sent the study’s data— frequently steered audiences’ interpretations in a
direction contrary to the investigators’ intentions.
I would like to stress an additional point. This chapter offers a critical
analysis of the ways in which the cluster study has been communicated
and interpreted since 1982 and, following Moss’s 1988 letter to the New
York Review of Books, emphasizes the implausibility of some of its un-
derlying hypotheses based on what would subsequently become known
about AIDS and HIV in the years after the study’s completion. In do-
ing so, there is no wish to diminish the work undertaken by the CDC
investigators. Indeed, their efforts at this early stage, on a limited bud-
get and under high pressure, are a testament to public health workers
who under take challenging assignments with fragmentary information
in high- stake situations.
Mid- Twentieth- Century Shifts in Epidemiology
From its early work in malaria control, and its later acquisition of the
US Public Health Service’s Venereal Disease Division, the CDC had
for decades been associated with control of infectious disease. While it
82
chapter 2
did stretch its efforts into the realms of prevention and chronic illnesses
in the 1970s, the organization’s history and its efforts to eradicate fi rst
syphilis, then smallpox, kept it fi rmly associated with infectious disease
outbreaks in the public mind.8 Alexander Langmuir, the CDC’s chief of
epidemiology from 1949 to 1971, had done much during his tenure to
secure this link by promoting the activities of the agency’s best- known
representatives: the Epidemic Intelligence Service (EIS) offi cers. After
completing the CDC’s six- week training course, these medical doctors
were posted around the country, ready to assist state health offi cials with
any challenging local outbreaks, often at a moment’s notice.9 Due to the
combination of their good work and Langmuir’s talent for publicity, the
EIS offi cers became known as infectious “disease detectives” in popu-
lar accounts from the 1950s onward, most notably in the short stories of
the New Yorker writer Berton Roueché. The detective image was not
imprinted solely in the popular mind- set. The EIS self- consciously em-
braced this characterization too, adopting as its symbol a tie emblazoned
with a well- worn shoe— indicative of detectives, or “gum- shoes,” chas-
ing down leads in the fi eld.10 Harold Jaffe, an infectious disease special-
ist who was one of the lead members of the KS/OI Task Force that con-
ducted the initial investigations into AIDS, would later refl ect, “I have
often been asked what it was like to be one of the early AIDS investiga-
tors. To me, it all began as a medical mystery. I was caught up in being
a ‘medical detective’ without much thought of the broader implications
of what we were investigating. As time went on, however, I gradually be-
gan to see that what we were studying was much bigger than I had fi rst
imagined.”11
The detective image also infl uenced, in an oppositional manner, the
professional identity of a separate group of epidemiologists. From the
midcentury onward, these researchers saw the techniques of their fi eld
shift away from those employed by the CDC disease investigators. This
8. For a discussion of the CDC’s expansion into broader areas of health prevention, see
Etheridge, Sentinel for Health, 276– 320.
9. For a historical view of the CDC’s less widely celebrated yet indispensable public
health advisors, see Beth E. Meyerson, Frederick A. Martich, and Gerald P. Naehr, Ready
to Go: The History and Contributions of U.S. Public Health Advisors (Research Triangle
Park, NC: American Social Health Association, 2008).
10. Ibid., 36– 48; Wald, Contagious, 23– 25.
11. Harold W. Jaffe, “The Early Days of the HIV- AIDS Epidemic in the USA,” Nature
Immunology 9, no. 11 (2008): 1203.
The Cluster Study 83
group of epidemiologists began to deal with the increasingly prevalent
cancers and chronic conditions which formed the bulk of American
morbidity and mortality in the post– World War II period.12 As historian
Allan Brandt has argued persuasively with regard to researchers assess-
ing the harms of smoking, a “new epidemiology” emerged in the middle
of the twentieth century. Investigators at this time revisited the use of
statistical inference and population studies, methods that had been pop-
ular in the nineteenth century before the powerful ascendancy of ideas
related to germ theory. From the 1880s onward, scientists who followed
the bacteriologist Robert Koch’s postulates of causation— searching for
specifi c mechanisms as the causes of specifi c diseases— had achieved
great successes in the laboratory. However, the chronic diseases emerg-
ing as a concern in the mid- twentieth century required a different set
of analytical tools. Researchers returned to the use of statistics— an ap-
proach displaced for decades by germ theory’s emphasis on laboratory
evidence of a disease’s cause— to develop ever more sophisticated means
of investigating disease causation.13 Andrew Moss’s training, fi rst in sta-
tistics and then for a PhD in cancer epidemiology, might be seen as in-
dicative of the fi eld’s developing professional divide.14
While this shift toward population- based epidemiology was under
way, it is important to recognize that at the CDC’s Venereal Disease
Branch, and at other city, county, and state public health departments
around the country, the traditional approach of laboratory- supported
person- to- person disease investigation— “contact epidemiology”— was
still going strong, particularly in the control of venereal disease. In 1957,
the Washington- based VD Division of the Public Health Service was re-
located to Atlanta, absorbed by a growing CDC. In the large- scale na-
tional campaign to eradicate syphilis that followed a special task force
12. Mervyn Susser, “Epidemiology in the United States after World War II: The Evo-
lution of Technique,” Epidemiologic Reviews 7, no. 1 (1985): 147– 77; Mervyn Susser and
Ezra Susser, “Choosing a Future for Epidemiology: I. Eras and Paradigms,” American
Journal of Public Health 86, no. 5 (1996): 668– 73.
