Diagnostic accuracy of ancillary tests for death by neurologic criteria: a systematic review and meta-analysis

Abstract

PurposeAncillary tests are frequently used in death determination by neurologic criteria (DNC), particularly when the clinical neurologic examination is unreliable. Nevertheless, their diagnostic accuracy has not been extensively studied. Our objective was to synthesize the sensitivity and specificity of commonly used ancillary tests for DNC.SourceWe performed a systematic review and meta-analysis by searching MEDLINE, EMBASE, Cochrane databases, and CINAHL Ebsco from their inception to 4 February 2022. We selected cohort and case–control studies including patients with 1) clinically diagnosed death by neurologic criteria or 2) clinically suspected death by neurologic criteria who underwent ancillary testing for DNC. We excluded studies without a priori diagnostic criteria and studies conducted solely on pediatric patients. Accepted reference standards were clinical examination, four-vessel conventional angiography, and radionuclide imaging. Data were directly extracted from published reports. We assessed the methodological quality of studies with the QUADAS-2 tool and estimated ancillary test sensitivities and specificities using hierarchical Bayesian models with diffuse priors.Principal findingsOverall, 137 records met the selection criteria. One study (0.7%) had a low risk of bias in all QUADAS-2 domains. Among clinically diagnosed death by neurologic criteria patients (n = 8,891), ancillary tests had similar pooled sensitivities (range, 0.82–0.93). Sensitivity heterogeneity was greater within (σ = 0.10–0.15) than between (σ = 0.04) ancillary test types. Among clinically suspected death by neurologic criteria patients (n = 2,732), pooled ancillary test sensitivities ranged between 0.81 and 1.00 and specificities between 0.87 and 1.00. Most estimates had high statistical uncertainty.ConclusionStudies assessing ancillary test diagnostic accuracy have an unclear or high risk of bias. High-quality studies are required to thoroughly validate ancillary tests for DNC.Study registrationPROSPERO (CRD42013005907); registered 7 October 2013.