Quality of reporting of bioelectrical impedance analysis (BIA) studies evaluating body fluid volumes: The need for standardization

Abstract

Aim: To assess the quality of reports where bioelectrical impedance analysis and dilution techniques were used to estimate body fluid volumes. Methods: From electronic searches of MEDLINE, the Cochrane Library, CINAHL, BIOSIS, EMBASE, and PROQUEST we identified 50 studies published between 2002 and 2007. The quality of reporting was assessed by using 40 items from the score system described in 2002 [1] which defines quality based in 4 headings: a. Study design, b. Subject selection, c. Procedures and d. Calculations and statistical analysis. Each heading can get 100 points. The maximum score is 400. Results: The average score was 258.81 ± 44.9 SD, representing 64.7% of 400 possible points. There has been a 10% improvement of reporting from 2002 to 2007, due mainly to advances in study design and subject selection report (sections A and B). However, the overall percentage of reporting was insufficient for the heading Procedures (34.9 %). The reporting of methodological performance for BIA was particularly poor for the items environment temperature, removal of metallic objects, test of BIA reproducibility, doing strenuous exercise before the test, consideration of menstrual cycle, use of non-conductive surface, no random exclusion of patients and cleaning of the skin with alcohol. Dilution procedures reports were also poor (40%) and failed to inform about discarding the first sample of urine when used to measure body tracer (12%) and storing the samples in airtight tubes (14%). Conclusions: The quality of publications reporting BIA and dilution procedures have improved but still is suboptimal and there is room for improvement. It is critical that, in reporting BIA studies, authors follow standardized procedures as well as consider other variables such as using different bioimpedance analyzers and report them. © 2009 Springer-Verlag.