You are here

Article Text

Article menu

Download PDFPDF

Epidemiology
Epidemiology of syphilis in Kenya: results from a nationally representative serological survey
  1. Boaz Otieno-Nyunya1,
  2. Eddas Bennett2,
  3. Rebecca Bunnell1,
  4. Sufia Dadabhai3,
  5. Anthony Gichangi A1,
  6. Nelly Mugo4,
  7. John Wanyungu5,
  8. Isaack Baya6,
  9. Reinhard Kaiser1,
  10. for the Kenya AIDS Indicator Survey Study Team
  1. 1Center for Global Health, Division of Global HIV/AIDS, Centers for Disease Control and Prevention, Nairobi, Kenya
  2. 2Center for Global Health, Division of Global HIV/AIDS, Centers for Disease Control and Prevention, Atlanta, USA
  3. 3Department of Epidemiology and Biostatistics, University of California. San Francisco, USA
  4. 4Kenyatta National Hospital and University of Nairobi, Nairobi, Kenya
  5. 5National AIDS and STD Control Program, Kenya Ministry of Health, Nairobi, Kenya
  6. 6Kenya National Public Health Laboratory Services, Nairobi, Kenya
  1. Correspondence to Dr Reinhard Kaiser, Center for Global Health, Division of Global HIV/AIDS, Centers for Disease Control and Prevention Kenya, P.O. Box 606-00621 Village Market Nairobi, Kenya; r_kaiser{at}gmx.de

Abstract

Objectives The authors used data from the Kenya AIDS Indicator Survey (KAIS) 2007 to determine the prevalence of syphilis and associated risk factors among adults aged 15–64 years.

Methods KAIS was a nationally representative population-based sero-survey that examined demographic and behavioural indicators and serological testing for syphilis, HIV and herpes simplex virus type 2 (HSV-2) in adults aged 15–64 years. The authors analysed data from 8935 women and 6727 men with complete syphilis results. Logistic regression models stratified by sex were used to assess potential factors associated with syphilis sero-prevalence.

Results Overall, 262 adults tested positive for syphilis (1.8%, 95% CI 1.5% to 2.1%); sero-prevalence was similar among women and men (1.7%, 95% CI 1.3% to 2.0% and 1.9%, 95% CI 1.5% to 2.3%, respectively). Syphilis prevalence was the highest among men with HIV (6.4%, 95% CI 3.1% to 9.7%) and HSV-2 (4.5%, 95% CI 3.4% to 5.7%) infection. Independent risk factors for syphilis included HIV (men only, adjusted OR (AOR) 3.4, 95% CI 1.6% to 7.1%), HSV-2 (women, AOR 3.5, 95% CI 2.1% to 5.8%; men AOR 2.2, 95% CI 1.3% to 3.7%), lack of male circumcision (AOR 2.2, 95% CI 1.3% to 3.7%), poorest or poorer versus richest wealth index (women, AOR 2.0, 95% CI 1.0% to 4.2%; men AOR 2.5, 95% CI 1.4% to 4.9%) and no primary versus secondary or more education in men (AOR 4.8, 95% CI 2.0% to 11.7%).

Conclusions Syphilis prevalence in the general population in Kenya is relatively low and eradication could be possible but would require intensified syphilis prevention and control efforts, including routine screening in HIV, sexually transmitted infection and antenatal care clinics as well as in family planning and male circumcision settings.

  • Syphilis
  • genital ulcer disease
  • Kenya
  • national prevalence
  • risk factors

https://doi.org/10.1136/sextrans-2011-050026

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Introduction

    Understanding the epidemiology of syphilis, including incidence, prevalence and associated risk factors, is important in the prevention and control of syphilis, congenital syphilis and other sexually transmitted infections (STIs) including HIV. While syphilis testing is common among women attending selected antenatal care (ANC) clinics, test results are rarely presented as part of HIV surveillance. In general, there is a need for more nationally representative STI prevalence estimates in Africa, including syphilis.1 Community-based surveys have shown wide variations in syphilis prevalence. In a study of four African cities, prevalence varied from 2% in Cotonou, Benin, to 4% in Kisumu, Kenya, 6% in Yaounde, Cameroon, and 14% in Ndola, Zambia.2 In this study, first sex before the age of 15 years, more than one non-spousal partner in the past year, more than three lifetime partners, female gender, being married, alcohol use and lower educational level were associated with increased syphilis prevalence.2 In South Africa, decline in national syphilis prevalence was reported among antenatal clinic attenders from 11.2% in 1997 to 4.9% in 2000 and to 1.9% in 2009.3 Among African blood donors, prevalence ranged from 0.4% in Djibouti to 12.7% in Tanzania.4 5

