See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/13454777

Home Sample Collection Tests for HIV Infection

Article in JAMA The Journal of the American Medical Association · December 1998
DOI: 10.1001/jama.280.19.1699 · Source: PubMed

CITATIONS READS
71 84

1 author:

Bernard Branson
Scientific Affairs LLC
161 PUBLICATIONS 11,863 CITATIONS

SEE PROFILE

All content following this page was uploaded by Bernard Branson on 26 May 2014.

The user has requested enhancement of the downloaded file.

 Toward Optimal Laboratory Use
Home Sample Collection Tests
for HIV Infection
Bernard M. Branson, MD

Context.—Home sample collection (HSC) tests allow persons to test themselves METHODS
for human immunodeficiency virus (HIV) infection at home without medical super- Data from both HSC kit manufacturers
vision. Characterizing the use of such tests can help assess their potential effect were included in this analysis. Demo-
on public health efforts to prevent and control HIV. graphic questions asked by the 2 manu-
Objective.—To describe use of HIV HSC tests. facturers were comparable, but the meth-
Design.—Retrospective descriptive analysis from data collected by test manu- ods used to elicit information differed.
facturers during 1996 and 1997. Home Access Health Corporation
Setting.—United States. (HAHC, Hoffman Estates, Ill) requires us-
ers to register their test kit before sub-
Participants.—Volunteer sample of consumers who used either of 2 HSC tests. mitting a specimen. During this tele-
Main Outcome Measures.—Demographic and behavioral aspects of users. phone registration, users are asked to
Results.—During the first year of availability, 174 316 HIV HSC tests were sub- provide demographic and behavioral in-
mitted to the manufacturers for analysis; 0.9% of the results were positive for HIV, formation via Touch-Tone telephone. Dur-
and 97% of all users called to learn test results. Survey responses from 70 620 ing a second call to obtain test results, us-
HIV-negative and 865 HIV-positive users revealed that most were men, white, and ers with negative results receive them and
aged 25 to 34 years; HIV prevalence was highest among nonwhites, aged 35 to counseling via prerecorded message un-
44 years, men who have sex with men, and injection drug users. Bisexual men ac- less they elect to speak with a counselor.
counted for a large proportion of HIV-positive users. Nearly 60% of all users and Users whose specimens are positive, in-
49% of those who tested HIV positive had never been tested before. Telephone determinate, or unsuitable for testing are
connected to a counselor who may elicit ad-
counselors found that 23% of HIV-positive users already had a source of follow-up ditional demographic and risk informa-
care, 65% accepted referrals, and 12% had tested themselves to evaluate the ef- tion and who also enters a summary of the
fects of antiretroviral therapy. telephone interaction in a call log. Home
Conclusions.—Home sample collection tests for HIV were used by persons who Access Health Corporation provided line-
were at risk for HIV and by persons who did not use other testing. Most HIV-positive listed data for July 1996 through Septem-
users either had a source of medical care or received referrals. ber 1997 for this analysis, including both
JAMA. 1998;280:1699-1701 the demographic survey and the call log.
Direct Access Diagnostics (DAD,
IN 1996, the Food and Drug Administra- most receive a prerecorded message with Bridgewater, NJ) used a pamphlet in-
tion (FDA) approved 2 products for home their result and counseling information. cluded with the HSC test kit to provide
sample collection (HSC) for human im- All users who test positive or indetermi- pretest counseling information. The us-
munodeficiency virus (HIV) testing. The nate and those whose specimen was un- ers’ first contact with the company took
HSC kits are marketed directly to con- suitable for testing are connected to place when they called for results. Users
sumers. Users perform a finger-stick to a counselor who provides test results, were asked for demographic and behav-
obtain a dried-blood spot specimen on fil- counseling, and referrals or follow-up ioral information after they received their
ter paper. The specimen is identified by instructions. test results. Direct Access Diagnostics
an anonymous code number and mailed Considerable debate preceded ap- surveyed only HIV-positive users from
directly to a laboratory for HIV antibody proval of HSC HIV testing. Proponents May 1996 through mid February 1997 and
testing by enzyme immunoassay. Reac- maintained that its availability would in- began to survey all users in February
tive specimens are confirmed by West- crease testing,1 but opponents raised 1997. On June 26, 1997, the company an-
ern blot or immunofluorescence assay. concerns about negative consequences nounced the withdrawal of its product
The user calls a toll-free telephone num- after people received results by tele- due to lack of demand, but continued to
ber to obtain test results, counseling, and phone and the possible negative effect provide test results and administer the
referrals. Persons with negative results on public health surveillance and con- survey until August 19, 1997. Informa-
may choose to speak with a counselor, but trol.2 Under regulations that permit the tion on the total number of tests, test re-
postmarketing evaluation of products sults, and sample adequacy was available
From the National Center for HIV, STD, and TB that are brought to market before all out- for May 1996 to August 1997, but demo-
Prevention, Centers for Disease Control and Preven- standing questions are resolved, the graphic and behavioral information com-
tion, Atlanta, Ga.
FDA required HSC product manufac- prises only those users who called from
The mention of trade names, commercial products,
or organizations does not imply endorsement by the US turers to collect data on users and to February 10 through August 29, 1997. Di-
government. share this information with public health rect Access Diagnostics provided sum-
Corresponding author: Bernard M. Branson, MD, Di-
officials. mary information aggregated by quarter
vision of HIV/AIDS Prevention, Centers for Disease
Control and Prevention, 1600 Clifton Rd, Mail Stop E-46, This article describes users during the and did not provide access to the call log.
Atlanta, GA 30333 (e-mail: BMB2@cdc.gov). first year of HSC test availability and
Reprints: National Center for HIV, STD, and TB Pre-
vention, Office of Communications, Mailstop E-06, Cen- compares them with clients of publicly Toward Optimal Laboratory Use section editor:
ters for Disease Control and Prevention, Atlanta, GA 30333. funded testing programs. George D. Lundberg, MD, Editor, JAMA.

