Thiamine utilization and the lack of prescribing standardization: A critical examination

Abstract

Objectives

Thiamine is often prescribed for thiamine deficiency during hospitalization despite the lack of US-based clinical guidelines. This study aims to evaluate thiamine prescribing patterns and key characteristics associated with the deficiency to address gaps in care.

Methods

Data were obtained from electronic health records of hospitalized patients between September 1, 2021, and March 30, 2022. Alcohol use disorder (AUD) was defined by a positive Clinical Institute Withdrawal Assessment score or a positive serum alcohol level upon admission. Geriatric patients were defined as age ≥65. Cohort 1 was defined as: AUD, albumin <4 g/L, INR >1.5, and total bilirubin >3 mg/dL. Cohort 2 was defined as: age >65, albumin <4 g/L, hemoglobin <15 g/dL, and folate <4 ng/mL. A multivariable LASSO regression model was used to identify characteristics associated with higher thiamine dosing (>100 mg/day).

Results

Among 780 patients, 520 (66.7%) were identified as AUD, of which 265 (50.1%) were between the ages of 45–64 years. The AUD cohort was significantly different (p < 0.05) in the mean serum albumin 4.16 g/L (IQR: 3.8–4.5), AST 73.55 U/L (23.75–82.00), ALT 52.57 U/L (17.00–57.00), total bilirubin 0.98 (0.3–1.0), and INR 1.1 (0.99–1.12), compared to non-AUD patients with a mean serum albumin 3.75 g/L (3.3–4.2), AST 35.07 U/L (11.00–42.00), ALT 32.77 U/L (5.00–34.00), total bilirubin 0.89 (0.2–0.9), and INR 1.21 (1.0–1.22). In the geriatric cohort, 136 patients (17%) had a mean serum albumin 3.77 g/L (3.4–4.2), AST 38.66 U/L (14.0–41.0), ALT 29.36 U/L (9.0–37.0), total bilirubin 0.62 mg/dL (0.30–0.90), and direct bilirubin 0.12 mg/dL (0.00–0.20), compared to the non-geriatric cohort with a mean serum albumin 4.10 g/L (3.8–4.40), AST 66.44 U/L (21.0–75.0), ALT 50.03 U/L (16.00–53.75), total bilirubin 1.02 mg/dL (0.30–1.00), and direct bilirubin 0.31 mg/dL (0.00–0.20). In cohort 1, 40.6% patients were between 51 and 64 years old, (66.5%) male, and had a BMI <25 (36.4%). In cohort 2, 52.6% were between 65 and 70 years old, (57.9%) male, and had a BMI <25 (57.9%). Cohort 1 were prescribed a dose of 100 mg (47.7 %), oral (63.5%), intramuscular (18.2%), daily (58.9%), one-day duration (49.4%) most frequently. Cohort 2 were prescribed a dose of 100 mg (56.0%), oral (77.2%), daily (77.2%), one-day duration (29.8%) most frequently. The AUD was significantly associated with having a higher dosage (e.g., >100 mg) of thiamine prescribed per day OR 1.62 (1.11–2.37) (p < 0.01).

Conclusions

This study confirms that thiamine prescribing patterns vary during hospitalization and suggest specific laboratory findings may aid in identifying cohorts associated with the deficiency.