Intravenous Thyroxine Administration in Hospitalized Patients, a Common but Unreported Practice: a Single Institution Recent Experience

Background: Intravenous levothyroxine (IVT4) is FDA-approved for the treatment of myxedema coma (ME). ATA guidelines also acknowledge other rare situations, mostly such where oral/enteral access is compromised for prolonged periods, in which IVT4 may be appropriate. We noticed that at our hospital, IVT4 is administered more frequently than expected. Aim of study: To assess the extent of IVT4 administration, the indications for such a treatment, and its outcome at a tertiary facility. Study design and Methods: A retrospective study of IVT4 administered to adult inpatients at Tel Aviv-Sourasky Medical Center between January 2017 and July 2020. A list of dispensed T4 vials during the period of interest was generated from the hospital pharmacy computerized database. Patients’ charts were searched for relevant clinical and laboratory data. Results: 107 patients (62 W/45 M), age 62.5±17.3 y (range 20-97) received IV T4, in the course of 113 hospitalizations. 94 subjects had primary hypothyroidism (PH), 10 had central hypothyroidism, while 3 subjects had no documented evidence of hypothyroidism. ME was likely in only 4 cases (3.5%). The leading stated indication for IVT4 was profound hypothyroidism in 57 instances (50.4%), jeopardized enteral route in 11 (9.7%), while no clear or justifiable indication was found in 39 cases (34.5%). An official endocrine consult backed treatment 74 times (65.5%). In subjects with PH, median serum TSH prior to treatment was 36.4 mIU/L (IQR 8-42), while free T4 was 0.4 ng/dl (IRQ 0.22-0.61, normal 0.8-1.7). In subjects with no ME, altered consciousness was present in 19%, bradycardia in 6.3% and 4.5% were hypothermic. The median initial dose of IV-T4 was 150 μg (range 20-500). Repeated administrations ranged from 1 to 29 times, with a median cumulative dose of 250 μg (IQR 150-400, range 20-3300). We could not identify adverse events directly attributable to IV-T4. Of the 113 admissions, 61 ended in patient’s recovery and discharge (54%), 22 (19.5%) in transfer to a rehab or nursing facility, while there were 30 cases of death (26.5%). Only one of the 4 patients with presumed ME died. In a logistic regression model, that also included age, gender, and ICU admission, the only variable that significantly predicted death was a need for artificial ventilation (OR:27.8, CI 3.5-189). In contrast, free T4, TSH, hospitalization length, altered consciousness, and other potential variables, were excluded from the equation. Conclusions: IVT4 administration is a common practice at our hospital. In a small minority of cases (13.2%), it is given for approved clinical conditions, while in all the others it appears to be unjustified. Reports on this practice are all but absent from the literature. Studies from other institutions are needed to determine its global extent, safety, and efficacy. Until it is proven safe and cost-effective, greater caution should be exercised before allowing it.