13. Allan M. Brandt, The Cigarette Century: The Rise, Fall, and Deadly Persistence of
the Product that Defi ned America (New York: Basic Books, 2007), 118– 23, 148– 52.
14. Andrew R. Moss, “AIDS Epidemiology: Investigating and Getting the Word Out,”
oral history interview conducted in 1992 by Sally Smith Hughes, in The AIDS Epidemic in
San Francisco: The Medical Response, 1981– 1984, Volume II, Regional Oral History Of-
fi ce, Bancroft Library, University of California– Berkeley, 1996, Online Archive of Cali-
fornia, 2009, http:// ark .cdlib .org/ ark: / 13030/ kt7b69n8jn.
84
chapter 2
report to the surgeon general in 1961, more than half of the CDC’s per-
sonnel were working on VD control.15 The report called for increased
funding to support exp
anded syphilis case- fi nding and public- education
efforts. The resulting case- fi nding program, Operation Pursuit, viewed
every index case— the fi rst sick individual brought to investigators’ atten-
tion— as a potential cause of an epidemic, requiring more health workers
to interview and reinterview individuals with syphilis for the names and
contact information of their sex contacts.16
VD Control and the Cluster Method
It was in this climate that Bill Darrow joined the CDC in 1963 as a pub-
lic health advisor, shortly after his graduation from university and fol-
lowing a yearlong contract of work as a venereal disease investigator
for the New York City Department of Health. His early work with the
CDC was representative of the optimistic aims of Operation Pursuit: to
eradicate the scourge of syphilis by tracking down every index case, in-
terviewing each of them for the names of potential contacts, and treat-
ing with penicillin those suspected of being infected. Young, educated,
male, and white (Darrow would later recall playing a role in challenging
this stereotype as a CDC recruiter), the public health advisor was viewed
as an essential player in the investigation of potential syphilis cases and
15. Etheridge, Sentinel for Health, 87– 92, 119.
16. Ibid., 121; John Parascandola, Sex, Sin, and Science: A History of Syphilis (West-
port, CT: Praeger, 2008), 140– 42; Amy L. Fairchild, “The Democratization of Privacy:
Public- Health Surveillance and Changing Perceptions of Privacy in Twentieth- Century
America,” in History and Health Policy in the United States: Putting the Past Back In,
ed. Rosemary A. Stevens, Charles E. Rosenberg, and Lawton R. Burns (London: Rutgers
University Press, 2006), 121– 22. A 1936 report on tuberculosis offers a useful working def-
inition of index case: “that person through whom attention was drawn to the household,
and may not be the only case in the household, or the initial case in the household in point
of time”; H. C. Stewart, R. S. Gass, and Ruth R. Puffer, “Tuberculosis Studies in Tennes-
see: A Clinic Study with Reference to Epidemiology within the Family,” American Journal
of Public Health and the Nation’s Health 26, no. 7 (1936): 689– 90. Note also the following
defi nition of the word case: “In epidemiology, a person in the population or study group
identifi ed as having the particular disease, health disorder, or condition under investiga-
tion”; A Dictionary of Epidemiology. ed. John Last (New York: Oxford University Press,
1983), s.v. “case.”
The Cluster Study 85
bringing them to treatment.17 Such investigators were “carefully selected
for their tact, integrity, intelligence, interviewing ability, personality, and
their understanding of the confi dential nature of their work.”18 Idealistic
in outlook, the young Darrow had felt as though President John F. Ken-
nedy “was talking to me when he said, ‘Ask not what your country can
do for you, but rather ask what you can do for your country.’”19
Darrow and other public health advisors like him had at their dis-
posal a newly developed technique of syphilis control: the cluster
method. Whereas venereal disease case fi nding had traditionally relied
on interviewing solely the sexual contacts named by the patient with
syphilis, “cluster testing” extended this traditional practice by having in-
vestigators interview and test the patient’s friends and acquaintances as
well.20 Incorporating a language that was strongly suggestive of detec-
tive fi ction, a syphilis cluster was expanded beyond the usual contacts
to include “suspects,” those acquaintances whom the person with syph-
ilis suspected might also have the disease, and “associates,” who “may
be a friend or social acquaintance who is named by or found in the com-
pany of sex contacts or ‘suspects.’”21 The cluster method emerged at a
time when certain populations in the United States were being increas-
ingly viewed as mobile disease threats requiring screening, identifi ca-
tion, and control. Sex workers and African Americans, for example,
had long and problematically been associated in the medical and popu-
lar imagination with higher levels of syphilis.22 To these traditional “res-
17. Despite the fact that women had performed contact tracing duties for decades, it
was thought that only men were able to undertake this risky occupation: see William W.
