Severe glenoid bone loss (SGBL) creates substantial technical difficulties. Alternative screw placement may be required for adequate glenoid implant fixation. While bone volumes for the spine and lateral pillars have been determined in the past, there is inadequate evidence addressing the distribution of screw placement for fixation in these areas in SGBL cases. This study aimed to see how much screw location varies. Researchers believed that understanding this variability and identifying common patterns of glenoid bone loss enabled preoperative planning, implant design, and implant selection recommendations. About 65 three-dimensional scapulae models demonstrating SGBL were identified when an internal registry of two high-volume shoulder and elbow surgeons was queried. Two 3.5mm x 30mm screws, one in the scapular spine (CS) bone volume and one in the inferior column (IS) bone volume were mimicked by a fellowship-trained shoulder and elbow surgeon. Using anatomical reference points, three orthogonal reference planes were created: the scapula trigonum, approximated glenoid center, and inferior pole.

The locations of the screws were plotted, and deviations from the reference planes were determined. The IS to CS mutual locations were also calculated. Nearly ten samples were chosen at random to test intraobserver reliability. Screw orientation distributions were given a median, 25th, and 75th percentile. For screw head positions, mean and standard deviations were recorded. Researchers were able to establish high intraobserver reliability (ICC 0.90-0.98). Around 50% of the CS had a retroversion of 10°±5° from the scapula plane and a 5°±5°  inclination. For IS, half of the participants were positioned 0°±4° from the scapula plane, with a slope of -33°±7°. In 49% of cases, the connection between the IS and the CS was medial and posterior, lateral and posterior in 45%, and lateral and anterior in 6%. The distance between the CS and IS heads was 25mm±4mm on average. For SGBL, sufficient glenoid implant fixation can be obtained with excellent repeatability by inserting screws in the spine and lateral columns. CS positioned -16° to -5° from the scapula plane, with 0° to 12° of inclination, and IS positioned -6° to 4° from the scapula plane, with -40° to -25° of inclination, should be accommodated in future implant designs. Furthermore, mutual screw positions showed anterior and inferior bone loss distributions. The potential benefits of augmentation to accommodate interscrew distances of 21-29mm and anatomic positions of the IS relative to the CS should be considered in future implant designs.