Claims data have become common during the past two decades. The electronic records include information
entered on bills (claims) submitted by healthcare providers to third-party payers. They are an attractive data source;
however, they contain limitations that threaten the validity of studies that use them. We reviewed 168 studies that
employed claims data, published during 2000-2005 in five healthcare journals, to investigate how claims data are being
used and whether their use is appropriate. Healthcare studies in our sample used claims data to select a sample, to
establish healthcare costs, to determine whether specific treatments or procedures had been provided, to ascertain the costeffectiveness
of services, and to establish their accuracy as a stand-in for other measures. Most studies appropriately used
claims data; however, there was a sizable percentage that used claims data in an inappropriate or questionable fashion.