    In Kenya, little is known about the prevalence of syphilis in the general population. Among women attending ANC clinics, syphilis prevalence has declined from 7.5% in 1994 to 3.8% in 19976 and further to 0.6% in 2006 (95% CI 0.5% to 0.8%), 0.9% in 2008 (95% CI 0.7% to 1.1%) and 0.4% in 2010 (95% CI 0.3% to 0.5%) (Davies Kimanga, Kenya National AIDS and STI Control Programme, personal communication, 2011). Studies among specific groups have shown rates ranging from 1.2% and 1.7% among blood donors in Nairobi7 8 to 3.1% among patients attending an HIV clinic in western Kenya.9 Higher prevalence ranging from 7–16% in 1997–1999 to 12.2% in 2000–2001 have also been reported among Kenyan sex workers.10

    The 2007 Kenya AIDS Indicator Survey (KAIS) offered a unique opportunity to study for the first time the epidemiology of syphilis in a nationally representative survey of adults aged 15–64 years.

    Methods

    Study population and study design

    KAIS was a cross-sectional household survey designed to be nationally representative of Kenya's adult population aged 15–64 years. KAIS was a stratified, two-stage cluster sample that was comparable with the design of the 2003 Kenya Demographic and Health Survey (DHS).11 The first stage involved selecting clusters from the same sampling frame that was used for the 2003 DHS, based on the 1999 national census, and the second stage involved the selection of households per cluster with equal probability of selection in the rural–urban strata within each district. From August to December 2007, a structured interviewer administered questionnaire was used to obtain participant demographics and sexual and behavioural characteristics. A venous blood draw was used to assess sero-prevalence of syphilis, HIV and herpes simplex virus type 2 (HSV-2). Further details of KAIS methods are published elsewhere.12 Ethical approval was obtained from the Kenya Medical Research Institute (KEMRI) Ethical Review Committee and the US Centers for Disease Control and Prevention (CDC) Institutional Review Board. All consenting participants signed written informed consent forms for the interviews and blood draws.

    Laboratory methods

    Laboratory tests were performed at the Kenya National Public Health Laboratory, and quality assurance was conducted at the CDC/KEMRI laboratory in Nairobi. For syphilis, a Treponema pallidum particle agglutination assay (TPPA) (Serodia-TPPA, Fujirebio Diagnostics Inc) test was used as an antibody screening test to identify previous exposure to syphilis. The test remains positive indefinitely even after treatment. Rapid plasma reaginin (RPR) (Macrovu-Vue RPR Card Test, BD, USA) was used on TPPA-positive specimens to identify active infections. Syphilis for prevalence estimation was defined as positive test results for TPPA and RPR. Participants who tested positive on the two tests received their results and were referred for treatment. HIV testing was performed using enzyme linked immunoassay (EIA) (Vironstika HIV-1/2) for screening and EIA (Murex HIV 1.2.0) for confirmation in a serial testing algorithm. Discordant samples were retested with the two assays. PCR (Roche HIV DNA v 1.5) tests were conducted on all samples with two sets of discordant results. For quality control, all HIV EIA positive specimens and 5% of negative specimens were retested in a different laboratory using the same testing algorithm. For HSV-2 testing, we used the Kalon HSV Type 2-specific immunoglobulin G (IgG) EIA (Kalon Biological, Ltd., Guildford, UK). This was a recombinant type 2 antigen (IgG-2) modified to eliminate reactivity arising from HSV-1 infection.

    Measures

    Wealth was a composite index of the living standard of a household calculated using data on a household's ownership of selected assets, materials used for housing construction, water access and sanitation facilities. The wealth index placed households on a continuous scale of relative wealth using principal components analysis. Individuals were categorised according to the score of their household and the sample was divided into quintiles, each with an equal number of individuals, ranging from the poorest to wealthiest.

