BackgroundWe aimed to determine the prevalence of pulmonary TB amongst the adult population (≥15 years) in 2016 in Kenya.MethodA nationwide cross-sectional survey where participants first underwent TB symptom screening and chest x-ray. Subsequently, participants who reported cough >2weeks and/or had a chest x-ray suggestive of TB, submitted sputum specimen for laboratory examination by smear microscopy, culture and Xpert MTB/RIF.ResultThe survey identified 305 prevalent TB cases translating to a prevalence of 558 [95%CI 455–662] per 100,000 adult population. The highest disease burden was reported among people aged 25–34 years (716 [95% CI 526–906]), males (809 [(95% CI 656–962]) and those who live in urban areas (760 [95% CI 539–981]). Compared to the reported TB notification rate for Kenya in 2016, the prevalence to notification ratio was 2.5:1. The gap between the survey prevalence and notification rates was highest among males, age groups 25–34, and the older age group of 65 years and above. Only 48% of the of the survey prevalent cases reported cough >2weeks. In addition, only 59% of the identified cases had the four cardinal symptoms for TB (cough ≥2 weeks, fever, night sweat and weight loss. However, 88.2% had an abnormal chest x-ray suggestive of TB. The use of Xpert MTB/RIF identified 77.7% of the cases compared to smear microscopy’s 46%. Twenty-one percent of the survey participants with respiratory symptoms reported to have sought prior health care at private clinics and chemists. Among the survey prevalent cases who reported TB related symptoms, 64.9% had not sought any health care prior to the survey.ConclusionThis survey established that TB prevalence in Kenya is higher than had been estimated, and about half of the those who fall ill with the disease each year are missed.

BackgroundThe prevalence of diseases other than TB detected during chest X-ray (CXR) screening is unknown in sub-Saharan Africa. This represents a missed opportunity for identification and treatment of potentially significant disease. Our aim was to describe and quantify non-TB abnormalities identified by TB-focused CXR screening during the 2016 Kenya National TB Prevalence Survey.MethodsWe reviewed a random sample of 1140 adult (≥15 years) CXRs classified as ‘abnormal, suggestive of TB’ or ‘abnormal other’ during field interpretation from the TB prevalence survey. Each image was read (blinded to field classification and study radiologist read) by two expert radiologists, with images classified into one of four major anatomical categories and primary radiological findings. A third reader resolved discrepancies. Prevalence and 95% CIs of abnormalities diagnosis were estimated.FindingsCardiomegaly was the most common non-TB abnormality at 259 out of 1123 (23.1%, 95% CI 20.6% to 25.6%), while cardiomegaly with features of cardiac failure occurred in 17 out of 1123 (1.5%, 95% CI 0.9% to 2.4%). We also identified chronic pulmonary pathology including suspected COPD in 3.2% (95% CI 2.3% to 4.4%) and non-specific patterns in 4.6% (95% CI 3.5% to 6.0%). Prevalence of active-TB and severe post-TB lung changes was 3.6% (95% CI 2.6% to 4.8%) and 1.4% (95% CI 0.8% to 2.3%), respectively.InterpretationBased on radiological findings, we identified a wide variety of non-TB abnormalities during population-based TB screening. TB prevalence surveys and active case finding activities using mass CXR offer an opportunity to integrate disease screening efforts.FundingNational Institute for Health Research (IMPALA-grant reference 16/136/35).

Objective and methods Worldwide, tuberculosis (TB) is the leading cause of death from a single infectious agent. In many countries, national TB prevalence surveys are the only way to reliably measure the burden of TB disease and can also provide other evidence to inform national efforts to improve TB detection and treatment. Our objective was to synthesise the results and lessons learned from national surveys completed in Africa between 2008 and 2016, to complement a previous review for Asia. Results Twelve surveys completed in Africa were identified: Ethiopia (2010–2011), Gambia (2011–2013), Ghana (2013), Kenya (2015–2016), Malawi (2013–2014), Nigeria (2012), Rwanda (2012), Sudan (2013–2014), Tanzania (2011–2012), Uganda (2014–2015), Zambia (2013–2014) and Zimbabwe (2014). The eligible population in all surveys was people aged ≥15 years who met residency criteria. In total 588 105 individuals participated, equivalent to 82% (range 57–96%) of those eligible. The prevalence of bacteriologically confirmed pulmonary TB disease in those ≥15 years varied from 119 (95% CI 79–160) per 100 000 population in Rwanda and 638 (95% CI 502–774) per 100 000 population in Zambia. The male:female ratio was 2.0 overall, ranging from 1.2 (Ethiopia) to 4.1 (Uganda). Prevalence per 100 000 population generally increased with age, but the absolute number of cases was usually highest among those aged 35–44 years. Of identified TB cases, 44% (95% CI 40–49) did not report TB symptoms during screening and were only identified as eligible for diagnostic testing due to an abnormal chest X‐ray. The overall ratio of prevalence to case notifications was 2.5 (95% CI 1.8–3.2) and was consistently higher for men than women. Many participants who did report TB symptoms had not sought care; those that had were more likely to seek care in a public health facility. HIV prevalence was systematically lower among prevalent cases than officially notified TB patients with an overall ratio of 0.5 (95% CI 0.3–0.7). The two main study limitations were that none of the surveys included people <15 years, and 5 of 12 surveys did not have data on HIV status. Conclusions National TB prevalence surveys implemented in Africa between 2010 and 2016 have contributed substantial new evidence about the burden of TB disease, its distribution by age and sex, and gaps in TB detection and treatment. Policies and practices to improve access to health services and reduce under‐reporting of detected TB cases are needed, especially among men. All surveys provide a valuable baseline for future assessment of trends in TB disease burden.

