Comparison of complication types and rates associated with anatomic and reverse total shoulder arthroplasty
These authors sought to determine the complications following anatomic (aTSA) and reverse (rTSA) Equinoxe; Exactech arthroplasty.
A total of 2224 aTSA (male-female, 1090:1134) and 4158 rTSA (male-female, 1478:2680) patients were enrolled in an international database of primary shoulder arthroplasty performed by 40 different surgeons in the United States and Europe.
The mean age of the aTSA patients at the time of surgery was 66 years with a mean follow-up of 34 months.
The mean age of the rTSA patients at the time of surgery was 72 years with a mean follow-up of 22 months.
For the aTSA patients, 239 individual adverse events and 124 revisions were reported. The most common complications for aTSA were
(a) rotator cuff tear/subscapularis failure (n = 69; 42 were revised),
(b) aseptic glenoid loosening (n 55; 43 were revised), and
(c) infection (n 28; 18 were revised).
For the rTSA patients, 372 adverse events and 104 revisions were reported. The most common complications for rTSA were
(a) acromial/scapular fracture/pain (n 102; no revisions were reported),
(b) instability (n 60; 40 were revised, and
(c) pain (n 49; 7 were revised).
Comment: This is an informative high volume case study that highlight the different complications experienced with the two different types of prostheses. While it may be tempting to conclude that aTSA is associated with a significantly greater complication and revision rate than rTSA, this may not be the case. When adjusted for the length of followup, the complication rates were 3.8%/year for the aTSA and 4.9%/year for the rTSA. When adjusted for the length of followup, the revision rates were 2.0%/year for the aTSA and 1.4%/year for the rTSA. Furthermore the complication and revision rates were not adjusted for patient age and diagnosis.
It is of note that for the rTSA acromial and spine fractures were more common that instability - this rank order is in contrast to that of many other series.
It would be of interest to know how the authors considered cases in which two complications occurred in the same shoulder, for example it is recognized that rotator cuff/subscapularis failure can lead to glenoid loosening. Were the 69 shoulders with cuff/subscapularis failure different than the 55 shoulders with glenoid loosening? Note that the revision rates were 1.9% for cuff/subscap failure and for glenoid loosening - were these revisions in the same patients?
Perhaps it would be a more meaningful comparison to compare the average complication rate per person to avoid double or triple counting patients that had more than one complication.
