Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

The risk of revision using tourniquet or not in primary total knee replacement : an observational study from the Swedish Knee Arthroplasty Register

W-Dahl, Annette LU ; Kärrholm, Johan ; Arani, Perna Ighani and Rolfson, Ola (2026) In Acta Orthopaedica 97. p.53-59
Abstract

Background and purpose —The use of a tourniquet in knee replacement surgery is debated. Given the conflicting evidence, we aimed to compare the risk of revision after total knee replacements (TKR) with or without the use of tourniquet. Methods — In this register based observational study, we included the 5 most common cemented primary TKR models due to osteoarthritis reported to the Swedish Arthroplasty Register 2010–2024 and followed them until December 31, 2024. The first revision for implant loosening was the primary outcome. We estimated the cumulative revision rate (CRR) with 95% confidence interval (CI) using the 1–Kaplan–Meier method. We examined the use of a tourniquet regarding the risk of revision using multiple Cox regression... (More)

Background and purpose —The use of a tourniquet in knee replacement surgery is debated. Given the conflicting evidence, we aimed to compare the risk of revision after total knee replacements (TKR) with or without the use of tourniquet. Methods — In this register based observational study, we included the 5 most common cemented primary TKR models due to osteoarthritis reported to the Swedish Arthroplasty Register 2010–2024 and followed them until December 31, 2024. The first revision for implant loosening was the primary outcome. We estimated the cumulative revision rate (CRR) with 95% confidence interval (CI) using the 1–Kaplan–Meier method. We examined the use of a tourniquet regarding the risk of revision using multiple Cox regression analysis to calculate the hazard ratio (HR) with CI and adjusted for potential confounding factors. Results — Of the 149,616 TKRs included, 65,570 (44%) were with tourniquet and 84,046 (56%) without tourniquet. The CRR was similar at all time-points for all causes and infection; however, CRR started to increase at 6–7 years for implant loosening with use of a tourniquet. In the Cox regression analysis, the use of a. tourniquet was associated with an increased risk of revision for implant loosening after 5 years (HR 1.56, CI 1.06–2.30). There was no difference in revision for all causes (HR 1.07, CI 0.99–1.15) or infection (HR 1.08, CI 0.97–1.21).Conclusion — The use of a tourniquet was associated with an increased risk of revision for implant loosening after 5 years, while no association was found for all-cause revision or infection. Our results do not support the use of a tourniquet in TKR as a strategy to reduce the risk of revision, either due to all causes, implant loosening, or infection.

(Less)
Please use this url to cite or link to this publication:
author
; ; and
organization
publishing date
type
Contribution to journal
publication status
published
subject
keywords
Arthroplasty, Knee, Osteoarthrosis
in
Acta Orthopaedica
volume
97
pages
7 pages
publisher
Taylor & Francis
external identifiers
  • pmid:41575540
  • scopus:105029273150
ISSN
1745-3674
DOI
10.2340/17453674.2026.45363
language
English
LU publication?
yes
id
3ea303c0-af57-420c-a9bf-c78fa3e183e4
date added to LUP
2026-02-20 13:46:04
date last changed
2026-02-21 03:00:08
@article{3ea303c0-af57-420c-a9bf-c78fa3e183e4,
  abstract     = {{<p>Background and purpose —The use of a tourniquet in knee replacement surgery is debated. Given the conflicting evidence, we aimed to compare the risk of revision after total knee replacements (TKR) with or without the use of tourniquet. Methods — In this register based observational study, we included the 5 most common cemented primary TKR models due to osteoarthritis reported to the Swedish Arthroplasty Register 2010–2024 and followed them until December 31, 2024. The first revision for implant loosening was the primary outcome. We estimated the cumulative revision rate (CRR) with 95% confidence interval (CI) using the 1–Kaplan–Meier method. We examined the use of a tourniquet regarding the risk of revision using multiple Cox regression analysis to calculate the hazard ratio (HR) with CI and adjusted for potential confounding factors. Results — Of the 149,616 TKRs included, 65,570 (44%) were with tourniquet and 84,046 (56%) without tourniquet. The CRR was similar at all time-points for all causes and infection; however, CRR started to increase at 6–7 years for implant loosening with use of a tourniquet. In the Cox regression analysis, the use of a. tourniquet was associated with an increased risk of revision for implant loosening after 5 years (HR 1.56, CI 1.06–2.30). There was no difference in revision for all causes (HR 1.07, CI 0.99–1.15) or infection (HR 1.08, CI 0.97–1.21).Conclusion — The use of a tourniquet was associated with an increased risk of revision for implant loosening after 5 years, while no association was found for all-cause revision or infection. Our results do not support the use of a tourniquet in TKR as a strategy to reduce the risk of revision, either due to all causes, implant loosening, or infection.</p>}},
  author       = {{W-Dahl, Annette and Kärrholm, Johan and Arani, Perna Ighani and Rolfson, Ola}},
  issn         = {{1745-3674}},
  keywords     = {{Arthroplasty; Knee; Osteoarthrosis}},
  language     = {{eng}},
  pages        = {{53--59}},
  publisher    = {{Taylor & Francis}},
  series       = {{Acta Orthopaedica}},
  title        = {{The risk of revision using tourniquet or not in primary total knee replacement : an observational study from the Swedish Knee Arthroplasty Register}},
  url          = {{http://dx.doi.org/10.2340/17453674.2026.45363}},
  doi          = {{10.2340/17453674.2026.45363}},
  volume       = {{97}},
  year         = {{2026}},
}