A systematic review and individual patient data meta-analysis of heart failure as a rare complication of traumatic arteriovenous fistulas

Florian A. Wenzl, Stephanie S. Miljkovic, Parinaz J. Dabestani, John J. Kessler, Tharun R. Kotaru, Louay D. Kalamchi, Sarah J. Aurit, Juan A. Asensio

Research output: Contribution to journalReview articlepeer-review

10 Scopus citations


Objective: Traumatic arteriovenous fistulas (AVFs) are rare. The vast majority occur secondary to penetrating injuries. High-output cardiac failure is a well-recognized serious complication of AVFs, associated with high morbidity and mortality. The objective of the present study was to identify predictors of heart failure (HF) in patients with traumatic AVF. Methods: Both PubMed/MEDLINE (Ovid) and CINAHL were searched (up to June 2019) for studies reporting individual patient data on the clinical and demographic characteristics of patients with AVF secondary to penetrating trauma. Exclusion criteria were age <18 years, no specification of symptoms, a cranial, spinal, or cardiac AVF location, and an iatrogenic mechanism of injury. The present study was performed in accordance with the PRISMA (Preferred Reporting Items for Systematic reviews and Meta-Analyses) guidelines. Results: A total of 274 AVF patients from 15 case series and 177 case reports were included. The median age at presentation was 32 years (interquartile range, 24-43 years), 90% were men. The most frequent mechanisms of injury were stab wounds (43%) and gunshot wounds (32%). The AVF location was the abdomen (n = 86; 31%), lower limb (n = 79; 29%), neck (n = 61; 22%), thorax (n = 38; 14%), and upper limb (n = 10; 4%). Of the 274 patients, 35 (13%) had presented with HF and 239 (87%) with other symptoms. The risk of HF increased with an increased feeding artery diameter (P < .001). On univariate analysis, HF was significantly associated with a longer median time from injury to presentation with AVF (11.2 years vs 0.1 years; P < .001), older median age at presentation (43 years vs 31 years; P = .002), involvement of a large feeding artery (ie, aorta, pulmonary artery, subclavian artery, external iliac artery; 40% vs 13%; P < .001), shrapnel injuries (11% vs 2%; P = .011), and injuries to the trunk or lower limb (94% vs 71%; P = .004). After adjusting for clinical and demographic patient characteristics, involvement of a large feeding artery (odds ratio, 3.25; 95% confidence interval, 1.26-8.42; P = .015) and every 6 years of delay to presentation (odds ratio, 1.30; 95% confidence interval, 1.03-1.63; P = .026) remained independent predictors for HF. Conclusions: HF occurs in a small but important fraction of traumatic AVF patients and develops after highly variable latency periods. Large feeding arteries and delayed presentation independently predicted HF in this cohort.

Original languageEnglish (US)
Pages (from-to)1087-1094.e8
JournalJournal of Vascular Surgery
Issue number3
StatePublished - Mar 2021

All Science Journal Classification (ASJC) codes

  • Surgery
  • Cardiology and Cardiovascular Medicine


Dive into the research topics of 'A systematic review and individual patient data meta-analysis of heart failure as a rare complication of traumatic arteriovenous fistulas'. Together they form a unique fingerprint.

Cite this