M A J O R A R T I C L E H I V / A I D S

What Is the Impact of Home-Based HIVCounseling and Testing on the Clinical Statusof Newly Enrolled Adults in a Large HIV CareProgram in Western Kenya?

Juddy Wachira,1,4 Sylvester Kimaiyo,4,5 Samson Ndege,4,6 Joseph Mamlin,2,4,5 and Paula Braitstein,2,3,4,5,7

1School of Health Physical Education and Recreation and 2School of Medicine, Indiana University, Bloomington, 3Regenstrief Institute, Indianapolis;4USAID-AMPATH (US Agency for International Development–Academic Model Providing Access to Healthcare) Partnership, 5Moi University School ofMedicine, 6Moi University School of Public Health, Eldoret, Kenya; and 7Dalla Lana School of Public Health, University of Toronto, Canada

(See the Editorial Commentary by Mills and Ford, on pages 282–4.)

Background. This article describes the effect point of entry into the human immunodeficiency virus (HIV) care

program had on the clinical status of adults presenting for the first time to USAID-AMPATH (US Agency for

International Development–Academic Model Providing Access to Healthcare) Partnership clinics for HIV care.

Methods. All patients aged $14 years enrolled between August 2008 and April 2010 were included. Points of

entry to USAID-AMPATH clinics were home-based counseling and testing (HBCT), provider-initiated testing and

counseling (PITC), HIV testing in the tuberculosis clinic, and voluntary counseling and testing (VCT). Tests for

trend were calculated, and multivariable logistic regression was used to compare the effect of HBCT versus other

points of entry on primary outcomes controlling for age and sex.

Results. There were 19 552 eligible individuals. Of these, 946 tested in HBCT, 10 261 in VCT, 8073 in PITC,

and 272 in the tuberculosis clinic. The median (interquartile range) enrollment CD4 cell counts among those who

tested HIV positive was 323 (194–491), 217 (87–404), 190 (70–371), and 136 cells/mm3 (59–266) for HBCT, VCT,

PITC, and the tuberculosis clinic, respectively (P , .001). Compared with those patients whose HIV infection was

diagnosed in the tuberculosis clinic, those who tested positive in HBCT were, controlling for age and sex, less likely

to have to have World Health Organization stage III or IV HIV infection at enrollment (adjusted odds ratio [AOR],

0.04; 95% confidence interval [CI], .03–.06), less likely to enroll with a CD4 cell count of ,200 cells/mm3 (AOR,

0.20; 95% CI, .14–.28), and less likely to enroll into care with a chief complaint (AOR, 0.08; 95% CI, .05–.12).

Conclusions. HBCT is effective at getting HIV-infected persons enrolled in HIV care before they become ill.

Human immunodeficiency virus (HIV) infection

continues to spread despite considerable efforts to

prevent new infections and reduce HIV mortality and

morbidity rates [1–3]. In Kenya, 1.4 million adults

(7.8%) are infected with the virus [1]. Even more

concerning, only 50% of adults are receiving

treatment for their HIV infection according to the

2010 World Health Organization (WHO) HIV treat-

ment guidelines [3]. Late presentation of HIV-in-

fected patients to care is an issue of paramount

importance because of its association with lower

treatment response rates, higher mortality rates,

higher treatment costs [4, 5] and possible increased

HIV transmission rates [5, 6]. In sub-Saharan Africa,

including Kenya, high percentages (�40%–73%) of

patients present for treatment at advanced stages of

disease (WHO stage III/IV [7–9]). To get persons into

care at earlier stages of their HIV disease, calls have

been made for intensified efforts to promote early di-

agnosis and treatment of HIV-infected cases [10, 11].

Received 10 May 2011; accepted 30 August 2011.Correspondence: Paula Braitstein, PhD, Indiana University School of Medicine,

1001 W 10th St, OPW-M200, Indianapolis, IN 46202 (pbraitstein@yahoo.com).

Clinical Infectious Diseases 2012;54(2):275–81� The Author 2011. Published by Oxford University Press on behalf of the InfectiousDiseases Society of America. All rights reserved. For Permissions, please e-mail:journals.permissions@oup.com.DOI: 10.1093/cid/cir789

HIV/AIDS d CID 2012:54 (15 January) d 275

by guest on Novem

ber 22, 2015http://cid.oxfordjournals.org/

Dow

nloaded from

In spite of the widespread availability of counseling and

testing services, including targeted mobile campaigns and fa-

cility-based services, HIV testing uptake in Kenya, as elsewhere

in sub-Saharan Africa, remains low, and only a fraction of those

living with the virus are aware of their status [1]. These forms of

HIV testing services utilize different approaches with regard to

who receives HIV testing at what point. Voluntary counseling

and testing (VCT) services are usually patient initiated and are

offered at either stand-alone testing centers in a clinical facility

or during targeted HIV campaigns as mobile VCT services [12].

