Although this may have helped keep costs down in one respect, it may have raised them in another because of the trade-off between length of stay and readmissions, they added.
All-cause 30-day readmissions rose from 4 percent to 5 percent for primary procedures and from 6 percent to 9 percent for revision procedures.
Readmissions for adverse events were relatively stable over the two decades studied for primary knee replacement but that was not the case for revisions.
Revision procedures were associated with a more than doubling in readmission rates for wound infection from 1 percent to 3 percent and a more than 100 percent increase in readmissions for hemorrhage, sepsis, and heart attack.
Patients who returned to the hospital soon after their procedure were older and more likely to be male, black, and sicker with comorbidities.
Notably, centers that did more knee replacements were associated with lower readmission rates for both primary and revision procedures.
A significant percentage of TKAs are performed by surgeons doing fewer than 12 cases a year, the editorialists noted.
"Therefore, careful consideration should be given to whether the majority of these cases should be shifted toward high-volume centers, which often have the infrastructure and the experience needed to develop the highly coordinated care pathways necessary to optimize the quality outcomes and efficiency of the episode of care for complex patients," they suggested.
The researchers acknowledged that their findings may not extrapolate to the 40 percent of the TKA population not under fee-for-service Medicare and noted the limitations of administrative data including lack of outcome data such as functional status and patient satisfaction.