Imprecision of Laboratory Determinations and Diagnostic Accuracy: Theoretical Considerations^{[*] [**]}

Recent studies suggest the determinations of clinical laboratories must be made more precise than at present. This paper presents a means of examining benefits of improvement in precision. To do this we use a mathematical model of the effect upon the diagnostic process of imprecision in measurements and the influence upon these two of Importance of Diagnosis and Prevalence of Disease. The interaction of these effects is grossly non-linear. There is therefore no proper intuitive answer to questions involving these matters. The effects can always, however, be calculated.

Including a great many assumptions the modeling suggests that improvements in precision of any determination ought probably to be made in hospital rather than screening laboratories, unless Importance of Diagnosis is extremely high.

Kürzlich durchgeführte Studien haben gezeigt, daß die in klinischen Laboratorien gemachten Bestimmungen genauer werden müßten, als sie es zur Zeit sind. Diese Arbeit dient dazu, die durch eine erhöhte Genauigkeit erzielten Vorteile zu untersuchen. Dazu benutzen wir ein mathematisches Modell über die Wirkung der Messungsgenauigkeit auf das diagnostische Verfahren und den Einfluß der Bedeutung der Diagnose und der Krankheitsprävalenz auf diese beiden. Die Wechselwirkung dieser Effekte ist keineswegs linear. Aus diesem Grimde kann es keine eigentlich intuitive Antwort auf diese Fragen geben. Die Wirkungen können jedoch immer errechnet werden.

Unter Einschluß sehr vieler Annahmen deutet das Modell an, daß Verbesserungen in der Bestimmungsgenauigkeit wohl eher in klinischen als in Vorsorgeuntersuchungs-Laboratorien erfolgen sollten, es sei denn, die Bedeutung der Diagnose ist außerordentlich groß.

Key-Words

Clinical Laboratory - Imprecision in Measurements - Diagnostic Accuracy Model

Schlüssel-Wörter

Klinisches Laboratorium - Messungsgenauigkeit - Modell der diagnostischen Genauigkeit

^* Presented in part at the Joint Meeting of the American Association of Clinical Chemists and the Canadian Society of Clinical Chemists, Seattle, Washington, August 8—13, 1971.

^** This work was supported in part by TJSPHS Research Grant HS 00014 from the National Center for Health Services Research and Development.

• References

• 1 Ahlvin R. C. Biochemical screening-a critique. New Engl. J. Med 283 1970; 1084-1086.
• 2 Barnett R. N. Medical significance of laboratory results. Amer. J. clin. Path 50 1968; 671-676.
• 3 Lord Cohen of Birkenhead. Williams E. T, McLacklan G. (Eds) Screening in Medical Care: Reviewing the Evidence. A collection of Essays. Publ. for Nuffield Provincial Hospitals Trust, Oxford University Press; London: 1968
• 4 Collen M. D, Feldman R, Siegelaub A. B. et al. Dollar cost per positive test for automated multiphasic screening. New Engl. J. Med 2S3 1970; 459-463.
• 5 Cotlove E, Harris E. K, Williams G. Z. Biological and analytic components of variation in long-term studies of serum constituents in normal subjects. III. Physiological and medical implications. Clin. Chem 16 1970; 1028-1032.
• 6 Flagle C. D. Automated multiphasic health testing and services: total systems analysis and design. Metli. Inform. Med 10 1971; 201-206.
• 7 Garfield S. R. Multiphasic health testing and medical care as a right. Now Engl. J. Med 283 1970; 1087-1089.
• 8 Good I. J, Card W. I. The diagnostic process with special reference to errors. Meth. Inform. Med 10 1971; 176-188.
• 9 Gowenlock A. H. The influence of accuracy and precision on the normal range. Ann. clin. Biochem 06 1969; 3-5.
• 10 Harris E. K, Kanofsky P, Shakarji G. et al. Biological and analytic components of variation in long-term studies of serum constituents in normal subjects. II. Estimating biological components of variation. Clin. Chom 16 1970; 1022-1027.
• 11 Jacquez J. A. (Ed.) The Diagnostic Process. Proceedings of a Conference held at the University of Michigan May 9-11, 1963. University of Michigan. 1964
• 12 Jacquez J. A. (Ed). Computer Diagnosis and Diagnostic Methods. Thomas; Springfield/Ill: 1972
• 13 Lindberg D. A. B. The Computer and Medical Care. Thomas; Springfield/Ill: 1968
• 14 Lusted L. B. Introduction to Medical Decision Maldng. Thomas; Springfield/Ill: 1968
• 15 Ojemann R. G, Montgomery W. W, Weiss A. D. Evaluation and surgical treatment of acoustic neuroma. New Engl. J. Med 2S7 1972; 895-899.
• 16 Rao C. R. Advanced Statistical Methods in Biometric Research. Chapter 8. Wiley; New York: 1952
• 17 Snyder J. M. An index of suspicion for acoustic neuroma. Letter to the editor. New Engl. J. Med 280 1973; 751.
• 18 Tones D. B. A study of the accuracy and precision of clinical chemistry determinations in 170 Canadian laboratories. Clin. Chem 09 1963; 217-233.
• 19 Williams G. Z, Young D. S, Stein M. R. et al. Biological and analytic components of variation in long-term studies of serum constituents in normal subjects. I. Objectives, subject selection, laboratory procedures and estimation of analytic deviation. Clin. Chem 16 1970; 1016-1021.
• 20 Young D. M, Drake N, Weir R. The advent of chemical screening techniques. Can. med. Ass. J 9S 1968; 868-870.