group receiving the continuous wound instillation of had less postoperative pain at rest and on mobilization, less use of rescue medication, and a shorter hospital stay, than the group receiving intravenous analgesics.  However, the intravenous medication was discontinued after 24 hours, while the continuous wound instillation was continued for 55 hours.  The investigators noted that this was the only study of continuous wound instillation of local anesthetic after hip or knee arthroplasty, and that further studies may be necessary to confirm the efficacy of this new pain management strategy.

Axelsson, et al. (2003) reported on a study involving 30 patients undergoing arthroscopic subacromial decompression who were randomized into three groups: group 1 received a preoperative bolus of intra-bursal anesthesia plus a patient-controlled infusion of anesthesia via a simple elastomeric balloon pump into the subacromial space; group 2 received a preoperative bolus of intra-bursal saline plus a patient-controlled infusion of anesthesia into the subacromial space via balloon pump; group 3 received a preoperative bolus of saline plus a patient-controlled infusion via balloon pump of saline into the subacromial space.  Postoperative pain at rest and on movement was significantly lower in group 1 than in group 2 or 3 during the first 30 minutes postoperatively, suggesting that the difference among groups in pain relief was due to the preoperative bolus of anesthesia rather than the postoperative intra-bursal anesthesia.  Two patients in group 1 required supplemental morphine postoperatively, compared to 6 persons in group 2 and 9 persons in group 3.  After the first hour the pain at rest decreased in all three groups, so that from the fourth postoperative hour, the VAS scores were between 1 and 2 cm in all groups.  No significant differences were found between all three groups in the verbal rating score (VRS) during the first 24 hours after the operation.  The investigators also assessed pain relief before and after a patient-controlled infusion.  Pain at rest decreased in all groups in all three groups, with no significant differences between groups.  Pain on movement decreased from an average of 5.9 pre-infusion to 4.7 post-infusion in group 1, 6.1 to 4.8 in group 2, and 6.3 to 6.1 in group 3.  Although the pain relief was statistically significant after anesthetic infusion in groups 1 and 2, the average VAS scores remained just below 5 in groups 1 and 2, indicating that the anesthetic infusion provides inadequate pain relief.  There were no significant differences among the three groups in nausea, vomiting, or pruritus among the groups. 

Rosseland, et al. (2004) reported that pain after knee arthroscopy is modest and short-lived and can successfully be treated with intraarticular saline as placebo in a randomized controlled study (n = 60).  In this study, 60 patients who developed moderate-to-severe pain after knee arthroscopy were randomly assigned to infusion of either 10 ml or 1 ml of intraarticular saline.  The investigators reported that pain intensity remained low and use of rescue medication and other pain outcome measures were similar during the 36-hour outcome period.  The investigators found that patients experienced equally good pain relief after intraarticular injection of saline.  The investigators concluded that this finding of a major placebo effect of intraarticular saline has implications for the interpretation of previously published placebo-controlled intraarticular analgesia studies.

Barber & Herbert (2002) reported on a randomized controlled clinical study of 50 consecutive patients undergoing arthroscopic shoulder surgery who were randomly assigned to either a saline or anesthetic solution via an infusion pump following surgery.  Although subjects assigned to anesthetic had lower pain scores than subject assigned to saline, there was no statistically significant difference between groups in use of postoperative oral medication.  Functional outcome measures were not assessed in this study. 

Harvey, et al. (2004) reported on a randomized, controlled clinical study of 24 patients undergoing arthroscopic subacromial decompression, 19 of whom completed the study.  Subjects were randomly assigned to continuous subacromial infusions of either anesthetic or saline.  Subjects assigned to anesthetic had less pain than subjects assigned to saline.  However, there were no significant differences between groups in the amount of supplemental hydrocodone consumption.  Functional outcome measures were not assessed in this study.    

Savoie, et al. (2000) reported on 62 consecutive patients undergoing arthroscopic subacromial decompression who were randomized to receive continuous intra-lesional infusions of either anesthetic or saline postoperatively.  Subjects assigned to anesthetic infusion reported modest but statistically significant reductions in pain scores postoperative days 1 through 5.  Subjects assigned to anesthetic infusion also had less use of supplemental narcotics.  Functional outcomes were not assessed.