Provider-initiated testing and counseling (PITC) is sometimes

called diagnostic testing and counseling. It is offered through

clinical settings (hospitals) and typically uses the opt-out ap-

proach, testing any patient seeking treatment or care in a hos-

pital in a high HIV-burden area for HIV as part of routine care.

It forms an important part of the diagnostic process especially

for patients with tuberculosis because of the strong interactions

between HIV and tuberculosis [13]. Testing in antenatal clinics

and in labor and delivery wards is also part of provider-initiated

services, as pregnant mothers seeking antenatal care are routinely

tested for HIV, with those found positive enrolled into pre-

vention of mother-to-child transmission (PMTCT) services [14].

Recently, home-based counseling and testing (HBCT) for HIV

infection has been advocated to promote early diagnosis of HIV

[15, 16]. Unlike other counseling and testing services, HBCT is

provided at home by specially trained HIV counselors, po-

tentially overcoming some of the challenges inherent in vol-

untary or institutional testing sites. As countries in Africa

embrace universal testing through HBCT [15, 16], the benefits

can be experienced only if they are coupled with access to timely

treatment of HIV infection. Although limited studies have

documented the efficacy of HBCT in increasing HIV testing

uptake [17, 18], no studies, to our knowledge, have assessed its

impact in promoting timely access to care in comparison to the

other testing points. The primary objective of this study was to

describe the clinical status of HIV-infected adults presenting for

the first time to USAID-AMPATH (US Agency for International

Development Academic Model Providing Access to Healthcare)

Partnership clinics for HIV care from the different testing points.

METHODS

Study AreaAMPATH was initiated in 2001 as a joint partnership between

Moi University School ofMedicine (Eldoret, Kenya), the Indiana

University School of Medicine (Indianapolis, Indiana), and the

Moi Teaching and Referral Hospital (Eldoret, Kenya). In 2004,

USAID became a major partner and the USAID-AMPATH

Partnership was inaugurated. Since 2001, the program has

enrolled more than 130 000 HIV-infected adults and children

in 25 Ministry of Health facilities and numerous satellite clinics

in western Kenya. All HIV- and tuberculosis-related care and

treatment are provided free at the point of care for patients

through AMPATH and the Kenyan National Leprosy, Tuber-

culosis and Lung Disease Program. A more detailed description

of the program can be found elsewhere [19, 20].

AMPATH provides comprehensive HIV care services in-

cluding HIV counseling and testing services through VCT, PITC

(inpatient and outpatient clinical departments, antenatal clinics,

and tuberculosis clinics) and HBCT strategies. Currently, HBCT

services are being rolled out in western Kenya through this

partnership. HBCT targets residents/household members who

are $13 years old; a more detailed description of the ongoing

HBCT program is available elsewhere [16]. At the time of the

study, HBCT had been implemented in 3 catchment area-

s—namely, Kosirai Division and Turbo Division in Rift Valley

Province and Bungoma East District in Western Province.

Study DesignThis was a retrospective observational study based on routine

clinical data obtained between August 2008 and April 2010.

Ethical approval was obtained from the Institutional Research

and Ethics Committee at Moi University School of Medicine,

Eldoret, Kenya, and Indiana University’s Institutional Review

Board in Indianapolis, Indiana.

Study PopulationEligible for inclusion in the analysis were adults aged $14 years

who enrolled in HIV care in any of the AMPATH clinics from

August 2008 to April 2010 and who had their point of entry

documented on their initial clinical encounter form.

Data CollectionClinicians complete standardized forms capturing demographic,

clinical, and pharmacologic information at each patient visit.

These data are then entered by trained data clerks into the

AMPATH Medical Record System, a secure, computerized da-

tabase designed for clinical management, with data entry vali-

dated by random review of 5% of the forms entered. At the time

of registration, patients are provided with a unique identifying

number. For this study, all data were stripped of identifying

information prior to analysis. The point of entry into care is

captured on an initial encounter form. The HBCT program has

been described in detail elsewhere [21]. In brief, after extensive

community consultations and mobilizations, HIV counselors

work in teams going door-to-door, village by village, offering

HIV counseling and testing to all individuals aged$13 years, as

well as children aged,13 years if their mother is HIV positive or

has an unknown HIV status or if the mother has died or her vital

status is unknown.

