In considering the proportion of phonological errors, although Table 2 shows half the participants made 11% or fewer phonological errors while half made 12% or over, we would not suggest using a number between these as the exact cut-off score. Further research investigating nature of difficulty

and outcome of intervention is necessary. From this study we suggest those with a small percentage of phonological errors (up to and including 5%) are not likely to have a phonological production deficit that results in generalised therapy effect. Those for whom 20% or more of errors are phonological are likely to have such a deficit. All participants except D.C., discussed above, and D.J. fall into one of these Pifithrin-�� in vitro two groups. The results for treated items replicate previous research which has shown intervention involving cues can aid naming in adults with aphasia (Nickels, 2002). The study shows that change can occur from intervention once a week for 8 weeks. The

outcomes do not relate straightforwardly to traditional aphasia classification. For example, from Fig. 1, it is clear that, of the two participants who made least change in naming, one had fluent aphasia (S.C.) and the other had non-fluent aphasia (G.B.). Likewise, the participant in the first study who named the most extra items (P.H.) had anomic aphasia; in contrast, the participant in the Health Service based study who named the most extra items (F.A.) had non-fluent aphasia. Thus, the results Ibrutinib do not relate to traditional aphasia classification or even the distinction between fluent and non-fluent

aphasia. It is, therefore, unlikely that the extent of improvement in picture naming of treated items would relate to lesion site, although this remains to be explored. This disassociation between outcome and traditional aphasia classification is also in line with other studies treating written and spoken naming (e.g., Carlomagno et al., 2001; Leonard et al., Guanylate cyclase 2C 2008). The introduction outlined three stages of processing in spoken language production. We return to these and relate them to findings from other studies which have investigated levels of deficit in relation to outcome and to the data from this study. Stages 1–3, outlined in the introduction, are illustrated to the left of Fig. 4 which displays assessment findings and not the nature of intervention provided.3 The figure includes only studies where detailed background assessment enables the link between level of deficit and outcome of intervention to be explored. The participants with anomia with a deficit at stage 1 (accessing word meaning) or stage 2 (accessing word form) do not show generalisation to untreated items from therapy directed at their anomia.