Real-time ultrasound-guided versus landmark techniques for thoracic epidural placement in elderly patients: A randomized comparative study Authors: Karuna Sutthibenjakul, Jatuporn Pakpirom, Pannawit Benjhawaleemart, Kwanruetai Naruanart, Sirikarn Siripruekpong, Netchanok Nokchan, Riam Kosem Published: 2026 ABSTRACT Thoracic epidural placement (TEP) in elderly patients is technically challenging, and evidence comparing real-time ultrasound guidance (USG) with the conventional landmark (LM) technique remains limited. This study aimed to compare the technical performance and success rates between two approaches. Seventy-six elderly patients who were receiving thoracic epidural analgesia were prospectively enrolled and randomized to the USG (n = 38) or LM (n = 38) group. The primary outcome was first-pass success (FPS). The secondary outcomes included the number of needle passes and skin punctures, TEP time, preparation time, and total procedure time. The median age was 70 years in the LM group and 71 years in the USG group. Overall, TEP success was greater with US than with LM [94.7% vs. 81.6%], although the difference was not statistically significant (p = 0.15). Compared with the LM group, the USG group presented significantly greater FPS [55.3% vs. 26.3%; p = 0.02], fewer needle passes and skin punctures, and a trend toward shorter TEP times. The preparation time was longer in the USG group, but the total procedure time was similar between the groups. Real-time ultrasound guidance significantly improves procedural efficiency in thoracic epidural placement in elderly patients, particularly in term of first-pass success and needle-related performance. However, no statistically significant different in overall procedural success was observed compared with landmark-based technique.Register: Thai Clinical Trials Registry No: TCTR20220913004 Register date: 13/09/2022. Introduction Thoracic epidural analgesia (TEA) is widely used for perioperative pain control and is recommended following major thoracic and upper abdominal surgery1. In addition to improving postoperative analgesia relative to systemic opioids, TEA has been associated with reductions in pulmonary, cardiovascular, metabolic, inflammatory, and neurohormonal complications1–3; enhanced bowel recovery4; and superior analgesia at rest and during movement compared with both systemic opioids and other regional techniques4,5. Although paravertebral blocks provide comparable analgesic efficacy for thoracic procedures, TEA continues to be favored in many clinical settings despite its higher incidence of technical challenges and procedure-related complications6. Thoracic epidural catheter placement (TEP) remains technically demanding because of the limited interlaminar space and steep caudal angulation of the thoracic spinous processes. These difficulties are amplified in elderly patients, who frequently present with age-related degenerative changes such as calcification of the ligamentum flavum, osteophyte formation, reduced interlaminar windows, and spinal deformities, including kyphosis and scoliosis7. As a result, elderly patients commonly require multiple needle attempts, exhibit higher primary failure rates when the epidural space cannot be reliably identified8, and face increased risks of complications such as epidural hematoma9,10, spinal cord injury11, and pleural puncture12. Ultrasound guidance has gained attention as a valuable adjunct for neuraxial procedures, enabling visualization of spinal structures and supporting more precise needle trajectory planning. While real-time ultrasound guidance has demonstrated benefits in lumbar neuraxial blocks—improving success rates and reducing needle passes13–15—evidence supporting its use in thoracic epidural placement remains limited. The existing trials of ultrasound-guided thoracic epidural techniques are rare, with inconsistent findings, and are often limited to preprocedural scanning rather than real-time needle guidance16. Although ultrasound may offer limited benefit in patients with normal anatomy, it may be advantageous in individuals with predicted technical difficulty, such as those with deformities or degenerative spinal changes17. However, real-time ultrasound-guided thoracic epidural placement has not been adequately evaluated in elderly patients, a population in which technical difficulty is common. Therefore, this study aimed to compare the first-pass success rates and procedural outcomes of landmark-based versus real-time ultrasound-guided thoracic epidural placement in elderly