Electronic medical records (EMRs) can be used to identify and categorize weight loss in older adults, but research has not scrutinized methods for doing so. Through a modified PRISMA protocol, we systematically reviewed published methods for quantifying weight change from EMRs. Articles (all available through July 2016) were identified through PubMed and SCOPUS searches, screened, and evaluated. We abstracted relevant data and tabulated the methods to assess weight change. The 13 selected articles showed little consistency in the approach to key methodological issues: 1) time ranges assessed; 2) removal of spurious values; 3) metrics to quantify weight change; 4) number of measures needed to estimate change; 5) threshold for significant weight change; and 6) relation to ideal weight. There was essentially no consensus around how to identify and categorize weight loss. Further investigation is needed to establish scientifically validated and clinically useful algorithms, accounting for the six issues above.
View details for DOI 10.1080/21551197.2017.1282390
View details for PubMedID 28207372