Article Text

Download PDFPDF

Ultrasound-guided transmuscular quadratus lumborum block for elective cesarean section significantly reduces postoperative opioid consumption and prolongs time to first opioid request: a double-blind randomized trial
  1. Christian K Hansen1,
  2. Mette Dam1,
  3. Gudny E Steingrimsdottir1,
  4. Gunnar Hellmund Laier2,
  5. Morten Lebech3,
  6. Troels Dirch Poulsen1,
  7. Vincent W S Chan4,
  8. Morné Wolmarans5,
  9. Thomas Fichtner Bendtsen6 and
  10. Jens Børglum1
  1. 1 Department of Anesthesiology, Zealand University Hospital, Roskilde, Denmark
  2. 2 Region Zealand, Sorø, Denmark
  3. 3 Department of Gynecology and Obstetrics, Zealand University Hospital, Roskilde, Denmark
  4. 4 Anesthesia, Toronto Western Hospital, Toronto, Ontario, Canada
  5. 5 Anaesthesiology, Norfolk and Norwich University Hospital NHS Trust, Norwich, UK
  6. 6 Anesthesiology, Aarhus Universitetshospital, Aarhus, Denmark
  1. Correspondence to Dr Jens Børglum, Department of Anesthesiology, Zealand University Hospital, University of Copenhagen, DK-4000 Roskilde, Denmark; jens.borglum{at}


Background Elective cesarean section (ECS) can cause moderate to severe pain that often requires opioid administration. To enhance maternal recovery, and promote mother and baby interaction, it is important to reduce postoperative pain and opioid consumption. Various regional anesthesia techniques have been implemented to improve postoperative pain management following ECS. This study aimed to investigate the efficacy of bilateral ultrasound-guided transmuscular quadratus lumborum (TQL) block on reducing postoperative opioid consumption following ECS.

Methods A randomized double-blind trial with concealed allocation was conducted in 72 parturients who received bilateral TQL block with either 30 mL ropivacaine 0.375% or saline. TQL block injectate was deposited in the interfascial plane between the quadratus lumborum and psoas major muscles, posterior to the transversalis fascia. Primary outcome was opioid consumption, which was recorded electronically. Pain scores and time to first opioid request were also evaluated.

Results Opioid consumption (oral morphine equivalents, OME) was significantly reduced in group ropivacaine (GRO) in the first 24 hours compared with group saline (65 mg OME vs 94 mg OME) with a mean difference of 29 mg OME; 95% CI 3 to 55, p<0.03. Time to first opioid request was significantly prolonged in GRO, p<0.003. Numerical rating scale pain scores were significantly lower in GRO in the first 6 hours after surgery, p<0.03.

Conclusions Bilateral TQL block significantly reduced 24 hours’ opioid consumption. Further, we observed significant prolongation in time to first opioid, and significant reduction of pain during the first 6 postoperative hours.

  • truncal blocks
  • postoperative pain
  • obstetrics

Statistics from


A new mother should ideally be able to bond and care for her newborn baby with minimal pain and negligible opioid side effects. Thus, reduction of pain and opioid consumption following elective cesarean section (ECS) is very important.1 2 Ultrasound provides visualization of the quadratus lumborum (QL) and psoas major (PM) muscles in the lumbar paravertebral area (figures 1 and 2)3 4 and the transmuscular QL (TQL) block technique has been previously described.5 6 Our cadaveric study found spread of dye into the thoracic paravertebral space (TPVS) as the mode of action following a TQL block injection.7 Furthermore, we found dye spread caudal to the diaphragm around the T12-L1 branches. Preliminary case study data suggest that TQL block reduces pain after ECS.8 While other QL block techniques have also been described to alleviate pain following ECS, their modes of action are clearly different from the TQL block.9–13 The aim of our current study was to evaluate the analgesic effect of ultrasound-guided (USG) TQL block in reducing opioid consumption in the first 24 hours following ECS.