The points of entry considered were HBCT, PITC, tuberculosis

clinic, and VCT. PITC includes individuals tested in the antenatal

clinic, inpatient wards, labor and delivery, and some outpatient

276 d CID 2012:54 (15 January) d HIV/AIDS

by guest on Novem

ber 22, 2015http://cid.oxfordjournals.org/

Dow

nloaded from

clinics, such as the tuberculosis clinic. We hypothesized that the

patients from the tuberculosis clinic would be more likely to have

advanced HIV disease than others tested through PITC, and we

thus examined patients from the tuberculosis clinic separately

from PITC.

As per Kenya Ministry of Health guidelines, tested clients are

typically issued with a card indicating their HIV status. Patients

are requested to present this card during enrollment to care as

evidence of their HIV status. In the event that patients were

tested at multiple testing points, the last point of testing is in-

dicated in their initial medical encounter form. Clinicians are

allowed to choose only one of the point-of-testing options on the

clinical encounter forms.

Data AnalysisThe primary dependent variable was the patient’s CD4 cell count

and WHO clinical stage at enrollment, and the secondary de-

pendent variables were having been hospitalized in the past year,

having a chief complaint (a symptom the patient self-reports at

the clinical encounter as being the main reason for coming to the

clinic) at enrollment, being in a discordant relationship, and

being pregnant (for women only). Other variables considered for

their potential explanatory or confounding relationship to the

primary dependent variable included age and sex. We identified

the point of entry to USAID-AMPATH clinics and compared age,

sex, pregnancy in women, discordant couples, the WHO clinical

stage at enrollment, CD4 cell count at enrollment, hospitalization

in the past year, and having a chief complaint at enrollment

between these points of entries and testing points. Patients whose

point of entry was documented as ‘‘other’’ were excluded from

the analysis. Proportions of missing data were calculated for each

variable and compared for different points of entry.

The data were analyzed using Stata/SE software (version 10).

Comparisons were done using Pearson’s v2 test for categorical

variables and the Kruskal–Wallis test for continuous variables.

Medians and interquartile ranges were also calculated for con-

tinuous variables (ie, age and CD4 cell count). For ordered

(categorical) variables, we used tests for trends to obtain P values.

With controlling for age and sex, logistic regression was con-

ducted to assess the impact of the point of entry into care on

primary and secondary outcomes. We conducted a subanalysis

comparing those entering from HBCT with those entering from

VCT, hypothesizing that these 2 testing points are the most

similar (because they are not dependent on the individual

seeking healthcare), and we asked whether patients entering from

HBCT had better clinical status at enrollment compared with

those entering from VCT.

RESULTS

There were 28 556 adults enrolled between August 2008 and

April 2010. Of these, 22 974 (80%) had point of entry indicated

on their initial encounter form and were eligible for analysis. We

excluded 3422 adults who had points of entry that were docu-

mented as ‘‘other.’’ Thus there were 19 552 adults included in the

analysis, including 946 from HBCT, 8073 from PITC, 272 from

tuberculosis, and 10 261 from VCT.

As shown in Table 1, there were no major differences in the

median age between the groups. From all the testing points

except the tuberculosis clinic, there was a higher proportion of

women enrolled in care than men (P 5 .001). The highest

proportion (11%) of HIV-positive pregnant women enrolled

had been tested in HBCT and the lowest in VCT and the

tuberculosis clinic. The highest percentage (24%) of individuals

in discordant couples was identified among persons entering via

HBCT, compared with the other points of entry.

As summarized in Table 2, the median (interquartile range)

CD4 cell count at enrollment was the highest for patients who

Table 1. Sociodemographic Characteristics of Patients Aged ‡14 Years Enrolling in US Agency for International Development AcademicModel Providing Access to Healthcare Clinics, Stratified by Point of Entry (N 5 19 552)

Variable HBCT (n 5 946) VCT (n 5 10 261) PITC (n 5 8073) Tuberculosis Clinic (n 5 272) P Value

Age median (IQR), years 37 (30–46) 36 (29–44) 36 (29–44) 36 (30–44) ,.022

Sex, No. (%) ,.001

Male 268 (28) 3537 (35) 3050 (38) 136 (50)

Female 678 (72) 6724 (66) 5023 (62) 136 (50)

Pregnant, No. (%)a 75 (11) 360 (5) 321 (6) 7 (5) ,.001b

HIV discordance, No. (%)

Member of discordant couple 214 (24) 649 (7) 466 (6) 13 (5) .045b

Missing data 63 (7) 619 (6) 659 (8) 17 (6)

No data were missing for age, sex, and pregnancy status.