patients who underwent major abdominal or thoracic surgery. We hypothesized that real-time ultrasound guidance would increase first-pass success and improve technical performance relative to the conventional technique. Methods Study design and ethical approval This prospective, randomized, comparative study was conducted at a university hospital and was approved on 9/09/2022 by the Human Research Ethics Committee of the Faculty of Medicine, Prince of Songkla University. The study followed the principles of the Declaration of Helsinki. The protocol was registered prospectively with the Thai Clinical Trials Registry (number: TCTR20220913004) on 13/09/2022. The first case enrollment was on 20/09/2022, and data were collected until August 2023. Patient recruitment and randomization Patients aged ≥ 65 years who were scheduled to undergo major abdominal or thoracic surgery requiring thoracic epidural analgesia were screened for eligibility (Fig. 1). The exclusion criteria included refusal to participate, contraindications to neuraxial block, thoracic spine deformity, previous thoracic spine surgery, allergy to local anesthetics, or severe comorbidities that could interfere with outcome assessment. Eligible patients were approached one day before surgery, and written informed consent was obtained. The participants were randomized at a 1:1 ratio to receive either landmark-based thoracic epidural placement (LM group) or real-time ultrasound-guided thoracic epidural placement (USG group). Randomization was performed via computer-generated sequential numbers, which were sealed in opaque envelopes and prepared by an independent third party. Owing to the nature of the intervention, blinding of patients, anesthesiologists, and procedural outcomes assessor was not feasible. However, postoperative outcomes in the post-anesthetic care unit (PACU) - including surgical wound block coverage, epidural bolus requirement, sensory blockade and patient satisfaction- were evaluated by a blinded assessor. To minimize the bias the primary and secondary outcomes consisted of objective measurements. In addition, the independent outcome assessor was trained to avoid guiding patients when obtaining sensory blockade and discomfort score. Fig. 1Consort flow diagram. Consort flow diagram. Preoperative preparation and epidural placement protocol All patients fasted for at least 8 h and did not receive premedication except for their routine medications. Upon arrival at the operating theatre, standard monitoring and intravenous access were established. Patients were instructed to rate procedural discomfort via a 0–10 verbal numeric rating scale (0 = very comfortable, 10 = worst pain imaginable). Supplemental oxygen (1–2 L/min) was administered, and intravenous fentanyl (50 µg) with or without midazolam (1 mg) was given for mild sedation. Thoracic epidural placement (TEP) was performed under full aseptic precautions with the patient in the lateral decubitus position. The procedures were performed by one of three anesthesiologists: PJ who completed 3 years of regional anesthesia training and had experience approximately 50 cases of real-time ultrasound-guided TEP, and SK and NK who had just completed one year of training and were in the early stages of learning this technique. The intervertebral level was selected according to the planned surgical procedure and identified via a counting-up or counting-down method from the first rib or 12th rib to the targeted intercostal space, which corresponds to the targeted intervertebral level. A Perifix® epidural kit (18-gauge Tuohy needle, 20-gauge multiorifice catheter) was used in both groups. Loss of resistance (LOR) to saline was used to identify the epidural space. After LOR, 5 mL of normal saline was injected to prefill the epidural space via Touhy needle before a pressure transducer was connected for epidural pressure waveform analysis. The catheter was then advanced 4–5 cm into the epidural space, tunneled, and secured with a transparent dressing. A test dose of 3 mL of 2% lidocaine with adrenaline (5 µg/mL) was administered to exclude intravascular or subarachnoid placement. Sensory blockade was assessed by pinprick sensation 15 min later. Landmark-based thoracic epidural technique After the target intervertebral space was identified, a skin mark was placed approximately 1 cm lateral and caudal to the intended puncture site on the dependent side (Fig. 2A). Using a paramedian approach, the Tuohy needle was advanced toward the lamina and then “walked” cranially until it entered