Accuracy of video head impulse test in stroke diagnosis during acute vestibular syndrome: a systematic review with meta-analysis

Background: Vertigo is a common symptom in people with acute stroke, leading to difficulty in the diagnostic process. This study aims to systematically report the accuracy of the video head impulse test (vHIT) for diagnosing stroke in individuals presenting to emergency departments with acute vertigo. Methods: Data from MEDLINE, CENTRAL, CINAHL, SCOPUS, and gray literature were searched from inception to July 2024 without language limits. Independent researchers created a contingency table to record the number of true positives, false positives, true negatives, and false negatives. A bivariate analysis was performed to estimate the summary sensitivity and specificity. Subgroup and sensitivity analyses were conducted to explore the source of heterogeneity. A quality assessment was performed using the Quality Assessment of Diagnostic Test Accuracy Studies 2 tool. Results: 13 studies (905 participants) were included. EyeSeeCam (6/13) and Otometrics (6/13) were the most represented vHIT devices. The studies were of good quality, with the flow and timing being the more prevalent domain presenting a high risk of bias (7/13); however, there were no concerns about applicability. The vHIT had a specificity of 85% (95% CI, 61%–95%), sensitivity of 84% (95% CI, 70%–92%), positive likelihood ratio (LR) of 5.56 (95% CI, 2.00–15.47), negative LR of 0.18 (95% CI, 0.10–0.35), and diagnostic odd ratio of 30.23 (95% CI, 8.96–102.06), making this test particularly useful for ruling out strokes. Conclusion: This review highlights the value of vHIT as a complementary tool in early diagnosis of stroke. Thus, expertise in its use and interpretation should be promoted, and future primary studies are expected to refine and update this systematic review. Prospero registration CRD42024532776.