    Data analysis

    We analysed syphilis sero-prevalence in women and men and examined potential risk factors for syphilis, including socio-demographic characteristics, HIV and HSV-2. To account for the complex survey design, data were analysed using survey procedures in SAS V.9.2 (SAS Institute, Inc) and SUDAAN V.10 (Research Triangle Institute, Cary, North Carolina, USA). Estimates were weighted for design and non-response. Population estimates were calculated based on the 2007 projected Kenyan population.13

    We used the Rao-Scott χ2 test to examine associations between the categorical variables and syphilis sero-status. The Rao-Scott χ2 is a survey design-adjusted version of the Pearson χ2 statistic. Logistic regression models stratified by sex were used to assess potential factors associated with syphilis sero-prevalence, including HIV and HSV-2 status, sex, age, marital status, education, wealth index, number of lifetime sex partners, male circumcision and interaction between these variables. Factors with p values <0.10 in bivariate analyses were included in multivariable models. ORs from multivariable models were adjusted for all other variables that remained significant in the model, and p values <0.05 were considered statistically significant. Estimates with a relative SE >30% were considered unreliable and should be interpreted with caution.

    Results

    KAIS consisted of 9691 households and 19 840 eligible adults and had an overall response rate of 90% for the questionnaire (women 93%, men 87%) and 80% for blood draws (women 83%, men 77%). There were no substantial differences between the individuals who provided a blood sample and those who did not with respect to major demographic and behavioral factors. We analysed data from 8935 women and 6727 men with complete syphilis results. Overall syphilis prevalence was 1.8% (95% CI 1.5% to 2.1%) with similar, not significantly different (p=0.4913) rates among women (1.7%, 95% CI 1.3% to 2.0%) and men (1.9%, 95% CI 1.5% to 2.3%) (table 1). Results indicated that syphilis prevalence increased with age, lower education, lower wealth index and more lifetime sex partners, for both women and men. Nyanza province had the highest prevalence in women and the second highest prevalence in men. Men aged 50 years and older had a significantly higher prevalence than men younger than 50 years (p<0.0001). Syphilis prevalence was also higher among women and men who tested positive for HIV or HSV-2, compared with those who tested negative and was higher among uncircumcised men compared with circumcised men (table 1). Condom use during last sexual activity was associated with lower syphilis prevalence in men but not in women; however, these estimates were unreliable, due to the small number of respondents with syphilis among condom users (relative SE >30%) (table 1). Estimates by marital status and HIV-syphilis co-infection (data not shown) were also unreliable, for similar reasons. Syphilis prevalence did not differ by urban/rural residence or perceived risk of HIV infection (data not shown). Syphilis prevalence was relatively low (0.3%) among women who had given birth in the prior 4 years or were pregnant at the time of the survey. Because of the very small number (only one syphilis positive respondent in this group), this finding should be interpreted with caution.

    View this table:
    Table 1

    Characteristics of survey respondents by syphilis prevalence and sex, Kenya AIDS Indicator Survey 2007

    In final multivariable models, separated, divorced or widowed versus never married, poorest or poorer versus richest wealth index, four or more versus one lifetime sex partner, and HSV-2 infection were independently associated with syphilis sero-prevalence in women (table 2). In men, increasing age, no and incomplete primary versus secondary or higher education, poorest or poorer versus richest wealth index, HIV infection and HSV-2 infection remained statistically significant (table 2).

    View this table:
    Table 2

    Factors independently associated with syphilis, Kenya AIDS Indicator Survey 2007

    Discussion

    For the first time in Kenya, KAIS 2007 provided estimates of syphilis sero-prevalence from a nationally representative population-based survey. A sero-prevalence of 1.8% was equivalent to approximately 300 000 Kenyan adults with active syphilis infection at the time of the survey. Antenatal sentinel surveillance in Kenya has shown a declining trend in pregnant women with recent prevalence estimates that were below 1%. Similar trends were also reported from South Africa.3 KAIS 2007 did not show a significant difference in syphilis prevalence between women and men, in contrast with a review from South Africa that showed mostly higher rates in women than men in the general population.1 In our study, men with HIV and HSV-2 infection had the highest syphilis prevalence and the largest differences in prevalence compared with persons without those infections (prevalence ratio of four and five, respectively). The relatively high syphilis prevalence in Nyanza province may be related to the relatively low proportion of circumcised men (48% compared with 85% nationally12) and the substantial burden of HIV infection (15% compared with 7% nationally12).