Darrow, “A Few Minutes with Venus, A Lifetime with Mercury,” in The Sex Scientists, ed.
Gary G. Brannigan, Elizabeth R. Allgeier, and A. Richard Allgeier (New York: Addison,
Wesley, Longman, 1998), 162– 63.
18. William J. Brown, Thomas F. Sellers, and Evan W. Thomas, “Challenge to the Pri-
vate Physician in the Epidemiology of Syphilis,” Journal of the American Medical Associ-
ation 171, no. 4 (1959): 392.
19. Darrow, “Few Minutes,” 157.
20. William J. Brown, “Cluster Testing— A New Development in Syphilis Case Find-
ing,” American Journal of Public Health 51, no. 7 (1961): 1043– 48; Etheridge, Sentinel for Health, 91– 92.
21. W. Brown, “Cluster Testing,” 1045.
22. See, for example, Elizabeth Fee, “Sin versus Science: Venereal Disease in
Twentieth-
Century Baltimore,” in Fee and Fox, AIDS: The Burdens of History, 125;
Brandt, No Magic Bullet, 31– 32; James H. Jones, Bad Blood: The Tuskegee Syphilis Exper-
iment, rev. and expanded ed. (New York: Free Press, 1993), 16– 29.
86
chapter 2
ervoirs” of VD— the terminology employed by public health workers at
the time, evoking connections with polluted wells and a residual social
pool of infection— were added two other marginal and frequently mobile
groups: migrant laborers from Mexico and homosexual men.
In 1959, as part of a series that profi led the CDC’s work, a Wash-
ington Post reporter outlined how offi cials were “Meeting the Health
Threat of Mexican Migrants.”23 He described how each year roughly
four hundred thousand Mexican “braceros,” or hand laborers, legally en-
tered the United States to work on farms “from the Rio Grande border
as far north as Michigan.” These workers, he wrote, had until recently
“posed an annual threat to this Nation’s efforts to control venereal dis-
eases.” Without elaborating on why authorities thought there might be
a higher prevalence of venereal disease in this population, the reporter
warned that an unfettered US need for laborers had resulted in these
Mexican workers being hired without attention to their health. In his
view, the result of such an indiscriminate search for labor was clear: “the
ready reception was perpetuating self- imported contagions of syphilis,
gonorrhea and other venereal diseases.” The reporter quoted William
J. Brown, the director of Venereal Disease Control at the CDC, who
praised the “newly developed technique” of cluster testing, which he as-
sured the nation could halt the rising incidence of syphilis. According to
Brown, by interviewing and testing those in the “living and travel envi-
ronments” of the individuals with syphilis, investi
gators could raise the
ratio of uncovered cases from 34 for every 100 checked using conven-
tional methods to 54 for every 100 checked. One might wonder whether
the approach simply allowed medical authorities to test based on race,
class, and later sexual behavior, providing a scientifi c scaffold onto
which cultural stereotypes could be erected. Perhaps seeking to neutral-
ize such criticism, the reporter pointed out that Mexico benefi ted from
the inspections too, “since the names and addresses of the VD- infected
and their home community sex contacts are supplied to the appropriate
Mexican local health departments.”24
23. Nate Haseltine, “Meeting the Health Threat of Mexican Migrants,” Washington
Post, February 25, 1959, B7.
24. All quotations in the paragraph in ibid. Brown headed the Venereal Disease Divi-
sion of the CDC from 1957 to 1971, working for years to promote the need to seek out and
treat venereal disease in most of the American population. He would subsequently receive
less favorable attention for his role in the continuation of the Tuskegee syphilis experi-
ment: see Jones, Bad Blood, 180– 209.
The Cluster Study 87
Brown published a separate article dealing with the diffi culties that
mobile populations presented for VD control efforts, in which he out-
lined the modern solutions that disease control workers had at their dis-
posal. While the article’s text acknowledged the effects of mobile Amer-
icans in spreading disease— particularly those taking part in the massive
postwar increase in leisure travel by automobile—
the accompanying
photographs suggested a racialized understanding of disease spread, one
which viewed national health security as under threat. White male health
offi cials out “in the fi eld,” resplendent in white shirts and dark ties, ex-
tended the virtues of science by performing rapid syphilis tests, in one
photo on a darker- skinned seaman and in another on a group of Mexican
workers. A map positioned above the photograph of these workers sug-
gests a slow, northward spread of VD, unchecked but for the unrelenting
efforts of the CDC’s scientifi c testing and treatment (see fi g. 2.2).25
“Reservoirs of Infectious Venereal Disease”
To this group was added another population that had begun to attract
Patient Zero and the Making of the AIDS Epidemic Page 15