Methods for systematic and efficient classification of medical practices.
Larry J. Kimbell, John H. Lorant · PubMed · 1973
An efficient classification is a collection of attributes and attribute subdivisions that provides enough distinctions for critical analysis while eliminating enough distinctions for statistical reliability and manageability.It is easy to develop, on a purely logical basis, highly discriminating cross classifications of medical practices (by legal form of practice, by number of physicians, by re- gion, by specialty distribution, by method of payment, etc.).But the process of adding more and more attributes or, equivalently, of using finer and finer sub- divisions of given attributes leads quickly to the result that almost every firm is unique.What is needed is a classification scheme that maintains the distinc- tions crucial to analysis while stopping short of a scheme that implies unreliable statistical estimates and incomprehensible detail.The present inquiry into this question is based on experience in developing a classification scheme to help study group and solo nmedical practices as part of an ongoing project, being conducted jointly at the Human Resources Research Center of the University of Southern California and at the Center for Health Services Research and Development of the American Medical Association, to study major aspects of private medical practice.Specifically required is the development of a methodology for the subdivision of group practices in a man- ner that will permit comparative evaluations between group practices and feefor-service solo practices of similar characteristics.The methodological prob- lems are the focus of the report; empirical results referred to serve solely to illustrate the nature of these problems.The definitive empirical application of the methods developed must await the completion of surveys currently being undertaken by the American Medical Association.Classification procedures are needed for tNvo basic purposes, which are quiteStudies performed under Contract HSM 110-70-354 between the National Center for Health Services Research and Development and the University of Southern California.