    HSV-2 infection and the lowest wealth group, compared with the highest wealth group, were the only independent risk factors for syphilis infection in both women and men. The adjusted odds of testing syphilis positive were 3.5 and 2.5 times higher in women and in men with HSV-2 than in those who were HSV-2 uninfected, respectively. The highest other adjusted odds were 2.5 times higher in men with HIV infection than in men who were HIV uninfected, more than four times higher in men with no primary education than in men with secondary or higher education, and eight times higher in men aged 50–64 years compared with men aged 15–29 years. The unadjusted odds for uncircumcised men to test positive for syphilis were twice as high as that for circumcised men; however, male circumcision did not remain significantly associated with syphilis after multivariable adjustment. Being currently married appeared to protect against syphilis in men; however, these results need to be interpreted with caution, due to relative SEs >30%.

    Our study was not powered to assess syphilis prevalence among pregnant women. Also, we did not conduct titration of RPR or solicit treatment histories to understand the likelihood that a syphilis sero-positive woman could have transmitted syphilis to her newborn. A recent study in rural Tanzania found a syphilis sero-prevalence of 1.6% among pregnant women aged 15–49 years,14 which is similar to our finding. If congenital syphilis is to be eliminated in Kenya, screening and treatment will need to be extended to all pregnant women. Syphilis screening at ANC clinics should be routine as a part of the national syphilis control policy in Kenya15 16 and continuous efforts are needed to ensure high quality of ANC services including adequate staff, equipment, supplies and ANC attendance.16–19

    Both HSV-2 and syphilis infection are biological cofactors in HIV acquisition and transmission.20 21 In our study, HSV-2 was independently associated with syphilis infection, highlighting the importance of increasing our understanding of this epidemiological synergy and developing policies and guidelines for prevention of HSV-2, syphilis and HIV infections. While HSV-2 testing is currently not widely available in Kenya, syphilis can be diagnosed. A serological test for syphilis should be performed on all pregnant women at the first prenatal visit and persons with signs of genital ulcer disease.22 23 Data suggest that consistent and correct use of latex condoms reduces the risk for HSV-2 and syphilis when the infected area or site of potential exposure is covered.24 25 Male circumcision has been shown to prevent HSV-2 and syphilis infections.26 STI clinics should routinely refer HIV uninfected men for male circumcision.

    Apart from limited power for some subgroup analyses, our study had other limitations. Cross-sectional surveys do not allow for determination of the sequence of cause and effect which complicates interpretation of associations. Furthermore, some of the factors used in the analysis were self-reported allowing for potential reporting bias. Finally, endemic trepanematoses (yaws, bejel and pinta) may have also been detected by TPPA and RPR testing because of cross reactivity. Some portion of the prevalence, especially in older adults, may be explained by residual antibody to endemic trepanematoses. Through a concerted effort to eliminate yaws made by WHO from 1948 to 1970,27 it has become a less common disease in Africa occurring primarily in a band stretching across equatorial Africa from Liberia to southern Ethiopia and Somalia.28 While yaws, which is caused by Treponema pertenue, a subspecies of T pallidum, was endemic in Kenya in the 1950s, to our knowledge, there is currently no residual yaws in Kenya. The prevalence reported in this study, in respondents under 50 years, is therefore most likely due to syphilis infection.

    The relatively low syphilis sero-prevalence in KAIS 2007 gives hope that elimination of syphilis is a possibility in Kenya. Elimination, however, will demand intensified syphilis prevention and control efforts. Screening for syphilis should be routine in HIV, STI and ANC clinics as well as in family planning and male circumcision settings. Improving antenatal syphilis screening in concert with antenatal HIV screening will also allow targeting elimination of congenital syphilis.

    Key messages

    • The national syphilis sero-prevalence in the Kenya AIDS Indicator Survey 2007 was 1.8% (95% CI 1.5% to 2.1%).

    • Prevention and control efforts need to be intensified to advance towards syphilis elimination in Kenya.

    • Screening for syphilis should be routine in HIV, sexually transmitted infection and antenatal clinics as well as in family planning and male circumcision settings.

    • Improving antenatal syphilis screening will allow targeting the elimination of congenital syphilis in Kenya.

    Acknowledgments

    We thank Drs Willi McFarland and George Rutherford of the University of California, San Francisco, for reviewing the manuscript and offering useful suggestions. We also wish to thank all our study respondents.

