P: 54 Diagnosis of Covert Hepatic Encephalopathy Is Influenced by Multiple Non-cognitive Variables That Varies by Testing Strategy

Abstract

BACKGROUND: Covert hepatic encephalopathy (CHE) is a serious complication on cirrhosis that manifests as an amnestic cognitive dysfunction. Diagnosis is based on examining cognitive functions, but results are influenced by multiple non-cognitive factors such as sleep and medications but data regarding the influence of other non-cognitive covariates is limited. We aimed to examine the potential non-cognitive variables that could influence testing on the psychometric hepatic encephalopathy score (PHES) and the encephalapp stroop. METHODS: Patients between ages 20-75 years were enrolled from clinic between 2012–2018. Those with severe uncontrolled psychiatric comorbidities were excluded as were those actively abusing alcohol or drugs. All patients underwent testing for CHE based on 2 testing strategy as recommended by the AASLD/EASL guidelines. Demographics and details of comorbidities were collected. CHE was diagnosed based on published norms. Appropriate t tests and logistic regression were done. Dependent variable was CHE on PHES and Stroop. RESULTS: We enrolled a total of 257 patients with mean age 61.1 ± 8.3, 72% were males, HCV was he predominant etiology 96 (37.3%). Median BMI was 29 (26, 34). Median education was 13 (12, 16) years. Median MELD was 11 (7.25, 15) with median childs score of 6 (5, 8). Charlson comorbidity score (CCI) was 5 (4, 6). Eighty five (33%) had a history of prior OHE and were on lactulose (13), rifaximin (9) or both (63), 95 (37%) had ascites with 58 (61%) controlled on diuretics and 17 (18%) had a h/o SBP. 41 (16%) had a h/o variceal bleeding. In terms of comorbidities 34.2% had Diabetes Mellitus, 58% had Hypertension, 11% had Coronary artery disease, 14% had hypothyroidism, 2% had CHF, 2.3% had COPD, 30% had depression and 87% were controlled on medications. 2.3% had post traumatic stress disorder (PTSD) and were on medications. 14% were on chronic narcotic medications. On cognitive testing 109 (42.4%) had CHE with a median PHES score of −3 (−7, 0) whereas 206 (80%) tested positive on encephalapp stroop. On univariable analysis hypertension was found to be significant only for stroop. On multivariable analysis age and prior OHE were predictive for both tests but for the Encephalapp stroop hypertension was found to be independently predictive (Table 1). CONCLUSION: The diagnosis of CHE can be influenced by other non-cognitive variables and these vary between individual testing strategies probably due to the differential effects that these systemic conditions have on cerebral/subcortical functions. Physicians must take into consideration these covariates while interpreting CHE testing based on these 2 tests.