Shoulder arthroplasty in complex acute and chronic proximal humeral fractures

From 1983 to 1988, 42 shoulder arthroplasties were performed on comminuted acute or chronic proximal humeral fractures. Patients were categorized according to the post-fracture operative delay; there were 15 four-part fractures, with median post-fracture delay of 13 days (range: 7 to 21), and 27 chronic fractures, including 11 four-part fractures, 9 three-part fractures, and 7 two-part fractures, with median post-fracture delay of 14 months (range: 4 to 72). Follow up was approximately 2 years in both groups (range: 1 to 5). All patients were evaluated according to a modified Neer score-system and classified into four groups. Pain relief was satisfactory in the acute group, but was unpredictable in the chronic group. The results in the acute group were significantly superior (P less than .05). In the acute group, 3 (20%) patients had an excellent result and 6 patients (40%) had a good result, compared to a good result obtained by 6 patients (22%) in the chronic group. Two patients (13%) in the acute group and 11 patients (40%) in the chronic group had a poor result. There were no statistical differences between two-part, three-part, or four-part fractures in the chronic group. Five cases of persistent instability were seen in shoulders formerly treated with osteosynthesis (one acute and four chronic cases). Two of these cases developed an infection. Good results can be expected after prosthetic replacement in acute proximal humeral fractures. Failed primary treatment reduces the possibility of a good result with revision arthroplasty.