BACKGROUND: Measures of diagnostic test accuracy provide evidence of how well a test correctly identifies or rules-out disease. Commonly used diagnostic accuracy measures (DAMs) include sensitivity and specificity, predictive values, likelihood ratios, area under the receiver operator characteristic curve (AUROC), area under precision-recall curves (AUPRC), diagnostic effectiveness (accuracy), disease prevalence, and diagnostic odds ratio (DOR) etc. Most available analysis tools perform accuracy testing for a single diagnostic test using summarized data. We developed a SAS macro for evaluating multiple diagnostic tests using individual-level data that creates a 2 × 2 summary table, AUROC and AUPRC as part of output. METHODS: The SAS macro presented here is automated to reduce analysis time and transcription errors. It is simple to use as the user only needs to specify the input dataset, "standard" and "test" variables and threshold values. It creates a publication-quality output in Microsoft Word and Excel showing more than 15 different accuracy measures together with overlaid AUROC and AUPRC graphics to help the researcher in making decisions to adopt or reject diagnostic tests. Further, it provides for additional variance estimation methods other than the normal distribution approximation. RESULTS: We tested the macro for quality control purposes by reproducing results from published work on evaluation of multiple types of dried blood spots (DBS) as an alternative for human immunodeficiency virus (HIV) viral load (VL) monitoring in resource-limited settings compared to plasma, the gold-standard. Plasma viral load reagents are costly, and blood must be prepared in a reference laboratory setting by a qualified technician. On the other hand, DBS are easy to prepare without these restrictions. This study evaluated the suitability of DBS from venous, microcapillary and direct spotting DBS, hence multiple diagnostic tests which were compared to plasma specimen. We also used the macro to reproduce results of published work on evaluating performance of multiple classification machine learning algorithms for predicting coronary artery disease. CONCLUSION: The SAS macro presented here is a powerful analytic tool for analyzing data from multiple diagnostic tests. The SAS programmer can modify the source code to include other diagnostic measures and variance estimation methods. By automating analysis, the macro adds value by saving analysis time, reducing transcription errors, and producing publication-quality outputs.

Many statistical models have been proposed in the literature for the analysis of longitudinal data. One may propose to model two or more correlated longitudinal processes simultaneously, with a goal of understanding their association over time. Joint modeling is then required to carefully study the association structure among the outcomes as well as drawing joint inferences about the different outcomes. In this study, we sought to model the associations among six nutrition outcomes while circumventing the computational challenge posed by their clustered and high-dimensional nature. We analyzed data from a 2  ×  2 randomized crossover trial conducted in Kenya, to compare the effect of high-dose and low-dose iodine in household salt on systolic blood pressure (SBP) and diastolic blood pressure (DBP) in women of reproductive age and their household matching pair of school-aged children. Two additional outcomes, namely, urinary iodine concentration (UIC) in women and children were measured repeatedly to monitor the amount of iodine excreted through urine. We extended the model proposed by Mwangi et al. (2021, Communications in Statistics: Case Studies, Data Analysis and Applications, 7(3), 413-431) allowing flexible piecewise joint models for six outcomes to depend on separate random effects, which are themselves correlated. This entailed fitting 15 bivariate general linear mixed models and deriving inference for the joint model using pseudo-likelihood theory. We analyzed the outcomes separately and jointly using piecewise linear mixed-effects (PLME) model and further validated the results using current state-of-the-art Jones and Kenward methodology (JKME model) used for analyzing randomized crossover trials. The results indicate that high-dose iodine in salt significantly reduced blood pressure (BP) compared to low-dose iodine in salt. Estimates for the random effects and residual error components showed that SBP and DBP had strong positive correlation, with effect of the random slope indicating that significantly related outcomes are strongly associated in their evolution. There was a moderately strong inverse relationship between evolutions of UIC and BP both in women and children. These findings confirmed the original hypothesis that high-dose iodine salt has significant lowering effect on BP. We further sought to evaluate the performance of our proposed PLME model against the widely used JKME model, within the multivariate joint modeling framework through a simulation study mimicking a 2 × 2 crossover design. From our findings, the multivariate joint PLME model performed exceptionally well both in estimation of random-effects matrix (G) and Hessian matrix (H), allowing satisfactory model convergence during estimation. It allowed a more complex fit to the data with both random intercepts and slopes effects compared to the multivariate joint JKME model that allowed for random intercepts only. When a hierarchical viewpoint is adopted, in the sense that outcomes are specified conditionally upon random effects, the variance-covariance matrix of the random effects must be positive definite. In some cases, additional random effects could explain much variability in the data, thus improving precision in estimation of the estimands (effect size) parameters. The key highlight in this evaluation shows that multivariate joint JKME model is a powerful tool especially while fitting mixed models with random intercepts only, in crossover design settings. Addition of random slopes may lead to model complexities in most cases, resulting in unsatisfactory model convergence during estimation. To circumvent convergence pitfalls, extention of JKME model to PLME model allows a more flexible fit to the data (generated from crossover design settings), especially in the multivariate joint modeling framework.