the interlaminar space, where the LOR was applied to identify the epidural space. Real-time ultrasound-guided technique A KONICA MINOLTA SONIMAGE HS1 ultrasound system (Philips Ultrasound Inc., Tokyo, Japan) equipped with either a curvilinear (5–2 MHz) or linear (11–3 MHz) transducer was used depending on patient habitus. The intervertebral level was identified by counting upward from the last rib (Fig. 2B). The probe was then moved cranially to the corresponding intercostal space (Fig. 2C) and gradually shifted medially to visualize the transverse process (Fig. 2D), articular process (Fig. 2E), and ultimately the interlaminar space by tilting the probe toward the midline (Fig. 2F). The procedure followed a previously described three-steps technique16,18: (1) identification of the target thoracic level and interlaminar window (Figs. 2B–F and 3A–C); (2) advancement of the Tuohy needle under real-time ultrasound visualization toward the interlaminar space (Fig. 3D); and (3) application of LOR to confirm entry into the epidural space (Fig. 3E). Although needle advancement toward the interlaminar space was performed under real-time ultrasound visualization, the transducer was removed after prior to LOR confirmation. Therefore, the technique represents a hybrid approach combining real-time guidance during trajectory planning with conventional identification of epidural space. Fig. 2Demonstrates how to use real-time USG for TEP. A: Skin marked in LM-gr; the desired intervertebral level was counted from the T7 spinous process (SP), which correlates with the scapular tip (red dotted line). The needle was inserted 1 cm laterally and 1 cm caudally (red dot) B-F: Ultrasound image from the USG-gr B: Rib 12th which is the last rib of the thoracic cage and floating rib, was used as a reference point to count up to the desired intercostal space, which corresponds to the intervertebral level. C: Ultrasound transducer-oriented paramedian sagittal scan (small picture) at the rib level. D: From the C position, the ultrasound transducer was moved medially (small picture) to visualize the transverse process of T9--11. E: Slightly medial direction from D. The ultrasound transducer is depicted, and the paramedian sagittal oblique view shows the articular process. F: From E, the ultrasound transducer was tilted to midline, allowing visualization of the interlaminar space, the opening space for epidural access. TEP: thoracic epidural placement, LOR: loss of resistance, NSS: normal saline, PSM: paraspinal muscle, TP: transverse process Demonstrates how to use real-time USG for TEP. A: Skin marked in LM-gr; the desired intervertebral level was counted from the T7 spinous process (SP), which correlates with the scapular tip (red dotted line). The needle was inserted 1 cm laterally and 1 cm caudally (red dot) B-F: Ultrasound image from the USG-gr B: Rib 12th which is the last rib of the thoracic cage and floating rib, was used as a reference point to count up to the desired intercostal space, which corresponds to the intervertebral level. C: Ultrasound transducer-oriented paramedian sagittal scan (small picture) at the rib level. D: From the C position, the ultrasound transducer was moved medially (small picture) to visualize the transverse process of T9--11. E: Slightly medial direction from D. The ultrasound transducer is depicted, and the paramedian sagittal oblique view shows the articular process. F: From E, the ultrasound transducer was tilted to midline, allowing visualization of the interlaminar space, the opening space for epidural access. TEP: thoracic epidural placement, LOR: loss of resistance, NSS: normal saline, PSM: paraspinal muscle, TP: transverse process Intraoperative and postoperative epidural management Following the confirmation of correct catheter function, general anesthesia was induced, and controlled ventilation was initiated. Once hemodynamic stability was established, an additional bolus of 2–3 mL of 0.2% bupivacaine with fentanyl (2 µg/mL) was administered, followed by a continuous epidural infusion at 0.1 mL/kg/h. The infusion rates were adjusted at the discretion of the attending anesthesiologist. After surgery, the infusion was changed to 0.1% bupivacaine with fentanyl for postoperative analgesia. In the postanaesthesia care unit (PACU), sensory level assessments were repeated. Catheters that failed to produce sensory blockade were removed and classified as primary failures, and no further investigate to determine the catheter tip position. Thoracic epidural analgesia was managed by acute