Figure 1

3-D anatomy figure relevant for transmuscular quadratus lumborum (TQL) block. The psoas major (PM) and quadratus lumborum (QL) muscles constitute the posterior abdominal wall. The transversalis fascia (TF) (green color) lines the deep surface of the rectus abdominis, transversus abdominis, and QL and PM muscles (left). The TF is cranially continuous with the endothoracic fascia (not shown). The diaphragm (DI, red color) covers the anterior and most cranial parts of the PM and QL muscles; that is, medial and lateral arcuate ligaments, respectively. The PM and QL muscles attach inside the thoracic cage to the vertebral bodies of T12-L1 and the 12th rib, respectively. With the TQL block (right), the block needle (blue color needle) is inserted in an oblique posterior to anterior direction through the QL muscle to penetrate the anterior epimysium of the QL muscle. The local anesthetic (LA—pink color above the DI and purple color below the DI) is injected in the interfascial plane between the PM and QL muscles, posterior to the TF. The LA spreads cranially between the QL and PM muscles posterior to the DI into the thoracic paravertebral space (Modified excerpt from VH Dissector with permission from Touch of Life Technologies, Built on real anatomy from the National Library of Medicine Visible Human Project).

Figure 2

The Shamrock sign. Sonographic model image of transmuscular quadratus lumborum (TQL) block. Anatomy relevant for the execution of the TQL block. The needle (white arrow) is inserted in plane to the curvilinear probe and advanced through the quadratus lumborum (QL) muscle to reach the interfascial plane between the QL and psoas major (PM) muscles. Injectate spreads in the interfascial plane and depresses the PM muscle. Red arrow denotes lumbar plexus; white small arrows denote the aponeurosis of the transversus abdominis muscle; white triangles refer to Gerota’s fascia. EO, external oblique muscle; IO, internal oblique muscle; IVC, inferior vena cava; L4, vertebral body of L4; TP, transverse process of L4.

The primary hypothesis was that bilateral TQL block with ropivacaine would result in a minimum relevant difference of 50% reduction in opioid consumption during the first 24 postoperative hours compared with bilateral TQL block with saline.


Ethical approval was obtained from the regional ethics committee of Region Zealand, Denmark (2017/SJ-594). The study was approved by the Danish Medicines Agency (EudraCT 2016-004594-41) and the Danish Data Protection Agency (REG-004-2017). The study was monitored by the regional Good Clinical Practice Unit and registered at (NCT03068260). The study was conducted according to the Consolidated Standards of Reporting Trials statement.14

This single-center, randomized, placebo-controlled and double-blind study was conducted at the Department of Anesthesiology and Intensive Care Medicine at Zealand University Hospital, University of Copenhagen, Denmark, from April 2017 to December 2017, and funded by the department. Written informed consent was obtained from all participants before enrolling for the study.

Study design

Inclusion criteria were American Society of Anesthesiologists class II, age ≥18 years and participants scheduled for ECS performed under spinal anesthesia. Exclusion criteria were inability to cooperate, lack of Danish language skills, allergy to local anesthetics (LA) or opioids, daily regular intake of opioids, local infection at the site of injection or systemic infection and difficult anatomy resulting in poor ultrasound visualization of muscular and fascial structures necessary for correct block administration.

All participants received spinal anesthesia with 2 mL hyperbaric bupivacaine 5 mg/mL with the addition of 0.5 mL sufentanil 5 µg/mL except one due to technical failure (figure 3). Sufentanil is the standard opioid used for spinal anesthesia at our institution. Intrathecal morphine is not applied since an internal review showed less pruritus, nausea and vomiting with addition of sufentanil and not morphine to the spinal anesthesia. All ECS procedures were performed with the Joel-Cohen incision as per routine practice. The Joel-Cohen incision is placed 3 cm caudad to the anterior intercristal line; that is, slightly more cephalad than the Pfannenstiel incision. When the Pfannenstiel incision is used, the subcutaneous layer, fascia and the parietal peritoneum are usually dissected sharply, whereas with the Joel-Cohen incision, the subcutaneous layer is only incised medially, followed by a manual separation of the lateral tissue. The fascia and peritoneum are dissected bluntly.

Figure 3

Consolidated Standards of Reporting Trials (CONSORT) flow diagram. TQL, transmuscular quadratus lumborum.

Following completion of the surgical procedure, participants were transferred to the postanesthesia care unit (PACU) and the TQL blocks were administered. During TQL block administration and the time spent in the PACU, patients were monitored with three-lead ECG, pulse oximetry and non-invasive blood pressure.


