1. INTRODUCTION

Pain after cardiac surgery is caused by extended incisions, wound retraction, and chest tube drainage, leading to reduced coughing and breathing strength, decreased patient mobility, prolonged hospitalization, and increased morbidity.1 In addition to these immediate postoperative challenges, patients who undergo cardiac surgery are at an increased risk of developing persistent pain after the surgery and may become more reliant on opioids for pain management in the long term.2,3

Regional analgesia techniques are advised for the prevention of persistent postoperative pain, the reduction of opioid consumption, and the minimization of related side effects after cardiac surgery.4 Although epidural or paravertebral blocks (PVB) provide effective analgesia, their use in cardiac surgery has decreased owing to the associated risk of complications. The erector spinae plane (ESP) block has gained acceptance for its simplicity in execution and capability to afford substantial analgesia following cardiac procedures,5–7 with a comparatively lower risk of complications than PVBs and fewer issues pertaining to anticoagulant therapy.8

Previous investigations into the use of the ESP block for cardiac surgery have predominantly employed single-shot techniques and focused on pain and opioid use shortly after surgery, typically within the first 24 hours, while offering limited data on respiratory function. The aim of this study is examining the outcomes of a continuous ESP block over a 72-hour postoperative period, comparing its impact on opioid consumption and respiratory function improvement against that of conventional care in cardiac surgery.

2. METHODS

This retrospective cohort study was approved by the Institutional Review Board (IRB) of Taichung Veterans General Hospital with a waiver of informed consent on January 28, 2022 (IRB CE22029B) and reported items following the Strengthening the reporting of observational studies in epidemiology (STROBE) checklist (Supplement File 1, https://links.lww.com/JCMA/A244).

2.1. Patient selection and data collection

We included patients aged between 20 and 80 years, undergoing elective cardiac procedures such as coronary artery bypass graft surgery, cardiac valve surgery, and robotic-assisted cardiac surgery. These patients were categorized based on whether they received pre-incision unilateral or bilateral ESP catheter insertion, or no ESP catheter at all. Exclusion criteria included emergent surgeries and re-do cardiac surgeries.

Between January 2021 and July 2022, we enrolled 262 patients in the entire cohort. Among these, 53 patients who received preincisional ESP block were designated as the ESP group, while the remaining 209 patients, who did not receive an ESP block, were labeled as the pre-match without ESP group. Using propensity score matching at a 1:1 ratio, we selected 53 patients from the without ESP group to serve as the control group (n = 53) (matched cohort). The surgical approach distribution was equivalent across both groups, with 42 patients undergoing sternotomy and 11 undergoing thoracotomy. All preoperative demographic and intraoperative data were collected from electronic medical records at Taichung Veterans General Hospital.

2.2. Propensity score matching

Propensity score matching was used to balance confounding factors including age, gender, surgical approach (sternotomy or thoracotomy), and EuroSCORE II (European system for cardiac operative risk evaluation), a risk rating system for predicting mortality after cardiac surgery. Covariates were chosen based on published literature indicating their correlation to postoperative pain trajectory, including age, gender, surgical approach.9,10 Additionally, EuroSCORE II was included as a more comprehensive indicator of patient physical status than the conventional American Society of Anesthesiologists (ASA) physical status classification, which was usually class III in cardiac surgical patients.11 The propensity score, representing the probability of receiving ESP or not receiving ESP, was calculated using logistic regression with these covariates as the independent variables. The dependent variable was the treatment status (receiving ESP or not). We opted for the nearest neighbor matching method without replacement, and a match tolerance threshold of 0.2 was applied to each matched pair.