Cross-tabulations are a simple but important tool for understanding the distribution of socio-demographic characteristics among participants in epidemiological studies. We developed a generic SAS macro, %svy_freqs, to create publication-quality tables from cross-tabulations between a factor and a by-group variable given a third variable using survey or non-survey data. The macro also performs two-way cross-tabulations and provides extra features not available in existing procedures such as ability to incorporate parameters for survey design and replication-based variance estimation methods, performing validation checks for input parameters, transparently formatting variable values from character into numeric and allowing for generalizability. We demonstrate the macro using the 2013-2014 National Health and Nutrition Examination Survey (NHANES), a complex survey designed to assess the health and nutritional status of adults and children in the United States.

Introduction In epidemiological studies, cross-tabulations are a simple but important tool for understanding the distribution of socio-demographic characteristics among study participants. They become more useful when comparisons are presented using a by-group variable such as key demographic characteristic or an outcome status; for instance, sex or the presence or absence of a disease status. Most available statistical analysis software can easily perform cross-tabulations, however, output from these must be processed further to make it readily available for review and use in a publication. In addition, performing three-way cross-tabulations of complex survey data such as those required to show the distribution of disease prevalence across multiple factors and a by-group variable is not easily implemented directly using available standard procedures of commonly used statistical software.Methods We developed a generic SAS macro, %svy_freqs, to create quality publication-ready tables from cross-tabulations between a factor and a by-group variable given a third variable using survey or non-survey data. The SAS macro also performs classical two-way cross-tabulations and refines output into publication-quality tables. It provides extra features not available in existing procedures such as ability to incorporate parameters for survey design and replication-based variance estimation methods, performing validation checks for input parameters, transparently formatting character variable values into numeric ones and allowing for generalizability.Results We demonstrate the application of the SAS macro in the analysis of data from the 2013-2014 National Health and Nutrition Examination Survey (NHANES), a complex survey designed to assess the health and nutritional status of adults and children in the United States (U.S.).Conclusion The SAS code use to develop the macro is simple yet comprehensive, easy to follow, straightforward for the end user and simple for a SAS programmer to extend. The SAS macro has shown to shorten turn-around time for statistical analysis, eliminate errors when preparing output, and support reproducible research.

Introduction Reproducible research is increasingly gaining interest in the research community. Automating the production of research manuscript tables from statistical software can help increase the reproducibility of findings. Logistic regression is used in studying disease prevalence and associated factors in epidemiological studies and can be easily performed using widely available software including SAS, SUDAAN, Stata or R. However, output from these software must be processed further to make it readily presentable. There exists a number of procedures developed to organize regression output, though many of them suffer limitations of flexibility, complexity, lack of validation checks for input parameters, as well as inability to incorporate survey design.Methods We developed a SAS macro, %svy_logistic_regression, for fitting simple and multiple logistic regression models. The macro also creates quality publication-ready tables using survey or non-survey data which aims to increase transparency of data analyses. It further significantly reduces turn-around time for conducting analysis and preparing output tables while also addressing the limitations of existing procedures.Results We demonstrate the use of the macro in the analysis of the 2013-2014 National Health and Nutrition Examination Survey (NHANES), a complex survey designed to assess the health and nutritional status of adults and children in the United States. The output presented here is directly from the macro and is consistent with how regression results are often presented in the epidemiological and biomedical literature, with unadjusted and adjusted model results presented side by side.Conclusions The SAS code presented in this macro is comprehensive, easy to follow, manipulate and to extend to other areas of interest. It can also be incorporated quickly by the statistician for immediate use. It is an especially valuable tool for generating quality, easy to review tables which can be incorporated directly in a publication.

