Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

Contrast-enhanced CT—Nephrotoxic or not? A critical review of propensity-score-adjusted studies

Nyman, Ulf LU ; Björk, Jonas LU orcid ; Hellström, Mikael ; Leander, Peter LU ; Liss, Per ; Sterner, Gunnar and Brismar, Torkel B. (2026) In Journal of Internal Medicine 300(1). p.5-25
Abstract

Retrospective, propensity-score-adjusted (PSA) studies have indicated to practitioners in the field that contrast-induced acute kidney injury (AKI) after contrast-enhanced CT (CECT) is of minor importance or even non-existent. This has led to controversies between clinicians demanding CECT not to miss any diagnoses and radiologists reluctant or even refusing CECT in risky patients. A critical review of AKI/dialysis risk post-CECT in PSA studies was performed based on a Cochrane Library/Embase/PubMed search up to December 2024 with stratification for various populations. Among unselected emergency-/inpatients with chronic kidney disease, CECT was associated with significantly higher AKI incidence than in controls in 8/10 cohorts... (More)

Retrospective, propensity-score-adjusted (PSA) studies have indicated to practitioners in the field that contrast-induced acute kidney injury (AKI) after contrast-enhanced CT (CECT) is of minor importance or even non-existent. This has led to controversies between clinicians demanding CECT not to miss any diagnoses and radiologists reluctant or even refusing CECT in risky patients. A critical review of AKI/dialysis risk post-CECT in PSA studies was performed based on a Cochrane Library/Embase/PubMed search up to December 2024 with stratification for various populations. Among unselected emergency-/inpatients with chronic kidney disease, CECT was associated with significantly higher AKI incidence than in controls in 8/10 cohorts (weighted absolute risk increase 4.0 percentage points (pp), cohort range −4 to 16.9 pp at glomerular filtration rate (GFR) <30 mL/min/1.73 m2) and dialysis in 5/6 cohorts (weighted absolute risk increase 1.8 pp, cohort range 1.0–5.9 pp at GFR <30 mL/min/1.73 m2). Corresponding figures regarding cohorts with presumed/potentially unstable renal function were 1/10 and 3/8 cohorts, respectively. No increased risk was found in the five miscellaneous cohorts, but with a few patients with low GFR. Associations between individual contrast medium (CM) doses and AKI/dialysis were generally lacking. Marked selection bias was evident in one cohort. Thus, the analyses support an association between CECT and AKI in emergency-/inpatients, including a non-negligible dialysis risk, especially at GFR <30 mL/min/1.73 m2. PSA studies reporting no increased risk of CECT must be interpreted with caution due to risk for selection bias and the potential masking effects of acute kidney functional recovery. Studies on CM dose or gram-iodine/GFR ratio correlation with AKI/dialysis are warranted.

(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
acute kidney injury, computed tomography, contrast media, glomerular filtration, meta-analysis, propensity score, rate
in
Journal of Internal Medicine
volume
300
issue
1
pages
5 - 25
publisher
Wiley-Blackwell
external identifiers
  • scopus:105036676455
  • pmid:42002788
ISSN
0954-6820
DOI
10.1111/joim.70095
language
English
LU publication?
yes
id
e7ec85b5-7fcb-43e0-883c-2cbbc1687f32
date added to LUP
2026-05-29 11:02:11
date last changed
2026-09-06 01:04:08
@article{e7ec85b5-7fcb-43e0-883c-2cbbc1687f32,
  abstract     = {{<p>Retrospective, propensity-score-adjusted (PSA) studies have indicated to practitioners in the field that contrast-induced acute kidney injury (AKI) after contrast-enhanced CT (CECT) is of minor importance or even non-existent. This has led to controversies between clinicians demanding CECT not to miss any diagnoses and radiologists reluctant or even refusing CECT in risky patients. A critical review of AKI/dialysis risk post-CECT in PSA studies was performed based on a Cochrane Library/Embase/PubMed search up to December 2024 with stratification for various populations. Among unselected emergency-/inpatients with chronic kidney disease, CECT was associated with significantly higher AKI incidence than in controls in 8/10 cohorts (weighted absolute risk increase 4.0 percentage points (pp), cohort range −4 to 16.9 pp at glomerular filtration rate (GFR) &lt;30 mL/min/1.73 m<sup>2</sup>) and dialysis in 5/6 cohorts (weighted absolute risk increase 1.8 pp, cohort range 1.0–5.9 pp at GFR &lt;30 mL/min/1.73 m<sup>2</sup>). Corresponding figures regarding cohorts with presumed/potentially unstable renal function were 1/10 and 3/8 cohorts, respectively. No increased risk was found in the five miscellaneous cohorts, but with a few patients with low GFR. Associations between individual contrast medium (CM) doses and AKI/dialysis were generally lacking. Marked selection bias was evident in one cohort. Thus, the analyses support an association between CECT and AKI in emergency-/inpatients, including a non-negligible dialysis risk, especially at GFR &lt;30 mL/min/1.73 m<sup>2</sup>. PSA studies reporting no increased risk of CECT must be interpreted with caution due to risk for selection bias and the potential masking effects of acute kidney functional recovery. Studies on CM dose or gram-iodine/GFR ratio correlation with AKI/dialysis are warranted.</p>}},
  author       = {{Nyman, Ulf and Björk, Jonas and Hellström, Mikael and Leander, Peter and Liss, Per and Sterner, Gunnar and Brismar, Torkel B.}},
  issn         = {{0954-6820}},
  keywords     = {{acute kidney injury; computed tomography; contrast media; glomerular filtration; meta-analysis; propensity score; rate}},
  language     = {{eng}},
  number       = {{1}},
  pages        = {{5--25}},
  publisher    = {{Wiley-Blackwell}},
  series       = {{Journal of Internal Medicine}},
  title        = {{Contrast-enhanced CT—Nephrotoxic or not? A critical review of propensity-score-adjusted studies}},
  url          = {{http://dx.doi.org/10.1111/joim.70095}},
  doi          = {{10.1111/joim.70095}},
  volume       = {{300}},
  year         = {{2026}},
}