2.3. ESP catheter placement and application

Patients in the ESP group received ESP catheter placement by a single anesthesiologist before surgery. Patients were positioned in the lateral-decubitus or prone position. After light sedation and skin sterilization, the anesthesiologist used ultrasound guidance with a linear transducer (6-13 mHz) positioned in a paramedian sagittal plane at the T5 level to identify the posterior lateral edge of the transverse process and the interfascial plane between the intertransverse ligaments and the erector spinae muscle. A 19G 100 mm Tuohy needle (Sonolong Nanoline, Pajunk®, Geisingen, Germany) was inserted in plane from the caudal to cranial direction, and 5 to 10 mL normal saline was injected to confirm adequate interfascial spreading. Then, a catheter (20 G, SonoLong Echo, echogenic catheter; Pajunk®) was inserted 4 to 6 cm further, placing the tip close to the T4 transverse process. 0.5% ropivacaine (Nang Kung®, Tainan, Taiwan) 0.3 mL·kg−1 (ideal body weight) was given in each catheter 30 minutes before the surgical incision and another same dose of 0.5% ropivacaine before sternum closure. After surgery, each catheter was connected to a Sapphire Multi-Therapy infusion pump (Eitan Medical®, Netanya, Israel). The pump was set to deliver an intermittent automatic bolus of 0.3 mL·kg−1 0.16% ropivacaine every 4 hours. We programmed a cumulative dose limit for each 4-hour interval, ensuring only one demand bolus of the same dosage (0.3 mL·kg−1) was available for each side in the time span. A 30-minute lockout time was also set to prevent inadvertent extra bolus administration. Patients who underwent the thoracotomy approach received the same treatment as those in the sternotomy group, but the ESP catheter was only placed ipsilaterally according to the surgical side. After the initial 48-hour period postsurgery, the continuation of ESP block analgesia could be determined based on individual patient preference and clinical indications, with a maximum duration not exceeding 5 days to reduce the likelihood of catheter-related infections.

2.4. Intraoperative and postoperative management

We induced anesthesia by intravenous administration of 2 µg·kg−1 fentanyl and 1 to 2 mg·kg−1 propofol. The choice and dosage of the neuromuscular agent were based on the attending anesthesiologists’ preferences. Anesthesia depth was monitored using the bispectral index (BIS™ sensor; Covidien, Boulder, CO), which was maintained between 40 and 50. Propofol was infused during surgery (target control infusion [TCI] mode; Schnider model, Ce 1-4 µg·mL−1) to maintain anesthesia. Analgesics used were remifentanil infusion (TCI mode, Minto model, Ce 1-7 ng·mL−1), alfentanil infusion (TCI mode; Scott model, Ce 20-50 ng·mL−1), or fentanyl infusion 1 µg·kg−1·h−1 based on the preference of the anesthesiologists. Muscle relaxation was achieved with a cisatracurium infusion at 1 to 3 µg·kg−1·min−1. Mechanical ventilation was performed with a tidal volume of 8 mL·kg−1 (predicted body weight), positive end-expiratory pressure of 5 cmH2O and a respiratory rate that maintained normocapnia. During the skin closure phase, dexmedetomidine was administered to all patients at a dose of 0.2 to 0.4 µg·kg−1·h−1 and continued during their stay at the Intensive care unit (ICU), where they were intubated and mechanically ventilated. Pain assessment for these intubated patients was conducted using the Critical-Care Pain Observation Tool (CPOT). A CPOT score exceeding three necessitated titration of dexmedetomidine and the provision of intravenous rescue analgesia to maintain comfort. Subsequent to achieving clear consciousness, hemodynamic stability, and muscle power recovery, the clinical team proceeded with weaning and extubation protocols. Postextubation, patient-reported pain levels were evaluated using the numerical rating scale (NRS, 0-10). Intravenous rescue analgesia, with either tramadol 75 mg or morphine at 30 to 50 µg·kg−1, was administered in response to CPOT or NRS scores of 3 or above, or when patients reported distressing pain.