pain services until catheter removal. Outcome measurement The primary outcome was first-pass success (FPS), defined as successful epidural placement on the first needle pass within a single skin puncture. Successful TEP requires eliciting LOR, acquiring epidural waveform analysis (Fig. 3F), successfully threading the catheter, and achieving a sensory block after the test dose. Procedures lasting > 30 min were classified as failures. The secondary outcomes included the number of needle passes and skin punctures, TEP time, preparation time, total procedural time, complications, difficulty of catheter advancement, discomfort scores, and epidural waveform quality. The preparation time was measured from patient positioning to completion of skin infiltration. The TEP time was defined as the interval from needle insertion to successful LOR. Total procedural time combined with preparation and TEP time. Epidural waveform quality was classified as strong, moderate, weak, or absent. Discomfort was rated on a 0–10 numeric scale. Catheter advancement was categorized as (1) easy to trace, (2) resistance but successful, or (3) significant resistance requiring > 2 attempts. Sensory coverage in the PACU was recorded as absent, partial, or complete. Fig. 3Steps of real-time USG for TEP. A: Patient position in the left lateral decubitus position B: Ultrasound transducer placing paramedian sagittally at the transverse process of the desired intervertebral level. C: the required interlaminar identification, and the target interlaminar gap moves to the right side of the ultrasound screen. D: Needle insertion in-plane approach (small picture); the needle passes from caudal to cranial toward the targeted interlaminar space (red arrow). E: Once the needle tip reaches the interlaminar space, the ultrasound transducer is removed, and the LOR syringe assembly is connected to the Touhy needle. F: The monitor shows an epidural pressure pulsatile waveform (red color); after eliciting the LOR, 5 ml of normal saline was injected through the Touhy needle, and the needle was connected to the pressure transducer. ESM: erector spinae muscle. Steps of real-time USG for TEP. A: Patient position in the left lateral decubitus position B: Ultrasound transducer placing paramedian sagittally at the transverse process of the desired intervertebral level. C: the required interlaminar identification, and the target interlaminar gap moves to the right side of the ultrasound screen. D: Needle insertion in-plane approach (small picture); the needle passes from caudal to cranial toward the targeted interlaminar space (red arrow). E: Once the needle tip reaches the interlaminar space, the ultrasound transducer is removed, and the LOR syringe assembly is connected to the Touhy needle. F: The monitor shows an epidural pressure pulsatile waveform (red color); after eliciting the LOR, 5 ml of normal saline was injected through the Touhy needle, and the needle was connected to the pressure transducer. ESM: erector spinae muscle. Statistical analysis The sample size was based on prior data indicating differences in FPS between techniques16. A sample of 68 patients (34 per group) provided 80% power with α = 0.05 for two independent proportions19. To account for potential dropout, 38 patients were recruited per group. Data analysis was performed via R version 3.5.3. The normality of continuous variables was assessed via the Shapiro–Wilk test. Depending on the distribution, the data are summarized as the means ± SDs or medians [IQRs] and were compared via Student’s t test or the Wilcoxon rank-sum test. Categorical variables were analyzed via the chi-square test or Fisher’s exact test. A p value < 0.05 was considered statistically significant. Results A total of 92 patients were screened for eligibility, and 16 were excluded (six did not meet the inclusion criteria, eight had changes in the planned surgical procedure, and two were excluded for other reasons). Seventy-six patients were ultimately randomized, with 38 allocated to the landmark group (LM group) and 38 to the real-time ultrasound-guided group (USG group) (Fig. 1). Seven patients in the LM group and two in the USG group had their epidural catheters removed in the PACU because of the absence of sensory blockade despite reassessment. Baseline characteristics, including age, sex, BMI, surgical category, and ASA physical status, were similar between the groups, although endocrine disorders were more common in the LM group. The median age was 70 years in the LM group and 71 years in the USG