A study by Gottschalk, et al. (2003) examined the effectiveness of continuous intra-lesional anesthetic infusion in 45 patients undergoing shoulder surgery.  Subjects were assigned to three groups: group 1 received a single dose wound infiltration of saline plus continuous postoperative wound infiltration with saline.  Groups 2 and 3 received a single dose wound infiltration with anesthetic, plus continuous postoperative wound infiltration with either lower dose or higher dose anesthetic.  Because of the design of this study, one cannot discern the contributions of single dose wound infiltration and postoperative continuous wound infiltration to outcomes.  Postoperative pain was less in the group receiving higher dose anesthetic than lower dose anesthetic or saline during the 48 hour duration of the study.  Cumulative supplemental analgesic consumption was less in the subjects receiving intra-lesional anesthetic.  Functional outcomes were not assessed.

Klein, et al. (2003) compared the effectiveness of interscalene brachial plexus block followed by continuous intra-articular infusion to interscalene brachial plexus block followed by continuous interscalene infusion in 17 patients who were undergoing outpatient rotator cuff repair.  The investigators reported similarly high VAS scores at rest and with movement and similarly high narcotic consumption between the two groups. The investigators noted that, overall, between 50 percent and 70 percent of all patients reported suboptimal analgesia, and that neither group was consistently able to achieve satisfactory analgesia (VAS less than 2) with supplemental oral narcotics.  The investigators concluded that “[t]he high VAS scores and need for additional medical care suggest that intra-articular administration may not be reasonable for this magnitude of surgery.”

Klein, et al. (2001) examined the effects of intra-articular analgesia with a continuous infusion of local anesthetic in 40 patients undergoing shoulder arthroscopy.  Patients were randomly assigned to postoperative intra-articular infusion of anesthetic or saline.  Subjects assigned to anesthetic had lower postoperative pain scores and less consumption of supplemental narcotics.  Functional measures were not assessed.   

A study by Lau, et al. (2001) of 44 persons undergoing inguinal hernia repair is of weaker design because it is non-blinded with no sham infusion pump treatment given to the control group. The investigators reported significant differences in pain scores in favor of the pump group lasting through the first day following surgery. They also reported none of the 20 subjects assigned to intralesional infusion pumps required analgesics, compared to 6 of 24 subjects in the control group. Because of the unblinded nature of this study, it is uncertain whether these differences may be attributable to placebo effects. There were no differences between groups in post-operative recovery, including time to resume ambulation, time to resume voiding, and return to normal activities. The investigators reported that the main drawbacks to the use of an intralesional pump were its high cost and the frequent seepage of blood-stained anesthetic fluid into the wound dressing, which occurred in a quarter of subjects assigned to intralesional anesthetic pumps.

A study by Cheong, et al. (2001) of 70 persons undergoing laparotomy for major colorectal surgery is also non-blinded.  Patients were randomly assigned to patient-controlled analgesia (PCA) or to intralesional anesthesia.  The investigators reported that there was no statistically significant difference in post-operative pain scores at rest and with movement between the two groups, except the first post-operative day, where the median pain scores in the intralesional anesthesia group were higher than those in the PCA group.  The investigators reported that the median amount of morphine used was significantly greater in the subjects assigned to PCA than in subjects assigned to intralesional anesthesia.  This difference may be attributable to the non-blinded nature of this study and the fact that subjects assigned to PCA could self-administer morphine on demand, whereas subjects assigned to intralesional anesthesia had to request morphine  administered via a subcutaneous injection.   The investigators noted that none of the patients in either group was unduly sedated or confused owing to either form of analgesia during the study.  The investigators reported no significant differences in time to return of bowel movement, time to post-operative mobilization, and time to discharge from hospital.  It should be noted that four patients in the intralesional anesthesia group developed wound infection, compared to one patient in the PCA group.  

A study by Morrison and Jacobs (2003) is also of weaker design in that it is non-blinded, non-randomized retrospective consecutive case series of 49 mastectomy patients treated over a 5-year period, with comparisons before and after introduction of intralesional anesthetic infusion pumps.  Factors other than the use of an intralesional infusion pump (e.g., improvements in surgical techniques, rehabilitation protocols, etc.) may have accounted for differences in use of post-operative pain medication, length of hospital stay, and post-operative stay in post-anesthesia care unit (PACU) before and after they began using intralesional anesthetic pumps at the study institution.

A study by Chew, et al. (2003) is of weaker design than many of the previously described studies in that it is a nonrandomized study that uses historical controls rather than randomly assigned concurrent controls. 