2.5. Outcome measurement

Outcomes were stratified by the type of surgical approach—sternotomy or thoracotomy—owing to the presence of both methods at our institution. The primary outcome of this study was the total oral morphine equivalent (OME) dose received from the time of patient extubation until 72 hours thereafter. OME was calculated using a conversion toolkit within our hospital’s electronic medical record system, which standardizes opioid analgesic doses to OMEs according to established guidelines.12 Secondary outcomes included: (1) the frequency of daily NRS scores of 3 or higher; (2) the volume of daily incentive spirometry; and (3) incentive spirometry performance, expressed as a percentage of the preoperative baseline, within the first 72 hours postextubation.

Postoperative complications were identified through medical record analysis, adhering to the definitions set forth by the European Joint Taskforce’s guidelines for perioperative clinical outcomes (EPCO).13 Recorded complications included pneumonia, pleural effusion, pulmonary edema, atelectasis, acute kidney injury, surgical site infection, surgical bleeding, delirium, ileus, and newly diagnosed arrhythmias during the postoperative period. Additionally, for patients receiving an ESP block, catheter-related complications such as puncture site hematoma or infection were monitored, documented, and reported by the nursing staff.

2.6. Statistical analyses

Data were analyzed using (IBM® SPSS Statistics 26.0, Chicago, IL, USA) 1989 - 2019. Categorical variables were assessed with Pearson’s chi-square test or Fisher’s exact test for low expected frequencies. Continuous variables were compared using the Mann–Whitney U test, with results presented as median (interquartile range [IQR]). Significance was established at a p value <0.05. Balanced baseline covariates between groups were confirmed using both p values and the standardized mean difference (SMD), with an SMD <0.1 indicating satisfactory balance. Effect size and 95% CI of the difference were reported to demonstrate clinically meaningful differences among outcome variables. For the Mann–Whitney U test, we used effect size r, where an r value of <0.3 represented a small effect, between 0.3 and 0.5 indicated a medium effect, and more than 0.5 suggested a large effect. For the chi-square test, we used the Φ (phi) coefficient, with values from 0.1 to 0.2 considered weak, greater than 0.2 as moderate, over 0.4 as relatively strong, and exceeding 0.6 as strong. To compare repeated measures like NRS ≥3 incidence and incentive spirometry performance over time between the ESP and control groups, we used a generalized estimating equation (GEE) model with an exchangeable correlation structure to account for temporal factors. Non-normally distributed variables underwent log transformation for appropriate analyses.

3. RESULTS

Table 1 presents the baseline demographic and clinical characteristics for both the ESP and control groups. In the matched cohorts, there were no statistically significant differences in terms of gender, age, body mass index (BMI), comorbidities, or type of surgery. Further stratification of demographic characteristics by surgical approach—sternotomy or thoracotomy—revealed no statistically significant differences between the ESP and control groups for either patient subset. Additional information regarding demographic characteristics in the pre-matched cohort is reported in Supplement File 2, https://links.lww.com/JCMA/A245.

