Humeral Shaft Nonunion

Excluding infection properly. "Exclude infection" is the instruction; these are the specifics, and the shoulder girdle has its own trap.

• Normal inflammatory markers do not exclude it. CRP and ESR are insensitive for the low-grade infection that causes a nonunion, so a normal CRP in a patient with previous surgery is reassurance you have not earned. Treat any nonunion after prior internal fixation as infected until proven otherwise.
• Take multiple separate deep cultures at surgery - conventionally five or more, each with fresh instruments and sent separately - because a single positive is usually contamination while two or more matching isolates indicate genuine infection.
• Ask the laboratory to hold cultures for 14 days. This is the point specific to the humerus and shoulder: Cutibacterium acnes is the region's characteristic low-virulence organism, it grows slowly, and a standard 48-hour culture will miss it. Failing to request extended culture is how a "presumed aseptic" nonunion gets plated over an occult infection.
• Expect some unexpected positives. A proportion of clinically aseptic nonunions grow an organism on intra-operative culture, which is why samples are taken even when infection is not suspected - the result changes antibiotic management afterwards even if the reconstruction proceeds.
• Send tissue for histology alongside culture where suspicion is real.

The metabolic screen the "optimise patient factors" line refers to. Where a nonunion is genuinely unexplained, the yield of screening is high enough that it is negligent to skip it. In the series that established this, 84% of carefully selected patients had at least one previously undiagnosed metabolic or endocrine abnormality - most often vitamin D deficiency (68%), along with calcium imbalance, central hypogonadism, and thyroid and parathyroid disorders. Most strikingly, eight patients united with medical treatment alone and no further surgery.

Read that number carefully, because it is easy to misquote: the 84% applies to selected patients meeting specific criteria - only 37 of 683 consecutive nonunions were referred - not to nonunions in general. The criteria are what to remember:

The practical screen is vitamin D and calcium, parathyroid hormone, thyroid function, and testosterone in men, with endocrinology referral rather than self-management when something is abnormal. A transverse humeral shaft fracture that was distracted in a brace is a mechanical explanation and does not need this; a well-fixed, well-reduced humerus that will not unite does.

The biology is fine - provide rigid stability and compression (compression plating). Bone graft is often not essential but may be added. Union is highly reliable once stability is restored.

Add biology: thorough debridement of avascular ends, autogenous bone graft (iliac crest), and consider biological adjuncts (e.g. BMP) for large/recalcitrant defects; structural graft or bone-transport/Masquelet techniques for segmental loss.

Eradicate the infection: debride, obtain cultures, culture-directed antibiotics, and often a staged approach (debridement/antibiotic cement spacer, then reconstruction); stabilise (plate or, if needed, external fixation) and graft once the infection is controlled.

Convert to compression plating + bone graft - do NOT rely on exchange nailing, which is unreliable in the humerus. Remove the nail, freshen the nonunion, compress and plate, and graft.