Fluoroscopically Guided Lateral Approach Hip Injection

Methods

The institutional review board at the VA Ann Arbor Healthcare System reviewed and granted approval for this study. One of 5 interventional pain physician staff members at the VA Ann Arbor Healthcare System performed fluoroscopically guided hip injections. Interventional pain fellows under the direct supervision of board-certified physicians performed the procedures for the study cases. Supervising physicians included both physiatrists and anesthesiologists. Images were reviewed and evaluated without corresponding patient biographic data.

For cases using the lateral approach, the patients were positioned supine on the fluoroscopy table. In anterior-posterior and lateral views, trajectory lines are drawn using a long metal marking rod held adjacent to the patient. With pulsed low-dose fluoroscopy, transverse lines are drawn to identify midpoint of the femoral head in lateral view (Figure 1A, x-axis) and the most direct line from skin to lateral femoral head neck junction joint target (Figure 1B, z-axis). Also confirmed in lateral view, the z-axis marked line drawn on the skin is used to confirm that this transverse plane crosses the overlapping femoral heads (Figure 1A, y-axis).

The cross-section of these transverse and coronal plane lines identifies the starting point for the most direct approach from skin to injection target at femoral head-neck junction. Using the coaxial technique in the lateral view, the needle is introduced and advanced using intermittent fluoroscopic images to the lateral joint target. Continuing in this view, the interventionalist can ensure that advancing the needle to the osseous endpoint will place the tip at the midpoint of the femoral head at the target on the lateral surface, avoiding inadvertent advance of the needle anterior or posterior the femoral head. Final needle placement confirmation is then completed in antero-posterior view (Figure 2A). Contrast enhancement is used to confirm intra-articular spread (Figure 2B).

Cases included in the study were performed over an 8-month period in 2017. Case images recorded in IntelliSpace PACS Radiology software (Andover, MA) were included by creating a list of all cases performed and documented using the major joint injection procedure code. The cases reviewed began with the most recent cases. Two research team members (1 radiologist and 1 interventional pain physician) reviewed the series of saved images for each patient and the associated procedure report. The research team members documented and recorded de-identified study data in Microsoft Excel (Redmond, WA).

Imaging reports, using the saved images and the associated procedure report, were classified for technical approach (anterior, lateral, or inconclusive), success of joint injection as evidenced by appropriate contrast enhancement within the joint space (successful, unsuccessful, or incomplete images), documented use of sedation (yes, no), patient positioning (supine, prone), radiation exposure dose, radiation exposure time, and additional comments, such as “notable pannus” or “hardware present” to annotate significant findings on imaging review.

Statistical Analysis

The distribution of 2 outcomes used to compare rates of complication, radiation dose, and exposure time was checked using the Shapiro-Wilk test. Power analysis determined that inclusion of 30 anterior and 30 lateral cases results in adequate power to detect a 1-point mean difference, assuming a standard deviation of 1.5 in each group. Both radiation dose and exposure time were found to be nonnormally distributed (W = 0.65, P < .001; W = 0.86, P < .001; respectively). Median and interquartile range (IQR) of dose and time in seconds for anterior and lateral approaches were computed. Median differences in radiation dose and exposure time between anterior and lateral approaches were assessed with the k-sample test of equality of medians. All analyses were conducted using Stata Version 14.1 (College Station, TX).