# Low all‐cause 90‐day mortality after revision knee arthroplasty: A study based on the Swedish Perioperative Register (SPOR), 2017–2022

## TL;DR

This study found that 90-day mortality after knee revision surgery in Sweden is low and stable, with older age being the main risk factor.

## Contribution

The study provides the first detailed analysis of early mortality after revision knee arthroplasty in Sweden using national register data.

## Key findings

- Ninety-day mortality after revision knee arthroplasty was 0.57% and remained stable from 2017 to 2022.
- Age over 80 years was the only significant predictor of 90-day mortality after adjustment for other factors.
- Advanced age, higher ASA class, and revision due to infection were dominant risk factors for mortality.

## Abstract

Revision knee arthroplasty (rKA) is performed when a primary implant fails and is associated with a higher perioperative risk than primary arthroplasty. Early postoperative mortality after rKA in Sweden is poorly described. We aimed to estimate all‐cause mortality within 90 days after rKA in Sweden and to explore associations with patient characteristics and calendar year.

Register‐based cohort study using data from the Swedish Perioperative Register (SPOR). All adults undergoing rKA (national classification of surgical procedures and the kod for knee revision atrhroplasty) during 2017–2022 were included (n = 2616). The primary outcome was all‐cause mortality within 90 days. Group comparisons used χ
2 tests and analysis of variance; multivariable logistic regression adjusted for calendar year, sex, age group, American Society of Anaesthesiologists (ASA) category and indication for revision.

Overall, 90‐day mortality was low (15/2616; 0.57%; 95% confidence interval [CI]: 0.34–0.92) with no significant variation by calendar year (p = 0.100). Thirty‐day mortality was 0.23% (6/2616; 95% CI: 0.10–0.47). In unadjusted analyses, age >80 years and ASA Class III/IV were associated with higher odds of 90‐day mortality; in adjusted analyses, age >80 years remained the only statistically significant predictor (adjusted odds ratio [OR] 13.63; 95% CI: 3.97–46.77). Sex, ASA class and indication were not statistically significant in the adjusted model. Age, ASA class and indication showed substantial interaction in relation to 90‐day mortality.

Ninety‐day mortality after rKA in Sweden was low and remained stable between 2017 and 2022. Advanced age, higher ASA class and revision due to periprosthetic joint infection appeared to be the dominant risk factors. Preoperative preparation, medical optimization and, where feasible, prehabilitation of older and frail patients should be considered.

Not applicable, quality register‐based study.