Force 1983– 85, Darrow Papers.
84. James W. Curran, “Briefi ng, Task Force on Kaposi’s Sarcoma and Opportunistic
Infections,” April 2, 1982, folder: AIDS Task Force: 1982, Darrow Papers.
106
chapter 2
not violate their trust, and assure them that although I was not a medi-
cal doctor and couldn’t help them at all, that whatever I learned I would
try to see to it [that] it would be benefi cial for them and other people that
they were concerned about.”85 Darrow also had years of experience in
terms of prompting VD patients to name their sexual partners, obtain-
ing important information by inference, and maintaining the commit-
ment to the legwork required in VD control. Still, there was a key dif-
ference in this instance: “I had worked on problems related to syphilis,
gonorrhea, herpes, and lots of other sexually transmitted diseases that
were very annoying and disturbing but rarely killed people. And sud-
denly we were fi nding a disease that was fatal, and for the fi rst time I was
going to come face- to- face with some people who . . . were dying from
this disease.”86 At a time when declarations of homosexual activity could
lead to felony charges in several states, the CDC investigators were well
aware of many possible reasons for noncooperation. Darrow and Auer-
bach were struck by the men’s concern, their desire to help researchers
learn more about the syndrome, and the willingness of most to speak
freely. They later gave special thanks to these men for their openness
and trust.87 Through their interviewing efforts in March and April 1982,
Auer bach and Darrow were able to establish sexual links between nine
out of thirteen of the twenty- six earliest reported KS/OI cases in south-
ern California. To their excitement, four men named the same out- of-
town patient— the CDC’s Case 57, whose details Darrow already knew
since he was an individual with an early date of symptom onset— as a
sexual contact. These four men reported that they or their acquain-
tances had had sexual exposures with this man between nine and eigh-
teen months before the onset of their symptoms.88 Locating and reinter-
viewing this individual became a top priority.
Given subsequent misunderstandings, it might be useful at this point
to emphasize how unlikely it would be that the cases reported to the
CDC by the spring of 1982 would represent all existent cases of the dis-
ease in California at that time. It is worth drawing on our current knowl-
85. Darrow, recording C1491/21, tape 1, side A.
86. Ibid.
87. Auerbach et al., “Cluster of Cases,” 491.
88. “Documentation for MMWR Article: ‘A Cluster of Cases of Kaposi’s Sarcoma and
Pneumocystis Carinii Pneumonia among Homosexual Male Residents of Los Angeles and
Orange Counties, California,’” May 12, 1982, p. 8, folder: AIDS Task Force: 1982, Darrow
Papers.
The Cluster Study 107
edge of HIV infection to consider the hurdles a prospective case subject
would need to clear to be included in the cluster study. Such an individ-
ual would have had to be infected with HIV some time before, likely
for several years. Subsequently he (and it was almost entirely men being
investigated during this period) would have had to feel suffi ciently un-
well to seek medical attention. If he received a medical examination, the
health care staff would have needed to be aware of the CDC’s case defi -
nition, recognize that his presenting signs and symptoms conformed to
it, and then report this individual to the CDC. Failing such a report, the
CDC’s task force members would have had to uncover the case in their
retrospective surveillance activities. Finally, the task force members
would have needed to be able to gather suffi cient information about the
patient’s sexual history to link him to the cluster. In short, it would be a
mistake to interpret, as some observers subsequently did, the twenty- six
earliest cases reported in California by April 1982 as the twenty- six fi rst
cases of infection, in absolute terms, caused by the transmissible agent
that would subsequently be identifi ed as HIV.
Letters and Numbers
The Los Angeles cluster study, as this work became known, was fi rst pub-
lished on June 20, 1982, in the CDC’s Morbidity and Mortality Weekly
Report and would soon be cited as evidence supporting the theory that
a sexually transmissible agent caused the immune suppression lead-
ing to AIDS.89 While there was no suggestion in the initial report that
the epidemic had begun in California, subsequent communications be-
tween KS/OI Task Force members suggested that they believed the clus-
ter might lead them to the root of the problem. “In order to discover the
source of the current outbreak of acquired cellular immunodefi ciency,”
Darrow and a colleague wrote in an internal proposal to extend the clus-
ter, “the interconnected series of cases uncovered in southern Califor-
nia should be followed as far as it may go.”90 This distance proved to be
89. Task Force on Kaposi’s Sarcoma and Opportunistic Infections, CDC, “A Cluster,”
305– 7.
90. “Relationships among Cases of and Contacts to Kaposi’s Sarcoma and Opportunis-
tic Infections in New York City: A Proposal,” fi nal draft of report, July 8, 1982, p. 1, folder:
KSOI: Cases and Contacts in New York City, Darrow Papers; emphasis added.