BACKGROUND: Nationally-representative surveys provide an opportunity to assess trends in recent HIV infection based on assays for recent HIV infection. METHODS: We assessed HIV incidence in Kenya in 2018 and trends in recent HIV infection among adolescents and adults in Kenya using nationally representative household surveys conducted in 2007, 2012 and 2018. To assess trends, we defined a recent HIV infection testing algorithm (RITA) that classified as recently infected (<12 months) those HIV-positive participants that were recent on the HIV-1 limiting antigen (LAg)-avidity assay without evidence of antiretroviral use. We assessed factors associated with recent and long-term (≥12 months) HIV infection versus no infection using a multinomial logit model while accounting for complex survey design. FINDINGS: Of 1,523 HIV-positive participants in 2018, 11 were classified as recent. Annual HIV incidence was 0.14% in 2018 (95% confidence interval [CI] 0.057-0.23), representing 35,900 (95% CI 16,300-55,600) new infections per year in Kenya among persons aged 15-64 years. The percentage of HIV infections that were determined to be recent was similar in 2007 and 2012 but fell significantly from 2012 to 2018 (adjusted odds ratio [aOR]=0.31, p<0.001). Compared to no HIV infection, being aged 25-34 versus 35-64 years (aOR=4.2, 95% CI 1.4-13), having more lifetime sex partners (aOR=5.2, 95% CI 1.6-17 for 2-3 partners and aOR=8.6, 95% CI 2.8-26 for ≥4 partners versus 0-1 partners), and never having tested for HIV (aOR=4.1, 95% CI 1.5-11) were independently associated with recent HIV infection. INTERPRETATION: Though HIV remains a public health priority in Kenya, HIV incidence estimates and trends in recent HIV infection support a significant decrease in new HIV infections from 2012 to 2018, a period of rapid expansion in HIV diagnosis, prevention, and treatment.

BACKGROUND: For assessing the HIV epidemic in Kenya, a series of independent HIV indicator household-based surveys of similar design can be used to investigate the trends in key indicators relevant to HIV prevention and control and to describe geographic and sociodemographic disparities, assess the impact of interventions, and develop strategies. We developed methods and tools to facilitate a robust analysis of trends across three national household-based surveys conducted in Kenya in 2007, 2012, and 2018. METHODS: We used data from the 2007 and 2012 Kenya AIDS Indicator surveys (KAIS 2007 and KAIS 2012) and the 2018 Kenya Population-based HIV Impact Assessment (KENPHIA 2018). To assess the design and other variables of interest from each study, variables were recoded to ensure that they had equivalent meanings across the three surveys. After assessing weighting procedures for comparability, we used the KAIS 2012 nonresponse weighting procedure to revise normalized KENPHIA weights. Analyses were restricted to geographic areas covered by all three surveys. The revised analysis files were then merged into a single file for pooled analysis. We assessed distributions of age, sex, household wealth, and urban/rural status to identify unexpected changes between surveys. To demonstrate how a trend analysis can be carried out, we used continuous, binary, and time-to-event variables as examples. Specifically, temporal trends in age at first sex and having received an HIV test in the last 12 months were used to demonstrate the proposed analytical approach. These were assessed with respondent-specific variables (age, sex, level of education, and marital status) and household variables (place of residence and wealth index). All analyses were conducted in SAS 9.4, but analysis files were created in Stata and R format to support additional analyses. RESULTS: This study demonstrates trends in selected indicators to illustrate the approach that can be used in similar settings. The incidence of early sexual debut decreased from 11.63 (95% CI: 10.95-12.34) per 1,000 person-years at risk in 2007 to 10.45 (95% CI: 9.75-11.2) per 1,000 person-years at risk in 2012 and to 9.58 (95% CI: 9.08-10.1) per 1,000 person-years at risk in 2018. HIV-testing rates increased from 12.6% (95% CI: 11.6%-13.6%) in 2007 to 56.1% (95% CI: 54.6%-57.6%) in 2012 but decreased slightly to 55.6% [95% CI: 54.6%-56.6%) in 2018. The decrease in incidence of early sexual debut could be convincingly demonstrated between 2007 and 2012 but not between 2012 and 2018. Similarly, there was virtually no difference between HIV Testing rates in 2012 and 2018. CONCLUSIONS: Our approach can be used to support trend comparisons for variables in HIV surveys in low-income settings. Independent national household surveys can be assessed for comparability, adjusted as appropriate, and used to estimate trends in key indicators. Analyzing trends over time can not only provide insights into Kenya's progress toward HIV epidemic control but also identify gaps.