Table 1 - Demographic characteristics of ESP and control groups in the matched cohort, stratified by sternotomy and thoracotomy approaches Matched cohort Sternotomy Thoracotomy ESP (n = 53) Control (n = 53) p ESP (n = 42) Control (n = 42) p ESP (n = 11) Control (n = 11) p Gender 0.840 0.825 0.586  Male 33 (62.3) 34 (64.15) 25 (59.5) 24 (57.1) 8 (72.7) 10 (90.9)  Female 20 (37.7) 19 (35.85) 17 (40.5) 18 (42.9) 3 (27.3) 1 (9.1) Age, y 61 (52.0-70.5) 63 (53-71) 0.615 61.0 (51.8-70.3) 62.0 (52.8-71) 0.632 62.0 (52-72) 65.0 (55-69) 0.834 BMI, kg·m−2 23.9 (22.4-26.1) 25.3 (21.7-28.4) 0.205 23.9 (22.2-25.9) 24.8 (21.6-28.3) 0.348 23.6 (22.7-26.3) 25.9 (22.3-29.8) 0.332 EuroSCORE II 2.7 (1.2-4.1) 2.9 (1.6-5.0) 0.355 2.9 (1.5-4.9) 3.8 (1.7-5.3) 0.395 1.6 (0.7-3.1) 1.7 (1.2-2) 0.606 Pulmonary function  FEV1, L 2.1 (1.6-2.6) 2.4 (1.8-2.8) 0.269 2.1 (1.5-2.6) 2.3 (1.7-2.7) 0.347 2.5 (2.1-3.2) 2.7 (2.3-3.0) 0.604  FVC, L 2.8 (1.9-3.6) 3.0 (2.3-3.6) 0.523 2.7 (1.8-3.3) 2.9 (2.2-3.3) 0.577 3.1 (2.4-3.9) 3.4 (2.6-3.8) 0.764  FEV1/FVC 0.8 (0.7-0.8) 0.8 (0.8-0.9) 0.142 0.8 (0.7-0.8) 0.8 (0.8-0.9) 0.139 0.8 (0.8-0.9) 0.8 (0.8-0.8) 0.842 Comorbidity  Hypertension 22 (41.5) 21 (39.6) 0.843 19 (45.2) 17 (40.5) 0.659 3 (27.3) 4 (36.4) >0.999  Diabetic 15 (28.3) 22 (41.5) 0.154 13 (31.0) 18 (42.9) 0.258 2 (18.2) 4 (36.4) 0.635  CRF 9 (17.0) 9 (17.0) >0.999 8 (19.0) 9 (21.4) 0.786 1 (9.1) 0 (0.0) >0.999  Heart failure 14 (26.4) 13 (24.5) 0.824 14 (33.3) 12 (28.6) 0.637 0 (0.0) 1 (9.1) >0.999  COPD 2 (3.8) 3 (5.7) >0.999 2 (4.8) 3 (7.1) >0.999 0 (0.0) 0 (0.0) -  Asthma 2 (3.8) 2 (3.8) >0.999 14 (33.3) 6 (14.3) >0.999 0 (0.0) 0 (0.0) -  Stroke 3 (5.7) 2 (3.8) >0.999 2 (4.8) 2 (4.8) >0.999 1 (9.1) 0 (0.0) >0.999  MDD 1 (1.9) 0 (0.0) >0.999 1 (2.4) 0 (0.0) >0.999 0 (0.0) 0 (0.0) -  Anxiety 1 (1.9) 0 (0.0) >0.999 0 (0.0) 0 (0.0) - 1 (9.1) 0 (0.0) >0.999 Surgery type 0.123 0.128 0.392  CABG 15 (28.3) 25 (47.2) 11 (26.2) 17 (40.5) 4 (36.4) 6 (54.5)  CABG with valve 6 (11.3) 1 (1.9) 5 (11.9) 3 (7.1) 0 (0.0) 0 (0.0)  Valve only 25 (47.2) 25 (47.2) 20 (47.6) 17(40.5) 7 (63.6) 5 (45.5)  Valve with aorta 2 (3.8) 2 (3.8) 2 (4.8) 5 (11.9) 0 (0.0) 0 (0.0)  Other 5 (9.4) 0 (0.0) 4 (9.5) 0 (0.0) 0 (0.0) 0 (0.0) Surgical approach 0.807  Sternotomy 42 (79.3) 42 (79.3)  Thoracotomy 11 (20.8) 11 (20.8) Continuous data were compared using the Mann–Whitney U test and are presented as the median (IQR). Categorical data were compared using the Chi-squared test and are presented as numbers (%). The matched cohort was generated from the entire cohort using propensity score matching. This was done to balance covariates, including age, gender, surgical approach, and EuroSCORE II, based on findings from published literatures.9–11

BMI = body mass index; CABG = coronary artery bypass graft; COPD = chronic obstructive pulmonary disease; CRF = chronic renal failure; ESP = erector spinae plane; EuroSCORE II = European system for cardiac operative risk evaluation; FEV1 = forced expiratory volume in 1 s; FEV1/FVC = proportion of FEV1 to FVC, normal value > 0.75; FVC = forced vital capacity; IQR = interquartile range; MDD = major depression disorder.