JAMA, November 18, 1998—Vol 280, No. 19 Home Sample Collection Tests for HIV Infection—Branson 1699
©1998 American Medical Association. All rights reserved.

Downloaded From: http://jama.jamanetwork.com/ on 02/25/2013

 Table 1.—Home Collection Users, by Test Result, May 1996 Through September 1997* percentage may be even higher because
No. (%) of Users
the DAD survey did not distinguish bi-
No. (%) of Tests Who Called for Results sexual men from other men who have
sex with men. Considering data from
HIV Result DAD HAHC DAD HAHC
only HAHC users, bisexual men ac-
Positive 858 (0.9) 636 (0.8) 818 (95) 610 (96)
counted for 38% of the positive results.
Negative 90 662 (94) 73 011 (93) 87 251 (96) 71 263 (98)
Home Access Health Corporation
Indeterminate† 74 (0.07) 1 (0.001) 71 (96) 1 (100)
asked users whether they had been tested
Unsuitable sample 4361 (0.05) 4713 (0.06) 4172 (96) 4455 (95)
previously for HIV and what they ex-
Total 95 955 78 361 92 312 76 329
pected the test results to be. Overall, 58%
*HIV indicates human immunodeficiency virus; DAD, Direct Access Diagnostics; and HAHC, Home Access Health of users reported they were testing for
Corporation. the first time (Table 3). The HIV preva-
†For confirmation of repeatedly reactive immunoassays, DAD used Western blot assay and HAHC used immu-
nofluorescence assay. lence was slightly higher (0.8%) among
first-time testers than among users who
Table 2.—Characteristics of Persons Using Home Table 3.—Testing History and HIV Test Result* had previously tested negative (0.7%). Of
Sample Collection HIV Tests* HIV-positive users, nearly half had never
All HIV-Positive
All HIV-Positive HIV Testing History Users, No. (%) Users, No. (%) been tested before. Only 17% of those who
Category Users Users Prevalence No previous test 29 812 (58) 224 (49) tested positive for HIV had anticipated
Sex Previous test that their results would be positive, 46%
Female 38 15 0.4 Negative 21 292 (41) 155 (34)
Indeterminate 282 (0.5) 8 (2)
expected their results to be negative, and
Male 62 85 1.4
Race Positive 83 (0.2) 57 (13) 37% were uncertain. Among HAHC us-
African American 5 16 2.7 Unknown 218 (0.4) 11 (2) ers found to be HIV negative, 2% had ex-
White 85 69 0.7 Total 51 687 455
pected a positive test result.
Hispanic 5 12 2.6
Other 5 3 0.6 *Home Access Health Corporation surveyed users Content analysis of the HAHC call log
Age, y about their testing history, but Direct Access Diagnostics revealed that 65% of HIV-positive users
did not. HIV indicates human immunodeficiency virus.
18-24 23 7 0.3 who called to receive results accepted
25-34 41 48 1.1
35-44 24 33 1.3
referrals for medical care and psychoso-
45-54 10 9 0.8 and 4888 persons who submitted samples cial services, 23% refused referrals be-
.55 3 3 0.9 unsuitable for testing. Response rates dif- cause they had a source of follow-up care,
Reported risk factor and 12% were receiving antiretroviral
Heterosexual 77 23 0.3
fered for the 2 HSC products and also for
Bisexual male† 8 27 3.1 specific questions. Response rates aver- therapy but had tested to see whether
MSM 10 38 3.5 aged 67% among 78 361 HAHC users sur- their HIV status had changed. During
IDU 1 6 4.1 veyedduringpretestcounseling.Moreus- the telephone sessions, counselors noted
MSM/IDU 3 6 1.7
ers provided information on sex (70%) and that 21% of persons discussed notifying
*Reported as percentage of total users who responded age (68%) than ZIP code (51%). Response their partners, 8% asked the counselor
to demographic questions (N = 76 373 for sex; numbers rates for all questions were higher (aver- to discuss the test result with their part-
slightly lower for other categories). HIV indicates human
