Imaging-based assessment of fracture stability does not reliably predict outcomes in patients with two-part proximal humeral fractures and may lead to unnecessary surgeries.

Two-part surgical neck fractures are the most common displaced proximal humeral fractures in the elderly. Most fractures can be categorized into varus-impacted or medially translated fracture patterns. The natural healing process often involves secondary displacement and partial resorption of the humeral head, but these changes are poorly correlated with shoulder function and patient satisfaction. Randomized trials have been unable to identify any benefits from surgery but report a high proportion of implant-related complications. In large prospective cohort studies, patients aged 60 or older with two-part surgical neck fractures treated non-operatively report shoulder function and quality of life close to the background population six months post-injury. Evidence-based and eminence-based approaches to interventions for osteoporotic proximal humeral fractures appear to collide.