Adverse outcomes in older adults attending emergency department: systematic review and meta-analysis of the Triage Risk Stratification Tool.

The objective of this study was to assess the predictive ability of the Triage Risk Stratification Tool (TRST) in identifying older adults at risk of adverse outcomes [return to emergency department (ED), hospitalization, or a composite outcome] within 30 and 120 days following discharge from ED. A systematic search was conducted to identify studies validating the TRST in older adults aged at least 65 years discharged from ED. The methodological quality of selected studies was assessed using the Quality Assessment of Diagnostic Accuracy Studies tool. A score of at least 2 was used to identify high-risk patients. A bivariate random effects model was applied to generate pooled estimates of sensitivity and specificity. The discriminative ability of the tool was assessed using the c statistic. Heterogeneity was assessed graphically and statistically using the variance of the logit-transformed sensitivity and specificity. Six studies (3233 patients) were included. With a cutoff of at least 2, the pooled sensitivity of the TRST for a composite outcome at 30 days was 0.58 (95% confidence interval 0.52-0.65), the specificity was 0.61 (95% confidence interval 0.58-0.64). Similar values were obtained for the composite outcome at 120 days and the individual outcomes at both timepoints. The c statistic was consistently low across the outcomes and timepoints. There was little heterogeneity across studies. In conclusion, the TRST is limited in its ability to discriminate between older adults with or without an adverse outcome following discharge from ED. The low levels of heterogeneity gives us confidence that the pooled estimates reflect the predictive ability of the TRST in clinical practice.