immunodeficiency virus; MSM, men who have sex with age, 75%) among HIV-positive HAHC ner, and 5% disclosed that they already
men; and IDU, injection drug use. users because counselors asked for fur- knew their partner was HIV positive.
†From Home Access Health Corporation survey only.
Direct Access Diagnostics did not survey this risk factor. ther demographic information when giv- However, such information was not ac-
ing test results and making referrals. Re- tively elicited from all callers.
Data in this analysis reflect the num- sponse rates for 49 712 DAD users, sur- The HAHC call log was also used to as-
ber of tests, not individual users, because veyed when they called for test results, sess the psychological distress of HIV-
HSC tests are anonymous, and it is im- averaged 43%. Like HAHC users, fewer positive users when they received their
possible to ascertain if some users are re- DAD users provided ZIP code informa- results. Information about coping was not
peat testers. No tests of statistical signifi- tion, but response rates for HIV-positive systematically elicited from each user, but
cance were performed for comparisons, and HIV-negative DAD users were simi- counselors recorded that 7% of the HIV-
because although all data available from lar. Because of higher response rates and positive users expressed shock and dis-
HSC test users were included, there is the longer survey period, HAHC users may at an unexpected positive test re-
little assurance that users who re- account for approximately 72% of demo- sult; 5% hung up immediately on
sponded to the survey constitute a repre- graphicandbehavioralinformation.How- receiving a positive test result, without
sentative sample. ever,therewerenosignificantdifferences counseling; and 3% asked someone else
in demographic or behavioral character- to call for their results (the counselor ac-
RESULTS istics between HAHC and DAD users. commodated only after obtaining per-
Data were available for 174 316 tests, of Tests were submitted from all 50 mission from the user). One of the 610
which 95 955 (55%) were performed by states, Washington, DC, Puerto Rico, HIV-positive users expressed suicidal
DAD and 78 361 (45%) by HAHC (Table and the US Virgin Islands. Most users ideation; the counselor noted that this
1). Overall, 97% of users called for their were heterosexual; the largest percent- person was with a friend, who agreed to
results. A total of 165 194 specimens (95%) ages were men, white, and aged 25 to 34 facilitate mental health support.
were suitable for testing; 9073 (5%) were years (Table 2). The HIV prevalence was Of the 71 263 HIV-negative HAHC us-
not tested because they were contami- highest among African Americans and ers who called for test results, 82% re-
nated, contained an insufficient amount of Hispanics, those aged 35 to 44 years, men ceived the recorded message only, 29%
blood, or were submitted too long after who have sex with men, and injection called more than once to hear their test
they were obtained. The HIV prevalence drug users. Although only 5% of users results and counseling, and 12% elected to
among HSC test users was 0.9%. were African American and 5% His- speak with a counselor.
Demographicandbehavioraldatawere panic, they accounted for 16% and 12%,
available from 76 373 persons (59%) who respectively, of positive results. Bi- COMMENT
responded to at least 1 survey question: sexual men constituted 8% of all users Comparison of HSC test users with
70 620 HIV-negative, 865 HIV-positive, tested and 27% of positive results; this those tested at publicly funded HIV sites

1700 JAMA, November 18, 1998—Vol 280, No. 19 Home Sample Collection Tests for HIV Infection—Branson

©1998 American Medical Association. All rights reserved.

