Origin of terms "sensitivity" and "specificity"

Question

Who coined the terms "sensitivity" and "specificity"—meaning the complements of false positives, and false negatives, respectively in tests and measures—and when did they first do so?

Bonus points if there is a motivating etymology.

Answer 1

Binney, N., C. Hyde and P. M. Bossuyt. ‘On the origin of sensitivity and specificity’. Annals of Internal Medicine, 174 (2021): 401–7. 10.7326/M20-5028.

Answer 2

tl;dr (this answer is a bit of a book report): The terms sensitivity and specificity have their origin in the roots of immunology in the first years of the 20th century, and work towards antibody testing for syphilis.

As @NickCox and others in the published literature have suggested (e.g., Lilienfield, 2007), Yerushalmy was an exponent of the biostatistical concepts of sensitivity and specificity in a 1947 article on radiological tests for tuberculosis. However, it is reasonably easy to find earlier uses of these terms in the context of medical tests (e.g., Gilman, Boerner, and Lukens, 1940) using academic search engines.

In the 2021 article by Binney, Hyde, and Bossuyt cited by @PeterMilne (and well worth the read!), the authors trace the genesis of these terms' use to immunology, and

the development of serology in the early 20th century. They are closely linked to the development of the Wassermann reaction as a test for syphilis, and to notions of species of pathogen and of specific disease entities. […Where it was] used to distinguish diseased from non-diseased patients.

The terms were already in established use in the syphilis serology literature by the late 30s and early 40s. According to these authors, the concepts of sensitivity and specificity spread to disciplines like clinical epidemiology (and presumably from there to other epidemiologies :)—including for other diseases, such as tuberculosis, decision making in medicine, and disciplines where evaluating tests is important.

According to Binney, Hyde, & Bossuyt, the term specificity has its origins in experiments around 1901 demonstrating the capacity for blood serum from a living animal previously exposed to a known pathogen to successfully destroy that specific pathogen (in contrast to other pathogens). They go on to describe Jules Jean Baptiste Vincent Bordet's advancement of the immunological concepts of antibody and complement from this specific destructive power, where antibodies were described by him as a mechanism to sensitize an organism's immune response to a specific pathogen. From these definitions, Binney, Hyde, & Bossuyt argue, come the concept that “specific entities must be determined to identify infected individuals,” and the development of the Wasserman antibody test^* first embodied these concepts in technology. These authors cite work by Craig a few years later in 1910 presenting tables of numbers testing positive and negative among cases of syphilis, and percent testing positive among cases of syphilis, where the term specificity of the test was used to describe “healthy individuals in whom [syphilis] could be absolutely excluded, we tested seventy individuals, with a negative result in every instance.”

The concept of sensitivity of the test has a blurrier origin, with Binney, Hyde, and Bossuyt writing:

Through the 1920s, the term “sensitive” referred both to the fraction of true positives and, sometimes in the same article, to the delicacy of an immunologic reaction: its propensity to return positive reactions, whether these were true or false positives.

They note that this overloaded meaning extended through to the 40s in the literature, until 1942 when American immunologist Reuben Leon Kahn wrote “Strictly speaking, the term ‘sensitivity’ when used in connection with tests for syphilis should apply only to cases of syphilis.”

^* The Wasserman test is not actually specific immunologically, but is diagnostically useful.

References

Craig, C.F. (1910). Observations upon the Noguchi modification of the Wassermann complement fixation test in the diagnosis of lues in the military service. Journal of Experimental Medicine. 12, 726–45.

Gilman, R. L., Boerner, F., & Lukens, M. (1940). A Simplified Complement Fixation Technic for the Diagnosis and Treatment Of Syphilis: Its Sensitivity and Specificity. Archives of Dermatology and Syphilology, 41(1), 32–37.

Kahn, R. L. (1942). Serology in Syphilis Control: Principles of Sensitivity and Specificity, With an Appendix for Health Officers and Industrial Physicians. Williams & Wilkins.

Lilienfeld, D. E. (2007). Abe and Yak: the interactions of Abraham M. Lilienfeld and Jacob Yerushalmy in the development of modern epidemiology (1945-1973). Epidemiology. 18, 507–14.

Yerushalmy, J. (1947). Statistical problems in assessing methods of medical diagnosis, with special reference to X-ray techniques. Public Health Reports. 62,1432–1449.