The Effectiveness of Semantic Feature Analysis for Speech and Language Ability in Aphasia Patient: Systematic Review

Abstract

Background: Stroke patients have a high incidence. Aphasia is a co-occurring condition in approximately 25-50% of stroke patients. Patients with anomic aphasia present impairments in their semantic network. Semantic feature analysis (SFA) is a technique that improves the connections in the semantic system by adjusting the deficit. However, due to various factors in studies, such as differences in patient groups, comparisons of different techniques, study durations, participant, or rehabilitation procedures, it is necessary to investigate the patterns and effectiveness of SFA. Objective: The primary objective is to conduct a systematic review of the effectiveness of language and speech in stroke patients with aphasia, contrasting the SFA with other approaches. Method: A comprehensive search was conducted across MEDLINE, CENTRAL, and Embase up to September 27^th, 2023. Eligible studies included randomized controlled trials (RCTs), quasi-experimental studies, or clinical trials. Risk of bias was evaluated using the Cochrane Collaboration's tool, and data synthesis was performed using Review Manager (RevMan) version 5.4. Result: Out of 60 relevant studies identified, 6 met the inclusion criteria, and 2 were suitable for meta-analysis Data was synthesized with Review Manager (RevMan) version 5.4. The meta-analysis focused specifically on naming tasks and indicated no statistically significant difference in outcomes between the SFA and other methods. This applied both immediately post-training (SMD = 0.18, 95%CI: –0.19, 0.55, p = .34) and at the 3-month follow-up (SMD = 0.36, 95%CI: –0.01, 0.72, p = .06). The I² value of 0% indicates no inconsistency between the studies analyzed. Conclusion: SFA effectively improves language and speech skills in stroke patients with aphasia, yielding outcomes comparable to other speech-language therapy techniques. Professionals can confidently utilize SFA to support patients' naming abilities. However, due to limitations regarding the risk of bias in the included literature, further high-quality studies are required.