group, and the median BMI values were 23.8 and 22.6 kg/m², respectively (Table 1). Table 1 Baseline characteristics of patients ( data are presented as numbers (%) or medians [IQRs]). CharacteristicsLM group (n = 38)USG group (n = 38)p value Age (years) 70 [67–73]71 [68.2–73]0.45Gender; n (%)0.82Male18 (47.4)20 (52.6)Female20 (52.6)18 (47.4)BMI (kg/m²)23.8 [21.9–27.0]22.6 [21.6–23.6]0.09Type of surgery n (%)0.31Thoracic9 (23.7)8 (21.1)Upper abdomen21 (55.3)16 (42.1)Lower abdomen8 (21.1)14 (36.8)ASA physical status n (%)0.82II21 (55.3)23 (60.5)III17 (44.7)15 (39.5) Comorbidities Hypertension21 (63.6)26 (68.4)0.86Diabetes mellitus10 (30.3)13 (34.2)0.92Endocrine disease7 (21.2)0 (0) 0.01 Heart disease5 (15.2)6 (15.8)1.00Renal disease2 (6.1)3 (7.9)1.00Lung disease3 (9.1)5 (13.2)0.72 Table 1 Baseline characteristics of patients ( data are presented as numbers (%) or medians [IQRs]). CharacteristicsLM group (n = 38)USG group (n = 38)p value Age (years) 70 [67–73]71 [68.2–73]0.45Gender; n (%)0.82Male18 (47.4)20 (52.6)Female20 (52.6)18 (47.4)BMI (kg/m²)23.8 [21.9–27.0]22.6 [21.6–23.6]0.09Type of surgery n (%)0.31Thoracic9 (23.7)8 (21.1)Upper abdomen21 (55.3)16 (42.1)Lower abdomen8 (21.1)14 (36.8)ASA physical status n (%)0.82II21 (55.3)23 (60.5)III17 (44.7)15 (39.5) Comorbidities Hypertension21 (63.6)26 (68.4)0.86Diabetes mellitus10 (30.3)13 (34.2)0.92Endocrine disease7 (21.2)0 (0) 0.01 Heart disease5 (15.2)6 (15.8)1.00Renal disease2 (6.1)3 (7.9)1.00Lung disease3 (9.1)5 (13.2)0.72 Baseline characteristics of patients ( data are presented as numbers (%) or medians [IQRs]). The technical outcomes of TEP are summarized in Table 2. Compared with the LM group, the USG group demonstrated a significantly higher first-pass success rate (55.3%; 95% CI 39.4–71.1 vs. 26.3%; 95% CI 12.3–40.3; p = 0.02). USG group also required significantly fewer needle passes (1[1–2] vs. 3 [1.2–5.8]; p < 0.001) and fewer skin punctures (1[1–2] vs. 2[1–3]; p = 0.03). Although the overall success rate was greater in the USG group (94.7% vs. 81.6%), the difference was not statistically significant. Table 2Procedural and technical outcomes of thoracic epidural placement. (Data are presented as the number (%), mean ± SD, or median [IQR]).OutcomesLM group (n = 38)USG group(n = 38)p valueOverall TEP success, n (%)31 (81.6)36 (94.7)0.15395% CI[69.2–93.9][87.6–100.0]First-pass success, n (%)10 (26.3)21 (55.3) 0.020 95% CI[12.3–40.3][39.4–71.1] Number of needle passes 3 [1.2–5.8]1 [1–2] < 0.001 Success TEP by needle pass, n (%) 0.041 1 st pass10 (26.3)21 (55.3)2nd pass7 (18.4)8 (21.0)>2 passes21 (55.3)9 (23.7) Number of skin punctures 2 [1–3]1 [1–2] 0.030 Success TEP by skin puncture, n (%) 0.003 1 st puncture18 (47.4)24 (63.2)2nd puncture8 (21.0)13 (34.2)>2 punctures12 (31.6)1 (2.6)First puncture success rate, n (%)18 (47.4)24 (63.2)0.24995% CI[31.5–63.2][47.8–78.5] Preparation time (min) 5 [4–6]8 [7–11] < 0.001 TEP time (min) 4.5 [1–11]2 [1–5] 0.050 Total procedure time (min) 13 [9.2–21]16 [14–19]0.072 Table 2Procedural and technical outcomes of thoracic epidural placement. (Data are presented as the number (%), mean ± SD, or median [IQR]).OutcomesLM group (n = 38)USG group(n = 38)p valueOverall TEP success, n (%)31 (81.6)36 (94.7)0.15395% CI[69.2–93.9][87.6–100.0]First-pass success, n (%)10 (26.3)21 (55.3) 0.020 95% CI[12.3–40.3][39.4–71.1] Number of needle passes 3 [1.2–5.8]1 [1–2] < 0.001 Success TEP by needle pass, n (%) 0.041 1 st pass10 (26.3)21 (55.3)2nd pass7 (18.4)8 (21.0)>2 passes21 (55.3)9 (23.7) Number of skin punctures 2 [1–3]1 [1–2] 0.030 Success TEP by skin puncture, n (%) 0.003 1 st puncture18 (47.4)24 (63.2)2nd puncture8 (21.0)13 (34.2)>2 punctures12 (31.6)1 (2.6)First puncture success rate, n (%)18 (47.4)24 (63.2)0.24995% CI[31.5–63.2][47.8–78.5] Preparation time (min) 5 [4–6]8 [7–11] < 0.001 TEP time (min) 4.5 [1–11]2 [1–5] 0.050 Total procedure time (min) 13 [9.2–21]16 [14–19]0.072 Procedural and technical outcomes of thoracic epidural placement. (Data are presented as the number (%), mean ± SD, or median [IQR]). The TEP time was shorter in the USG group than in the LM group, whereas the total procedural time was slightly longer in the USG group, although the difference did not reach statistical significance (16 [14–19] min vs. 13 [9.2–21] min; p = 0.072) (Table 2). Epidural waveform patterns differed significantly between groups, with the USG showing a greater proportion of strong and moderate waveforms (p = 0.029) (Table 3). The USG group also demonstrated higher rates of complete sensory wound coverage in the PACU (75.7% vs. 43.2%; p = 0.012). Additionally, fewer patients in the USG group required supplemental epidural boluses immediately after surgery (16.2% vs. 48.6%; p = 0.006). Table 3Procedural information and intraoperative/PACU analgesia The data are presented as numbers (%), medians [IQRs], or means ± SDs.VariablesLM group (n = 38)USG group (n = 38)p value Sensory blockade after test dose 4.5 ± 2.55.6 ± 2.1 0.040 Epidural waveform grade, n 0.029 Strong/Moderate/Weak/Absent17/12/2/726/8/4/0 Mean epidural pressure (mmHg) 17 [12.5–22.5]15.5 [11–20]0.218Accidental dural puncture, n (%)0 (0)3 (7.9)0.240Hypotension requiring treatment, n (%)30 (78.9)35 (92.1)0.192 TEP by operator (PJ/SK/NK) 15/20/318/19/10.570 Thoracic level (T5–T8/T9–T12) 12/269/290.558 Epidural depth (cm) 6 [5–6]5 [5–6]0.098Difficulty of catheter insertion, n (%)0.309Some resistance but successful6 (16.2)3 (7.9)Easy to trace31 (83.8)35 (92.1)Sensory coverage in PACU, n (%) 0.012 No blockade7 (18.9)2 (5.4)Partial coverage14 (37.8)7 (18.9)Complete coverage16 (43.2)28 (75.7)Patients requiring epidural bolus, n (%)18 (48.6)6 (16.2) 0.006 Table 3Procedural information and intraoperative/PACU analgesia The data are presented as numbers (%), medians [IQRs], or means ± SDs.VariablesLM group (n = 38)USG group (n = 38)p value Sensory blockade after test dose 4.5 ± 2.55.6 ± 2.1 0.040 Epidural waveform grade, n 0.029 Strong/Moderate/Weak/Absent17/12/2/726/8/4/0 Mean epidural pressure (mmHg) 17 [12.5–22.5]15.5 [11–20]0.218Accidental dural puncture, n (%)0 (0)3 (7.9)0.240Hypotension requiring treatment, n (%)30 (78.9)35 (92.1)0.192 TEP by operator (PJ/SK/NK) 15/20/318/19/10.570 Thoracic level (T5–T8/T9–T12) 12/269/290.558 Epidural depth (cm) 6 [5–6]5 [5–6]0.098Difficulty of catheter insertion, n (%)0.309Some resistance but successful6 (16.2)3 (7.9)Easy to trace31 (83.8)35 (92.1)Sensory coverage in PACU, n (%) 0.012 No blockade7 (18.9)2 (5.4)Partial coverage14 (37.8)7 (18.9)Complete coverage16 (43.2)28 (75.7)Patients requiring epidural bolus, n (%)18 (48.6)6 (16.2) 0.006 Procedural information and intraoperative/PACU analgesia The data are presented as numbers (%), medians [IQRs], or means ± SDs. No significant differences were observed between the groups in the mean epidural pressure, or patient-reported discomfort during the procedure. Although the incidence of accidental dural puncture in USG was higher than conventional group but there was not statistically significant. Discussion This randomized comparative study demonstrated that real-time ultrasound guidance improved procedural efficiency in thoracic epidural placement in elderly patients undergoing major thoracic and abdominal surgery, particularly by increasing first-pass success and reducing needle manipulation. However, these technical advantages did not translate into statistically significant improvement in overall procedural success, under scoring the distinction between technical surrogate outcomes and clinically meaningful endpoints. In elderly patients, age-related anatomical changes-such as reduced interlaminar space and degenerative spinal alterations may contribute to procedural difficulty and variability in TEP outcomes20,21. Structural alterations such as ligamentum flavum calcification, osteophytes, and a reduced interlaminar space commonly complicate epidural needle advancement in elderly individuals20. These changes can obscure external anatomical landmarks and increase the likelihood of repeated needle attempts21. Real-time ultrasound guidance offers direct visualization of the interlaminar space and needle trajectory, which likely explains the marked reduction in needle redirections observed in this study. A greater proportion of complete postoperative wound coverage, together with stronger epidural pressure waveforms in the ultrasound group, suggest more accurate catheter placement and improved epidural spread. However, because the catheter position was not confirmed by epidurography or other imaging modalities in this study, these findings-along with the observed favorable clinical outcomes, should be interpreted cautiously. Nonetheless, they indicate that real-time ultrasound-guided thoracic epidural placement may enhance catheter positioning. This interpretation is supported by Kim et al.22 who performed real-time ultrasound-guided TEP at the T10-11 level and verified catheter location using fluoroscopy. They reported a 100% success rate, with all catheters correctly located within the epidural space. Furthermore, catheter tips were consistently positioned between