Participants were randomly assigned to either active or placebo bilateral TQL block. Randomization list was computer generated in nine blocks of eight assuring equal distribution in the two groups throughout the study period. A nurse anesthetist, not involved in the anesthesia of study participants, would open the sealed opaque envelope containing group allocation and (according to written instructions) prepare two 30 mL syringes of either 30 mL of ropivacaine 0.375% or saline 0.9%. Thereafter, another nurse anesthetist controlled the procedure and both signed the allocation paper before resealing the envelopes. Envelopes would only be opened by the good clinical practice (GCP) monitor and by the investigators at the end of the study. Thus, investigators, participants, care providers and those assessing outcomes were all blinded to group allocation.

TQL block procedure

Investigators (CKH, MD, JB) were all experienced in USG TQL block (>200 TQL blocks each). Blocks were performed immediately following surgery, when the patient had been transferred to the PACU.

The TQL block was performed with the patient in the lateral position using an ultrasound unit (X-Porte, FujiFilm, SonoSite, Bothell, WA) and a curvilinear transducer (2–5 MHz, C60xp). The skin was prepared with application of 2% chlorhexidine/70% isopropyl alcohol. The transducer was covered with a sterile plastic sheath (Safersonic Medizinprodukte Handels, Ybbs, Austria) and then positioned transverse above the iliac crest at the posterior axillary line where the Shamrock sign was identified (figure 2).3 4 A 21 G, 100 mm needle (Polymedic Ultrasound Evolution needle with 30° bevel; Temena SAS, Carrières-sur-Seine, France) was advanced obliquely through the QL muscle to penetrate the QL muscle epimysium to reach the interfascial plane between the QL and PM muscles, posterior to the transversalis fascia (TF) (figure 2).5–7 Once the needle tip had reached this sonographic endpoint and was confirmed with a small amount of injected saline, then 30 mL of 0.375% ropivacaine was injected on each side in group ropivacaine (GRO) patients (225 mg ropivacaine, which is within the accepted maximum dose limit for single shot blocks)15 and 30 mL of saline was injected on each side in group saline (GSA) patients. Correct distribution of injectate was confirmed by turning the probe 90° to a longitudinal orientation, thus observing injectate spread longitudinally between the QL and PM muscles from the iliac crest caudally to the 12th rib cranially and beyond.8

On average, the bilateral TQL block procedures take 7 min to perform (personal data).

Furthermore, on arrival at the PACU, all patients were connected to an intravenous patient-controlled analgesia (PCA) pump (Rhythmic Evolution pump; Micrel Medical, Athens, Greece) containing morphine 1 mg/mL with the following settings: 5 mg per dosage, lockout time of 20 min and a maximum 4-hour dosage of 40 mg. Patients were instructed to administer a bolus whenever numerical rating scale (NRS; 0–10/10) pain scores exceeded 3/10.

The standard postoperative analgesic regime consisted of 1 g paracetamol orally four times a day, 400 mg ibuprofen orally three times a day and if needed, or chosen instead of the PCA pump, supplemental opioids. This was most often oral morphine or oral oxycodone if adverse events from intravenous morphine were intolerable. The total postoperative opioid consumption was recorded and the total consumed intravenous morphine dose and the supplemental opioids were converted ( to oral morphine equivalents (OME).

Outcome measures

Primary outcome was total OME consumption during the first 24 postoperative hours. T0 was time zero at the end of nerve block administration. Secondary outcomes were NRS pain scores (0–10/10), OME consumption at 6-hour intervals until 48 postoperative hours, time to first opioid, time to first mobilization and morphine-related adverse events such as postoperative nausea and vomiting (PONV).

Patients had three different methods of measuring pain. The pain was NRS scored at predefined time settings throughout the 48 hours’ study period at T0, 15, 30, 45 min and 1, 2, 4, 6, 12, 24, 36 and 48 hours after T0. At each time point, the pain was scored at rest and at mobilization (from supine to sitting position). Finally, when patients activated the PCA device, they were instructed to register the NRS pain score directly on the display of the PCA pump, at the time of request for morphine. Thus, we registered the highest NRS pain score on the PCA pump as maximum NRS score.

