The purpose of this study was to evaluate predictors of fracture-induced humeral head ischemia. Between February 1998 and December 2001, 100 intracapsular fractures of the proximal humerus, treated by open surgery, were included in a prospective surgical evaluation protocol (mean age, 60 years; minimum, 21 years; maximum, 88 years; 45 men; 57 right shoulders). Fracture morphology was assessed following a structured questionnaire and based on radiographic and intraoperative findings. Perfusion was assessed intraoperatively by observation of backflow after a borehole was drilled into the central part of the head in all shoulders and by intraosseous laser Doppler flowmetry in 46. Good predictors of ischemia were the length of the metaphyseal head extension (accuracy, 0.84 for calcar segments < 8 mm), the integrity of the medial hinge (accuracy, 0.79 for disrupted hinge), and the basic fracture pattern (accuracy, 0.7 for combined types 2, 9, 10, 11, and 12). Moderate and poor predictors of ischemia were fractures consisting of four fragments (accuracy, 0.67), angular displacement of the head (accuracy, 0.62 for angulations over 45 degrees ), the amount of displacement of the tuberosities (accuracy, 0.61 for displacement over 10 mm), glenohumeral dislocation (accuracy, 0.49), head-split components (accuracy, 0.49), and fractures consisting of three fragments (accuracy, 0.38). When the above criteria (anatomic neck, short calcar, disrupted hinge) were combined, positive predictive values of up to 97% could be obtained. The most relevant predictors of ischemia were the length of the dorsomedial metaphyseal extension, the integrity of the medial hinge, and the basic fracture type determined with the binary description system.

Polls results

On a scale of 1 to 10, rate how much this article will change your clinical practice?

NO change
BIG change
84% Article relates to my practice (79/93)
9% Article does not relate to my practice (9/93)
5% Undecided (5/93)

Will this article lead to more cost-effective healthcare?

67% Yes (63/93)
23% No (22/93)
8% Undecided (8/93)

Was this article biased? (commercial or personal)

4% Yes (4/93)
91% No (85/93)
4% Undecided (4/93)

What level of evidence do you think this article is?

8% Level 1 (8/93)
32% Level 2 (30/93)
39% Level 3 (37/93)
16% Level 4 (15/93)
3% Level 5 (3/93)