Date Published: May 31, 2019
Publisher: Public Library of Science
Author(s): Ian Ellis McCoy, Glenn Matthew Chertow, Tara I-Hsin Chang, Emmanuel A. Burdmann.
To inform future outcomes research on diuretics, we sought to describe modern patterns of diuretic use in the intensive care unit (ICU), including diuretic type, combination, and dosing. We also investigated two possible quality improvement targets: furosemide dosing in renal impairment and inclusion of an initial bolus with continuous furosemide infusions.
In this descriptive study, we retrospectively studied 46,037 adult ICU admissions from a publicly available database of patients in an urban, academic medical center.
Diuretics were employed in nearly half (49%, 22,569/46,037) of ICU admissions. Mechanical ventilation, a history of heart failure, and admission to the post-cardiac surgery unit were associated with a higher frequency of diuretic use. Combination use of different diuretic classes was uncommon. Patients with severely impaired kidney function were less likely to receive diuretics. Furosemide was by far the most common diuretic given and the initial intravenous dose was only 20 mg in more than half of ICU admissions. Among patients treated with a continuous infusion, 30% did not receive a bolus on the day of infusion initiation.
Patterns of diuretic use varied by patient-specific factors and by ICU type. Diuretic dosing strategies may be suboptimal.
Fluid management is one of the most challenging clinical problems in the intensive care unit (ICU). While some patients present with fluid overload, other patients acquire fluid overload after admission to the ICU due to administration of intravenous fluid therapy, which is often the initial treatment maneuver for hypotension of any cause [1,2]. Diuretics are a mainstay for managing fluid overload and are commonly prescribed in the ICUs of all types [3–5]. However, there are few guidelines regarding the selection and combination of different diuretic classes, the choice of initial dosages, or the timing of initiation during a patient’s clinical course . Providers from different specialties may have significant variation in diuretic practice patterns.
We identified 46,037 adult ICU stays. Table 1 shows patient characteristics for these ICU stays within each of the five ICU types (Medical, Surgical, Post-Cardiac Surgical, Cardiac, and Trauma). The mean age was 64.2 years, 43.6% were female, 2.4%, 8.5%, and 3.4% were Asian, Black, and Hispanic respectively. More than one in eight (12.7%) patients had a history of chronic kidney disease (CKD), more than half (54.6%) had a history of hypertension, and more than a quarter (27.8%) had a history of heart failure. The mean (± SD) serum creatinine concentration on hospital admission was 1.3 ± 1.0 mg/dL. Cardiovascular and injury/poisoning diagnoses were the most common reasons for admission overall (35.1% and 16.4% respectively).
Using a large, contemporary database, we found that patterns of diuretic use varied by patient characteristics and by ICU type. Admission to the post-cardiac surgery ICU remained the strongest predictor of diuretic use, even after adjustment for clinical factors often associated with fluid overload, including mechanical ventilation and heart failure. Mechanical ventilation was strongly associated with diuretic use, especially carbonic anhydrase inhibitor use. The increased use of carbonic anhydrase inhibitors with mechanical ventilation may be related to concerns of adverse effects of metabolic alkalosis in this population , despite recent evidence that carbonic anhydrase inhibitors may not have clinical benefits such as decreased duration of mechanical ventilation [12,13]. The greater than 2-fold higher prevalence of carbonic anhydrase inhibitor usage in surgical ICUs as compared with medical ICUs cannot be entirely explained by the higher rate of mechanical ventilation in surgical ICUs and may reflect different paradigms for responding to metabolic alkalosis among specialties .