Each of the datasets has many different diagnosis and procedure codes to represent a patient’s condition. There are thousands of potential codes, and millions of potential combinations of codes. In order to use patient diagnosis and procedure information in statistical models, there has to be some form of compression as there are far too many to include all of them. Therefore, there has to be some method to compress codes. While such methods are discussed in detail in Cerrito (2009), they will be discussed briefly here. Information codes are used in billing and administrative data to define patient conditions, and also to define patient treatments. These codes are used to define patient severity indices. Therefore, it is absolutely essential to both understanding the severity indices, and to defining such severity indices to be able to work with these codes. The most difficult data to work with are contained within claims databases where different coding methods are used by different providers; the different codes must be reconciled in some manner.
TopBackground
There are a number of different approaches to compressing patient diagnosis and procedure codes. One of the most common is to decide upon specific inclusion/exclusion criteria and then to extract patients who have those specific criteria. (Bateman, Simpson, Bateman, & Simpson, 2006; Glance, et al., 2009; Goff, et al., 2007; Mountford, et al., 2007; Olsen, et al., 2008) Another, as discussed in previous chapters, is to find the most frequently occurring codes in a subpopulation of patients, and to use them in the analysis. While electronic medical records can make text analysis easier to perform to investigate health outcomes, emphasis continues to be on the benefits to the providers, and also on the standardization of language. (Doyle & Doyle, 2006; Jaspers, Knaup, Schmidt, & Jaspers, 2006; Shaw & Shaw, 2006; Yamamoto, Khan, Yamamoto, & Khan, 2006)
Another common method of compression is to define a patient severity score. (Burd, et al., 2008; Chung, Krishnan, & Chakravarty, 2007; Kuykendall, Ashton, Johnson, & Geraci, 1995; Ricciardi, et al., 2007; Rutledge, et al., 1997; West, Rivara, Cummings, Jurkovich, & Maier, 2000) There are several different methods available that define a patient score. Typically, a number of diagnosis codes are used to define the score, and patient information regarding these diagnoses are extracted in the same way that inclusion criteria are used. One common method is called the Charlson Index.
Still other types of compression are proprietary, and often rely upon physician consensus panels to define a patient risk assessment. A novel method discussed in detail in Cerrito (2009) and Cerrito (2007) is to use text analysis to define patient clusters of conditions. (P. B. Cerrito & Cerrito, 2008)