Evaluation of inpatient medication guidance from an artificial intelligence chatbot.

PURPOSE: To analyze the clinical completeness, correctness, usefulness, and safety of chatbot and medication database responses to everyday inpatient medication-use questions. METHODS: We evaluated the responses from an artificial intelligence chatbot, a medication database, and clinical pharmacists to 200 real-world medication-use questions. Answer quality was rated by a blinded group of pharmacists, providers, and nurses. Chatbot and medication database responses were deemed "acceptable" if the mean reviewer rating was within 3 points of the mean rating for pharmacists' answers. We used descriptive statistics for reviewer ratings and Kendall's coefficient to evaluate interrater agreement. RESULTS: The medication database generated responses to 194 (97%) questions, with 88% considered acceptable for clinical correctness, 76% considered acceptable for completeness, 83% considered acceptable for safety, and 81% considered acceptable for usefulness compared to pharmacists' answers. The chatbot responded to only 160 (80%) questions, with 85% considered acceptable for clinical correctness, 65% considered acceptable for completeness, 71% considered acceptable for safety, and 68% considered acceptable for usefulness. CONCLUSION: Traditional search methods using a drug database provide more clinically correct, complete, safe, and useful answers than a chatbot. When the chatbot generated a response, the clinical correctness was similar to that of a drug database; however, it was not rated as favorably for clinical completeness, safety, or usefulness. Our results highlight the need for ongoing training and continued improvements to artificial intelligence chatbots for them to be incorporated reliably into the clinical workflow. With continued improvement in chatbot functionality, chatbots could be a useful pharmacist adjunct, providing healthcare providers with quick and reliable answers to medication-use questions.

5 versus 10 Units of Intravenous Insulin for Hyperkalemia in Patients With Moderate Renal Dysfunction.

BACKGROUND: A reduced dose of 5 units of intravenous (i.v.) insulin has been widely accepted for treatment of hyperkalemia in those with end-stage renal dysfunction. However, there remains a dearth of data for patients with moderate renal dysfunction (estimated glomerular filtration rate 15-59 mL/min/m2). OBJECTIVE: Describe the incidence of hypoglycemia and relative change in serum potassium when using 5 vs. 10 units of insulin for hyperkalemia in patients with moderate renal dysfunction. METHODS: This was a single-center, retrospective study evaluating adult patients with moderate renal dysfunction who received i.v. insulin for treatment of hyperkalemia. Patients were analyzed based on whether they received 5 or 10 units of i.v. insulin. The primary outcome was the rate of hypoglycemia in each group. Secondary outcomes included rate of relative potassium-lowering effect and incidence of severe hypoglycemia. RESULTS: Hypoglycemia occurred in 12 patients who received 5 units of i.v. insulin and 16 patients who received 10 units of i.v. insulin (6.5% vs. 8.4%, p = 0.476). Serum potassium was significantly reduced when utilizing 10 units over 5 units of i.v. insulin (-0.9 mmol/L vs. -0.63 mmol/L, p = 0.001). Severe hypoglycemia was seen in two encounters in both the 5-unit and 10-unit groups (1.1% vs. 1.0%, p = 0.979). CONCLUSION: There was no difference in hypoglycemic events among patients with moderate renal dysfunction receiving 5 vs. 10 units of i.v. insulin for hyperkalemia. However, 10 units of i.v. insulin lowered serum potassium significantly more than 5 units of i.v. insulin.

Epsom Salt Ingestion Leading to Severe Hypermagnesemia Necessitating Dialysis.

BACKGROUND: Hypermagnesemia is an often overlooked electrolyte abnormality that has a myriad of presenting symptoms. It has been observed after both accidental and intentional ingestions of magnesium-containing compounds, and as in the case presented, Epsom salts, which are primarily magnesium sulfate. CASE REPORT: A 56-year-old man presented to the emergency department reporting weakness after an ingestion of Epsom salts used as a laxative and was found to be bradycardic and hypotensive. He subsequently developed altered mental status and respiratory depression necessitating intubation. His magnesium level was found to be > 3.91 mmol/L (> 9.5 mg/dL). He was given multiple doses of calcium gluconate and generous i.v. fluids with furosemide, with minimal improvement. However, his magnesium level corrected rapidly after initiation of dialysis, and 3 days later he was discharged home in good condition with normal neurologic function. WHY SHOULD AN EMERGENCY PHYSICIAN BE AWARE OF THIS?: Keeping a high level of suspicion for, and quickly recognizing, hypermagnesemia allows for prompt initiation of treatment, which can avoid significant hemodynamic or respiratory compromise. Mainstays of treatment are i.v. calcium and i.v. fluids. Loop diuretics may be given as an adjunct as well. Dialysis should be considered in cases of severe hypermagnesemia because it results in rapid correction of magnesium levels.

Incidence of Adverse Events During Peripheral Administration of Sodium Chloride 3.

