The group-randomized trial described in this paper is designed to evaluate the effectiveness and feasibility of integrating HIV prevention into the routine care of PLHIV attending care and treatment clinics in Kenya, Namibia, and Tanzania. Baseline data from the study participants provides information about the sociodemographic characteristics, health status, and HIV risk behavior of patients attending care. These data indicate that participants were similar in socioeconomic status (SES) to nationally representative samples in each of the countries 
, with low education and income levels. The lack of paid work and low income reported by many of the PLHIV in this study may affect their health and clinic attendance. For example, transportation costs are often reported as a reason for poor clinic attendance among PLHIV 
The mean age of participants in this sample was higher than national estimates of ages of PLHIV, especially for women 
. In addition, the majority of participants in this study were diagnosed with HIV within the past 3 years. Over half of participants had their most recent CD4 count less than 350 cells/mm3
, and participants who had been on ARVs less than one year (20% of those on ARVs) had very low CD4 counts (median
170). These data suggest that many participants accessed care late in their illness, a finding consistent with other studies from sub-Saharan Africa showing that many patients access care after developing advanced symptomatic disease 
and aren't benefiting from care and treatment interventions to improve morbidity and mortality. In addition, their declining CD4 counts and increased viral load are also placing their sexual partners at increased risk of HIV acquisition 
. Increased HIV testing and counseling (HTC) efforts with focused linkage to care and treatment services in both facilities (e.g., ANC, TB, STI, and out-patient departments) and communities (e.g., home-based testing and mobile testing) are needed to identify individuals with asymptomatic disease, especially men, and ensure earlier enrollment into appropriate prevention, care, and treatment services. In addition, as one-third of participants did not know their partners' HIV status, more intensive efforts are clearly needed to increase partner and couple testing to identify HIV-positive persons and discordant couples and to link them with prevention, care, and treatment services.
Men in HIV care and treatment clinics and in the study sample are under-represented compared to the population of PLHIV. Men also had significantly lower CD4 counts than women. Nearly two-thirds of men had CD4 counts less than 350 cells/mm3
compared to less than half of women (64% vs. 46%, p<0.0001). This is consistent with several other studies from sub-Saharan Africa which have found that men are less likely than women to enroll in care at an earlier clinical stage 
, and to remain in clinical care after enrollment 
. These findings highlight the importance of strengthening HIV testing and counseling programs targeting men, such as medical male circumcision programs and partner testing in ANC settings. Efforts to identify HIV-positive men earlier, link them to prevention, care and treatment services, and retain them in care are clearly needed.
The risk behavior data from this study identify HIV prevention needs among PLHIV attending HIV clinics. Specifically, nearly one-fifth of the participants were in a known discordant relationship, and over one-third did not know their partner's HIV status. In addition, 18% of the sample had not disclosed their HIV status to their partner and 24% were not consistently using condoms. These data indicate that many participants are in need of prevention services such as support for disclosure of HIV status, partner testing, and reducing unprotected sex. In addition, alcohol use was reported by about 20% of participants. Alcohol use has been associated with poor adherence 
and increased disease progression 
, as well as increased risky sexual behavior among PLHIV 
. Interventions to address these prevention issues in the context of comprehensive care and treatment services are indicated.
There were significant differences between countries in many of the reported HIV risk behaviors. For example, consistent condom use among PLHIV in care was 86% in Namibia, but significantly lower in Tanzania (69%). Similarly, only 60% of participants in Tanzania and 63% in Kenya knew their partner's HIV status, but 73% of Namibian participants knew their partner's status. There were differences in reported alcohol use as well, with Tanzanian and Kenyan participants less likely to drink alcohol than Namibian participants, and Tanzanian participants also less likely to report problem drinking. These participant baseline differences between countries highlight the importance of considering the local context when designing multi-country studies and the need to recognize that regional differences could affect generalizability of results from studies conducted in a single region or country.
This study has several limitations. Given that this was a clinic-based sample conducted in three countries with generalized HIV epidemics, the generalizability of these findings to PLHIV not enrolled in clinical care and/or to PLHIV in non-generalized epidemics may be limited. In addition, much of the data in this paper are from patient self-report, thus social desirability and recall bias may have affected participants' responses. As a result, some variables may be overestimated (e.g., condom use, disclosure, adherence) and others underestimated (e.g., alcohol use).
In summary, the baseline data from this study suggest that increased efforts are needed to identify PLHIV earlier, especially men, and to ensure they access prevention, care, and treatment services following diagnosis. Many PLHIV would also benefit from prevention interventions which address disclosure support, partner testing, alcohol and sexual risk reduction, as well as providing contraceptives and support for safer pregnancy. Integrating these services into the clinical care of HIV-positive persons may increase access to prevention interventions and improve retention in clinical care