108
chapter 2
to New York and beyond, as the follow- up article published in 1984 dem-
onstrated. In it, Auerbach and Darrow were able to link forty cases of
AIDS in ten cities in an implied web of transmission, though ultimately
they did not speculate as to the geographic point of entry for the trans-
missible agent.
Key to their success was the assistance of Case 57, an individual who
received the designation “Patient 0” in the 1984 article and whose per-
sonal sexual network extended to a large number of men in many cities.
This non- Californian resident was able to provide them with a list of
72 contacts from his personal address book, nearly 10 percent of the
750 partners he estimated he had had sex with in the previous three
years, through whom they were able to continue their cluster research.
His ability to provide names stood out sharply in contrast to the other
men linked to the cluster. More than 65 percent of these men had re-
ported more than 1,000 lifetime sexual partners, and more than 75 per-
cent had reported more than 50 partners in the year before they expe-
rienced symptoms. Yet most of these men were able to provide only a
handful of names. For example, in the year before the onset of his KS
in February 1978, the man known as Case 335 (LA 1 in the cluster) re-
ported 240 sexual partners and named 6. Case 32 (the cluster’s LA 6) re-
ported an estimated 500 sexual partners in the twelve months prior to his
becoming sick in November 1980; he was able to name 5 of them.91 Thus,
by comparison, the trove of information possessed by Case 57 was quite
remarkable. By 1982, the man was
spending more time in San Francisco
and had recently updated his book, discarding, to the disappointment of
some task force members, many older names from other cities.92
Both the desire to locate the source of the epidemic, implicit in con-
tact epidemiological practice and long part of its history, and the trans-
formative coding assigned to cluster cases appear to have worked against
the originally stated intentions of the study.93 Auerbach and Dar row
91. “Table 1: Cases of KS/OI Reported among Homosexual Men in LA or Linked to
LA Cluster,” in Auerbach and Darrow, “Los Angeles Cluster: Background,” May 12, 1982,
p. 6, folder: AIDS Task Force: 1982, Darrow Papers.
92. Some task force investigators, including Harold Jaffe, would subsequently wonder
whether discarding these names may have been an excuse to protect some of his contacts,
though there is no additional evidence to suggest that this was the case.
93. William Coleman describes the work of mid- nineteenth- century French and British
epidemiologists who emphasized the importance of attempting to locate the “initial vic-
tims” of epidemics of yellow fever and cholera to determine whether they shared exposures
The Cluster Study 109
needed to integrate numerical identifi ers from several sources into their
study and to simplify the complex webs of sexual contact that they had
uncovered. Each patient had been assigned a unique case number when
reported to the CDC but also likely had another unique coding in the
local public health district. For example, “Patient 0” had been CDC’s
“Case 57,” while at the same time in San Francisco, he was labelled
“D
K+ (Montreal, NY, SF)” by Selma Dritz, the city’s public health de-
partment epidemiologist who was investigating the outbreak there (see
fi g. 2.3).94 By May 1982, in an unpublished report leading up the June
MMWR article about the Los Angeles cluster, he became “Patient O”—
short for “Out[side]- of- California” or “non- Californian KS”— before ac-
quiring his fi nal, numerical designation. In early September 1982, Dar-
row wrote an internal report to his colleagues, summarizing his New
York investigations during July and August, in which he used the CDC’s
case numbers to designate individual cases and employed the prefi x Pa-
tient for each one. One month later, he annotated a separate copy of
the report by hand to be typed up for distribution beyond the CDC to
partner organizations in New York and elsewhere. On this report, he
replaced each cluster- linked case number with a sequential city- based
number in order of symptom onset date within each city’s cluster cases—
the style of designation which would eventually feature in the 1984 arti-
cle—and wrote the letter O in substitution for Patient 57’s number to ab-
breviate “Out- of- California.” What happened next may have been the
result of a secretarial error, or indeed due to the fact that the CDC’s
typewriters produced a capital letter O and a numeral 0 that were virtu-
ally indistinguishable. In any event, while Darrow and Auerbach would
continue to refer to “Patient O” using a letter in conversation (sounding
like “Patient Oh”) and in print, others discussing the investigation began
to use the numeral 0 (resulting in speech sounding like “Patient Zero”),
interpreting the ambiguous oval as a digit, alongside the other city case
numbers.95 Thus the term “Patient 0” was born, and it was this code that
to hazardous local conditions or to other sick individuals; see W. Coleman, Yellow Fever in
the North, 20– 21, 182– 87.
94. In the portrait of Dritz shown in fi gure 2.3, the positioning of the subject in front
of her work apparatus, representing the “proof” of her authority, links this photograph to
other works of scientifi c portraiture; see Ludmilla Jordanova, Defi ning Features: Scientifi c
and Medical Portraits 1660 – 2000 (London: Reaktion Books with National Portrait Gal-
lery, London, 2000).