Statistical analysis and sample size calculation

Sample size estimation was based on a thorough 1-year retrospective survey of >200 women who had undergone ECS at our hospital (from 2015 to 2016). This survey revealed a postoperative opioid consumption mean (SD) of 30 (21) mg OME during the first 24 postoperative hours. With power 80% and level of significance 5% (two sided), the calculated sample size including a 15% dropout rate was 36 participants in each group in order to detect a difference of 50% reduction in OME. Ordinal data and continuous data were analyzed using Student’s t-test and Wilcoxon tests as appropriate. For categorical data, we applied the χ2 test or Fisher’s exact test. Variables were presented as mean (SD), median (IQR) count (%), and range. Measurements were split into 6-hour time intervals from which we calculated the mean value for all scales, transforming ordinal measures to continuous measures. We applied log-rank tests in comparisons of Kaplan-Meier plots for duration of time until first opioid request. Level of statistical significance was 5%. All reported p values are two sided.


One hundred and forty parturients scheduled for ECS were screened for eligibility and 72 participants were enrolled (figure 3). Twenty-eight declined to participate, and 40 lacked sufficient Danish language skills. None were excluded due to poor sonographic visualization.

Two participants were excluded prior to intervention but after the sealed envelopes had been broken. One of these participants could not receive spinal anesthesia (ECS was performed under epidural anesthesia) and the other participant had massive intraoperative hemorrhage and was subsequently withdrawn from the study after advice from the surgeon, because of extended surgery. Another two enrolled participants were excluded shortly after the block when they had to return to theater for surgical complications under general anesthesia. In total, 68 patients completed the study (34 in each group) according to the set protocol.

Parturients were included from April 2017 to December 2017.

There were no statistically significant differences in patient characteristics between the two groups regarding age, weight or body mass index (table 1).

Table 1

Patient characteristics, primary and secondary outcomes

The mean (SD) 24 hours’ postoperative OME consumption was 65.3 (48) mg in GRO vs 94.3 (60) mg in GSA, with mean difference of 29 mg (95% CI 3 to 55, p=0.03). OME data are also presented in table 1.

NRS pain scores (median (IQR)) were significantly lower in GRO for all NRS scores at 0–6 hours; that is: (1) at rest 0.77 (0.33–2.00) vs 2.09 (0.83–2.92), p<0.01, (2) at activity 1.33 (0.83–2.00) vs 2.50 (1.33–3.58), p<0.003 and (3) maximum NRS pain scores registered with PCA pump 3.50 (0–4.00) vs 5.00 (4.00–6.00), p<0.002 (table 1). Even after adjusting p values with Bonferroni correction because of repeated measurements, the difference in pain scores between groups in the first 6 hours remained statistically significant. With all subsequent 6-hour intervals, we did not detect any statistical significant difference between groups regarding NRS pain scores (table 1).

Time to first opioid request was significantly shorter in GSA with a mean (SD) of 4.0 (2.5) hours vs 5.6 (3.3) in GRO. T-test showed a p value of <0.03 (table 1) and the Kaplan-Meier estimate with the log-rank test revealed a p<0.003 (figure 4). All 34 GSA patients required opioids, whereas 3/34 GRO patients (8.8%) had no requests for opioid.

Figure 4

Kaplan-Meier estimate of time to first opioid. Participant’s ‘survival’ ends with first administration of opioid.

There was no significant difference in incidence of PONV, time to first ambulation and time to first sitting up in bed.


Bilateral TQL block significantly reduced opioid consumption from 94.3 mg in GSA to 65.3 mg in GRO (31%) in patients undergoing ECS with a Joel-Cohen incision during the first 24 postoperative hours, prolonged time to first opioid request and reduced pain during the first 6 postoperative hours. Thus, the analgesic effect was more moderate than our expectations of 50% reduction. However, these analgesic benefits did not reduce PONV, nor time to first mobilization.

Importantly, none of the 72 participants experienced lower limb paralysis after TQL block and all study subjects were able to mobilize immediately after the spinal anesthesia effects wore off. This indicates proper LA placement without spread of injectate to the lumbar plexus within the PM. No bladder paralysis, hemodynamic instability or LA systemic toxicity occurred.

With the TQL block technique, LA injected in the lumbar paravertebral area (L3-L4 level) in the interfascial plane between the QL and PM muscles, posterior to the TF,5–8 spreads cephalad into the TPVS through the diaphragmatic openings (lateral and medial arcuate ligaments) to eventually reach the thoracic segmental nerves and the thoracic sympathetic trunk to exert its anesthetic effect. Caudal to the diaphragm, LA anesthetizes the ilioinguinal, iliohypogastric and subcostal nerves,7 but should never, if injected correctly, reach the femoral nerve, obturator nerve, nor the lumbar sympathetic trunk. Hence, we would not expect lower limb or bladder paralysis nor any detectable hemodynamic effects.