A study by Mallon, et al. (2000) is of weaker design in that it compares intra-articular anesthetic infusion to no infusion, and hence the study is non-blinded and lacks a placebo control group.  Studies by Rawal, et al. (1997), Ganapathy, et al. (2000) and Crawford, et al. (1997) are of weaker design in that they lack a control group.  A study by Yamaguchi, et al. (2002) is a report of a retrospective, uncontrolled case series.  A study by Vintar, et al. (2002) compared intralesional bupivacaine to intralesional ropivacaine in 60 patients who underwent inguinal hernia repair, and hence did not inform whether there are clinically significant benefits to the administration of intralesional anesthesia. 

Several studies after total knee arthroplasty (Klasen et al, 1999; Schwarz et al, 1999; Rautoma et al, 2000; DeWeese et al, 2001) and other surgical procedures (Adams et al, 1991; Forgach and Ong, 1995) have concluded that application of intraarticular or intralesional local anesthetics and/or morphine does not reduce analgesic requirements, and there have been no studies to prove beneficial effects on post-operative recovery and rehabilitation.

Nechleba, et al. (2005) examined the effectiveness of local, continuous infusion of bupivacaine for pain control following total knee arthroplasty. A total of 11 men and 19 women with an average age of 65 years (range of 43 to 83 years) randomly received either 0.25 % bupivacaine or normal saline by local infusion pump. Standard wound drainage also was implemented. Pain was assessed with a VAS along with patient-controlled analgesia demand, narcotic delivery, and NSAID administration. Drug lost to drainage also was assessed. Mean pre-operative VAS were similar between the saline and bupivacaine groups (6.5 +/- 1.4 and 6.1 +/- 2.0, respectively; p = 0.535). By the end of the second post-operative day, scores decreased to 3.4 +/- 3.2 for the saline group and 2.5 +/- 1.6 for the bupivacaine group. Although post-operative reductions were statistically significant (p = 0.007), the main treatment effect was not (p = 0.404). Mean narcotic demand and usage were 87 +/- 114.1 requests with usage of 11.8 +/- 12.3 mg for the saline group and 96 +/- 104.8 requests with usage of 7.5 +/- 3.8 mg for the bupivacaine group (p = 0.505). Cumulative ketorolac administration was 47 +/- 52.2 mg for the saline group and 83.6 +/- 64.9 mg for the bupivacaine group (p = 0.100). Hydrocodone-acetaminophen usage also was similar between the saline and bupivacaine groups (88 +/- 43.9 mg and 64.6 +/- 35 mg, respectively) (p = 0.112). Drug lost to drainage was estimated to be 27 %. These investigators concluded that their findings suggested continuous local analgesic infusion after total knee arthroplasty does not offer significant improvements in either pain relief or medication use. Drug loss from drainage may exceed 25 % and may compromise analgesic effectiveness.

Other recently published studies also demonstrate the inconsistencies in results of intralesional and intraarticular anesthetic pumps (see, e.g., Wu, et al., 2005; Kushner, et al., 2005; Morgan, et al., 2006; Baig, et al., 2006; Parker, et al., 2007).  In a prospective, double-blind, placebo-controlled, randomized study, Wu et al (2005) examined if a subfascial continuous infusion of local anesthetic in patients undergoing radical retropubic prostatectomy would result in a reduction in post-operative opioid requirements and an improvement in pain scores. A small catheter was placed subfascially at the end of the operation and attached to an elastomeric pump, which administered either 0.5 % bupivacaine or normal saline into the wound at a rate of 2 ml/hour until discharge on post-operative day 3. The outcomes assessed included the dosage of hydromorphone used by a patient-controlled analgesic system, a VAS for pain at rest and with activity, a VAS of nausea, and length of hospital stay. A total of 100 patients were successfully randomized, with all patients completing the protocol. No differences were found between the groups with regard to VAS pain at rest, VAS pain with activity, intravenous or oral analgesic consumption, or VAS nausea scores. The authors concluded that continuous subfascial infusion of local anesthetic did not result in a post-operative reduction in opioid requirements or an improvement in pain scores in patients undergoing radical retropubic prostatectomy.

