Diagnostic test accuracy meta-analysis of PRE-DELIRIC (PREdiction of DELIRium in ICu patients): A delirium prediction model in intensive care practice

Intensive Crit Care Nurs. 2020 Apr;57:102784. doi: 10.1016/j.iccn.2019.102784. Epub 2019 Dec 24.


Objectives: To review and examine the evidence on diagnostic test accuracy of PRE-DELIRIC (PREdiction of DELIRium in ICu patients) for predicting delirium risk in critically ill patients.

Research methodology: This meta-analysis included studies reporting the diagnostic performance of PRE-DELIRIC between 2012 and 2019. The Cochrane Library, MEDLINE, Embase, CINAHL and Chinese Electronic Periodical Services databases were searched for eligible diagnostic studies. Risk of bias was assessed using a standard procedure according to the Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2) criteria.

Results: We included seven studies involving a total of 7941 critically ill patients in intensive care units settings. Results indicated that PRE-DELIRIC had a summary sensitivity of 0.76 (95% CI 0.60-0.87), and specificity of 0.66 (95% CI 0.45-0.82), suggesting that diagnostic performance of PRE-DELIRIC is useful to predict delirium risk in ICU patients. The area under the summary receiver operator characteristics (SROC) curve was 0.78 (95% CI 0.74-0.81), which also confirmed good accuracy of PRE-DELIRIC.

Conclusion: We suggest that the PRE-DELIRIC model can be applied in the intensive care unit according to its good diagnostic test accuracy. However, this finding should be interpreted with caution due to the heterogeneity of this meta-analysis.

Keywords: Delirium; Diagnostic test; Meta-analysis; PRE-DELIRIC; Sensitivity and specificity.

Publication types

  • Meta-Analysis
  • Systematic Review

MeSH terms

  • Critical Illness / psychology
  • Delirium / diagnosis*
  • Delirium / physiopathology
  • Delirium / psychology
  • Diagnostic Techniques, Neurological / standards*
  • Diagnostic Techniques, Neurological / trends
  • Humans
  • Intensive Care Units / organization & administration
  • Predictive Value of Tests