Abbreviations: HBCT, home-based counseling and testing; HIV, human immunodeficiency virus; IQR, interquartile range; PITC, provider-initiated testing and

counseling; VCT, voluntary counseling and testing.a Restricted to women.b P value for trend across points of entry.

HIV/AIDS d CID 2012:54 (15 January) d 277

by guest on Novem

ber 22, 2015http://cid.oxfordjournals.org/

Dow

nloaded from

tested HIV positive in HBCT (median 323 cells/mm3), followed

by VCT and then PITC. Patients coming from the tuberculosis

clinic had the lowest median CD4 cell count (136 cells/mm3)

(P 5 .001). This trend was similar for patients with WHO stage

I/II disease: 86% of patients who had their HIV status identified

in HBCT had WHO stage I/II disease, compared with 62% in

VCT, 54% in PITC, and 21% in the tuberculosis clinic

(P5 .002). The trend was reversed for patients with WHO stage

III/IV disease. Patients who were tested in the tuberculosis clinic

had the highest likelihood of having WHO stage III/IV disease at

enrollment (79%), compared with PITC (46%), VCT (38%) and

HBCT (14%), (P 5 .002).

As expected, a higher proportion of patients testing HIV

positive in PITC (14%) and the tuberculosis clinic (13%) had

been hospitalized in the past year, compared with patients in

HBCT (2%) (P 5 .001). Patients coming from either the tu-

berculosis clinic (87%) or PITC were also much more likely to

present with a chief complaint at enrollment, compared with

those coming from VCT or HBCT (P 5 .017) (Table 2).

In logistic regression, adjusting for the potential confounding

effects of age and sex, individuals who tested HIV positive in

HBCT were less likely to be enrolled into care with a CD4 cell

count ,200 cells/mm3 than were those who tested positive in

the tuberculosis clinic (adjusted odds ratio [AOR], 0.20; 95%

confidence interval [CI], .14–.28). They were also less likely

to have WHO stage III/IV disease at enrollment (AOR, 0.04;

95% CI, .03–.06), less likely to have been hospitalized in the past

year (AOR, 0.14 95%; 95% CI, .07–.25), and less likely to

have a chief complaint (AOR, 0.08; 95% CI, .05–.12) (Table 3).

In a subanalysis comparing HBCT with VCT, HBCT was

consistently independently associated with improved clinical

status at enrollment into HIV care (Table 4).

DISCUSSION

Early diagnosis of HIV infection is critical to providing infected

individuals with antiretroviral therapy and effective prevention

and care services [7, 22, 23]. Late presentation to care as evi-

denced by advanced clinical HIV disease results in poorer clinical

outcomes [4, 5, 8]. This study suggests that population-based,

HBCT can be an effective strategy for engaging HIV-infected

persons into care earlier in the course of their HIV disease

compared with other testing strategies. HBCT was also effective

at identifying pregnant women in the community and linking

them to care. Although more research is needed, including cost-

effectiveness analyses, these data call for broad and rapid ex-

pansion of HBCT throughout countries with high HIV burdens

as a means of facilitating access to HIV care for HIV-infected

persons with all the resultant downstream positive effects (in-

cluding improved treatment outcomes and primary prevention

through viral load reduction).

Women represented a greater proportion of patients access-

ing HIV care from all testing points except the tuberculosis

clinic, where they were equally represented. These data support

findings by our group and others that men are more challenged

than women in accessing HIV care compared with women, even

after adjustment for HIV prevalence among men and women in

individual countries [21, 24]. Our data indicate that as the na-

ture of the access becomes more voluntary (ie, with testing in

HBCT or VCT rather than, for example, the tuberculosis clinic),

Table 2. Clinical Characteristics of Patients Aged ‡14 Years Enrolling in US Agency for International Development Academic ModelProviding Access to Healthcare Clinics Stratified by Point of Entry (N 5 19 552)

Variable

HBCT

(n 5 946)

VCT

(n 5 10 261)

