Introduction: Ammonia appears to play a major role in the pathophysiology of hepatic encephalopathy (HE), but its role in guiding management is unclear. We aimed to understand the impact of ammonia levels on inpatient HE management, hypothesizing that patients with elevated ammonia levels would receive more aggressive lactulose therapy than patients with normal ammonia or no ammonia level drawn. METHODS: We examined patients with cirrhosis older than 18 years admitted for management of HE from 2005 to 2015. We additionally used propensity matching to control for confounding by the severity of underlying disease. Patients with an ammonia level taken at time of HE diagnosis were further separated into those with normal or elevated ammonia levels. The primary endpoint was the total lactulose (mL) amount (or dose) given in the first 48 hours of HE management. RESULTS: One thousand two hundred two admissions with HE were identified. Ammonia levels were drawn in 551 (46%) patients; 328 patients (60%) had an abnormal ammonia level (>72 μmol/L). There were no significant differences in the Child-Pugh score, MELD, or Charlson Comorbidity Index in those with and without ammonia levels drawn. The average total lactulose dose over 48 hours was 167 and 171 mL in the no ammonia vs ammonia groups, respectively (P = 0.42). The average lactulose dose in patients with an elevated ammonia level was 161 mL, identical to the lactulose dose in patients with a normal ammonia level. There was no correlation between lactulose dose and ammonia level (R 2 = 0.0026). DISCUSSION: Inpatient management of HE with lactulose was not influenced by either the presence or level of ammonia level, suggesting that ammonia levels do not guide therapy in clinical practice.

Ogilvie's syndrome, also known as acute colonic pseudo-obstruction, refers to pathologic dilation of the colon without underlying mechanical obstruction, occurring primarily in patients with serious comorbidities. Diagnosis of Ogilvie's syndrome is based on clinical and radiologic grounds, and can be treated conservatively or with interventions such as acetylcholinesterase inhibitors (such as neostigmine), decompressive procedures including colonoscopy, and even surgery. Based on our clinical experience we hypothesized that conservative management yields similar, if not superior, results to interventional management. Therefore, we retrospectively examined all patients over the age of 18 with Ogilvie's syndrome who presented to the Medical University of South Carolina (MUSC). The diagnosis of Ogilvie's syndrome was confirmed by clinical criteria, including imaging evidence of colonic dilation ≥9 cm. Patients were divided and analyzed in 2 groups based on management: conservative (observation, rectal tube, nasogastric tube, fluid resuscitation, and correction of electrolytes) and interventional (neostigmine, colonoscopy, and surgery). Use of narcotics in relation to maximal bowel size was also analyzed. Over the 11-year study period (2005–2015), 37 patients with Ogilvie's syndrome were identified. The average age was 67 years and the average maximal bowel diameter was 12.5 cm. Overall, 19 patients (51%) were managed conservatively and 18 (49%) underwent interventional management. There was no significant difference in bowel dilation (12.0 cm vs 13.0 cm; P = .21), comorbidities (based on the Charlson Comorbidity Index (CCI), 3.2 vs 3.4; P = .74), or narcotic use (P = .79) between the conservative and interventional management groups, respectively. Of the 18 patients undergoing interventional management, 11 (61%) had Ogilvie's-syndrome-related complications compared to 4 (21%) of the 19 patients in the conservative management group (P < .01). There was no difference in overall length of stay in the 2 groups. Two patients, one in each group, died from complications unrelated to their Ogilvie's syndrome. We conclude that Ogilvie's syndrome, although uncommon, and typically associated with severe underlying disease, is currently associated with a low inpatient mortality. While interventional management is often alluded to in the literature, we found no evidence that aggressive measures lead to improved outcomes.

Predicting clinical outcomes in patients with cirrhosis is challenging, and is likely best accomplished with a combination of objective data (such as MELD and HVPG provide) in addition to the clinical course of cirrhosis.

Scoring systems such as Model for End-stage Liver Disease (MELD) and Child-Pugh are often used by clinicians to determine prognosis in patients with cirrhosis. Since clinical complications are important in determining cirrhosis outcomes, our goal was to use these to develop a novel prognostic staging model. Data from the Nationwide Inpatient Sample (NIS), years 2003-2011, were queried for records of patients over the age of 18 with cirrhosis excluding patients with prior or inpatient liver transplantation. The primary outcome was inpatient mortality with focus on cirrhosis-related complications: non-bleeding esophageal varices, variceal hemorrhage, ascites, hepatic encephalopathy (HE), spontaneous bacterial peritonitis (SBP), and hepatorenal syndrome (HRS). Of 59 862 903 hospitalizations, 824 783 (1.4%) with cirrhosis were identified. Overall mortality was 7% with two-thirds (66%) of deaths occurring in patients with a decompensating event, defined as variceal hemorrhage, ascites, HE, SBP, and/or HRS. Overall mortality rates decreased from 2003 to 2011 (9.0-6.0%), in both compensated and decompensated groups. Mortality was higher in patients with variceal haemorrhage (OR 1.56; p<0.05), HE (OR 1.75; p<0.05), SBP (OR 2.64; p<0.05) and HRS (OR 9.10; p<0.05) compared with patients with no complications. HRS had the highest mortality, whether alone or in combination with another event such as HE (OR 12.40; p<0.05) or SBP (OR 12.64; p<0.05). Cirrhosis inpatient outcomes are related to the severity of liver disease, with more severe complications such as HE, SBP, and HRS having the most significant effect on inpatient mortality, and are utilised in this novel four-stage clinical model.

No abstract