Malnutrition in hospitalized adults in the United States, 2016-2019

Author Department

Healthcare Quality; Medicine

Document Type

Article, Peer-reviewed

Publication Date

7-2024

Abstract

Background: Malnutrition in hospitalized patients is associated increased length of stay, cost, readmission, and death. No recent studies have examined trends in prevalence or outcomes of hospitalized patients with a diagnosis of malnutrition.

Objectives: To study the prevalence of malnutrition diagnostic codes and associated hospital outcomes in the United States between 2016 and 2019.

Methods: We conducted a retrospective trends study to identify use of malnutrition codes in hospitalizations in the National Inpatient Sample between 2016 and 2019. We used direct standardization by logistic regression to adjust outcomes of percutaneous gastrostomy tube placement, mechanical ventilation, and death for age, Gagne comorbidity score, and sex. We then used linear regression to test for trends over time by malnutrition type.

Results: Across all hospitalizations, codes for diagnoses of non-severe malnutrition and severe malnutrition were present in 3.7% and 4.1% of hospitalizations, respectively. Codes for any malnutrition increased over time, from 6.6% in 2016 to 8.6% in 2018 (p = .03). Codes for severe malnutrition increased from 3.3% to 4.7% (p = .01). Among hospitalizations with coded severe malnutrition diagnoses, there was a statistically significant decrease in adjusted rate of death over time (-0.54% per year, p = .03) which was not seen in hospitalizations without coded malnutrition diagnoses.

Conclusions: Use of malnutrition diagnosis codes increased significantly from 2016 to 2019. During this time, mortality among hospitalizations with a diagnosis code for severe malnutrition decreased. Though the increased prevalence of malnutrition codes may represent a change in the clinical characteristics of hospitalized patients, the decline in mortality suggests some of the increase may be due to lower threshold for coding and assignment of the diagnosis to less ill patients.

PMID

38982534

Share

COinS