targeted T9-10 vertebral level, with contrast demonstrating uniform cranial and caudal spread22. The improved first-pass success rate observed with ultrasound guidance is consistent with prior studies evaluating both preprocedural scanning and real-time ultrasound techniques16,23–28. Although the magnitude of benefit varies across studies—depending on operator expertise, patient complexity, and ultrasound technique—the overall evidence suggests that ultrasound is particularly advantageous in patients with challenging spinal anatomy, such as elderly individuals or those with degenerative spinal changes15,29–32. The reported first-pass and overall success rates from previous real-time ultrasound-guided thoracic epidural studies vary widely, ranging from approximately 60% to over 90%16,27,33,34. The success rate observed in our cohort falls within this expected range, particularly given the advanced age and anatomically challenging spinal characteristics of our study population. Notably, a recent study employing a real-time ultrasound-guided mid-thoracic approach using the paramedian sagittal oblique cross (PX) view demonstrated a remarkably high first-attempt success rate of 96% and an overall success rate of 100%27. This study found a relatively low first-pass success observed reflect the increased technical complexity associated with thoracic epidural placement in elderly individuals and the learning experience of the operators. Preparation time for real-time ultrasound-guided thoracic epidural placement was longer than that for the landmark-based technique, consistent with previous reports16,27. However, total procedure time did not differ significantly between the two approaches. In clinical practice, the increase preparation time is unlikely impact the operating room turn over, particularly when real-time ultrasound-guided TEP is performed in a dedicated procedure room outside the operating theater. This study demonstrated that the primary failure rate of landmark-based thoracic epidural placement was higher than that of real-time ultrasound guidance (18.4% vs. 5.2%). The observed primary failure incidence in the landmark-based group falls within the range reported in previous studies, which have documented failure rates up to 24%8,16,35. However, these reported rates vary considerably depending on the definition of failure, the technique used to confirm epidural catheter placement, operator experience, and the modalities employed to assist thoracic epidural placement36. In the present study, most catheters were successfully placed in situ; however, they were subsequently removed at the end of surgery due to the inability to achieve adequate sensory dermatome blockade. We did not perform further exploration or verification to determine whether the epidural catheters had been placed beyond the epidural space. The primary causes of failure are likely catheter misplacement or misidentification epidural space8. Notably, all seven failure cases in the landmark-based group demonstrated absent epidural waveform analysis, whereas the two failure cases in the real-time ultrasound-guided group exhibited weak and strong epidural waveform patterns, respectively. These findings suggest that epidural wave form analysis may serve as a reliable adjunct for identifying catheter misplacement and confirming correct epidural catheter position37,38. The use of LOR alone can produce false-positive results37,38; combining LOR with epidural pressure waveform analysis can enhance the accuracy of epidural space identification37. However, the presence of epidural pressure waveform analysis cannot exclude intravascular or intrathecal catheter insertion39. The real-time ultrasound-guided thoracic epidural placement is technically demanding procedure associated recognized learning curve40. The technical challenges inherent to this technique encompass several distinct skill domains: the ability to visualized the Touhy needle under real-time ultrasound guidance; the capacity to maintain a consistent needle trajectory during the transition from ultrasound guidance to LOR technique; the skill to advance the needle blindly while applying continuous pressure to LOR syringe assembly; and the tactile perception required to navigate the needle through the interlaminar space and identify the characteristic pressure change upon entry into the epidural space. Consequently, operator experience is a recognized determinant of procedural success and complication rates40. Proficiency in real-time ultrasound-guided neuraxial techniques