PITC

(n 5 8073)

Tuberculosis Clinic

(n 5 272) P Value

CD4 cell count

Median (IQR), cells/mm3 323 (194–491) 217 (87–404) 190 (70–371) 136 (59–266) ,.001

Missing data, No. (%) 221 (23) 4697 (45) 4209 (52) 103 (38)

WHO stage for HIV disease, No. (%)

Stage I/II 612 (86) 5271 (62) 3357 (54) 47 (21) .002a

Stage III/IV 97 (14) 3165 (38) 2843 (46) 172 (79) .002a

Missing data 237 (25) 1825 (18) 1873 (23) 53 (19)

Hospitalized in past year, No. (%)

Yes 18 (2) 843 (8) 1076 (14) 33 (13) ,.001a

Missing data 16 (2) 208 (2) 241 (3) 8 (3)

Had chief complaint at enrollment, No. (%)

Yes 264 (35) 6008 (67) 5044 (72) 205 (87) .017a

Missing data 196 (21) 1324 (13) 1051 (13) 36 (13)

Abbreviations: HBCT, home-based counseling and testing; HIV, human immunodeficiency virus; IQR, interquartile range; PITC, provider-initiated testing and

counseling; VCT, voluntary counseling and testing; WHO, World Health Organization.a P value for trend across points of entry.

278 d CID 2012:54 (15 January) d HIV/AIDS

by guest on Novem

ber 22, 2015http://cid.oxfordjournals.org/

Dow

nloaded from

men are proportionately less well represented. Much more at-

tention needs to be focused on helping HIV-infected men access

care in a timely manner, and more research is needed to un-

derstand the issues facing HIV-infected men in sub-Saharan

Africa that prevent them from accessing care.

With improved access to antiretroviral therapy in sub-Saharan

Africa [3, 25], rates of mother-to-child transmission of HIV have

been greatly reduced [3]. Even so, the coverage of PMTCT services

is not optimal, with most countries having not yet reached all

pregnant women with these services [3, 25]. Our findings revealed

that HBCTwas able to enroll twice the proportion of HIV-infected

pregnant women compared with VCT, PITC or tuberculosis entry

points. The implication is that HBCT could complement PMTCT

by identifying and referring pregnant women who are not other-

wise accessing antenatal or PMTCT services.

Differences in clinical status among persons who tested posi-

tive in PITC versus in VCT were rather attenuated, even com-

pared with the differences between VCT andHBCT. This suggests

that persons who test positive in VCT are more likely to come for

care only once their health begins to deteriorate, reducing the

effectiveness of VCT as a testing strategy in high-burden areas. It

may also be that individuals seek out VCT services if they suspect

they are infected, either because of the presence of symptoms or

for other reasons. The clinical characteristics of patients tested at

the tuberculosis clinic support the strong association between

HIV and tuberculosis and the efforts so far made to integrate

tuberculosis and HIV services in the region [26].

The benefits of couple counseling and testing, and most re-

cently the use of combination antiretroviral therapy, have been

widely documented in reducing HIV transmission rates among

discordant couples as well as reducing mother-to-child trans-

mission of HIV [27–29]. HBCT identified the highest pro-

portions of discordant couples, with the rest of the testing points

identifying almost identically low percentages. This could be

associated with the fact that HBCT services are offered at home,

presenting a better opportunity to capture couples together

for testing. With reports of low disclosure rates and the

challenges it presents [30, 31], VCT, PITC, and tuberculosis

testing sites might not be effective in promoting couple coun-

seling and testing if this concept is not incorporated within these

strategies.

This study has several strengths. The USAID-AMPATH Part-

nership provides a unique setting where a variety of counseling

and testing strategies are employed, including HBCT. Moreover,

HIV care and treatment services are provided at no cost to pa-

tients. Together these factors enabled a robust analysis without

bias related to whether or not patients could pay for care and

Table 4. Logistic Regression Subanalysis of Impact of Home-Based Counseling and Testing Versus Voluntary Counseling andTesting on Key Markers of Clinical Status at Enrollment, Adjustedfor Age and Sex

Marker at

Enrollment and

Point of Entry Into HIV Care AOR 95% CI P Value

WHO III/IV clinical stage disease

VCT Reference . .

HBCT 0.26 .21–.32 ,.001

CD4 cell count ,200 cells/mm3

VCT Reference . .

HBCT 0.38 .32–.45 ,.001

Having a chief complainta

VCT Reference . .

