674

Abstract

Introduction: Objective measures with precise guidelines are employed to ensure standardized and reproducible assessments of the mental status of patients in the medical intensive care unit (MICU). We hypothesize that, despite using similar clinical scoring tools, there is still discordance between nurse and physician evaluation of patients' mental status. Methods: An IRB-approved cross sectional trial evaluating the mental and sedation status of all patients in the MICU during the daytime was performed. Using our electronic medical record we recorded the last documented nursing (RN) assessment of each patient's mental and sedation status by the Glascow Coma Scale (GCS), Confusion Assessment Method for the ICU (CAM-ICU), and Richmond Agitation and Sedation Score (RASS). We then assessed the same evaluations by a study physician's (MD). We compared both evaluations. Results: One hundred and twenty seven patients were evaluated who were in the hospital for a mean ± SD of 11 ± 14 days. The mean age was 61 ± 16 years and 53 % were male. 44 % of patients were mechanically ventilated and 11 % were on vasopressors at the time of evaluation. Nine percent were on continuous sedation. Discrepancies between RN and MD scoring of the GCS occurred in 31% of patients; 32% of patients in RASS evaluation; and 24% of patients in CAM-ICU assessment. There was good overall correlation between RN and MD scoring with both GCS (r2 = 0.72, p< 0.001) and RASS (r2 = 0.74, p< 0.001). In patients who are awake (RAAS of 0), nurses and physicians had excellent agreement. However, in sedated patients (RAAS ≤ -1), the physicians score patients with a lower RASS than nurses. There was better agreement between MD and RN evaluations of GCS. Physicians were more likely to score a patient as CAM-ICU positive as compared to RN (58 vs. 8 patients, p<0.001). RN were more likely to document inability to complete the CAM-ICU (38 vs. 6 patients, p<0.001). Nurses documented inability to assess CAM-ICU across the spectrum of RASS scores, whereas physicians only documented inability to perform CAM-ICU in patients with a RASS -4 or -5. Conclusions: Significant heterogeneity exists in the results of standardized mental status assessments in the MICU. Objective evaluations suffer from inherent subjectivity of the evaluator and fluctuations in the mental status of critically ill patients. We hypothesize that the implications in management and patient outcomes may be very important.