Continuous local anesthetic infusion has also been employed at the iliac crest bone graft (ICBG) site following spinal arthrodesis. Singh et al (2005) examined the effects of post-operative continuous local anesthetic agent infusion at the ICBG harvest site in reducing pain, narcotic demand and usage, and improving early post-operative function after spinal fusion. A total of 37 patients were enrolled in a prospective, randomized, double-blind, parallel-designed study (28 had ICBG harvested for lumbar arthrodesis and 9 for cervical arthrodesis). During spinal arthrodesis surgery, patients were randomly assigned to receive 96 ml (2 ml/hour x 48 hours) of either normal saline (control group, n = 22) or 0.5 % Marcaine (treatment group, n = 15) delivered via a continuous infusion catheter placed at the ICBG harvest site. All patients received Dilaudid patient-controlled analgesia after surgery. Pain scores, narcotic use/frequency, activity level, and length of stay (LOS) were recorded. Physicians, patients, nursing staff, and statisticians were blinded to the treatment. Mean patient age was 60 years and similar between groups. Narcotic dosage, demand frequency, and mean VAS pain score were significantly less in the treatment group at 24 and 48 hours (p < 0.05). The average LOS was 4.1 days with no difference between the treatment group (4.3 days) and the control (group 3.9 days). No complications were attributed to the infusion-catheter system. The authors concluded that continuous infusion of 0.5 % Marcaine at the ICBG harvest site reduced post-operative parenteral narcotic usage by 50 % and decreased overall pain scores. No complications were attributed to the infusion-catheter system. They noted that the use of continuous local anesthetic infusion at the iliac crest may help in alleviating acute graft-related pain, hastening patient recovery and improving short-term satisfaction. This is in agreement with the findings of Cowan et al (2002) who stated that administration of local anesthetic is a safe and effective technique for pain relief at the iliac crest donor site in patients who have undergone cervical fusion (n = 14).

In contrast to the findings by Cowan et al (2002) and Singh et al (2005), Morgan and colleagues (2006) reported that continuous infusion of bupivacaine at ICBG sites during the post-operative period is not an effective pain control measure in hospitalized patients receiving systemic narcotic medication. In a prospective, double-blind, randomized clinical trial, Morgan et al examined if continuous infusion of 0.5 % bupivacaine into the iliac crest harvest site provides pain relief that is superior to the relief provided by systemic narcotic pain medication alone in patients undergoing reconstructive orthopedic trauma procedures. Patients (over 18 years of age) were randomized to the treatment arm (bupivacaine infusion pump) or the placebo arm. Post-operatively, all subjects received morphine sulfate with use of a patient-controlled analgesia pump. Subjects recorded the pain at the donor and recipient sites with use of a scale ranging from 0 to 10. The use of systemic narcotic medication was recorded. Independent-samples t tests were used to assess differences in perceived pain relief between the treatment and control groups at 0, 8, 16, 24, 32, 40, and 48 hours after surgery. Pain was also evaluated at 2 and 6 weeks post-operatively. A total of 60 patients were enrolled. Across all data points, except pain at the recipient site at 24 hours, no significant differences in the perception of pain were found between the bupivacaine group and the placebo group. It is interesting to note that on the average, patients in the treatment group reported more pain than those in the control group. No significant difference was found between the two groups with regard to the amount of narcotic medication used. The authors concluded that no difference in perceived pain was found between the groups. The results of this study indicated that continuous infusion of bupivacaine at ICBG sites during the post-operative period is not an effective pain-control measure in hospitalized patients receiving systemic narcotic medication. This is in agreement with the observation of Puri et al (2000) who stated that in view of the lack of improvement in pain relief and the risk of infection, local administration of bupivacaine at the iliac bone harvest site following cervical diskectomy/foot arthrodesis (n = 13) is not recommended for post-operative analgesia.

Polglase, et al. (2007) reported on a lack of efficacy of a continuous wound infusion of ropivacaine in conjunction with best practice postoperative analgesia after midline laparotomy for abdominal colorectal surgery. The investigators performed a randomized, participant and outcome assessor-blinded, placebo-controlled trial on patients presenting for major abdominal colorectal surgery. Subjects were allocated to receive ropivacaine 0.54 percent or normal saline via a dual catheter Painbuster Soaker continuous infusion device into their midline laparotomy wound for 72 hours postoperatively. A total of 310 patients were included in this study. The investigators found that the continuous wound infusion of ropivacaine after abdominal colorectal surgery conveys minimal benefit compared with saline wound infusion. The investigators found no statistically significant difference for: pain at rest, morphine usage, length of stay, mobility, nausea, or return of bowel function. There was a small, statistically significant difference in mean pain on movement on day 1 for the ropivacaine group (adjusted mean difference -0.6 (range, -1.08 to -0.13)). The investigators reported that, although this trend continued on days 2 and 3, the differences between groups were no longer statistically significant. The investigators concluded that delivery of ropivacaine to midline laparotomy wounds via a Painbuster Soaker device did not demonstrate any significant clinical advantage over current best practice.