INTRODUCTION: Reproducible research is increasingly gaining interest in the research community. Automating the production of research manuscript tables from statistical software can help increase the reproducibility of findings. Logistic regression is used in studying disease prevalence and associated factors in epidemiological studies and can be easily performed using widely available software including SAS, SUDAAN, Stata or R. However, output from these software must be processed further to make it readily presentable. There exists a number of procedures developed to organize regression output, though many of them suffer limitations of flexibility, complexity, lack of validation checks for input parameters, as well as inability to incorporate survey design. METHODS: We developed a SAS macro, %svy_logistic_regression, for fitting simple and multiple logistic regression models. The macro also creates quality publication-ready tables using survey or non-survey data which aims to increase transparency of data analyses. It further significantly reduces turn-around time for conducting analysis and preparing output tables while also addressing the limitations of existing procedures. In addition, the macro allows for user-specific actions to handle missing data as well as use of replication-based variance estimation methods. RESULTS: We demonstrate the use of the macro in the analysis of the 2013-2014 National Health and Nutrition Examination Survey (NHANES), a complex survey designed to assess the health and nutritional status of adults and children in the United States. The output presented here is directly from the macro and is consistent with how regression results are often presented in the epidemiological and biomedical literature, with unadjusted and adjusted model results presented side by side. CONCLUSIONS: The SAS code presented in this macro is comprehensive, easy to follow, manipulate and to extend to other areas of interest. It can also be incorporated quickly by the statistician for immediate use. It is an especially valuable tool for generating quality, easy to review tables which can be incorporated directly in a publication.

Western Kenya suffers a highly endemic and also very heterogeneous epidemic of human immunodeficiency virus (HIV). Although female sex workers (FSW) and their male clients are known to be at high risk for HIV, HIV prevalence across regions in Western Kenya is not strongly correlated with the fraction of women engaged in commercial sex. An agent-based network model of HIV transmission, geographically stratified at the county level, was fit to the HIV epidemic, scale-up of interventions, and populations of FSW in Western Kenya under two assumptions about the potential mobility of FSW clients. In the first, all clients were assumed to be resident in the same geographies as their interactions with FSW. In the second, some clients were considered non-resident and engaged only in interactions with FSW, but not in longer-term non-FSW partnerships in these geographies. Under both assumptions, the model successfully reconciled disparate geographic patterns of FSW and HIV prevalence. Transmission patterns in the model suggest a greater role for FSW in local transmission when clients were resident to the counties, with 30.0% of local HIV transmissions attributable to current and former FSW and clients, compared to 21.9% when mobility of clients was included. Nonetheless, the overall epidemic drivers remained similar, with risky behavior in the general population dominating transmission in high-prevalence counties. Our modeling suggests that co-location of high-risk populations and generalized epidemics can further amplify the spread of HIV, but that large numbers of formal FSW and clients are not required to observe or mechanistically explain high HIV prevalence in the general population. Copyright © 2018

BACKGROUND: WHO recommends viral load (VL) as the preferred method for diagnosing antiretroviral therapy (ART) failure; however, operational challenges have hampered the implementation of VL monitoring in most resource-limited settings. This study evaluated the accuracy of dried blood spot (DBS) VL testing under field conditions as a practical alternative to plasma in determining virologic failure (VF). METHODS: From May to December 2013, paired plasma and DBS specimens were collected from 416 adults and 377 children on ART ≥6 months at 12 clinics in Kenya. DBS were prepared from venous blood (V-DBS) using disposable transfer pipettes, and from finger-prick capillary blood using microcapillary tube (M-DBS) and directly spotting (D-DBS). All samples were tested on Abbott m2000 platform; V-DBS was also tested on Roche-CAP/CTM Version 2.0 platform. VF results were compared at three DBS thresholds (≥1000, ≥3000 and ≥5000 copies/ml) and a constant plasma threshold of ≥1000 copies/ml. RESULTS: On Abbott platform, at ≥1000 copies/ml threshold, sensitivities, specificities, and Kappa values for VF determination were ≥88.1%, ≥93.1% and ≥0.82 respectively for all DBS methods and it had the lowest percentage of downward misclassification compared to higher thresholds. V-DBS performance on CAP/CTM had significantly poorer specificity at all thresholds (1000-33.0%, 3000-60.9%, 5000-77.0%). No significant differences were found between adults and children. CONCLUSION: VL results from V-DBS, M-DBS and D-DBS were comparable to plasma for determining VF using the Abbott platform but not with CAP/CTM. A 1000 copies/ml threshold was optimal and should be considered for VF determination using DBS in adults and children.