Table 2 delineates intraoperative and postoperative data, also stratified by surgical approach. No significant differences were observed between the ESP and control groups in terms of intraoperative propofol and opioid use, anesthesia duration, duration of mechanical ventilation, or hospital stay, irrespective of the surgical method used. Additionally, before extubation, both the pain intensity (CPOT) and the need for rescue analgesics showed no statistically significant differences between the ESP and control groups, regardless of whether a sternotomy or thoracotomy approach was employed. The ESP group reported a lower incidence of composite postoperative pulmonary complications (PPCs), a difference particularly noticeable in thoracotomy patients (sternotomy: 50% vs 71.4%, p = 0.044, effect size ϕ = 0.22; Thoracotomy: 27.3% vs 90.9%, p = 0.003, effect size ϕ = 0.65). Other postoperative complications, including bleeding, ileus, postoperative cognitive dysfunction, acute kidney injury, and surgical site infection occurred at similar rates in both groups. No ESP procedure-related complications, such as hematoma or injection site infection, were reported.

Table 2 - Intraoperative and postoperative variables in different surgical approaches between the ESP and control groups Sternotomy Thoracotomy ESP (n = 42) Control (n = 42) p ESP (n =11) Control (n = 11) p Intraoperative  Propofol, mg 1935.1 (1622.7-2521.5) 2086.2 (1725.1-2459.5) 0.661 2190 (1571-2381.6) 3243.9 (2035.8-4100.2) 0.088  Opioid, µga 232.3 (156.7-680.9) 524 (145.9-985.5) 0.154 199.7 (134.5-299.1) 592.4 (219-1441.9) 0.098  Patients receive remifentanil 17 (40.5) 27 (64.3) 0.029 3 (27.3) 5 (45.5) 0.385   Remifentanil, µgb 654.2 (353.1-1222.9) 796.0 (407.9-1280.8) 0.650 296 (134.5-1842.8) 977.9 (621.5-1598.3) 0.297   Remifentanil TWA, µg/kg/minb 0.018 (0.013-0.029) 0.021 (0.016-0.028) 0.588 0.010 (0.003-0.049)b 0.023 (0.018-0.034) 0.456  Anesthesia duration, h 9.7 (8.2-10.8) 9.0 (7.8-11.5) 0.522 10.9 (10-12.5) 10.6 (10.1-12.8) 0.861 Postoperative  Pain scale (CPOT) 1.5 (1.2-1.7) 1.6 (1.4-1.7) 0.405 1.3 (1.2-1.5) 1.5 (1.2-1.7) 0.699  Patients receive rescue analgesics 9 (21.4) 14 (33.3) 0.224 3 (27.3) 6 (54.5) 0.203  Length of mechanical ventilation, d 0.7 (0.5-0.8) 0.7 (0.5-0.8) 0.847 0.5 (0-0.6) 0.5 (0.1-0.6) 0.961  Postoperative length of stay, d 10.0 (8.0-14.0) 10.0 (7-14) 0.794 9.0 (7.0-10.0) 8.0 (6.0-16.0) 0.809 Complications  Composite PPCs, n (%) 21 (50) 30 (71.4) 0.044 3 (27.3) 10 (90.9) 0.003   Respiratory infection, n (%) 4 (9.5) 12 (28.6) 0.024 0 (0.0) 6 (54.5) 0.005   Pleural effusion, n (%) 15 (35.7) 25 (59.5) 0.026 1 (9.1) 8 (72.7) 0.007   Pulmonary edema, n (%) 7 (16.7) 11 (26.2) 0.287 3 (27.3) 3 (27.3) >0.999   Atelectasis, n (%) 2 (4.8) 5 (11.9) 0.433 0 (0.0) 0 (0.0) -  Surgical bleeding, n (%) 0 (0.0) 2 (4.8) 0.494 1 (9.1) 0 (0.0) >0.999  New-onset arrhythmia, n (%) 6 (14.3) 7 (16.7) 0.763 1 (9.1) 2 (18.2) >0.999  Acute kidney injury, n (%) 0 (0.0) 1 (2.4) >0.99 0 (0.0) 0 (0.0) -  Surgical site infection, n (%) 0 (0.0) 1 (1.2) >0.99 0 (0.0) 1 (9.1) >0.999  Delirium, n (%) 2 (4.8) 0 (0.0) 0.494 0 (0.0) 0(0.0) -  Ileus, n (%) 0 (0.0) 1 (2.4) >0.99 0 (0.0) 0 (0.0) -