Downloaded From: http://jama.jamanetwork.com/ on 02/25/2013

 suggests that HSC is used by persons who United States.4 Publicly funded clinics ac- 5. What have we learned about evalu-
are at risk for HIV and those who do not count for approximately 2.5 million (15%) ating consumer-use testing? The course
seek other testing. The 0.9% HIV preva- of these tests and HSC accounts for ap- negotiated by the FDA during consider-
lence among HSC test users is nearly 3 proximately 200 000 (1%). Cost of the test ations of HSC testing offers some insight
times the estimated prevalence in the gen- ($30-$40) may present a barrier to the use about whether the process for evaluat-
eral population3 and similar to that ob- of HSC. In a survey of 2370 HIV-negative ing the safety and effectiveness of HSC
served in 1996 at publicly funded clinics persons in 9 states, 76% of respondents tests will be useful for the introduction of
among persons tested for the first time said they might use a free HSC test within other new technologies. The HSC tests
(1.1%) and those whose previous result was the next year, but only 40% said they were approved 10 years after the first
negative (0.9%) (Centers for Disease Con- might use one if it cost $40.6 product was submitted, during which the
trol and Prevention Client Record Data- 2. Can consumers correctly use the FDA grappled with its mandate to en-
base, unpublished data, 1997). Although HSC test? Of specimens submitted by con- sure the safety and effectiveness of home
58% of HSC test users and 44% of those sumers, 5% were unsuitable for HIV test- diagnostic products for HIV. For HSC
tested at publicly funded sites were be- ing, compared with 4.8% of dried-blood tests, it seems to have been possible to col-
ing tested for the first time, the propor- spot samples obtained by health profes- lect sufficient data to evaluate safety and
tion of HIV-positive HSC test users who sionals.7 This suggests that consumers can effectiveness of the device itself as well
had never been tested before (49%) is con- obtain an adequate sample for testing. as the consequences of its approved use,
siderably higher than that (29%) at pub- However, many persons are unwilling to such as the psychological responses of us-
licly funded sites. The high proportion provide finger-stick specimens,8 which ers. More detailed information also ex-
(97%) of users who called for results com- may limit the acceptability of HSC tests. ists about the cost and feasibility of the
pares favorably with publicly funded clin- 3. Do HSC test users suffer because they data collection and postmarketing stud-
ics, where the return rate in 1996 aver- receive no face-to-face counseling? Oppo- ies that were imposed on HSC product
aged 67%, ranging from 51% at sexually nents of HSC tests feared that telephone sponsors as conditions of approval. The ap-
transmitted disease clinics to 83% at vol- counseling would lead to emotional dis- proval of HSC tests for HIV may pave the
untary HIV counseling and testing sites.4 tress for persons who tested HIV posi- way for the approval of other technolo-
It is important to recognize the limi- tive and that many might commit suicide gies, such as 1-step rapid HIV tests, which
tations of the data on user characteris- or fail to obtain follow-up medical care and can provide results in 10 to 15 minutes
tics, as only 59% of users responded to psychological support. The evidence to without the need for a laboratory, and even
questions about demographics and be- date, based on small numbers of HIV- true home tests,10 which consumers could
havior. Because HSC testing is anony- positive users, is mixed. In the first year perform and interpret at home. These
mous, no information is available about of use, only 1 caller expressed suicidal ide- could dramatically change HIV testing,
nonrespondents. This inability to char- ation. More worrisome, 5% of callers who and HIV prevention strategies based on
acterize persons who may have used tested HIV positive hung up without coun- immediate access to test results could
HSC testing hampers our understand- seling, and no follow-up information is greatly affect the AIDS epidemic.
ing of the ultimate role of HSC in HIV available for them. The author is grateful to Katarzyna Kruzel, MPH,
testing strategies. The data indicate that many HIV- Clinical Research Associate at Home Access Health
Two reviews1,5 posed questions about positive HSC test users have a source of Corporation, and Ernie Lewis, data manager at the