    References

    1. ↵
      1. Johnson L,
      2. Coetzee D,
      3. Dorrington R
      . Sentinel surveillance of sexually transmitted infections in South Africa: a review. Sex Transm Infect 2005;81:287–93.
      OpenUrlAbstract/FREE Full Text
    2. ↵
      1. Buve A,
      2. Weiss HA,
      3. Laga M,
      4. et al
      . The epidemiology of gonorrhoea, chlamydial infection and syphilis in four African cities. AIDS 2001;(15 Suppl 4):S79–88.
    3. ↵
      Department of Health of South Africa, 2010. National Antenatal Sentinel HIV and Syphilis Prevalence Survey in South Africa. 2009. http://www.doh.gov.za/docs/reports/2010/hiv_syphillis_09.pdf (accessed 17 Jun 2011).
    4. ↵
      1. Massenet D,
      2. Bouh A
      . Aspects of blood transfusion in Djibouti. Med Trop (Mars) 1997;57:202–5.
      OpenUrlPubMed
    5. ↵
      1. Matee MI,
      2. Lyamuya EF,
      3. Mbena EC,
      4. et al
      . Prevalence of transfusion-associated viral infections and syphilis among blood donors in Muhimbili Medical Centre, Dar es Salaam, Tanzania. East Afr Med J 1999;76:167–71.
      OpenUrlPubMedWeb of Science
    6. ↵
      1. Temmerman M,
      2. Fonck K,
      3. Bashir F,
      4. et al
      . Declining syphilis prevalence in pregnant women in Nairobi since 1995: another success story in the STD field? Int J STD AIDS 1999;10:405–8.
      OpenUrlCrossRefPubMedWeb of Science
    7. ↵
      1. Abdalla F,
      2. Mwanda OW,
      3. Rana F
      . Comparing walk-in and call-responsive donors in a national and a private hospital in Nairobi. East Afr Med J 2005;82:531–5.
      OpenUrlPubMed
    8. ↵
      1. Makokha EP,
      2. Otsyula MBV,
      3. Mining SK,
      4. et al
      . Prevalence of Hepatitis B and C, Syphilis and HTLV-1 in HIV-positive voluntary donor blood in western Kenya. Abstract International Conference on AIDS. Abstract MoPeB3287. Bangkok, Thailand, 12 July 2004. http://www.aids2004.org/admin/images/upload/424.pdf (accessed 5 Sep 2011).
    9. ↵
      1. Thumbi M
      . Testing time for syphilis in Kenya. Biomed Sci 2007;824–5.
    10. ↵
      1. Kaul R,
      2. Kimani J,
      3. Nagelkerke NJ,
      4. et al
      . Risk factors for genital ulcerations in Kenyan sex workers: the role of human immunodeficiency virus type 1 infection. Sex Transm Dis 1997;24:387–92.
      OpenUrlCrossRefPubMedWeb of Science
    11. ↵
      Central Bureau of Statistics (CBS) Kenya, Ministry of Health (MOH) Kenya and ORC Macro. Kenya Demographic and Health Survey 2003. Calverton, MD: CBS, MOH and ORC Macro, 2004.
    12. ↵
      NASCOP. 2007 Kenya AIDS Indicator Survey: Final Report. Nairobi, Kenya: National AIDS and STI Control Programme, 2009.
    13. ↵
      Kenya National Bureau of Statistics. Revised Population Projections for Kenya 2000-2010, Nairobi Kenya: Kenya National Bureau of Statistics, 2006.
    14. ↵
      1. Yahya-Malima KI,
      2. Evjen-Olsen B,
      3. Matee MI,
      4. et al
      . HIV-1, HSV-2 and syphilis among pregnant women in a rural area of Tanzania: prevalence and risk factors. BMC Infect Dis 2008;8:75.
      OpenUrlCrossRefPubMed
    15. ↵
      Republic of Kenya Ministry of Health. National Guildelines for Reproductive Tract Infection Services. 1st edn. Nairobi, Kenya: Republic of Kenya Ministry of Health, 2006.
    16. ↵
      1. Gichangi P,
      2. Fonk K,
      3. Sekadde-Kigondu C,
      4. et al
      . Partner notification of pregnant women infected with syphilis in Nairobi, Kenya. Int J STD AIDS 2000;11:257–61.
      OpenUrlAbstract/FREE Full Text
    17. ↵
      Central Bureau of Statistics Ministry of Planning and National Development and Ministry of Health. Kenya Demographic and Health Survey 2008-2009 Preliminary Report. Nairobi, Kenya: Central Bureau of Statistics Ministry of Planning and National Development and Ministry of Health, 2010.
    18. ↵
      1. van Eijk AM,
      2. Bles HM,
      3. Odhiambo F,
      4. et al
      . Use of antenatal services and delivery care among women in rural western Kenya: a community based survey. Reprod Health 2006;3:2.
      OpenUrlCrossRefPubMed
    19. ↵
      1. Mwaniki PK,
      2. Kabiru EW,
      3. Mbugua GG
      . Utilisation of antenatal and maternity services by mothers seeking child welfare services in Mbeere District, Eastern Province, Kenya. East Afr Med J 2002;79:184–7.
      OpenUrlPubMed
    20. ↵
      1. Abu-Raddad LJ,
      2. Magaret AS,
      3. Celum C,
      4. et al
      . Genital herpes has played a more important role than any other sexually transmitted infection in driving HIV prevalence in Africa. PLoS ONE 2008;3:e2230.
      OpenUrlCrossRefPubMed
    21. ↵
      1. Tobian AA,
      2. Quinn TC
      . Herpes simplex virus type 2 and syphilis infections with HIV: an evolving synergy in transmission and prevention. Curr Opin HIV AIDS 2009;4:294–9.
      OpenUrlCrossRefPubMed
    22. ↵
      WHO. Guidelines for the Management of Sexually Transmitted Infections. 2003. http://www.who.int/reproductive-health (accessed 17 Jun 2011).
    23. ↵
      Centers for Disease Control and Prevention. 2010 Guidelines for Treatment of Sexually Transmitted Diseases. MMWR 2006;59:1–110.
      OpenUrl
    24. ↵
      1. Martin ET,
      2. Krantz E,
      3. Gottlieb SL,
      4. et al
      . A pooled analysis of the effect of condoms in preventing HSV-2 acquisition. Arch Intern Med 2009;169:1233–40.
      OpenUrlCrossRefPubMed
    25. ↵
      1. Koss CA,
      2. Dunne EF,
      3. Warner L
      . A systematic review of epidemiologic studies assessing condom use and risk of syphilis. Sex Transm Dis 2009;36:401–5.
      OpenUrlCrossRefPubMed
    26. ↵
      1. Tobian AA,
      2. Serwadda D,
      3. Quinn TC,
      4. et al
      . Male circumcision for the prevention of HSV-2 and HPV infections and syphilis. N Engl J Med 2009;360:1298–309.
      OpenUrlCrossRefPubMedWeb of Science
    27. ↵
      World Health Organization. Yaws and Endemic Trepanematoses. Epidemiological Situation. http://www.who.int/yaws/epidemiology/en/index.html (accessed 23 Feb 2011).
    28. ↵
      World Health Organization. Yaws and Other Endemic Trepanematoses. Historical Perspective. http://www.who.int/yaws/en/ (accessed 23 Feb 2011).