The Mann–Whitney U test was used to compare continuous data, which are presented as the median (IQR). The Chi-squared test and the Fisher’s exact test were used to compare categorical data, and the results are presented as numbers (%). p value comparing ESP vs control group.

CPOT = Critical-Care Pain Observation Tool; ESP = erector spinae plane; IQR = interquartile range; length of mechanical ventilation (d), duration of mechanical ventilation; postoperative length of stay (d), duration of stay in hospital after surgery; TWA = time-weighted average, equals to cumulative remifentanil (µg)/body weight (kg)/anesthesia duration (min).

aAll intraoperative opioids (remifentanil, alfentanil, and fentanyl) were converted to equivalent intravenous fentanyl doses in a ratio of remifentanil:alfentanil:fentanyl = 1:20:1 (eg, remifentanil 20 µg = alfentanil 200 µg = fentanyl 20 µg). This conversion is based on established opioid equivalency ratios.14–16

bData were calculated among patients who received remifentanil and presented as median (IQR). For data with limited cases (eg, only three cases in the cohort), values are presented as median (lowest value-highest value).

Table 3 details a comparison of the primary and secondary outcomes in the ESP and control groups, stratified by surgical approach. The ESP group displayed a significantly reduced cumulative OME dose within the first 72 hours postextubation. Specifically, a median decrease of 113 mg (95% CI: 60-157.5 mg, p < 0.001, effect size r = 0.48, indicating a medium effect) was observed in sternotomy patients and 172.5 mg (95% CI: 45-285 mg, p = 0.010, effect size r = 0.54, indicating a large effect) in thoracotomy patients. Patients who underwent sternotomy in the ESP group reported significantly fewer instances of a maximum NRS ≥3 on day 1 (11.9% vs 40.5%, p = 0.003, effect size ϕ = 0.32) and day 3 (9.5% vs 38.1%, p = 0.002, effect size ϕ = 0.34). In the thoracotomy subset, the ESP group had significantly lower instances of NRS ≥3 on day 1 (18.2% vs 72.7%, p = 0.010, effect size ϕ = 0.55) and day 2 (9.1% vs 72.7%, p = 0.008, effect size ϕ = 0.65). While preoperative baseline incentive spirometry volume were comparable between the ESP and control group, the ESP group demonstrated higher incentive spirometry volumes on day 2 and day 3 in sternotomy patients (day 2, 1000 vs 750 mL, p = 0.024, effect size r = 0.26; day 3, 1250 vs 750 mL, p = 0.025, effect size r = 0.26) and higher volume on day1 in thoracotomy patients (1000 vs 500 mL, p = 0.010, effect size r = 0.54). In terms of performance comparison as a percentage of baseline value (spirometry %baseline), the ESP group consistently had a higher incentive spirometry performance as spirometry %baseline, particularly in thoracotomy patients over the first 3 days (day 1, 58.8% vs 25%, p = 0.002, effect size ϕ = 0.69; day 2, 58.8% vs 40%, p = 0.022, effect size ϕ = 0.53; day 3, 60% vs 50%,