OBJECTIVE: There is evidence of substantial subnational variation in the HIV epidemic. However, robust spatial HIV data are often only available at high levels of geographic aggregation and not at the finer resolution needed for decision making. Therefore, spatial analysis methods that leverage available data to provide local estimates of HIV prevalence may be useful. Such methods exist but have not been formally compared when applied to HIV. DESIGN/METHODS: Six candidate methods - including those used by the Joint United Nations Programme on HIV/AIDS to generate maps and a Bayesian geostatistical approach applied to other diseases - were used to generate maps and subnational estimates of HIV prevalence across three countries using cluster level data from household surveys. Two approaches were used to assess the accuracy of predictions: internal validation, whereby a proportion of input data is held back (test dataset) to challenge predictions; and comparison with location-specific data from household surveys in earlier years. RESULTS: Each of the methods can generate usefully accurate predictions of prevalence at unsampled locations, with the magnitude of the error in predictions similar across approaches. However, the Bayesian geostatistical approach consistently gave marginally the strongest statistical performance across countries and validation procedures. CONCLUSIONS: Available methods may be able to furnish estimates of HIV prevalence at finer spatial scales than the data currently allow. The subnational variation revealed can be integrated into planning to ensure responsiveness to the spatial features of the epidemic. The Bayesian geostatistical approach is a promising strategy for integrating HIV data to generate robust local estimates.

BACKGROUND: A clinical decision support system (CDSS) is a computer program that applies a set of rules to data stored in electronic health records to offer actionable recommendations. We aimed to establish whether a CDSS that supports detection of immunological treatment failure among patients with HIV taking antiretroviral therapy (ART) would improve appropriate and timely action. METHODS: We did this prospective, cluster randomised controlled trial in adults and children (aged ≥18 months) who were eligible for, and receiving, ART at HIV clinics in Siaya County, western Kenya. Health facilities were randomly assigned (1:1), via block randomisation (block size of two) with a computer-generated random number sequence, to use electronic health records either alone (control) or with CDSS (intervention). Facilities were matched by type and by number of patients enrolled in HIV care. The primary outcome measure was the difference between groups in the proportion of patients who experienced immunological treatment failure and had a documented clinical action. We used generalised linear mixed models with random effects to analyse clustered data. This trial is registered with ClinicalTrials.gov, number NCT01634802. FINDINGS: Between Sept 1, 2012, and Jan 31, 2014, 13 clinics, comprising 41 062 patients, were randomly assigned to the control group (n=6) or the intervention group (n=7). Data collection at each site took 12 months. Among patients eligible for ART, 10 358 (99%) of 10 478 patients were receiving ART at control sites and 10 991 (99%) of 11 028 patients were receiving ART at intervention sites. Of these patients, 1125 (11%) in the control group and 1342 (12%) in the intervention group had immunological treatment failure, of whom 332 (30%) and 727 (54%), respectively, received appropriate action. The likelihood of clinicians taking appropriate action on treatment failure was higher with CDSS alerts than with no decision support system (adjusted odds ratio 3·18, 95% CI 1·02-9·87). INTERPRETATION: CDSS significantly improved the likelihood of appropriate and timely action on immunological treatment failure. We expect our findings will be generalisable to virological monitoring of patients with HIV receiving ART once countries implement the 2015 WHO recommendation to scale up viral load monitoring.

