High risk for revision after shoulder arthroplasty for failed osteosynthesis of proximal humeral fractures

Figures & data

Figure 1. Flowchart of included cases. 299 were reported to the registry, 285 were included in this study.

Table 1. Type of osteosynthesis in 285 cases

	n	%
Locking plate	208	73.0
K-wires	17	6.0
Non-locking plate	15	5.3
Screws only	10	3.5
Intramedullary nail	9	3.2
Helix wire	4	1.4
Intramedullary nail (thin)	3	1.1
Missing	19	6.7

Table 2. Mode of osteosynthesis failure and outcome

		Mean	Revisions
Mode	n (%)	WOOS	n (%)
Avascular necrosis	107 (38)	46	9 (29)
Fracture displacement	49 (17)	45	6 (19)
Pseudarthrosis/non-union	36 (13)	49	7 (23)
Secondary arthrosis	12 (4)	46	2 (6)
Infection	9 (3)	32	3 (10)
Screw penetration	9 (3)	51	0 (0)
Pain	8 (3)	32	1 (3)
Rotator cuff tear	6 (2)	48	0 (0)
Implant failure	6 (2)	50	0 (0)
New fracture	2 (1)	41	0 (0)
Missing	41 (14)	47	3 (10)

Table 3. Causes of revision in cases (secondary arthroplasties) and controls (primary arthroplasties)

	Cases	Controls
Cause	n = 31	n = 34
Dislocation	8	8
Rotator cuff tears	5	10
Infection	5	4
Pain	5	1
Other	2	5
Glenoid erosion	1	3
Technical failure	2	1
Loosening	1	1
Missing	2	1

Figure 2. Implant survival functions of primary arthroplasty (green) and arthroplasty after failed osteosynthesis (blue).

Figure 3. Implant survival functions of primary arthroplasty (green) and arthroplasty after failed locking plates (blue).

Table 4. Mode of failure of locking plates and outcome

		Mean	Revisions
Mode	n (%)	WOOS	n (%)
Avascular necrosis	82 (39)	43	6 (33)
Fracture displacement	41 (20)	42	6 (33)
Pseudarthrosis/non-union	27 (13)	53	4 (22)
Secondary arthrosis	4 (2)	22	0 (0)
Infection	5 (2)	30	1 (6)
Screw penetration	8 (4)	45	0 (0)
Pain	8 (4)	30	1 (6)
Rotator cuff tear	4 (2)	52	0 (0)
Implant failure	6 (3)	46	0 (0)
New fracture	0 (0)	–	0 (0)
Missing	23 (11)	47	0 (0)