Article Text

Download PDFPDF

Analgesic efficacy of adding the IPACK block to a multimodal analgesia protocol for primary total knee arthroplasty
  1. Jason Ochroch1,
  2. Victor Qi1,
  3. Ignacio Badiola1,
  4. Taras Grosh1,
  5. Lu Cai1,
  6. Veena Graff1,
  7. Charles Nelson2,
  8. Craig Israelite2 and
  9. Nabil M Elkassabany1
  1. 1 Department of Anesthesiology and Critical Care, University of Pennsylvania, Philadelphia, Pennsylvania, USA
  2. 2 Department of Orthopedic Surgery, University of Pennsylvania, Philadelphia, Pennsylvania, USA
  1. Correspondence to Dr Nabil M Elkassabany, Anesthesiology & Critical Care, University of Pennsylvania, Philadelphia, PA 19104, USA; nabil.elkassabany{at}


Background and objectives Peripheral nerve blocks have been integrated into most multimodal analgesia protocols for total knee arthroplasty (TKA). The adductor canal block (ACB) has gained popularity because of its quadriceps muscle sparing. Similarly, local anesthetic injection between the popliteal artery and the posterior capsule of the knee, IPACK block, has been described to provide analgesia to the posterior capsule of the knee with motor-sparing qualities. This prospective randomized controlled trial aimed to assess the analgesic efficacy of adding the IPACK block to our current multimodal analgesic regimen, including the ACB, in patients undergoing primary TKA.

Methods 119 patients were randomized to receive either an IPACK or a sham block in addition to multimodal analgesia and an ACB. We were set to assess pain in the back of the knee 6 hours after surgery. Other end points included quality of recovery after surgery, pain scores, opioid requirements, and functional measures.

Results Patients who received the IPACK block had less pain in the back of the knee 6 hours after surgery when compared with the sham block: 21.7% vs 45.8%, p<0.01. There was marginal improvement in other pain measures in the first 24 hours after surgery. However, opioid requirements, quality of recovery and functional measures were similar between the two groups.

Conclusion The IPACK block reduced the incidence of posterior knee pain 6 hours postoperatively.

  • analgesia
  • analgesics
  • opioid
  • anesthesia
  • local
  • nerve block
  • outcome assessment
  • health care

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Total knee arthroplasty (TKA) is among the most commonly performed orthopedic procedures, with 700 000 knee replacements performed annually in the USA.1 Unfortunately, this surgery is known to cause significant postoperative pain. Multiple multimodal analgesic regimens have been used to alleviate pain and improve overall recovery, of which regional anesthetics have become an integral part.2 However, due to the complex innervation of the knee, nature of the surgical procedure, and rehabilitative needs, the optimal analgesic block for TKA is still elusive.3 The adductor canal block (ACB) popularity is due to its quadriceps muscle sparing, which may lead to better performance during physical therapy and possibly better functional outcomes.2 3

Likewise, recent techniques have sought to block innervation to the posterior knee via the sciatic nerve while still sparing distal motor function. Selective tibial nerve block has been proposed as an alternative. However, potential spread to the peroneal nerve and postoperative foot drop were major drawbacks.4 More recently, a novel ultrasound-guided nerve block targeting the interspace between the popliteal artery and the capsule of the posterior knee (IPACK) has been described to provide analgesia to the posterior knee with these motor-sparing qualities.5 The IPACK block is purported to anesthetize the popliteal plexus which is formed by the interdigitations of the articular branches of the tibial nerve and posterior division of the obturator nerve that entwine around the popliteal artery and vein.6 This prospective randomized controlled trial aimed to assess the analgesic efficacy of adding the IPACK block to our current multimodal analgesic regimen (including ACB) in patients undergoing primary TKA. We hypothesized that adding this block to our current multimodal regimen would improve the incidence of posterior knee pain and overall analgesia.


Patients with American Society of Anesthesiologists (ASA) status I-III undergoing primary TKA at Penn Presbyterian Medical Center (Philadelphia, PA) from November 2018 to July 2019, age 18–80 years were eligible for enrollment into this prospective randomized trial. All enrolled subjects provided written informed consent. Patients were excluded from the study if they had an allergy to any of the study medications, body mass index (BMI) >45, coagulopathy, chronic kidney disease or recent chronic opioid therapy, defined as the use of regular daily doses of systemic opioids for the past 3 months prior to the surgery. Revision knee replacement procedures were also excluded.