BACKGROUND: Kenya is home to several high-performing internationally-accredited research laboratories, whilst most public sector laboratories have historically lacked functioning quality management systems. In 2010, Kenya enrolled an initial eight regional and four national laboratories into the Strengthening Laboratory Management Toward Accreditation (SLMTA) programme. To address the challenge of a lack of mentors for the regional laboratories, three were paired, or 'twinned', with nearby accredited research laboratories to provide institutional mentorship, whilst the other five received standard mentorship. Objectives: This study examines results from the eight regional laboratories in the initial SLMTA group, with a focus on mentorship models. METHODS: Three SLMTA workshops were interspersed with three-month periods of improvement project implementation and mentorship. Progress was evaluated at baseline, mid-term, and exit using the Stepwise Laboratory Quality Improvement Process Towards Accreditation (SLIPTA) audit checklist and scores were converted into a zero- to five-star scale. RESULTS: At baseline, the mean score for the eight laboratories was 32%; all laboratories were below the one-star level. At mid-term, all laboratories had measured improvements. However, the three twinned laboratories had increased an average of 32 percentage points and reached one to three stars; whilst the five non-twinned laboratories increased an average of 10 percentage points and remained at zero stars. At exit, twinned laboratories had increased an average 12 additional percentage points (44 total), reaching two to four stars; non-twinned laboratories increased an average of 28 additional percentage points (38 total), reaching one to three stars. CONCLUSION: The partnership used by the twinning model holds promise for future collaborations between ministries of health and state-of-the-art research laboratories in their regions for laboratory quality improvement. Where they exist, such laboratories may be valuable resources to be used judiciously so as to accelerate sustainable quality improvement initiated through SLMTA.

BACKGROUND: Populations with higher risks for HIV exposure contribute to the HIV epidemic in Kenya. We present data from the second Kenya AIDS Indicator Survey to estimate the size and HIV prevalence of populations with high-risk characteristics. METHODS: The Kenya AIDS Indicator Survey 2012 was a national survey of Kenyans aged 18 months to 64 years which linked demographic and behavioral information with HIV results. Data were weighted to account for sampling probability. This analysis was restricted to adults aged 18 years and older. RESULTS: Of 5088 men and 6745 women, 0.1% [95% confidence interval (CI): 0.03 to 0.14] were persons who inject drugs (PWID). Among men, 0.6% (CI: 0.3 to 0.8) had ever had sex with other men, and 3.1% (CI: 2.4 to 3.7) were males who had ever engaged in transactional sex work (MTSW). Among women, 1.9% (CI: 1.3 to 2.5) had ever had anal sex, and 4.1% (CI: 3.5 to 4.8) were women who had ever engaged in transactional sex work (FTSW). Among men, 17.6% (CI: 15.7 to 19.6) had been male clients of transactional sex workers (TSW). HIV prevalence was 0% among men who have sex with men, 6.3% (CI: 0 to 18.1) among persons who injected drugs, 7.1% (CI: 4.8 to 9.4) among male clients of TSW, 7.6% (CI: 1.8 to 13.4) among MTSW, 12.1% (CI: 7.1 to 17.1) among FTSW, and 12.1% (CI: 5.0 to 19.2) among females who ever had engaged in anal sex. CONCLUSIONS: Population-based data on high-risk populations can be used to set realistic targets for HIV prevention, care, and treatment for these groups. These data should inform priorities for high-risk populations in the upcoming Kenyan strategic plan on HIV/AIDS.

BACKGROUND: The Kenyan Ministry of Health initiated a voluntary medical male circumcision (VMMC) program in 2008. We used data from 2 nationally representative surveys to estimate trends in the number, demographic characteristics, and sexual behaviors of recently circumcised and uncircumcised HIV-uninfected men in Kenya. METHODS: We compared the proportion of circumcised men between the first and second Kenya AIDS Indicator Survey (KAIS 2007 and KAIS 2012) to assess the progress of Kenya's VMMC program. We calculated the number of uncircumcised HIV-uninfected men. We conducted descriptive analyses and used multivariable methods to identify the variables independently associated with HIV-uninfected uncircumcised men aged 15-64 years in the VMMC priority region of Nyanza. RESULTS: The proportion of men who reported being circumcised increased significantly from 85.0% in 2007 to 91.2% in 2012. The proportions of circumcised men increased in all regions, with the highest increases of 18.1 and 9.0 percentage points in the VMMC priority regions of Nyanza and Nairobi, respectively. Half (52.5%) of HIV-uninfected and uncircumcised men had never been married, and 84.6% were not using condoms at all times with their last sexual partner. CONCLUSIONS: VMMC prevalence has increased across Kenya demonstrating the success of the national program. Despite this accomplishment, the Nyanza region remains below the target to circumcise 80% of all eligible men aged 15-49 years between 2009 and 2013. As new cohorts of young men enter into adolescence, consistent focus is needed. To ensure sustainability of the VMMC program, financial resources and coordinated planning must continue.