requires the integration both visual and tactile perception of operators. Furthermore, the operator must maintain precise control of the Touhy needle angulation throughout the interlaminar space and ligamentum flavum. Operators experience also bears directly on complications avoidance. In particularly awareness that the needle should not be advanced more than approximately one centimeter beyond the interlaminar entry point is essential to prevent accidental dural puncture. Although all operators in this study had established proficiency in regional anesthesia and conventional thoracic epidural placement, variability in experience with real-time ultrasound-guided thoracic technique may have influenced procedural efficiency and complication rate. The learning curve specific to real-time ultrasound-guided thoracic epidural placement (TEP) has not yet been formally assessed. However, Weil G et al.41 evaluated the learning curves of 18 anesthesiology residents performing thoracic epidural analgesia using the learning curve cumulative sum (LC-CUSUM) method, with a predefined endpoint of 90% success rate. They reported that residents performed a median of 19 (range 5–41) procedures each, yet only two of eighteen residents achieved the target success rate within the study period. In a separate study42 examining ultrasound-guided caudal epidural block, residents were found to achieve procedural self-confidence after approximately 13 procedures. Taken together, these findings suggest that the learning curve for real-time ultrasound-guided TEP is likely to require a considerably greater number of procedures than those reported for either conventional thoracic epidural placement or simpler ultrasound-guided neuraxial techniques, given the additional complexity of integrating real-time sonographic guidance with the loss-of-resistance technique in the thoracic region. This constitutes an important confounding factor and may partly explain the observed variability in outcomes. Of note, the higher incidence of accidental dural puncture in the ultrasound group appeared predominantly during the learning curve phase of two less-experience operator; although this difference did not reach statistical significance, it is likely attributable to the limit exposure of these individuals to the real-time ultrasound-guided technique. This study has several limitations. First, its single-center design may limit the generalizability of finding, particularly to settings with varying levels of ultrasound expertise. Second, the sample size was powered to detect differences in primary outcome (First-pass success; FPS) rather than the clinical meaningful outcomes; the overall success rate, serious complications such as epidural hematoma or neurological injury. Hence to refer the difference of overall success rate and it’s complications are required to adjust. Third, blinding of operators and patients was not feasible due to the nature of the intervention which may introduce bias. Nevertheless, the objective outcome measures were employed wherever possible. Procedural outcomes were recorded by non-blinded staff who were not directly involved in the study, whereas PACU outcomes were assessed by blinded personnel. Fourth, the operator in this study had varying level of experience with real-time ultrasound guided thoracic epidural placement, which may represent a confounding factor, and limit the applicability of this finding to setting with a more homogenous operator skill set. Finally, confirmation of the epidural space relies on loss of resistance supplemented by epidural pressure waveform analysis37,38 rather than fluoroscopic or contrast-based confirmation, which may introduce variability in verifying the final catheter position. Conclusion Real-time ultrasound guidance significantly enhances the technical performance of thoracic epidural placement in elderly patients by improving first-pass success, reducing needle passes and skin puncture, and providing shorter needling time. Although this technique requires additional preparation time, the overall procedure duration remains similar to that of the landmark technique, with some potential clinical advantages including improving surgical wound coverage and epidural spreading of local anesthetic. Ultrasound guidance should therefore be considered an adjunct rather than a replacement for landmark-based thoracic epidural placement technique in elderly patients. Larger multicenter studies are warranted to validate these results and refine implementation and training strategies.