    Footnotes

    • Kenya AIDS Indicator Survey Study Group: Michael Arnold, Godfrey Baltazar, Isack Baya, John Bore, Rebecca Bunnell, Sufia Dadabhai, Helen Dale, Jared Ichwara, Allen Hightower, Tura Galgalo, Catherine Gichimu, Anthony Gichangi, Reinhard Kaiser, Timothy Kellogg, George Kichamu, Andrea Kim, Evelyn Kim, Samuel Kipruto, George K'opiyo, Ernest Makokha, Barbara Marston, Lawrence Marum, Margaret Mburu, Jonathan Mermin, Joy Mirjahangir, Rex Mpazanje, Ibrahim Mohamed, Patrick Muriithi, James Muttunga, Mary Mwangi, Carol Ngare, Raymond Nyoka, Linus Odawo, Samuel Ogola, Christopher Omolo, Tom Oluoch, Ray Shiraishi, John Wanyungu and Anthony Waruru.

    • Funding This work was supported by the President's Emergency Plan for AIDS Relief (PEPFAR) through the US Centers for Disease Control and Prevention (CDC) and the USA Agency for International Development and by the Joint United Nations Programme on HIV/AIDS. Several authors are CDC employees. The findings and conclusions in this manuscript are those of the authors and do not necessarily represent the views of the Centers for Disease Control and Prevention.

    • Competing interests None.

    • Patient consent Obtained.

    • Ethics approval Kenya Medical Research Institute (KEMRI) Ethical Review Committee and the US Centers for Disease Control and Prevention (CDC) Institutional Review Board.

    • Provenance and peer review Not commissioned; externally peer reviewed.