Why Physicians Delay in the Diagnosis and Treatment of Valproic Acid Toxicity

This study aims to review valproic acid (VA) toxicity and treatment at an urban safety-net hospital to determine the timeliness and appropriateness of treatment and outcomes.

We gathered data on 159 patients with the diagnosis of VA toxicity who were admitted to the Detroit Medical Center (USA) from January 2005 to 2015. Patients with known liver disease due to alcohol intake and viral infections were excluded. We extracted the following parameters and compared them: demographic variables, time at presentation, time elapsed between presentation and diagnosis, time elapsed between diagnosis and treatment, time elapsed between treatment and discharge, the total hospital length of stay (LOS), diagnosis, symptoms, treatments given, concomitant medications, and laboratory values. The analysis of all variables was descriptive, and simple correlations were done between dependent and independent variables.

A total of 159 patients with VA toxicity were identified. The majority of patients (53% male, 47% female) were African Americans (75%), with the remainder Caucasian (22%) and others (3%). According to our data, one-third (33%) of the total cases comprised patients who had multiple disorders (schizophrenia, depression, bipolar, seizure), 36 (22.6%) had schizophrenia, 34 (21%) had bipolar disorder, 23 (14%) had seizure disorder, 11 (7%) had major depression, and two (1%) were without any previous diagnosis. Data showed that 72 patients (45.3%) were found with altered mental status, 49 (31%) with GI and other CNS symptoms, and 38 (24%) without any symptoms. There were 58 (37%) patients who committed a suicidal attempt, but more than half were asymptomatic (59%) on initial presentation. Disposition included discharge from ED in 22 (15%), admission to psychiatry in 11 (7%), floors in 108 (68%), and ICU in 18 (11%). The mean time from presentation to diagnosis was three hours, with a range of 0.1-10 hrs. Supportive care alone was given to 62 (39%), charcoal to 35 (22%), and pharmacological treatment to 62 (39%). Of those receiving pharmacologic treatment, 47 received only levocarnitine; 15 received lactulose followed by levocarnitine. The average ammonia level at presentation among those who received pharmacologic treatment was 89 ± 63 mcg/dL. Those treated with lactulose (15/62) had an initial ammonia level of 97 ± 62 mcg/dL, which fell to 72 ± 58 mcg/dL at 20 hrs. Once levocarnitine was added, ammonia decreased to 56 ± 25 mcg/dL at 25 hrs. Those who were treated only with levocarnitine had an initial ammonia level of 87 ± 63 mcg/dL, which fell to 64 ± 48 mcg/dL at 14 hrs. Of 159 patients, 44 (28%) had known alcohol abuse. In those who required pharmacologic treatment, the time from diagnosis to appropriate treatment with levocarnitine was 7.8 hrs in alcoholics and 4.3 hrs in patients who were not alcoholics. No patient with elevated alcohol on admission was diagnosed with hepatic encephalopathy on discharge. The average hospital LOS for those who received pharmacologic treatment was 92.4 hrs. Patients who received lactulose had a LOS of 98.9 hrs. The total LOS for those who did not receive lactulose was 90.3 hrs.

Early diagnosis of VA toxicity is necessary for timely, effective, high-quality care. Patients who initially received lactulose for hyperammonemia had slower recovery. Recognition of VA toxicity and prompt initiation of levocarnitine are associated with faster recovery and shorter LOS.