IntroductionCommunity Health Workers (CHWs) have been utilised for various primary health care activities in different settings especially in developing countries. Usually when utilised in well defined terms, they have a positive impact. To support Kenya's policy on engagement of CHWs for tuberculosis (TB) control, there is need to demonstrate effects of utilising them.ObjectivesThis study assessed TB treatment adherence among patients who utilised CHWs in management of their illness in comparison to those who did not in urban and rural settings.MethodsA retrospective cohort study was conducted in selected health facilities using standard clinical records for each TB patient registered for treatment between 2005 to 2011. Qualitative data was collected from CHWs and health care providers.ResultsThe study assessed 2778 tuberculosis patients and among them 1499 (54%) utilized CHWs for their TB treatment. The urban setting in comparison with the rural setting contributed 70% of patients utilising the CHWs (p<0.001). Overall treatment adherence of the cohort was 79%. Categorizing by use of CHWs, adherence among patients who had utilized CHWs was 83% versus 68% among those that had not (p<0.001). In comparison between the rural and urban settings adherence was 76% and 81.5% (p<0.001) respectively and when categorized by use of CHWs it was 73% and 90% (p<0.001) for the rural and urban set ups respectively. Utilisation of CHWs remained significant in enhancing treatment adherence in the cohort with unadjusted and adjusted ORs; OR 2.25, (95% 1.86–2.73) p<0.001 and OR 1.98 (95% 1.51–2.5) p<0.001 respectively. It was most effective in the urban set-up, OR 2.65 (95% 2.02–3.48, p<0.001) in comparison to the rural set up, OR 0.74 (95% 0.56–0.97) p = 0.032.ConclusionUtilisation of CHWs enhanced TB treatment adherence and the best effects were in the urban set-up.

It is well established that intentions to quit smoking is the strongest predictor of future quit attempts. However, most studies on quit intentions have been conducted in high-income countries with very few in low- and middle-income countries particularly in Africa. This is the first population-based study to compare factors associated with quit intentions among smokers in two African countries. Data were from the International Tobacco Control (ITC) Kenya and Zambia Surveys (2012), face-to-face surveys of nationally representative samples of 2291 adult smokers (Kenya = 1103; Zambia = 1188). Multivariate logistic regression analyses were conducted to identify predictors of quit intentions. Most Kenyan (65.1%) and Zambian (69.1%) smokers had quit intentions of which 54.8% planned to quit within the next 6 months. Five factors were significantly associated with quit intentions in both countries: being younger, having tried to quit previously, perceiving that quitting is beneficial to health, worrying about future health consequences of smoking, and being low in nicotine dependence. The predictive strength of these factors did not differ in the two countries. Four additional factors were significant predictors in Zambia only: having a quit attempt lasting six months or more, lower smoking enjoyment, having a negative opinion about smoking, and concern about cigarette expenses. The factors predicting quit intentions were similar to those in other ITC countries including Canada, US, UK, China and Mauritius. These findings highlight the need for stronger tobacco control policies in Kenya and Zambia including increased taxation, greater access to cessation services, and anti-smoking campaigns denormalizing tobacco use.

Community-based screening for tuberculosis (TB) could improve detection but is resource intensive. We set out to evaluate the accuracy of computer-aided TB screening using digital chest X-ray (CXR) to determine if this approach met target product profiles (TPP) for community-based screening. CXR images from participants in the 2016 Kenya National TB Prevalence Survey were evaluated using CAD4TBv6 (Delft Imaging), giving a probabilistic score for pulmonary TB ranging from 0 (low probability) to 99 (high probability). We constructed a Bayesian latent class model to estimate the accuracy of CAD4TBv6 screening compared to bacteriologically-confirmed TB across CAD4TBv6 threshold cut-offs, incorporating data on Clinical Officer CXR interpretation, participant demographics (age, sex, TB symptoms, previous TB history), and sputum results. We compared model-estimated sensitivity and specificity of CAD4TBv6 to optimum and minimum TPPs. Of 63,050 prevalence survey participants, 61,848 (98%) had analysable CXR images, and 8,966 (14.5%) underwent sputum bacteriological testing; 298 had bacteriologically-confirmed pulmonary TB. Median CAD4TBv6 scores for participants with bacteriologically-confirmed TB were significantly higher (72, IQR: 58–82.75) compared to participants with bacteriologically-negative sputum results (49, IQR: 44–57, p<0.0001). CAD4TBv6 met the optimum TPP; with the threshold set to achieve a mean sensitivity of 95% (optimum TPP), specificity was 83.3%, (95% credible interval [CrI]: 83.0%—83.7%, CAD4TBv6 threshold: 55). There was considerable variation in accuracy by participant characteristics, with older individuals and those with previous TB having lowest specificity. CAD4TBv6 met the optimal TPP for TB community screening. To optimise screening accuracy and efficiency of confirmatory sputum testing, we recommend that an adaptive approach to threshold setting is adopted based on participant characteristics.