Centers for Disease Control and Prevention, for
home testing for which data now suggest care and that those who do not are will- providing assistance with the data and to Rick
some answers. The following 5 questions ing to accept referrals. Information is in- Steketee, MD, for his review of the manuscript.
relate specifically to HSC and also to sufficient to determine whether HSC test
References
more general issues about consumer in- users actually seek medical care or
1. Bayer R, Stryker J, Smith MD. Testing for HIV
fluence in the context of HIV testing. whether telephone counseling is effec- infection at home. N Engl J Med. 1995;332:1296-1299.
1. Do HSC tests expand access to test- tive for encouraging users to change their 2. SkelleyM.TestimonyonbehalfoftheAssociationof
ing? Proponents argue that HSC tests en- risk behaviors. State and Territorial Health Officials, Council of State
and Territorial Epidemiologists, National Alliance of
courage persons to be tested who might not 4. How do HSC tests affect public health State and Territorial AIDS Directors, Association of
otherwise.Evidencefromthisanalysissup- practices?Concern about the effect of State and Territorial Public Health Laboratory Direc-
ports this view: more than half of the us- anonymous HSC testing on name report- tors,FDABloodProductsAdvisoryCommittee; June
ers had not been tested previously. The ing and partner notification led to at- 22, 1994; Washington, DC.
3. Karon JM, Rosenberg PS, McQuillan G, et al.
data also show that HIV-infected bi- tempts by several states to block the sale Prevalence of HIV infection in the United States,
sexual males, African Americans, and His- of HSC HIV tests, citing state laws that 1984 to 1992. JAMA. 1996; 276:126-131.
panicsareusingHSCtests,suggestingthat mandated the reporting of names of HIV- 4. Centers for Disease Control and Prevention. HIV
Counseling and Testing in Publicly Funded Sites:
it is an acceptable option for persons who infected persons. However, under the fed- 1995 Summary Report. Atlanta, Ga: US Dept of
may have less access to testing or who may eral Food, Drug, and Cosmetic Act, FDA Health and Human Services; 1997.
be particularly concerned about confiden- approval of HSC tests preempted these 5. Schopper D, Vercauteren G. Testing for HIV at
tiality. Although relatively few African more restrictive state laws.9 No data yet home: what are the issues? AIDS. 1996;10:1255-1265.
6. Colfax GJ, Lehman J, Hecht FM, et al. Likelihood
Americans and Hispanics used HSC tests, exist on whether the availability of HSC of at-risk individuals using home HIV test collection
the risk for HIV seemed to be high among tests hinders effective and timely surveil- kits. Paper presented at: Society of General Inter-
those who did. African Americans and His- lance or partner notification, but such in- nal Medicine, 20th Annual Meeting; May 2, 1997;
Washington, DC.
panics each accounted for 5% of HSC test formation is also meager about other 7. Hoxie NJ, Vergeront JM, Pfister JR, et al. Improv-
users (compared with 30% and 18%, re- anonymous and private-sector testing. ing estimates of HIV-1 seroprevalence among child-
spectively, at publicly funded sites), but However, given the high proportion of us- bearing women: use of smaller blood spots. Am J Pub-
their HIV prevalence (2.7% and 2.6%, re- ers with sources of care or accepting re- lic Health. 1992;82:1370-1373.
8. Malor CJ, Read SE, Coates RA, et al. Comparison
spectively) was similar to that for publicly ferrals and the fact that health care pro- of saliva and blood for human immunodeficiency virus
funded test site users (2.6% and 2.0%).4 fessionals are required to report new cases, prevalence testing. J Infect Dis. 1991;163:699-702.
The overall effect of HSC tests on ac- HSC tests are likely to improve these ef- 9. Schultz WB. Advisory Opinion. Washington, DC:
cess to testing may be small. Excluding forts: a high percentage of HIV-positive US Food and Drug Administration; 1996.
10. Kassler WJ. Advances in HIV testing technology
blood donors, an estimated 16.6 million persons are learning of their HIV infec- and their potential impact on prevention. AIDS Educ
HIV tests are performed annually in the tion from HSC as their first test. Prev. 1997;9(suppl B):27-40.

JAMA, November 18, 1998—Vol 280, No. 19 Home Sample Collection Tests for HIV Infection—Branson 1701
©1998 American Medical Association. All rights reserved.

DownloadedViewFrom:
publicationhttp://jama.jamanetwork.com/
stats on 02/25/2013