Evaluating pain, sedation, and delirium in the neurologically critically ill-feasibility and reliability of standardized tools: a multi-institutional study.

OBJECTIVES

To assess the feasibility and reliability of systematic evaluations of analgesia, sedation level, and delirium features in the neurologically critically ill and to determine whether delirium features are linked to clinical outcomes in this population.

DESIGN

Multicentered prospective observational study.

SETTING

Neurological, Neurosurgical, Neurosciences or Surgical Trauma ICUs from three hospitals (two in Canada and one in the United States).

PATIENTS

A convenience sample of adult NICU or neurologic, neurosurgical, neurosciences, or surgical trauma ICU patients admitted for greater than 12 hours from November 2011 to April 2012.

INTERVENTIONS

Systematic assessments were simultaneously and independently performed by a neurologist, intensivists, or trauma surgeon, and a nurse in three multispecialty ICUs. Pain was evaluated with the numeric rating scale or behavioral pain scale. Sedation was assessed using the Richmond Agitation-Sedation Scale. Patients with Richmond Agitation-Sedation Scale greater than or equal to -4 were screened for features of delirium with the Intensive Care Delirium Screening Checklist. Intraclass correlation coefficient was used to evaluate inter-rater reliability between the nurse and the physician for pain and sedation scales, and the kappa coefficient was calculated for concordance of the Intensive Care Delirium Screening Checklist items.

MEASUREMENTS AND MAIN RESULTS

151 patients had 439 assessments. Pain and sedation were always assessable with excellent inter-rater reliability (numeric rating scale intraclass correlation coefficient, 0.92; behavior pain scale intraclass correlation coefficient, 0.83; and Richmond Agitation-Sedation Scale intraclass correlation coefficient, 0.92). Patients were sufficiently alert for delirium screening 3/4 of the time; Intensive Care Delirium Screening Checklist items had good concordance (kappa coefficients between 0.58 and 0.91 for the eight Intensive Care Delirium Screening Checklist items). Nonevaluable items were most often orientation, hallucinations, and speech or mood content. Furthermore, each additional Intensive Care Delirium Screening Checklist item present in proportion to the total evaluable Intensive Care Delirium Screening Checklist score was associated with a 10% increase in ICU length of stay.

CONCLUSIONS

Pain and sedation can be systematically assessed in the neurologically critically ill; the majority can also be screened for delirium features with excellent inter-rater reliability. Increased proportion of Intensive Care Delirium Screening Checklist items is associated with worse outcomes.