A computer-generated randomization table was used for patient allocation to one of the two study groups: the ACB+IPACK group, or the ACB+sham block group. Randomization was done in blocks of 10 patients each. Patients’ assignments were written in a sealed envelope that was opened only after the patient consented for the study. The study was blinded such that patients, nurses on the floor, the research coordinator, and the physical therapists were not aware of the nature of the assignment. Access to analgesic medications was not restricted and prescription of analgesic medication was done according to our institutional protocol.

Study procedure

All patients received a multimodal perioperative pain protocol (MP3) that included the placement of a preoperative ultrasound-guided continuous ACB. The MP3 protocol is described in table 1.

Table 1

Multimodal perioperative pain protocol for TKA

Once consented, the patient’s group assignment was revealed to the anesthesiologist performing the block. Patients monitoring and sedation were according to standard operating procedure. All blocks were done using ultrasound (US) guidance. Sonosite (Bothell, Washington) Edge II machines were used with a high-frequency linear US probe (6–13 MHz).

The anesthesiologist scanned the medial part of the thigh, halfway between the anterior superior iliac spine and the patella. In a short-axis view, the superficial femoral artery was identified underneath the sartorius muscle, with the vein just inferior and the saphenous nerve just lateral to the artery. An 18-gauge Tuohy (B-Braun, Melsungen, Germany) needle was introduced using an in-plane approach, and 2 to 3 mL of local anesthetic (LA) bolus was used to verify correct placement of the needle in the vicinity of the saphenous nerve in the adductor canal, deep to the sartorius muscle and deep to the vaso-adductor membrane lateral to the neurovascular sheath. A 20-gauge catheter was introduced and advanced 2–3 cm beyond the needle tip under US visualization. The needle was withdrawn over the catheter and a bolus of 5 mL of ropivacaine 0.5% was injected through the catheter while observing the spread of LA under US to verify correct catheter placement. The remaining volume (12–13 mL) of the LA was injected through the catheter. The catheter hub was affixed to the upper lateral thigh.

Block success was defined as a change in cutaneous sensation to temperature with an alcohol pad in the saphenous nerve distribution over the medial lower leg within 30 min after injection. Subjects with successful catheter placement were retained in the study. Subjects with a failed catheter insertion or misplaced catheter indicated by a lack of sensory changes had their catheter replaced or were withdrawn from the study. Infusion rates were programmed to deliver ropivacaine 0.2% at a basal rate of 8 mL/hour with a patient-controlled demand bolus of 5 mL every 30 min. The catheter was removed on the second postoperative day (POD 2).

Patients who were randomized into the IPACK group underwent an additional block using a low-frequency curvilinear US probe (2–5 MHz). With the patient in a supine position and leg slightly externally rotated, the transducer was placed at the medial knee joint to identify the femoral condyle (figure 1). The transducer was scanned proximally to identify the popliteal artery as it courses posterior to the superior aspect of the condyle. At this point, the needle was inserted in plane at the anterior aspect of the transducer and advanced in a medial to lateral direction parallel to the posterior border of the femoral condyle until the needle tip located was between the bone and popliteal artery. Twenty milliliters of ropivacaine 0.5% was injected (figure 1). Sham blocks involved superficial injection of LA to create a skin weal of the medial side of the knee. None of the study patients received periarticular infiltration (PAI).

Figure 1

(A) The IPACK block is performed with the patient supine with an externally rotated leg, with a low-frequency curvilinear probe placed along the medial. A needle is inserted in-plane at the anterior aspect of the transducer (*). (B) Ultrasound image showing a needle’s (dotted line) approach between the popliteal artery (#) and the posterior aspect of the femoral condyle (F).

Choice of anesthesia (spinal vs general anesthesia) was left to the discretion of the attending anesthesiologist. All patients received prophylaxis for postoperative nausea and vomiting during surgery, including 4 mg of dexamethasone after induction of anesthesia and 4 mg of ondansetron 20 min before recovery from anesthesia. Dexamethasone was withheld in patients with blood glucose above 250 mg/dL.


The primary outcome was the presence of posterior knee pain 6 hours after surgery. Patients were presented with a chart to display the five possible pain locations (anterior, medial, lateral, posterior, and thigh) (figure 2). Secondary outcomes included Post Anesthesia Care Unit (PACU) length of stay, pain scores and location at 6-hour intervals, opioid consumption, mobility and functionality, quality of pain management and overall quality of recovery up to 1 week after surgery. Patients were surveyed using the 15-item Patient-Related Quality of Recovery Questionnaire (QoR-15) at 24 hours, 48 hours, and 1 week after surgery.7 The quality of pain management was assessed using the revised American Pain Society Patient Outcome Questionnaire (APS-POQ-R).8 9

Figure 2

Knee pain diagram used by patients to indicate where their knee pain was located.