HBCT 0.25 .21–.29 ,.001

Hospitalized in past year

VCT Reference . .

HBCT 0.21 .13–.34 ,.001

Abbreviations: AOR, adjusted odds ratio; CI, confidence interval; HBCT,

home-based counseling and testing; HIV, human immunodeficiency virus;

VCT, voluntary counseling and testing; WHO, World Health Organization.a A symptom the patient self-reports at the clinical encounter as being the

main reason for coming to the clinic.

Table 3. Logistic Regression Analysis of Impact of Point of EntryInto HIV Care on Key Markers of Clinical Status at Enrollment,Adjusted for Age and Sex

Marker at

Enrollment and

Point of Entry Into HIV Care AOR 95% CI P Value

WHO clinical stage III/IV disease

Tuberculosis clinic Reference . .

PITC 0.24 .17–.33 ,.001

VCT 0.17 .12–.23 ,.001

HBCT 0.04 .03–.06 ,.001

CD4 cell count ,200 cells/mm3

Tuberculosis clinic Reference . .

PITC 0.62 .45–.86 .004

VCT 0.52 .38–.72 ,.001

HBCT 0.20 .14–.28 ,.001

Having a chief complainta

Tuberculosis clinic Reference . .

PITC 0.39 .27–.57 ,.001

VCT 0.31 .21–.46 ,.001

HBCT 0.08 .05–.12 ,.001

Hospitalized in past year

Tuberculosis clinic Reference . .

PITC 1.11 .77–1.61 .567

VCT 0.64 .44–.93 .018

HBCT 0.14 .07–.25 ,.001

Abbreviations: AOR, adjusted odds ratio; CI, confidence interval; HBCT, home-

based counseling and testing; HIV, human immunodeficiency virus; PITC,

provider-initiated testing and counseling; VCT, voluntary counseling and

testing; WHO, World Health Organization.a A symptom the patient self-reports at the clinical encounter as being the

main reason for coming to the clinic.

HIV/AIDS d CID 2012:54 (15 January) d 279

by guest on Novem

ber 22, 2015http://cid.oxfordjournals.org/

Dow

nloaded from

treatment. We chose to include patients enrolling at all AMPATH

clinics in this analysis, because many patients may test positive in

one place but may choose to attend a clinic in another catchment

because of stigma or other issues. This approach to the analysis

strengthens our findings, because the effect estimates of point of

entry are averaged across all clinics and can thus be considered an

analysis of effectiveness (ie, under routine circumstances), rather

than efficacy (ie, under ideal circumstances), of HBCT.

The study also presented some limitations. There is a poten-

tial for some selection bias due to missing entry points and those

classified as ‘‘other’’ points of entry, both of which were ex-

cluded from the analysis. However, we know of no systematic

reason why clinicians would fail to document the point of entry,

thus reducing the potential impact of any selection bias that may

have occurred. There was also the potential for random mis-

classification bias if clinicians incorrectly documented point of

entry or the patient lied for whatever reason (unlikely but dif-

ficult to substantiate). Finally, there was a significant amount of

missing data on enrollment CD4 cell count. We postulate that

this was particularly so among patients who had tested positive

in PITC because, given that they were coming from PITC, they

were probably quite ill at enrollment. It is likely that clinicians

attempt to address the urgent clinical issues before obtaining

a baseline CD4 cell count. Thus, our findings would likely be

even stronger if the CD4 cell count data were complete. These

data should be interpreted with some caution.

In conclusion, our findings provide strong evidence that

HBCT can result in earlier diagnosis of HIV infection with

consequent earlier access to care. Our data demonstrate that

HBCT is effective at getting HIV-infected persons to enroll in

HIV care at much earlier stages of HIV disease and while their

immune systems are still relatively intact. If the ‘‘treatment as

prevention’’ paradigm is to be successfully rolled out in sub-

Saharan Africa, HBCT and other community support programs

could play an instrumental role in identifying HIV-infected

persons in the community, linking them with care, and helping

to retain them care. These data suggest that efforts to roll out

HBCT in Kenya and elsewhere should be intensified even as other

testing points continue to provide services. Cost-effectiveness

studies are desperately needed to determine the most cost-

effective testing strategy, taking into account the downstream

consequences of late presentation for HIV care.

