J. Robert Beck

A multivariate data reduction system

The problem of predictive diagnosis based on laboratory data is approached from a mathematical standpoint. A descriptive system is introduced which examines the current information about a clinical problem and identifies best predictors of the problem. Algorithms are described for the assessment of current diagnostic ability, the evaluation of new laboratory tests, and the identification of patients to study for the development of new procedures. The laboratory approach to the predictive diagnosis of iron deficiency is chosen as an example of the system.

A multivariate approach to laboratory practice

The application of multivariate statistical technics to laboratory data analysis is a recent development in clinical pathology. The proliferation in number and type of diagnostic tests requires a simple method to extract predictors of clinical states from data generated by existing laboratory procedures, as well as a method to assess the usefulness of new tests as they are proposed. A series of algorithms can select a “best predictor” subset of laboratory tests and verify this reduced groups predictive diagnostic worth. The established predictive ability of existing laboratory tests can serve as a reference scale when a newly available laboratory test is evaluated. The proposed new test is included with the best predictors already identified and confirmed. In multivariate analysis the usefulness of the proposed new procedure in the predictive diagnosis scheme is determined.

The effect of decision analysis on clinical uncertainty at tumor board

Decision analysis is a formal approach to clinical decision making in the setting of uncertainty. Computer assisted clinical decision analysis was used at tumor board to help examine difficult oncological cases. Patients were identified early in the week and the decisions, possible outcomes, and uncertainties of each decision were modeled. These models were then presented at tumor board where the calculated expected utilities of each decision could be discussed. A description of decision analysis and three representative cases are presented here. These exercises were valuable for at least two reasons. First, modeling the problems confronting practicing oncologists was instructive as such. Second, the mathematical treatment of the probabilities and the utilities of the various outcomes for each decision stimulated a lively debate among the board members, forcing them to quantify probabilities and outcomes that they had previously ascertained qualitatively.

Cost-Effectiveness Analysis of Thyroid Testing Strategies

Over the past 14 years the techniques of clinical decision analysis and cost-effectiveness analysis have been introduced to medicine, and within the last decade these tools have been adopted by academic laboratory scientists. In today’s environment, the laboratory director must consider the introduction of new or enhanced laboratory tests as a comparative problem: how does the novel procedure or analyte fit into both the clinical information stream and in the milieu of existing tests?1 Although a considerable literature deals with problems of technical accuracy and precision in analytic laboratory medicine, this is not true for problems of clinical efficacy and economy.

Book Reviews : Clinical Decisions and Laboratory Use Donald P. Connelly, Ellis S. Benson, M. Desmond Burke, and Douglas Fenderson, Minneapolis: University of Minnesota Press, 1982

discourage such reflexive mimicry and instead seek to render its user a critical thinker, who would act to initiate change rather than reacting reflexively because of change. I feel that this section would have been improved if the authors had spent more effort explaining the advantages and disadvantages of the probabilistic approach, instead of trying to prove its superiority. This book is not written just for physicians. The authors’ engaging, nontechnical writing style will make the case histories enjoyable and provocative reading for patients as well. In addition, I wholeheartedly recommend Chapters 1, 5, and 9-15 to medical students, as an introduction to clinical medicine and the prevailing philosophies of patient management. These chapters should be a therapeutic contrast to their clerkship experiences in tertiary care centers, and truly inspirational to students who wonder whether a physician can still be personally involved with patients in today’s increasingly technological medical system. Strip away from these case studies terms like &dquo;probabilistic paradigm,&dquo; &dquo;Heisenberg’s uncertainty principle,&dquo; and &dquo;multicausal stochastic,&dquo; and one finds &dquo;common sense,&dquo; &dquo;critical judgment,&dquo; and &dquo;sensitivity&dquo; mixed together with a large dose of &dquo;humility&dquo;-perhaps revolutionary ideas for some, and almost certainly a valuable reminder to all.