Diagnostic Accuracy and Agreement Between Computed Tomography and a Single Frequency Bioelectrical Impedance Analysis
Teresa Ellen Brown1, Oliver Ritchie2, Judy Bauer3
1School of Human Movement and Nutrition Sciences, The University of Queensland, St Lucia, Queensland, Australia; Dietetics and Food Services, Royal Brisbane & Women's Hospital, Brisbane, Queensland, Australia.
Background:
Loss of muscle mass (sarcopenia) and malnutrition are associated with adverse patient outcomes, highlighting the crucial role of nutrition management in patients with head and neck cancer. Accurate assessment of body composition is necessary to determine skeletal muscle depletion and adequacy of nutrition interventions; however, current bedside assessment tools have limitations.
Objective:
To evaluate the diagnostic accuracy of a single-frequency bioelectrical impedance analysis (SF-BIA) device for identifying skeletal muscle depletion, and the agreement between SF-BIA estimates and computed tomography (CT)-derived estimates of fat-free mass (FFM) and fat mass (FM) in patients with head and neck cancer.
Design:
Secondary analysis of a prospective cohort of patients with head and neck cancer undergoing treatment of curative intent. Body composition was assessed at baseline and 3 months posttreatment using SF-BIA and CT analysis at the third lumbar vertebra by a single trained analyst.
Participants/Setting:
Participants with head and neck cancer (N = 131) referred for prophylactic gastrostomy placement before concurrent chemoradiotherapy were recruited from a tertiary hospital in Australia (September 2012 to June 2015). For this analysis, patients were included if baseline diagnostic CT scans were available for body composition analysis (n = 99) and a CT scan at 3 months posttreatment (n = 87).
Main Outcome Measures:
Diagnostic accuracy of an SF-BIA device for identifying CT-defined muscle mass depletion; agreement between CT and SF-BIA-derived estimates of FFM and FM at baseline and 3 months posttreatment.
Statistical Analyses Performed:
Diagnostic accuracy was evaluated using sensitivity, specificity, positive predictive value, and negative predictive value. Agreement between body composition analysis methods was compared using Bland-Altman analysis.
Results:
Patients were predominantly men (91%), with a mean age of 61.2 ± 9.8 years, and most had oropharyngeal cancer (86%). Compared with CT, SF-BIA demonstrated high specificity (89.1%) but low sensitivity (22.6%) for identifying skeletal muscle depletion (Class I and II), with a positive predictive value of 0.708 and a negative predictive value of 0.500. At baseline, the SF-BIA device underestimated FFM (bias -7.7 kg) and overestimated FM (3.6 kg), with negative proportional bias. The change between baseline and 3 months posttreatment showed the SF-BIA device overestimated percentage change in both FFM (6.3%) and FM (7.8%) with positive proportional bias.
Conclusions:
The SF-BIA device demonstrates poor diagnostic accuracy for identifying skeletal muscle depletion and poor agreement with CT-derived measures of body composition at both the individual and group levels. These findings indicate that SF-BIA device estimates should not be used in isolation for diagnosing skeletal muscle depletion or considered interchangeable for body composition measurement in patients with head and neck cancer.