95. Copies of the original, the annotated, and the retyped versions of this memo are
110
chapter 2
Figure 2.3 Selma Dritz in her offi ce, November 15, 1982. Photograph by Jerry Telfer for
the San Francisco Chronicle. © Jerry Telfer / San Francisco Chronicle / Corbis. This por-
trait shows the public health offi cial standing in her small offi ce, next to a locker, pointing
to a blackboard that depicts known connections between early cases of AIDS in gay men.
On the right half of the blackboard behind her, Gaétan Dugas’s information is listed in a
rectangle at top left—“D
K
+ (Montreal, NY, SF)”— the fi rst, and perhaps most important,
example on her list, and the man whom she would confront that month about his continu-
ing sexual activity. Arrows linking cases were labeled as “RM” for “roommate” and “S?”
to query whether there had been sexual contact between them. These notations can be
compared with similar images in Dritz’s presentation slides, which are now held in the spe-
cial collections of the University of California– San Francisco. A cropped version of this
photograph, covering roughly one- quarter of the width of the newspaper’s page, accom-
panied an article by the Chronicle’s science editor: David Perlman, “A War against ‘Gay
Plague,’” San Francisco Chronicle, November 17, 1982, 5.
designated the CDC’s Case 57 in the most well- known publication of the
cluster study in 1984.
The consequences of this change cannot be overemphasized, given
the multitude of meanings for the word zero. Particularly noteworthy
defi nitions include: “a worthless thing or person,” “an absence or lack
in Darrow’s fi les: “Trip Report to New York City: July 12– 16 and August 3– 6, 1982,” Sep-
tember 3, 1982, folder: KSOI: Cases and Contacts in New York City. A copy of one page
from the retyped version, in the archived fi les of the New York City Department of Public
Health, is reproduced in Colgrove, Epidemic City, 112.
The Cluster Study 111
of anything,” “the initial point of a process or reckoning . . . the starting-
point, [and] the absolute beginning.”96 Such a shift was further compli-
cated by the rather nonspecifi c identifi cation of “Patient 0,” in the 1984
publication, as the “index case,” which in contemporary usage usually
meant the fi rst case to come to the investigators’ attention, as opposed
to the primary case— which was the earliest to occur in time— although
neither was true for “Patient 0” in this scenario.97 Judging from previ-
ous CDC- reported outbreaks, such as one of smallpox in Sweden in
1963, it appears to have been the agency’s practice to use the term index
case to refer to the “source” case— a case to which other, more specifi c
terms such as primary or initial might be applied. A report of the Swed-
ish smallpox outbreak in MMWR noted that “the fi rst case to be iden-
tifi ed occurred in an unvaccinated 19- year- old bricklayer (Case 7) who
had onset of fever, vomiting, and backache on May 5.” He later died, as
did three of the other twenty-
four confi rmed case subjects. Epidemio-
logic research determined that the “original source of the outbreak was
a 24- year- old seaman who after two weeks residence in Australia left
Darwin on March 22 on BOAC Flight #709,” stopping briefl y at Dja-
karta, Singapore, Rangoon, Calcutta, Karachi, Teheran, and Damascus
on the way to Zurich, where he boarded another plane and fl ew to Swe-
den. A fi gure accompanying reports of the smallpox investigation indi-
cated three generations of confi rmed and suspected “indigenous” cases
that stemmed from the twenty- four- year- old. The seaman was labeled
as the “index case” in the fi gure, even though he was not the fi rst pa-
tient to come to investigators’ attention in the outbreak. A tabular rep-
resentation of the investigation numbered the cases in order of the indi-
viduals’ symptom onset and listed the “presumed source of infection”
for each. While the indigenous cases were each linked to an earlier case
number (Cases 2 to 5 linked to Case 1, Cases 6 and 7 to Case 2, etc.), the
presumed source of Case 1’s infection was listed far more vaguely— and
problematically— as “Southeast Asia.” This historical example not only
demonstrates the ambiguities of the CDC’s use of the term index case
96. Oxford English Dictionary Online, s.v. “zero,” accessed February 22, 2017, http://
www .oed .com/ view/ Entry/ 232803 ?rskey = SMeMCR & result = 1.
97. Index case was defi ned as “the fi rst case in a family or other defi ned group to come
to the attention of the investigator” in Last, Dictionary of Epidemiology, s.v. “index case.”
It may not be not surprising, therefore, that patient zero is even now used interchangeably
as a term denoting index case and primary case.
112
chapter 2
but also indicates the ease with which reports of well- resourced epidemi-
ologic investigations in one country can, in a few words, construct whole
regions of the world as the source of an infectious outbreak.98
Returning to the cluster study, in mid- 1983 Darrow and Auerbach
were revising their work for publication and seeking peer review. One
experienced epidemiologist, who had seen Auerbach present the study
at a CDC conference that spring, entered into correspondence with the
Patient Zero and the Making of the AIDS Epidemic Page 19