Liu, et al. (2006) conducted a systematic evidence review of intralesional and intraarticular anesthetic pumps. The authors stated that they were motivated to conduct a systematic review of continuous wound catheters delivering local anesthetic because “there have been conflicting reports of the overall efficacy, and no single, large RCT has definitively assessed the risk of this modality.” Available randomized controlled clinical studies of continuous wound catheters are small considering the size of the eligible population. A primary problem with this systematic review is that it inappropriately combined studies involving heterogenous patient populations, anesthesia indications, catheter placement, and methods of continuous infusion in its overall and subgroup analyses. The authors noted that future large homogenous randomized controlled trials would be valuable to verify the findings of the systematic review and provide better quantitative data. In addition, the authors stated that they were not able to answer basic questions, including cost-effectiveness, site of catheter placement, or dosage, because of the variability among studies.

It remains unclear whether any analgesia produced by intra-articular and intralesional anesthetics is clinically useful.  Estimates of the impact of intralesional and intraarticular anesthetic pumps on duration of hospitalization were based upon very few studies. Few studies have examined the impact of intraarticular and intralesional anesthetic pumps on functional outcomes (reductions in disability, improvements in function, or faster recovery).  In addition, few studies have directly compared the effectiveness and safety of intra-articular or intralesional infusions with established methods of postoperative analgesia (see, e.g., Tran, et al., 2005).  Finally, available studies are small and not sufficiently powered to evaluate uncommon but clinically significant adverse effects of intralesional and intraarticular catheters (see, e.g., Hoeft, et al., 2006).

In a retrospective study, Bray and colleagues (2007) evaluated the effectiveness of a local anesthetic pain infusion pump in the management of post-operative pain in abdominoplasty patients. A total of 38 abdominoplasty patients with local anesthetic pain pumps and 35 abdominoplasty patients without pain pumps were included in this study.  Pain pumps were loaded with 0.25 % or 0.5 % bupivacaine and infused at a constant rate of 4 ml/hour.  All patients were admitted post-operatively and started on a narcotic patient-controlled analgesia (PCA).  Post-operative PCA narcotic use and pain scores were recorded every 2 hours by the nursing staff.  For the first 24 hours of post-operative hospital stay, pain medication, pain scores, and anti-emetic use were determined from the patients' charts.  Hospital stay was also reviewed.  In the pain pump group, there was a small but statistically non-significant reduction in pain medication use (2.65 versus 3.04 pain units) (p = 0.34).  Interestingly, pain scores were higher in the pain pump group but not significantly (2.73 versus 2.31) (p = 0.17).  There was no statistically significant difference in the use of anti-emetics (0.8 versus 0.6) (p = 0.60).  Hospital length of stay averaged 2.2 days in the pain pump group and 2.5 days in the group without pain pumps (p = 0.09).  The authors concluded that the post-operative use of pain pumps in abdominoplasty patients does not significantly improve pain management.  They stated that further investigation into this application of the pain pump is necessary before recommending their routine use in abdominoplasty patients.

Charous (2008) stated that management of post-operative pain can be critical to the success of a patient's recovery following head and neck surgery.  Various medications and delivery methods have been tried to maximize patients' comfort while minimizing many of the medications' potential side effects.  Continuous wound perfusion pain management systems are being used in various surgical specialties.  In a preliminary report, the author described the use of one such pain management system (On-Q) in thyroid and parotid surgeries.  Statistically significant less levels of pain, use of opioids and nausea/vomiting were noted in patients who used the On-Q system.  There were no complications.  The author concluded that the use of the On-Q system in various head and neck procedures is promising; further research, evaluation, and exploration of its possible uses are encouraged.