Additional outcomes included various markers of recovery during physical therapy, including ambulation distance on the first attempt, cumulative ambulation distance at 24 and 48 hours after surgery, and Timed Up and Go (TUG) on POD 1 and POD 2. TUG is the time in seconds, and it takes a patient to stand up from a chair, walk to a line 3 feet away, turn around, and get back to a sitting position in the chair. The use of an assisting device was allowed. Additionally, any incidence of foot drop was also recorded.

Sample size calculation

Based on the pilot data collected from our institution, there was a 60% incidence of patients complaining of pain at the posterior knee 6 hours after their TKA. Therefore, we sought that a clinically meaningful reduction in the incidence of posterior pain in the knee would be a 50% reduction. Assuming a power of 0.9 and an alpha of 0.05, we needed to enroll 46 patients per intervention. To account for drop out, we increased our sample size to 60 patients per intervention.

Statistical analysis

Statistical analyses were performed using STATA 15 statistical software (College Station, Texas). Age, BMI, surgery and PACU time were analyzed using Student’s t-test. Gender, race, anesthesia type, and American Society of Anesthesiologists status were analyzed using χ2 test. Pain scores were compared at each time interval using the Student’s t-test. We used Mann-Whitney U test to compare opioid requirements, and other paired comparisons at each time interval. Normally distributed data are presented as mean±SEM; non-normally distributed data are presented as median±quartiles (IQR); and categorical data are presented as raw data and as frequencies. The alpha level for our primary outcome was set as p<0.05. The area under the curve (AUC) is calculated by plotting pain scores against time, slicing the AUC into thin trapezoidal slices, calculating the area of each slice, and adding them all up.


Four hundred and sixty-four patients were screened for eligibility, with 120 patients ultimately enrolling in the study (figure 3). Of the 120 patients enrolled, 119 patients completed all necessary follow-up for inclusion into the study: 60 patients in the IPACK group and 59 patients in the sham group. There were no block failures.

Figure 3

Consort flow diagram for patient participation in the study. TKA, total knee arthroplasty.

There was no significant difference in patient demographics or surgical characteristics between the two groups (table 2).

Table 2

Patient demographics and surgical characteristics

For the primary outcome, patients randomized to the IPACK group had a significant reduction in the presence of posterior knee pain 6 hours after surgery when compared with the sham block group, 21.7% vs 45.8%, p<0.01. Pain in the back of the knee was equally present in patients in both groups in the remaining time intervals, except at 18 hours (23.3% vs 42.4%) (table 3).

Table 3

Presence of posterior knee pain

For our secondary outcomes, mean pain scores in the IPACK group were lower than the sham group at 6, 12, and 48 hours after surgery: 4.6 (SD 2.6; 95% CI 3.9 to 5.2) vs 5.8 (SD 2.9; 95% CI 5.0 to 6.5), p=0.01; 4.3 (SD 2.7; 95% CI 3.5 to 5) vs 5.3 (SD 2.8; 95% CI 4.6 to 6), p=0.03; and 4.2 (SD 2.9; 95% CI 3.4 to 5) vs 5.2 (SD 2.5; 95% CI 4.6 to 6), p=0.03, respectively (figure 4). Mean difference between pain scores at 6, 12, and 48 hours were −1.2,–1.1, and −1.1, respectively. Mean (SD) of the AUC for pain scores in the first 24 hours was lower in the IPACK group relative to the sham group: 95 (43) vs 116 (49), p=0.01. The AUC for pain scores between 0 and 48 hours were similar between the two groups: 218 (74.5) vs 251 (88), p=0.16. There was no difference in total opioid consumption between groups (table 4).

Table 4

Opioid consumption presented as oral morphine equivalent doses over the first 48 hours after TKA

Figure 4

Mean pain scores (95% CI) up to 48 hours after surgery with and without an IPACK block. ACB, adductor canal block; IPACK, interspace between the popliteal artery and the capsule of the posterior knee.

Additionally, there were no differences in QoR-15 scores at POD 1, 2, or 7 between groups (table 5).

Table 5

QoR-15 scores after TKA

Patients in the IPACK group did better in the pain intensity domain on the APS-POQ-R scores at POD 1, with significant reductions in average and worst pains, as well as the degree of pain interfering with activities performed out of bed (table 6).

Table 6

APS-POQ-R questionnaire results 24 hours after TKA

There were also no changes in ambulation distances or TUG scores (table 7). There were no major complications noted in this study, including foot drop.