Notes

Acknowledgments. The authors would like to thank all the clinicians

in all the AMPATH clinics, especially the clinical officers, medical officers,

pediatricians, nutritionists, outreach workers, and social workers, for their

dedication in caring for patients and their attentiveness in accurately re-

cording their patients’ data. We would also like to thank all the data entry

technicians, data managers, and administrative and clerical staff for en-

abling the collection, management, interpretation, and publication of these

data. We appreciate Dr Jane Carter and Dr Harsha Triamurthy, who both

made important contributions to earlier versions of the manuscript.

J. W. wrote the first draft of the manuscript and was primarily re-

sponsible for incorporating authors’ comments into subsequent drafts.

S. K., S. N., and J. M. are the principal investigators of the AMPATH HBCT

Program and contributed to the manuscript by revising drafts and assisting

in the interpretation of the findings. P. B. conceived the idea for the

analysis, conducted all statistical analysis, and took primary responsibility

for supervising and mentoring J. W. in developing the paper and making

the final decision about its submission.

Both the Indiana University Institutional Review Board and the Moi

University Institutional Research Ethics Committee provided approval for

retrospective analysis of routinely collected clinical data in AMPATH.

Financial support. The HBCT program was supported by grants from

Abbott Laboratories, the Purpleville Foundation, and the Global Business

Coalition. The United States Agency for International Development as part

of the President’s Emergency Plan for AIDS Relief (USAID-PEPFAR)

supported care for those found to be HIV positive, and the Abbott Fund

provided test kits and logistical support. AMPATH and the authors are

particularly grateful to the Rockefeller Foundation for funding the de-

velopment of the AMPATH Medical Records System, and the Kenyan

Division of Leprosy, TB and Lung Disease, formerly the Kenyan National

Leprosy and Tuberculosis Program, for its support.

Potential conflicts of interest. All authors: No reported conflicts.

All authors have submitted the ICMJE Form for Disclosure of Potential

Conflicts of Interest. Conflicts that the editors consider relevant to the

content of the manuscript have been disclosed.

References

1. Kenya Ministry of Health. Kenya AIDS indicator survey 2007. Nairobi,

Kenya: Ministry of Health, 2009.

2. Joint United Nations Programme on HIV/AIDS (UNAIDS). UNAIDS

report on the global AIDS epidemic 2010. Available at: http://www.

unaids.org/globalreport/global_report.htm. Accessed 22 August 2011.

3. WHO/UNAIDS/UNICEF. Towards universal access: scaling up priority

HIV/AIDS interventions in the health sector. 2009. Available at: http://

www.who.int/hiv/pub/tuapr_2009_en.pdf. Accessed 22 August 2011.

4. Badri M, Lawn SD, Wood R. Short-term risk of AIDS or death

in people infected with HIV-1 before antiretroviral therapy in South

Africa: a longitudinal study. Lancet 2006; 368:1254–9.

5. Sabin CA, Smith CJ, Gumley H, et al. Late presenters in the era of

highly active antiretroviral therapy: uptake of and responses to anti-

retroviral therapy. AIDS 2004; 18:2145–51.

6. Antinori A, Johnson M, Moreno S, et al. Introduction to late pre-

sentation for HIV treatment in Europe. Antivir Ther 2010; 15:1–2.

7. Kigozi IM, Dobkin LM, Martin JN, et al. Late-disease stage at pre-

sentation to an HIV clinic in the era of free antiretroviral therapy in

sub-Saharan Africa. J Acquir Immune Defic Syndr 2009; 52:280–9.

8. Braitstein P, Brinkhof MW, Dabis F, et al. Mortality of HIV-1–infected

patients in the first year of antiretroviral therapy: a comparison be-

tween low-income and high-income countries. Lancet 2006; 367:

817–24.

9. Wools-Kaloustian KD, Kimaiyo SD, Diero LD, et al. Viability and

effectiveness of large-scale HIV treatment initiatives in sub-Saharan

Africa: experience from western Kenya. AIDS 2006; 20:41–8.

10. Chesney MA, Smith AW. Critical delays in HIV testing and care: the

potential role of stigma. Am Behav Scientist 1999; 42:1162–74.

11. Samet JH, Freedberg KA, Stein MD, et al. Trillion virion delay: time

from testing positive for HIV to presentation for primary care. Arch

Intern Med 1998; 158:734–40.

12. Marum E, Taegtmeyer M, Chebet K. Scale-up of voluntary HIV

counseling and testing in Kenya. JAMA 2006; 296:859–62.