PURPOSE: Traditionally, sodium chloride 3% has been administered via a central venous line (CVL) because of the perceived risk of infiltration and tissue injury due to its high osmolarity. In clinical practice, sodium chloride 3% is commonly administered through peripheral venous catheters (PVCs) given the necessity of timely administration. However, there is no published data on the safety of administering sodium chloride 3% through PVCs in the adult population. The objective of this study was to evaluate the safety of peripheral venous administration of sodium chloride 3%. MATERIALS AND METHODS: A retrospective review was conducted in patients who received sodium chloride 3% in the intensive care unit (ICU). Patients were excluded if they had a CVL for the entire duration of the infusion or younger than 18 years at the time of administration. Baseline patient and infusion characteristics were collected. Infusion-related adverse events (IRAEs) were recorded, graded, and interventions required were noted. RESULTS: A total of 66 patients were included in the analysis. The most common indication was hyponatremia and majority of the patients were managed in the neurosurgical ICU. The most common risk factor for IRAEs was the presence of altered mental status. Four patients experienced an IRAE at an event rate of 6.1%. Patients who experienced an IRAE ranged from 38 to 82 years old. The IRAEs were grade 1 in severity, managed conservatively with removal of the PVC, and 2 of the 4 patients had their infusions restarted peripherally. The time to initial IRAE ranged from 2 to 94 hours. For the entire cohort, hospital and ICU length of stay were 8 and 4 days, respectively. CONCLUSIONS: The rate of IRAEs related to the infusion of sodium chloride 3% through PVCs appears to be similar to those reported with other hyperosmotic agents and could be considered for patients who need time-sensitive therapy.

Pharmacokinetics of Peramivir in an Adolescent Patient Receiving Continuous Venovenous Hemodiafiltration.

Critically ill patients requiring renal replacement therapy commonly experience pharmacokinetic alterations. This case report describes the pharmacokinetics of peramivir (Rapivab, BioCryst Pharmaceuticals, Inc, Durham, NC), the first US Food and Drug Administration-approved intravenous neuraminidase inhibitor for the treatment of influenza, in an adolescent patient receiving continuous renal replacement therapy (CRRT). A 49.5-kg, 17-year-old Caucasian female presented with fever, cough, and persistent hypoxia. She quickly progressed to acute respiratory and renal failure in the setting of viral septic shock as a result of a severe influenza H1N1 infection. On hospital day 3, therapy was switched from oseltamivir (Tamiflu, Roche Laboratories Inc, Nutley, NJ) to peramivir owing to the concern for inadequate enteral absorption. On the third day of peramivir treatment, at a dose of 200 mg daily, peramivir serum concentrations revealed a smaller peak concentration, larger volumes of distribution, similar 24-hour area under the curve, and a shorter half-life as compared to adult patients with normal renal function. This illustrated the significant differences in pharmacokinetics when administered in the setting of CRRT. The patient had resolution of viral infection as evidenced by negative respiratory viral panel polymerase chain reaction at hospital day 14 and was eventually discharged at her baseline.

Dexmedetomidine for refractory adrenergic crisis in familial dysautonomia.

OBJECTIVE: Adrenergic crises are a cardinal feature of familial dysautonomia (FD). Traditionally, adrenergic crises have been treated with the sympatholytic agent clonidine or with benzodiazepines, which can cause excessive sedation and respiratory depression. Dexmedetomidine is a centrally-acting α 2-adrenergic agonist with greater selectivity and shorter half-life than clonidine. We evaluated the preliminary effectiveness and safety of intravenous dexmedetomidine in the treatment of refractory adrenergic crisis in patients with FD. METHODS: Retrospective chart review of patients with genetically confirmed FD who received intravenous dexmedetomidine for refractory adrenergic crises. The primary outcome was preliminary effectiveness of dexmedetomidine defined as change in blood pressure (BP) and heart rate (HR) 1 h after the initiation of dexmedetomidine. Secondary outcomes included incidence of adverse events related to dexmedetomidine, hospital and intensive care unit (ICU) length of stay, and hemodynamic parameters 12 h after dexmedetomidine cessation. RESULTS: Nine patients over 14 admissions were included in the final analysis. At 1 h after the initiation of dexmedetomidine, systolic BP decreased from 160 ± 7 to 122 ± 7 mmHg (p = 0.0005), diastolic BP decreased from 103 ± 6 to 65 ± 8 (p = 0.0003), and HR decreased from 112 ± 4 to 100 ± 5 bpm (p = 0.0047). The median total adverse events during dexmedetomidine infusion was 1 per admission. Median hospital length of stay was 9 days [interquartile range (IQR) 3-11 days] and median ICU length of stay was 7 days (IQR 3-11 days). CONCLUSIONS: Intravenous dexmedetomidine is safe in patients with FD and appears to be effective to treat refractory adrenergic crisis. Dexmedetomidine may be considered in FD patients who do not respond to conventional clonidine and benzodiazepine pharmacotherapy.