University of Pittsburgh, PA.
VA Pittsburgh Healthcare System, PA.
Am J Speech Lang Pathol. 2022 Oct 25;31(5S):2366-2377. doi: 10.1044/2021_AJSLP-21-00296. Epub 2022 Mar 15.
Specifying the active ingredients in aphasia interventions can inform treatment theory and improve clinical implementation. This secondary analysis examined three practice-related predictors of treatment response in semantic feature verification (SFV) treatment. We hypothesized that (a) successful feature verification practice would be associated with naming outcomes if SFV operates similarly to standard feature generation semantic feature analysis and (b) successful retrieval practice would be associated with naming outcomes for treated, but not semantically related, untreated words if SFV operates via a retrieval practice-oriented lexical activation mechanism.
Item-level data from nine participants with poststroke aphasia who received SFV treatment reported in the work of Evans, Cavanaugh, Quique, et al. (2021) were analyzed using Bayesian generalized linear mixed-effects models. Models evaluated whether performance on three treatment components (facilitated retrieval, feature verification, and effortful retrieval) moderated treatment response for treated and semantically related, untreated words.
There was no evidence for or against a relationship between successful feature verification practice and treatment response. In contrast, there was a robust relationship between the two retrieval practice components and treatment response for treated words only.
Findings were consistent with the second hypothesis: Retrieval practice, but not feature verification practice, appears to be a practice-related predictor of treatment response in SFV. However, treatment components are likely interdependent, and feature verification may still be an active ingredient in SFV. Further research is needed to evaluate the causal role of treatment components on treatment outcomes in aphasia.
明确失语症干预措施中的活性成分可以为治疗理论提供信息，并改善临床实施。本二次分析考察了语义特征验证 (SFV) 治疗中与实践相关的三个预测治疗反应的因素。我们假设：(a) 如果 SFV 的操作类似于标准特征生成语义特征分析，那么成功的特征验证实践将与命名结果相关联；(b) 如果 SFV 通过检索实践导向的词汇激活机制起作用，那么成功的检索实践将与治疗过的、但与语义无关的未治疗词的命名结果相关联。
使用贝叶斯广义线性混合效应模型对 Evans、Cavanaugh、Quique 等人在 2021 年发表的工作中报告的 9 名中风后失语症患者的 SFV 治疗的项目级数据进行了分析。模型评估了三种治疗成分（促进检索、特征验证和费力检索）的表现是否调节了治疗过的和语义相关的未治疗词的治疗反应。
没有证据支持或反对成功的特征验证实践与治疗反应之间的关系。相比之下，在检索实践的两个成分与仅治疗词的治疗反应之间存在着强有力的关系。
研究结果与第二个假设一致：检索实践而不是特征验证实践似乎是 SFV 中治疗反应的一个与实践相关的预测因素。然而，治疗成分可能是相互依存的，特征验证可能仍然是 SFV 的一个活性成分。需要进一步的研究来评估治疗成分对失语症治疗结果的因果作用。
