SP239DEVELOPMENT OF A PROGNOSTIC INDEX FOR ASSESSING THE RISK OF DELIRIUM IN SUBJECTS WITH AKI UNDERGOING DIALYSIS

Abstract

INTRODUCTION AND AIMS: Kidney damage and brain involvement are frequent during acute diseases. In the acute setting AKI is strongly associated with morbidity and mortality, whereas delirium represents the most frequent neuropsychiatric syndrome, also associated with unfavorable outcomes. Few studies have examined the association between AKI and delirium or considered AKI as a risk factor for delirium. The aim of the study is to identify the factors associated with delirium in subjects with AKI undergoing dialysis; develop a clinical prediction rule for the risk estimation of delirium in the same population. METHODS: Retrospective study conducted analyzing the data in the medical record from January 2015 to February 2017. Fifty-nine patients admitted consecutively from the ER were enrolled. Inclusion criteria were: the presence of AKI requiring dialysis (AKI was defined according to the 2012 KDIGO guideline on AKI). Incident delirium defined by the following criteria: presence in the medical record of words as delirium, confusion, agitation, fluctuation of consciousness and a concomitant new prescription of antipsychotics or benzodiazepines within 48 hours from the onset of delirium. We detected systematically 33 standardized indicators related to the premorbid history and clinical-functional characteristics at baseline, the cause of AKI and the patient's clinical pathway. These indicators and blood examinations were considered as potential predictors for the onset of delirium. RESULTS: Incident delirium occurred in 21 (35.6%) patients. All the subjects had high clinical complexity.We found no difference between patients who developed delirium compared to patients without delirium, with the exception of age (76.4+10.2 vs. 68.9+15.3, p=0.002), and history of dementia (38.1%, vs. 13.2 p=0.03) and cardiac diseases (95.2%, vs. 60.5%p=0.01). Patients presenting delirium were significantlymore dependent with regards to both the basicmobility (67%vs. 32%, p= 0.01) and the Barthel Index Score at admission (38.6+32.7 vs. 68.7+23; p=0.00). The two groups didn't differ in the causes of AKI. All-causesmortality was 18.6%, significantly higher in delirium group (33.3%vs. 10.5%; p= 0.03). In themultivariable logistic regression analysis, the factors that retain an independent predictive value on the onset of delirium were reported in the table CONCLUSIONS: The analyses conducted found that the age> 80 years, the baseline functional dependence, CRP, and K+ are independent risk factors for the onset of delirum. The predictive magnitude of these clinical elements can be quantified in a score (DAKIR) that stratifies patients for a risk ranging from 4% (score=1) to 97% (score=6). The attention to risk factors and the use of the score DAKIR can support the decision-making processes on monitoring and treatment with different intensities of patients with AKI undergoing dialysis. More patients will be enrolled for the validation of the predictive model (Table presented).