B - 75 Recall Versus Recognition? A Systematic Review and Meta-Analysis of Recall Versus Recognition Testing to Differentiate Diagnostic Groups.

Abstract

A systematic review and meta-analysis to evaluate the clinical utility of separate recall and recognition testing in the identification of anterograde memory impairment. Articles published from 2010 onwards that reported outcomes for both recall and recognition testing in patients with a diagnosed memory disorder compared to community controls were eligible for inclusion. A search of PsycINFO, Medline, and EMBASE databases in June 2020 led to a total of 63 articles. Standardized mean differences between patients and controls were calculated for recall and recognition outcomes. Study quality was assessed using a modified QUADAS-2 form. Multivariate meta-analyses revealed that overall, patients performed significantly worse than controls on both recall (Hedges' g = 2.37, 95%CI: 1.88 to 2.87) and recognition tests (Hedges' g = 1.62, 95%CI: 1.26 to 1.99). A comparison of effect sizes between recall and recognition outcomes using multivariate meta-analysis showed that recall tests were significantly more impaired in patients versus controls (g = 0.76, 95%CI: 0.54 to 0.98) across all test categories and all diagnostic categories, except temporal lobe epilepsy, likely due to the small sample size for this group. The results were not influenced by test material (i.e., verbal or visual), the age or education of patients, or studies considered to have a high risk of bias. The results show that recall test scores provide a more effective method to identify anterograde memory impairment compared to recognition test scores, regardless of the underlying cause of memory impairment. In no diagnostic group examined did recognition testing confer any advantage.