Combined use of the Montreal Cognitive Assessment and Symbol Digit Modalities Test improves neurocognitive screening accuracy after cardiac arrest: A validation sub study of the TTM2 trial

<p>Aim</p> <p><br></p> <p>To assess the merit of clinical assessment tools in a neurocognitive screening following out-of-hospital cardiac arrest (OHCA).</p> <p>Methods</p> <p><br></p> <p>The neurocognitive screening that was evaluated included the performance-based Montreal Cognitive Assessment (MoCA) and Symbol Digit Modalities Test (SDMT), the patient-reported Two Simple Questions (TSQ) and the observer-reported Informant Questionnaire on Cognitive Decline in the Elderly-Cardiac Arrest (IQCODE-CA). These instruments were administered at 6-months in the Targeted Hypothermia versus Targeted Normothermia after Out-of-Hospital Cardiac Arrest (TTM2) trial. We used a comprehensive neuropsychological test battery from a TTM2 trial sub-study as a gold standard to evaluate the sensitivity and specificity of the neurocognitive screening.</p> <p>Results</p> <p><br></p> <p>In our cohort of 108 OHCA survivors (median age = 62, 88% male), the most favourable cut-off scores were: MoCA < 26; SDMT z ≤ -1; IQCODE-CA ≥ 3.04. The MoCA (sensitivity 0.64, specificity 0.85) and SDMT (sensitivity 0.59, specificity 0.83) had a higher classification accuracy than the TSQ (sensitivity 0.28, specificity 0.74) and IQCODE-CA (sensitivity 0.42, specificity 0.60). When using the cut-points for MoCA or SDMT in combination to identify neurocognitive impairment, sensitivity improved (0.74, specificity 0.81), area under the curve = 0.77, 95% CI [0.69, 0.85]. The most common unidentified impairments were within the episodic memory and executive functions domains, with fewer false negative cases on the MoCA or SDMT combined.</p> <p>Conclusion</p> <p><br></p> <p>The MoCA and SDMT have acceptable diagnostic accuracy for screening for neurocognitive impairment in an OHCA population, and when used in combination the sensitivity improves. Patient and observer-reports correspond poorly with neurocognitive performance.</p>