13. Chakaya JM, Mansoer JR, Scano F, et al. National scale-up of HIV

testing and provision of HIV care to tuberculosis patients in Kenya. Int

J Tuberc Lung Dis 2008; 12:424–9.

14. Anand A, Shiraishi RW, Sheikh AA, et al. Site factors may be more

important than participant factors in explaining HIV test acceptance in

280 d CID 2012:54 (15 January) d HIV/AIDS

by guest on Novem

ber 22, 2015http://cid.oxfordjournals.org/

Dow

nloaded from

the prevention of mother-to-child HIV transmission programme in

Kenya, 2005. Trop Med Int Health 2009; 14:1215–19.

15. Matovu JK, Kigozi G, Nalugoda F, Lutalo T, Gray RH. Evaluation of

a community-based VCT program in Rakai, Uganda. Paper presented

at the Sixth INDEPTH Annual General and Scientific Meeting; 18–22

September 2006; Ouagadougou, Burkina Faso.

16. Kimaiyo S, Ndege SK, Were MC, et al. Know your epidemic: home-

based HIV counseling and testing in western Kenya. East Africa Med J

2010; 87:100–8.

17. Were WA, Mermin JH, Wamai N, et al. Undiagnosed HIV infection,

couple HIV discordance among household members of HIV-infected

people receiving antiretroviral therapy in Uganda. J Acquir Immune

Defic Syndr 2006; 43:91–5.

18. Yoder PS, Katahoire AR, Akol Z, Bunnell R, Kaharuza F. Home-based

HIV testing and counseling in a survey context in Uganda. Calverton,

MD: ORC Macro, 2006.

19. Einterz RM, et al. Responding to the HIV pandemic: the power of an

academic medical partnership. Acad Med 2007; 82:812–8.

20. Wools-Kaloustian K, Kimaiyo S, Musick B, et al. The impact of the

President’s Emergency Plan for AIDS Relief on expansion of HIV care

services for adult patients in western Kenya. AIDS 2009; 23:195–201.

21. Ochieng-Ooko V, Ochieng D, Braitstein P, et al. Influence of gender on

loss to follow-up in a large HIV treatment programme in western

Kenya. Bull World Health Organ 2010; 88:681–8.

22. Milberg J, Sharma R, Scott F, Conviser R, Marconi K, Parham D.

Factors associated with delays in accessing HIV primary care in rural

Arkansas. AIDS Patient Care STDs 2001; 15:527–32.

23. Plitt SS, Mihalicz D, Singh AE, Jayaraman G, Houston S, Lee BE. Time

to testing and accessing care among a population of newly diagnosed

patients with HIV with a high proportion of Canadian aboriginals,

1998–2003. AIDS Patient Care STDS 2009; 23:93–9.

24. Braitstein P, Boulle A, Low N, et al. Gender and the use of antiretroviral

treatment in resource-constrained settings: findings from a multicenter

collaboration. J Womens Health 2008; 17:47–55.

25. John FN, Farquhar C, et al. Cost effectiveness of couple counselling

to enhance infant HIV-1 prevention. Int J STD AIDS 2008; 19:

406–9.

26. Howard AA, El-Sadr WM. Integration of tuberculosis and HIV services

in sub-Saharan Africa: lessons learned. Clin Infect Dis 2010; 50:238–44.

27. Lingappa JR, Lambdin B, Bukusi EA, et al. Regional differences in

prevalence of HIV-1 discordance in Africa and enrollment of HIV-1

discordant couples into an HIV-1 prevention trial. PLoS One 2008;

3:e1411.

28. Kumar A, Waterman I, Kumari G, Carter A. Prevalence and correlates

of HIV serostatus disclosure: a prospective study among HIV-infected

postparturient women in Barbados. AIDS Patient Care STDS 2006;

20:724–30.

29. Cohen MS, Chen YQ, McCauley M, et al. Prevention of HIV-1 infection

with early antiretroviral therapy. N Engl J Med 2011; 365:493–505.

30. Miller A, Rubin D. Motivations and methods for self-disclosure of HIV

seropositivity in Nairobi, Kenya. AIDS Behav 2007; 11:687–97.

31. Sullivan KM. Male self-disclosure of HIV-positive serostatus to sex

partners: a review of the literature. J Assoc Nurses AIDS Care 2005;

16:33–47.

HIV/AIDS d CID 2012:54 (15 January) d 281

by guest on Novem

ber 22, 2015http://cid.oxfordjournals.org/

Dow

nloaded from