Abstract Background and aims Diagnosing posterior circulation stroke in patients presenting with isolated acute dizziness remains clinically challenging. The HINTS Plus test is proposed to differentiate between central and peripheral causes of acute dizziness, showing high sensitivity and specificity in selected populations when performed by specialists. However, its performance in unselected populations and when applied by non-specialists remains uncertain. Methods The PROVIDE trial included 800 adults with acute continuous dizziness without focal neurological deficits. Nystagmus and gait imbalance were permitted, but not required for inclusion. Patients underwent neurological examination, HINTS Plus testing, and MRI with diffusion-weighted imaging. MRI scans were centrally reviewed. Functional outcome (modified Rankin Scale) and recurrent vascular events were assessed at 3-month follow-up. Final diagnoses were determined by an expert panel integrating clinical data, imaging findings, and follow-up information, serving as a construct reference standard. We assessed the diagnostic accuracy of the HINTS Plus test as performed by the initial treating physician. Results Compared with the expert panel diagnosis, the HINTS Plus test showed a sensitivity of 39%, specificity of 82%, PPV of 25%, NPV of 90%, LR+ of 2.11, and LR- of 0.75. Conclusions The real-world sensitivity of the HINTS Plus test is substantially lower than previously reported. This reduced performance is likely attributable to differences in patient population, performance by non-specialists, and potential overfitting of the test to its development cohort. Our study provides valuable insights from the Netherlands, likely representative of many European healthcare settings, and suggest that clinicians can not rely on the HINTS Plus test. Conflict of interest Merel Verhagen: nothing to disclose. Adriaan Ros: nothing to disclose. Vincent Klokman: nothing to disclose. Kim Jie: nothing to disclose. Meike Schoenmakers: nothing to disclose. Karlijn de Laat: nothing to disclose. Suzanne Cannegieter: nothing to disclose. Korné Jellema: nothing to disclose.
Verhagen et al. (Fri,) studied this question.