Diagnostic accuracy of pulse pressure variation, stroke volume variation and plethysmography variability index for prediction of fluid responsiveness in anesthetized and mechanically ventilated dogs: a systematic review and meta-analysis

Abstract

Objective: To assess the diagnostic accuracy of pulse pressure variation (PPV), plethysmographic variability index (PVI) and stroke volume variation (SVV) for predicting fluid responsiveness in dogs.

Study design: Systematic review and meta-analysis of diagnostic test accuracy studies.

Animals: Anesthetized dogs undergoing surgery, in critical care, or in experimental settings.

Methods: The study protocol was registered in Open Science Framework (https://doi.org/10.17605/OSF.IO/4DBGP). Databases (Medline, Embase, LILACS, Web of Science and Scopus) were searched from inception to 14 April 2025. Study quality was assessed using standardized criteria. Pooled sensitivity and specificity were calculated using a bivariate random-effects model. Likelihood ratios (LR), diagnostic odds ratios (DOR) and summary receiver operating characteristic (sROC) curves were generated. Confidence intervals (CIs) were calculated.

Results: Risk of bias was high in all studies of PPV, PVI and SVV. In four clinical PPV studies, sensitivity was 0.88 (95% CI: 0.65-0.97), specificity 0.97 (95% CI: 0.83-0.99), LR+ 30.99 (95% CI: 5.25-182.85), LR- 0.11 (95% CI: 0.02-0.40) and DOR 263 (95% CI: 26-1890). In five experimental PPV studies, sensitivity was 0.78 (95% CI: 0.64-0.87), specificity 0.83 (95% CI: 0.70-0.91), LR+ 4.63 (95% CI: 2.41-8.88), LR- 0.27 (95% CI: 0.16-0.46) and DOR 17 (95% CI: 6-48). For PVI (five studies), sensitivity was 0.83 (95% CI: 0.75-0.89), specificity 0.85 (95% CI: 0.78-0.90) and DOR 29 (95% CI: 13-64). For SVV (10 studies), sensitivity was 0.77 (95% CI: 0.69-0.84), specificity 0.81 (95% CI: 0.68-0.89) and DOR 14 (95% CI: 6-32). The sROC curve for clinical PPV showed wide prediction and CI.

Conclusions and clinical relevance: PPV, PVI and SVV show moderate to high diagnostic accuracy for predicting fluid responsiveness in dogs, with clinical PPV performing best. However, findings are limited by low precision and high risk of bias.

Keywords: PVI; PVV; QUADAS; SVV; canine; fluid responsiveness.