We identified a total of 188 patients who underwent UISIH between August 2014 and November 2018. First, we excluded 105 patients who showed no evidence of partial round ligament tear or degeneration on pre-procedural MRI, including two cases of complete round ligament tear. Next, we excluded 19 patients who displayed additional MRI findings that could be responsible for hip pain, namely femoroacetabular impingement, calcific tendinitis, trochanteric bursitis, ischiofemoral impingement, or femoral head stress fracture. Then, seven patients who had been treated with repeated UISIH and one patient who had undergone total hip replacement arthroplasty were excluded. Thirteen patients who had no post-procedural records and three patients who were followed up for longer than 90 days after the procedure were also excluded. Ultimately, 40 patients who met all criteria were included in this study (Fig. 1). The study population comprised 19 men and 21 women (52.5%), with a mean age of 48 years (range, 20 to 76 years) and an average follow-up duration of 43 days (range, 15 to 90 days). Thirteen patients underwent steroid injection into the left hip joint and 27 into the right.

We retrospectively reviewed pre-procedural MRI records analyzed by six musculoskeletal radiologists with 2 to 17 years of experience to determine if patients displayed round ligament tears or degeneration. Two musculoskeletal radiologists then came to a consensus regarding the presence of round ligament tearing or degeneration as follows. Partial tears, degeneration, and complete tears were characterized by different MRI findings. A partial tear was defined by torn fibers or marginal irregularity accompanied by focal insufficient thickness of the ligament compared to the contralateral side [2]. Degeneration was defined by a marked increase in signal intensity on proton density-weighted fat-suppressed imaging and T2-weighted fat-suppressed imaging or swelling relative to the contralateral side [2]. A complete tear was defined by the complete absence of ligamentous material or a full-thickness disruption of the normal pathway between the fovea capitis and the transverse ligament [2,14]. Normal round ligaments were defined by the presence of intact fibers with normal origin and insertion, normal thickness, homogeneous hypointense signal intensity with all pulse sequences, and smooth margins. Fig. 2 shows an example MRI image from one patient each with round ligament degeneration, a partial tear, a complete tear, and normal findings. In addition, we reviewed causes of hip pain other than round ligament tear or degeneration on MRI to exclude the compounding factors of femoroacetabular impingement, calcific tendinitis, trochanteric bursitis, ischiofemoral impingement, and femoral head stress fracture [15].

Each patient lay on a bed in the supine position with the hip slightly abducted and the knee pointed forward. All UISIHs were conducted by an experienced musculoskeletal radiologist (Y.C.Y.) using an iU22 ultrasound scanner (Philips, Bothell, WA, USA) equipped with a C5-1 MHz curved array transducer. A survey scan was first performed to determine the entry point of the needle. First, to identify the femoral head, the transducer was placed on the inguinal area parallel to the inguinal ligament. Then, the probe was rotated internally with the probe marker aimed toward the umbilicus until the femoral head, neck, and anterior capsular recess aligned in a single plane. The groin area was sterilized with a cotton ball soaked with 2% chlorhexidine-alcohol and covered with a sterile cotton surgical drape centered on the planned needle entry point. The transducer was covered with a sterile ultrasound probe cover. With the probe held in an oblique longitudinal orientation so that the femoral head and neck aligned in a single plane, 2-3 mL of local anesthetic (lidocaine hydrochloride, 200 mg/20 mL; Daihan Pharmaceutical Co., Ltd., Seoul, Korea) was injected at the planned needle entry point to minimize the pain of spinal needle insertion. Next, the spinal needle (Quincke Bevel, 22-gauge 3-1/2 inch; TaeChang Industrial Co., Ltd., Gongju, Korea) was inserted using an in-plane approach to target the anterior capsular recess under real-time ultrasound guidance. When the needle tip was positioned inside the joint capsule, 1 mL of saline loaded in a 5-mL syringe was injected to confirm successful targeting of the joint space. Following this, 1 mL of triamcinolone acetonide (Tamceton, 40 mg/1 mL; HanAll Biopharma Co., Ltd., Seoul, Korea), 2 cc of 1% lidocaine, and 5 mL of normal saline (sodium chloride, 180 mg/20 mL; Huons Co., Ltd., Seongnam, Korea) were flushed through the extension tube. To reduce the possibility of chondrotoxicity by local anesthetics, which can be impacted by exposure time and concentration, we chose short-acting 1% lidocaine and diluted it with 5 mL of normal saline. The injection was considered a technical success when the hyperechoic material generated by the steroid particles was observed below the joint capsule on real-time ultrasound imaging conducted after the procedure. We also checked for the presence of immediate complications such as hematoma on the post-procedural scan (Fig. 3).

Short-term relief from symptoms was retrospectively evaluated at the first follow-up appointment after UISIH. One author reviewed patients’ electronic medical records to assess their subjective experience of symptoms, which were classified on a 5-point scale as follows: 0 (aggravated), 1 (unchanged), 2 (symptoms improved but cannot be ignored), 3 (greatly improved; negligible symptoms), and 4 (no residual symptoms). The use of a 5-point scale was based on an article by Lee et al. [16], and the scale was constructed in reference to the Mankoski Pain Scale (2000) [17]. A score of 0 (aggravated) was assigned if the patient described experiencing any aggravation of pain. A score of 1 (unchanged) was defined as including mentions of the pain being the same as before the procedure or unchanged from before the procedure. A score of 4 (no residual symptoms) was assigned based on mentions of a lack of problems with hip joint pain and motion. Of the patients not already classified as 0, 1, or 4, those who used positive language and whose descriptions of remaining symptoms indicated that they were mild (with word choice such as "little" or "negligible") were assigned a score of 3 (greatly improved; negligible symptoms). The remaining patients were scored as 2 (symptoms improved but cannot be ignored). If descriptions were ambiguous or a local dialect was used, two radiologists assigned a score through consensus. Patients who had a reduction of at least 2 out of 5 points were determined to have experienced symptom reduction. The same author also assessed the incidence of complications after the procedure.

The only continuous variable analyzed in this study was follow-up duration, and it was expressed as mean±standard deviation. Age was categorized into three groups: 20 to 39 years, 40 to 59 years, and 60 years and older. Categorical variables were expressed as frequencies and percentages. The relationship between symptom relief and follow-up duration was analyzed using the t test. The relationships between symptom relief and categorical variables such as age, sex, and injection side were analyzed using the chi-square test and the Fisher exact test. A P-value <0.05 was considered to indicate statistical significance. Statistical analysis was conducted using SAS version 9.4 (SAS Institute, Cary, NC, USA).