In a prospective, randomized, double-blind study, Banerjee and associates (2008) assessed the effectiveness of continuous low-dose bupivacaine infiltration by infusion pump after arthroscopic rotator cuff repair.  A total of 60 patients undergoing arthroscopic rotator cuff repair received a bolus injection in the subacromial space of 35 ml of 0.25 % bupivacaine with 1:200,000 epinephrine at surgical closure and were randomized to one of three groups: (i) 0.25% bupivacaine at 2 ml/hr (n = 20), (ii) 0.25 % bupivacaine at 5 ml/hr (n = 20), or (iii) saline at 5 ml/hr (n = 20) via infusion pump into the subacromial space.  Pain was evaluated using the VAS and narcotic consumption was measured until 48 hours after surgery and converted to dose equivalents (DE).  Sixty patients used the infusion pump for a mean of 43.9 hours (range of 15.50 to 50.75 hours).  Mean total narcotic consumption, expressed in DEs, was 2.24 for the 2-ml group, 3.52 for the 5-ml group, and 2.32 for the placebo group.  Mean pain score was 2.9 for the 2-ml group, 3.6 for the 5-ml group, and 3.3 for the placebo group.  There were no differences in operating room time or infusion pump use time among groups.  The 2-ml group had a non-significant trend toward less pain and lower narcotic consumption.  The 5-ml group evidenced a non-significant trend toward more pain and higher narcotic consumption.  The findings of this study neither supported nor refuted the use of infusion pumps.  The authors hypothesized that the placebo group would experience greater pain than the 5-ml group; however, a non-significant trend toward the contrary occurred.  A trend toward less pain in the 2-ml group was not significant.

In a prospective, randomized, double-blind, controlled trial, Kazmier et al (2008) examined the effectiveness of the pain pump after cosmetic breast augmentation.  A total of 25 women were enrolled in the study; 5 were eliminated from analysis because of data inadequacy or device problems.  After bilateral augmentation, the remaining 20 patients received a 4-day continuous infusion of bupivacaine in one breast pocket and saline in the other.  Laterality of bupivacaine infusion was randomized and blinded to both the patient and the surgeon.  Patients completed a questionnaire on post-operative days 1, 2, 3, 4, and 7, rating their pain on a scale of 0 to 10, with 10 being worst.  On post-operative day 1, the mean pain score was 4.7 on the bupivacaine side versus 5.4 on the saline side (p = 0.36).  On post-operative days 2, 3, 4, and 7, the mean scores were 4.3 versus 4.6 (p = 0.63), 3.3 versus 3.8 (p = 0.50), 3.4 versus 3.6 (p = 0.78), and 3.4 versus 3.1 (p = 0.63) for the bupivacaine and saline sides, respectively.  The authors concluded that the pain pump appears to provide breast augmentation patients marginal improvement in pain control, although this advantage did not reach statistical significance in this study.  The benefit, if real, also appears to wane over the first post-operative week.

Ciccone and co-workers (2008) assessed the effectiveness of interscalene regional blocks and infusion pumps for post-operative pain control after arthroscopic subacromial decompression with or without arthroscopic rotator cuff repair.  A total of 76 patients were included in the prospective study.  Participants were randomized into four treatment groups: (i) interscalene regional block, (ii) infusion pump with 0.5 % bupivacaine, (iii) interscalene block combined with an infusion pump containing 0.5 % bupivacaine, and (iv) interscalene block combined with an infusion pump containing 0.9 % saline solution.  The interscalene regional block was performed with a nerve stimulator.  Infusion pump catheters were positioned in the subacromial space.  Visual analog scale data were collected pre-operatively, at 1 and 2 hours post-operatively, and daily for an additional 6 days post-operatively.  An analysis of variance with a Student-Newman-Keuls post hoc test was used to identify statistically significant (p < 0.05) differences in VAS scores between the groups at each time point.  Percentages of patients who took medication for pain management in the recovery room were compared between the 4 groups by use of chi(2) analysis.  Significant differences were noted in VAS scores post-operatively.  Group (ii) (pump only) had significantly higher scores than all other groups for the first 2 hours.  Furthermore, group (iv) (block and pump filled with saline solution) had significantly lower VAS scores than group (i) (block only) at 1 hour.  This difference was no longer significant by the second hour.  The percentage of patients who required oral narcotics or intravenous pain medication was significantly larger for group (ii) than for the other groups.  The authors concluded that the interscalene regional block provided more pain relief than infusion pumps immediately after arthroscopic shoulder surgery.  Moreover, infusion pumps did not significantly reduce pain levels after the blocks wore off.

