Low revision rate despite poor functional outcome after stemmed hemiarthroplasty for acute proximal humeral fractures: 2,750 cases reported to the Danish Shoulder Arthroplasty Registry

Figures & data

Figure 1. Annual number of SHAs for acute proximal humeral fractures in Denmark from 2006 to 2015.

Table 1. Hierarchy of reasons for revision where more than one reason was reported

Hierarchy of reasons for revision
I Infection – an infection that requires revision of the arthroplasty
II Periprostethic fracture – fracture that requires revision of the arthroplasty
III Dislocation and instability
IV Loosening – loosening of any arthroplasty component
V Rotator cuff problem
VI Others – glenoid wear, biomechanical problems including overstuffing, and pain with no other complication

Table 2. Demographics, proportion of revisions, and mean WOOS score overall and for each age group

	Total	< 55 years	55–74 years	≥ 75 years
Sex, n (%)
Male	570 (21)	104 (59)	310 (24)	156 (12)
Female	2,180 (79)	73 (41)	1,002 (76)	1,105 (88)
Prosthesis brand, n (%)
Depuy Global FX	952 (34)	48 (27)	495 (38)	409 (32)
Zimmer Bigliani-Flatow	840 (31)	67 (38)	372 (28)	401 (32)
Tornier Aequalis	268 (10)	22 (12)	141 (11)	105 (8)
Biomet Comprehensive	245 (9)	15 (9)	98 (7)	132 (11)
Others	445 (16)	25 (14)	206 (16)	214 (17)
Revision, n (%)	101 (4)	9 (5)	70 (5)	22 (2)
WOOS
Complete, n (%)	1,525 (60)	87 (52)	780 (62)	658 (58)
Score, mean (SD)	55 (26)	53 (24)	54 (26)	57 (25)

Figure 2. Distribution of WOOS scores. Red lines signify a WOOS score of 30 and 50.

Table 3. Risk of revision. Cox regression model^a

Age	Univariate relative risk (95% CI)	Adjusted relative risk (95% CI)
≥ 75	1.0 (reference)	1.0 (reference)
55–75	2.9 (1.8–4.7)	2.9 (1.8–4.7)
< 55	2.7 (1.3–5.9)	3.3 (1.5–7.5)

a Sex, prosthesis brand (Depuy Global FX, Zimmer Bigliani-Flatow, Tornier Aequalis Fx, Biomet Comprehensive Fx) and year of surgery (2006–2007, 2008–2009, 2010–2011, 2012–2013, 2014–2015) were included in the multiple model and showed no statistical significance.

Table 4. Reasons for revision. Values are frequency (percent)

	Total	< 55 years	55–75 years	≥ 75 years
Dislocation	21 (0.8)	2	11	8
Loosening	2 (0.1)	0	2	0
Infection	15 (0.5)	1	11	3
Fracture	1 (0.0)	0	1	0
Rotator cuff failure	30 (1.1)	4	19	7
Others ^a	21 (0.8)	2	17	2
Missing	11 (0.4)	0	9	2
Total	101 (3.7)	9	70	22

a Including pain of unknown cause.

Figure 3. The 10-year cumulative implant survival rate and CI for patients younger than 55 years (blue), patients between 55 and 74 years (green), and patients older than 75 years (red) (p < 0.01).