Translation of the Fugl-Meyer assessment into Romanian: Transcultural and semantic-linguistic adaptations and clinical validation.

Purpose: The Fugl-Meyer Assessment (FMA) scale, which is widely used and highly recommended, is an appropriate tool for evaluating poststroke sensorimotor and other possible somatic deficits. It is also well-suited for capturing a dynamic rehabilitation process. The aim of this study was to first translate the entire sensorimotor FMA scale into Romanian using the transcultural and semantic-linguistic adaptations of its official afferent protocols and to validate it using the preliminary clinical evaluation of inter- and intra-rater reliability and relevant concurrent validity.

Methods: Through three main steps, we completed a standardized procedure for translating FMA's official afferent evaluation protocols into Romanian and their transcultural and semantic-linguistic adaptation for both the upper and lower extremities. For relevant clinical validation, we evaluated 10 patients after a stroke two times: on days 1 and 2. All patients were evaluated simultaneously by two kinesi-physiotherapists (generically referred to as KFT1 and KFT2) over the course of 2 consecutive days, taking turns in the roles of an examiner and observer, and (inter-rater). Two scores were therefore obtained and compared for the same patient, i.e., being afferent to an inter-rater assay by comparing the assessment outcomes obtained by the two kinesi-physiotherapists, in between, and respectively, to the intra-rater assay: based on the evaluations of the same kinesi-physiotherapist, in two consecutive days, using a rank-based method (Svensson) for statistical analysis. We also compared our final Romanian version of FMA's official protocols for concurrent validity (Spearman's rank correlation statistical method) to both of the widely available assessment instruments: the Barthel Index (BI) and the modified Rankin scale (mRS).

Results: Svensson's method confirmed overall good inter- and intra-rater results for the main parts of the final Romanian version of FMA's evaluation protocols, regarding the percentage of agreement (≥80% on average) and for disagreement: relative position [RP; values outside the interval of (-0.1, 0.1) in only two measurements out of the 56 comparisons we did], relative concentration [RC; values outside the interval of (-0.1, 0.1) in only nine measurements out of the same 56 comparisons done], and relative rank variation [RV; all values within an interval of (0, 0.1) in only five measurements out of the 56 comparisons done]. High correlation values were obtained between the final Romanian version of FMA's evaluation protocols and the BI (ρ = 0.9167; = 0.0002) for FMA-upper extremity (FMA-UE) total A-D (motor function) with ρ = 0.6319 and for FMA-lower extremity (FMA-LE) total E-F (motor function) with = 0.0499, and close to the limit, with the mRS (ρ = -0.5937; = 0.0704) for FMA-UE total A-D (motor function) and (ρ = -0.6615; = 0.0372) for FMA-LE total E-F (motor function).

Conclusions: The final Romanian version of FMA's official evaluation protocols showed good preliminary reliability and validity, which could be thus recommended for use and expected to help improve the standardization of this assessment scale for patients after a stroke in Romania. Furthermore, this endeavor could be added to similar international translation and cross-cultural adaptations, thereby facilitating a more appropriate comparison of the evaluation and outcomes in the management of stroke worldwide.