An assessment by the Galacian Agency for Health Technology Assessment (AVALIA-T) (Acevedo Prado & Atenzio Merino, 2008) found no clear evidence of improved outcomes with continuous anesthetic infusion pumps versus other methods of managing postoperative pain. The assessment identified 10 clinical trials that met prespecified inclusion criteria. The investigators found that, in general, the results of these clinical trials did not consistently favor continuous anesthetic infusion pumps over standard methods of postoperative pain management. The assessment found that, in some of the clinical trials, there was a slight improvement in pain scores or reductions requirements for narcotic analgesics, but other studies found no such differences. One difficulty in interpreting studies that was noted in the assessment is the lack of common methodology in clinical trials.

In a prospective, randomized study, Järvelä and Järvelä (2008) evaluated the long-term effect of the use of a pain pump after arthroscopic subacromial decompression. A total of 50 patients were included in this study (25 had a 24-hr pain pump with 0.375 % ropivacaine infusion and a continuous rate of 5 ml/hr in the subacromial space after arthroscopic subacromial decompression, and 25 did not). Rehabilitation was similar in both groups. Evaluation methods were clinical examination, radiographical evaluation, and isometric elevation strength measurements, as well as the University of California, Los Angeles and Constant shoulder scores. All the operations were done by 1 experienced orthopedic surgeon, and all the evaluations at follow-up were done by 1 independent, blinded examiner. There were no differences between the study groups pre-operatively. Of the patients, 47 (94 %) were available at a minimum follow-up of 2 years (range of 24 to 32 months). Concerning the duration of sick leave (p = 0.053) and ability to return to work (p = 0.321), the group differences were not statistically significant. At follow-up, the shoulder scores (University of California, Los Angeles and Constant) were significantly better than pre-operatively (p < 0.001) in both groups, although no differences were found between the groups. The isometric elevation strengths of the operated shoulders were equally good in both groups (p = 0.976) as well, and no significant differences were observed between the operated shoulders and non-operated shoulders at follow-up. The authors concluded that the use of a pain pump after arthroscopic subacromial decompression did not have any long-term effects on the patients' recovery, return to work, or final result at the minimum 2-year follow-up.

In a randomized, participant and outcome assessor-blinded, placebo-controlled study, Coghlan et al (2009) examined the safety and effectiveness of ropivacaine infusion following arthroscopic or mini-incision rotator cuff surgery. Subjects, stratified by operative procedure (either arthroscopic decompression or rotator cuff repair), were given preemptive 1 % ropivacaine (20 ml) and intra-operative intravenous parecoxib (40 mg) and were randomly assigned to 0.75 % ropivacaine or placebo by elastomeric pump at 5 ml/hr. Pain at rest was reported on a verbal analog scale at 15, 30, and 60 mins and at 2, 4, 8, 12, 18, and 24 hrs. The use of alternative analgesia, delay in discharge, and adverse events, including development of stiff painful shoulder, infection, and leakage, were also assessed. A total of 84 participants received arthroscopic decompression (43 in the placebo arm and 45 in the ropivacaine arm) and 70 received rotator cuff repair (35 participants in each treatment arm). Compared with placebo, ropivacaine infusion resulted in a significant but clinically unimportant improvement in average pain in the first 12 hrs following both procedures (the average pain score was 1.62 and 2.16 for the ropivacaine and placebo arms, respectively, in the arthroscopic decompression group and 2.12 and 2.82 in the rotator cuff repair group, with a pooled difference between groups of 0.61; 95 % confidence interval [CI], 0.22 to 1.01; p = 0.003). When adjusted for opioid use, the pooled difference between groups was 0.49 (95 % CI, 0.12 to 0.86; p = 0.009). No difference was detected between groups with regard to the maximum pain in the first 12 hrs or the average or maximum pain in the second 12 hrs, with or without adjustment for opioid use, and no difference was found between groups with regard to the amount of oral analgesia used. No difference was detected between groups with regard to the prevalence of nausea and vomiting, catheter leakage, delayed discharge, or stiff painful shoulder, and no subject in either group developed post-operative infection. The authors concluded that there was minimal evidence to support the use of ropivacaine infusion for improving outcomes following rotator cuff surgery in the setting of preemptive ropivacaine and intra-operative parecoxib.

There is emerging evidence of a relationship between intraarticular administration of chondrotoxic anesthetics and postarthroscopic glenohumoral chondrolysis (McNickle, et al., 2009; Busfield, et al., 2009; Saltzman, et al., 2009; Bailie, et al., 2009; Hansen, et al., 2007; Gomoll, et al., 2006). The chondrotoxic effects of anesthetics bupivicane and epinepherine are thought to lead to cartilage damage.