Background Despite free diagnosis and treatment for tuberculosis (TB), the costs during treatment impose a significant financial burden on patients and their households. The study sought to identify the determinants for catastrophic costs among patients with drug-sensitive TB (DSTB) and their households in Kenya. Methods The data was collected during the 2017 Kenya national patient cost survey from a nationally representative sample (n = 1071). Treatment related costs and productivity losses were estimated. Total costs exceeding 20% of household income were defined as catastrophic and used as the outcome. Multivariable Poisson regression analysis was performed to measure the association between selected individual, household and disease characteristics and occurrence of catastrophic costs. A deterministic sensitivity analysis was carried using different thresholds and the significant predictors were explored. Results The proportion of catastrophic costs among DSTB patients was 27% (n = 294). Patients with catastrophic costs had higher median productivity losses, 39 h [interquartile range (IQR): 20–104], and total median costs of USD 567 (IQR: 299–1144). The incidence of catastrophic costs had a dose response with household expenditure. The poorest quintile was 6.2 times [95% confidence intervals (CI): 4.0–9.7] more likely to incur catastrophic costs compared to the richest. The prevalence of catastrophic costs decreased with increasing household expenditure quintiles (proportion of catastrophic costs: 59.7%, 32.9%, 23.6%, 15.9%, and 9.5%) from the lowest quintile (Q1) to the highest quintile (Q5). Other determinants included hospitalization: prevalence ratio (PR) = 2.8 (95% CI: 1.8–4.5) and delayed treatment: PR = 1.5 (95% CI: 1.3–1.7). Protective factors included receiving care at a public health facility: PR = 0.8 (95% CI: 0.6–1.0), and a higher body mass index (BMI): PR = 0.97 (95% CI: 0.96–0.98). Pre TB expenditure, hospitalization and BMI were significant predictors in all sensitivity analysis scenarios. Conclusions There are significant inequities in the occurrence of catastrophic costs. Social protection interventions in addition to existing medical and public health interventions are important to implement for patients most at risk of incurring catastrophic costs. Graphic abstract

BackgroundCommunity-based screening for tuberculosis (TB) could improve detection but is resource intensive. We set out to evaluate the accuracy of computer-aided TB screening using digital chest X-ray (CXR) to determine if this approach met target product profiles (TPP) for community-based screening.MethodsCXR images from participants in the 2016 Kenya National TB Prevalence Survey were evaluated using CAD4TBv6 (Delft Imaging), giving a probabilistic score for pulmonary TB ranging from 0 (low probability) to 99 (high probability). We constructed a Bayesian latent class model to estimate the accuracy of CAD4TBv6 screening compared to bacteriologically-confirmed TB across CAD4TBv6 threshold cut-offs, incorporating data on Clinical Officer CXR interpretation, participant demographics (age, sex, TB symptoms, previous TB history), and sputum results. We compared model-estimated sensitivity and specificity of CAD4TBv6 to optimum and minimum TPPs.ResultsOf 63,050 prevalence survey participants, 61,848 (98%) had analysable CXR images, and 8,966 (14.5%) underwent sputum bacteriological testing; 298 had bacteriologically-confirmed pulmonary TB. Median CAD4TBv6 scores for participants with bacteriologically-confirmed TB were significantly higher (72, IQR: 58-82.75) compared to participants with bacteriologically-negative sputum results (49, IQR: 44-57, p<0.0001). CAD4TBv6 met the optimum TPP; with the threshold set to achieve a mean sensitivity of 95% (optimum TPP), specificity was 83.3%, (95% credible interval [CrI]: 83.0%—83.7%, CAD4TBv6 threshold: 55). There was considerable variation in accuracy by participant characteristics, with older individuals and those with previous TB having lowest specificity.ConclusionsCAD4TBv6 met the optimal TPP for TB community screening. To optimise screening accuracy and efficiency of confirmatory sputum testing, we recommend that an adaptive approach to threshold setting is adopted based on participant characteristics.Take home messageCAD4TBv6 met the optimal WHO target product profile for a community TB screening tool. Specificity was lower in adults with previous TB and those aged 41 years or older; an adaptive approach to setting CAD thresholds will likely be required to optimize use.