As a primary cause of postural and gait disorders, lateropulsion must be systematically assessed after stroke. The current gold standard presenting unsatisfactory metrological properties, we propose a new scale optimized through 4 methodological steps:

– conception based on scientific rationale, a critical review of the literature, and local testing of feasibility an relevance of items retained (V0);

– improvement and validation of the content through a Delphi process with a panel of international experts (V1);

– monocentric study to adjust the SCALA construction (V2);

– final multicentric validation (V3). Here we present the V1 of the SCAle for LAteropulsion (SCALA), obtained after 3 years of work.

The V0 locally conceived was submitted to a panel of 21 international experts having clinical and scientific background of lateropulsion. At each round, their degree of agreement/disagreement was analyzed on Likert scales from 1 to 9 for general procedure and each item (separately for instructions, procedure, scoring). In addition an in-depth analysis of comments served to guide improvements between rounds.

Two rounds were needed to reach consensus, the 3rd one serving to finalize small adjustments in the instructions, and to get final approval. The final degree of agreement was high: median 8 [Q1=8 and Q3=9]/9. The SCALA V1 is a matrix comprising 4 components (spontaneous lateral tilt, self-initiated lateral pushing, resistance to passive lateral tilting, and unawareness of the lateropulsion) and 10 different postural tasks (sitting, transfers, standing, walking, in different postural and visual conditions). Items are scored with a four-grade scoring, except for one component where the score is binary. The total score ranges from 0 to 50 points.

Strong consensus and enthusiastic comments prejudges the success of this promising tool, designed to detect and evaluate contraversive and ipsiversive post-stroke lateropulsion, regardless its mechanisms.