Measurement methods to assess diastasis of the rectus abdominis muscle (DRAM): A systematic review of their measurement properties and meta-analytic reliability generalisation

Study design: Systematic literature review. Background: Diastasis of the rectus abdominis muscle (DRAM) has been linked with low back pain, abdominal and pelvic dysfunction. Measurement is used to either screen or to monitor DRAM width. Determining which methods are suitable for screening and monitoring DRAM is of clinical value. Objectives: To identify the best methods to screen for DRAM presence and monitor DRAM width. Methods: AMED, Embase, Medline, PubMed and CINAHL databases were searched for measurement property studies of DRAM measurement methods. Population characteristics, measurement methods/procedures and measurement information were extracted from included studies. Quality of all studies was evaluated using 'quality rating criteria'. When possible, reliability generalisation was conducted to provide combined reliability estimations. Results: Thirteen studies evaluated measurement properties of the 'finger width'-method, tape measure, calipers, ultrasound, CT and MRI. Ultrasound was most evaluated. Methodological quality of these studies varied widely. Pearson's correlations of r = 0.66-0.79 were found between calipers and ultrasound measurements. Calipers and ultrasound had Intraclass Correlation Coefficients (ICC) of 0.78-0.97 for test eretest, inter-and intra-rater reliability. The 'finger width'-method had weighted Kappa's of 0.73e0.77 for test-retest reliability, but moderate agreement (63%; weighted Kappa = 0.53) between raters. Comparing calipers and ultrasound, low measurement error was found (above the umbilicus), and the methods had good agreement (83%; weighted Kappa = 0.66) for discriminative purposes. Conclusions: The available information support ultrasound and calipers as adequate methods to assess DRAM. For other methods limited measurement information of low to moderate quality is available and further evaluation of their measurement properties is required. (C) 2015 Elsevier Ltd. All rights reserved.