Table 7

Physical therapy parameters after TKA


Our study found that the IPACK block reduced posterior knee pain 6 hours after surgery. Furthermore, the addition of this block also provided overall improved pain scores and pain intensity over the first 24 hours after surgery. Interestingly, this did not translate to reduced opioid consumption, particularly in the first 12 hours after surgery. The short-term benefits of the IPACK may correspond to the duration of action of the LA administered. The reduction in mean differences in pain scores were marginally clinically meaningful at 6, 12, and 48 hours after surgery.10 It is worth noting that this difference did not reach that threshold according to other studies.7 We examined the AUC for pain scores as a more comprehensive measure of pain score over time rather than the snapshot approach of measuring pain scores at specific time points. Additionally, there was some improvement in the pain intensity and pain interference with activity domains on the APS-POQ-R questionnaire.

It is possible that the no reduction in opioid consumption is due to the way opioids are prescribed and administered on the floor. In our institution, opioids are prescribed based on whether patients report mild, moderate, or severe pain. Escalating doses of opioids are prescribed accordingly. Difference in pain scores may not have translated into different dosing of opioids. The addition of the IPACK block in our study also did not appear to affect rehabilitation with physical therapy, or the quality of recovery after surgery.

Several recent studies have looked at IPACK blocks in TKA with similar results to our study.5 11–15 A recent prospective trial also evaluated the effect of adding an IPACK block to ACB, and found no statistically significant difference in median opioid consumption for 24 hours after surgery, but secondary outcomes suggested reductions in average and worst pain scores in the PACU.11 Other studies also found marginal benefits in pain scores, opioid requirements, and physical therapy outcomes, but they were either retrospective, non-randomized, or involved concomitant PAI.5 12–15

Our study, and others, have shown that the IPACK block is a promising new technique that is an improvement from the originally described selective tibial nerve blockade, as it more consistently avoids accidental blockade of the common peroneal nerve.4 Furthermore, the IPACK block has the potential to be an effective alternative to PAI by the surgeon especially in the area of the posterior capsule of the knee, although more studies are needed.13 PAI has not been found to be as effective as sciatic nerve blockade when added to FNB, and nor has the addition of liposomal bupivacaine in PAI provided the long-term benefits it seeks to provide.16 17

Altogether, this study found that the IPACK block improved pain scores and the incidence of posterior knee pain in the TKA population in the first 24 hours. However, the clinical significance of these benefits is unclear, as there were no improvements in our measured functional outcomes or the quality of recovery. Additionally, the reduction in pain scores did not translate into decreased opioid consumption after surgery.

This study has several limitations. First, many of the benefits of the IPACK block could have been limited by the duration of our block which did not include adjuvants. The benefits of an IPACK block could perhaps be better elucidated with a prolonged block, either through the addition of adjuvants or through the placement of continuous catheters. However, a continuous catheter may pose anatomic challenges or disruption of the sterile surgical field and will require further study. Second, we relied on patients to self-report the location of their pain. This assessment is subjective and sometimes patients reported that their pain is ‘everywhere’, and required an in-depth discussion to tease out specific locations of their pain (figure 2). Third, exclusion of patients with chronic pain syndromes may limit the generalizability of the study results to all TKA patients.


IPACK block reduced the incidence of posterior knee pain 6 hours postoperatively. Given the relative ease and safety profile, it may have a potential role as part of the multimodal analgesia after knee arthroplasty, particularly as a distinct alternative to sciatic nerve blockade that does not affect motor function. The IPACK block can also be considered as a more consistent and reproducible alternative to surgical PAI of the posterior capsule of the knee, but more studies are needed.

Supplemental material

Supplemental material


The authors would like to thank our research coordinators team (Rupa Chowdary, BS, Ashish Jain, BS, Anmol Madaan, BS, and Aliaksei Basatski, BS) for their work on data collection and cleaning the data.



  • Twitter @dr_tgro, @veenagraffmd, @nelkassabany

  • Correction notice This article has been corrected since it published Online First. Dr Lu Cai's name has been corrected.

  • Contributors All authors have seen, edited, and approved the manuscript before submission. All authors contributed to the study design and conduct, data analysis, and writing and editing the manuscript.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests Dr Elkassabany is a consultant for Foundry therapeutics. Dr Nelson declares working as a consultant for Acutive Medical, Allentwon, Pennsylvania and for Zimmer Biomet.

  • Patient consent for publication Not required.

  • Ethics approval After obtaining iInstitutional rReview bBoard approval at the University of Pennsylvania ( ID: NCT03703206: first posted 10/11/2018).

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data availability statement Data are available upon reasonable request. Raw data can be made available upon reasonable request.