The QL1 block or lateral QL block anesthetizes the ilioinguinal, iliohypogastric and subcostal nerves in the pararenal fat compartment.16 17 Krohg and colleagues concluded in a study of 40 parturients undergoing ECS with Pfannenstiel incision and in the absence of neuraxial morphine that this block significantly reduces postoperative ketobemidone consumption and pain intensity.11

The QL2 block or posterior QL block may exert its effect by LA spread to a network of sympathetic nerves in the thoracolumbar fascia,10 17 or as recently reported this block may also involve a spread of injectate to the lower parts of the TPVS.18

Another transmuscular technique is the anterior subcostal QL block.18 The spread of LA into the TPVS from this technique is very similar to what we have previously described with the TQL block; that is, both these techniques build on our initial description of the transmuscular approach and injection anterior to the QL muscle and posterior to the TF.6 7

Some may consider the TQL block a cumbersome technique; that is, probably because our technique—like the epidural and the thoracic paravertebral block—requires the patient to be in the lateral or sitting position. In addition, with the TQL block, it is necessary with a curvilinear transducer and the point of injection is rather deep. Our technique requires some training, and at our institution the residents are supervised for the first 20–30 block applications. We acknowledge that the anatomical complexity in the lumbar paravertebral region can be challenging. In particular, it can be difficult to distinguish between the QL muscle and the pararenal fat pad, due to their comparable ultrasonographic resemblance.19

Additionally, with the TQL block it is essential to inject precisely in the interfascial plane between the QL and PM muscles (figures 1 and 2), and not intramuscularly. Carline and coworkers clearly describe in their summary of block characteristics that they ‘pierced the PMM’ in all four attempts during TQL block injection, but failed to deposit the injected dye in the aforementioned interfascial plane between the QL and PM muscles.20 They also reported staining of the L1–L3 nerve roots and the femoral and obturator nerves in a single case. This is in direct contrast with the findings of Dam and colleagues.7 Thus, with the TQL block, we strongly recommend against intramuscular injection into the PM muscle and the QL muscle.20 21 In our current study, we have demonstrated a significant efficacy in improving postoperative pain management with bilateral TQL block without any motor impairment.

Future studies concerning a direct comparison with other regional anesthetic techniques are warranted.

There are some limitations to this study. First, we did not clinically evaluate block success (dermatomes and myotomes affected) due to the risk of unblinding the patient, staff or investigators to block allocation. It would also have been clinically impractical due to the prolonged analgesic and paralytic effect of the spinal anesthesia. Second, using a validated and more detailed scale for mobilization and related pain by protocol at predefined hours after block administration may have provided more information than our study has revealed. Finally, future studies should also focus on the role and potential advantages of the TQL block as part of enhanced recovery, as these issues are central for patients, healthcare providers and administrators.


Bilateral ropivacaine TQL block for patients undergoing ECS with a Joel-Cohen incision under spinal anesthesia resulted in significant reduction in postoperative opioid consumption during 24 hours postoperatively. Further, we observed significant prolongation in time to first opioid, and significant reduction of pain during the first six postoperative hours.



  • Contributors CKH, JB: planning, conducting, reporting, conception, design, acquisition of data, data analysis, interpretation of data and writing of the manuscript. MD: conducting, acquisition of data and writing of the manuscript. GES, ML, TRD: conception, design and writing of the manuscript. GHL: data analysis, interpretation of data and writing of the manuscript. VWSC, MW: interpretation of data and writing of the manuscript. TFB: planning, conception, design, interpretation of data and writing of the manuscript.

  • Funding This work was supported by the Department of Anesthesiology and Intensive Care Medicine, Zealand University Hospital, University of Copenhagen, Denmark, and by the Research Foundation of Region Zealand.

  • Competing interests None declared.

  • Patient consent for publication Not required.

  • Ethics approval Ethical approval was obtained from the regional ethics committee of Region Zealand, Denmark (2017/SJ-594). The study was approved by the Danish Medicines Agency (EudraCT 2016-004594-41) and the Danish Data Protection Agency (REG-004-2017).

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

  • Data availability statement Data are available upon reasonable request.

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.