Reliability and validity of the total cerebral small vessel disease score: a systematic review and meta-analysis

Guilherme Diogo Silva^1*Joao Paulo Mota Telles¹Carolina De Medeiros Rimkus¹Germana Titoneli Vieira¹Emily Figueiredo Vieira Neves Yuki¹Raymundo Soares Azevedo²Gisela Tinone¹Leandro Tavares Lucato¹Rosa Maria Pereira²Adriana Bastos Conforto²

• ¹Hospital das Clinicas da FMUSP, SAO PAULO, Brazil
• ²Faculty of Medicine, University of São Paulo, São Paulo, São Paulo, Brazil

Cerebral small vessel disease (CSVD) is a research priority to reduce the burden of stroke and dementia. The total cerebral small vessel disease (tSVD) score provides a global view of CSVD burden combining lacunes of presumed vascular origin, cerebral microbleeds, enlarged perivascular spaces, and white matter hyperintensities of presumed vascular origin. While its use in research is expanding, a systematic review of the tSVD score's reliability and validity had not yet been undertaken. We reviewed the inter-rater and intra-rater reliability for the tSVD score and its features.We also examined the associations between the tSVD score and age, hypertension, stroke and cognitive impairment.We performed a systematic review of studies on Pubmed/MEDLINE, Embase, and Scopus databases from inception until June 21st, 2024. We included manuscripts that reported at least one of the following metrics for the tSVD or for its components: inter-rater reliability, intra-rater reliability, or associations with age, hypertension, stroke, and/or cognitive impairment. We provided summary Cohen's kappa coefficients for inter and intra-rater reliability for each feature of the tSVD score. Subgroup analysis and meta-regression models were used to evaluate the impact of raters, MRI fields, age, and median tSVD score values in inter-rater reliability. We summarized studies reporting associations between the tSVD score, stroke and cognitive impairment.The summary Cohen's kappa values for inter-rater reliability ranged from 0.79 to 0.82 for each CSVD feature (13 studies, 8177 participants). We found a high heterogeneity between studies (I 2 = 94%), which may be explained by differences in rater, age, and median tSVD score. The summary Cohen's kappa values for intra-rater reliability ranged from 0.78 to 0.84 (four studies, 250 cases were randomized from 3654 participants). Heterogeneity was low. Seven studies (6,022 participants) reported associations between tSVD scores and either age or hypertension. Fifteen studies (6,941 participants) reported associations between tSVD scores and either stroke or cognitive impairment.