Use of classification and regression tree (CART), to identify hemoglobin A1C (HbA1C) cut-off thresholds predictive of poor tuberculosis treatment outcomes and associated risk factors.

BACKGROUND: Rifampin-based therapy potentially exacerbates glycemic control among TB patients who are already at high risk of hyperglycemia. This impacts negatively to the optimal care of TB- diabetes mellitus co-affected patients. Classification and regression tree (CART), a machine-learning algorithm impervious to statistical assumptions is one of the ideal tools for clinical decision-making that can be used to identify hemoglobin A1C (HbA1C) cut-off thresholds predictive of poor TB treatment outcomes in such populations. METHODS: 340TB smear positive patients attending two peri-urban clinics were recruited and prospectively followed up for six months. Baseline HbA1C and random blood glucose (RBG) levels were determined. CART was then used to identify cut-off thresholds and rank outcome predictors at end of therapy by determining Risk ratios (RR) and 95% confidence interval (CI) of each predictor threshold. Fractal geometry law explained effect of weight, while U-shaped curve explained effect of HbA1C on these clinical outcomes. RESULTS: Of the 340 patients enrolled: 84%were cured, 7% completed therapy and 9% had unfavorable outcomes out of which 4% (n = 32) had microbiologic failure. Using CART HbA1C identified thresholds were >2.95%, 2.95-4.55% and >4.55%, containing 8/11 (73%), 111/114 (97%) and 189/215 (88%) of patients who experienced favorable outcomes. RR for favorable outcome in patients with weight <53.25 Kg compared to >53.25 Kg was 0.61 (95% CI, 0.45-0.88) among patients with HbA1C >4.55%. Simulation of the CART model with 13 patients data failed therapy revealed that 8/11 (73%) of patients with HbA1C <2.95%, 111/114 (97%) with HbA1C between 2.95% and 4.55% and 189/215 (88%) of patients with HbA1c >4.55% experienced microbiologic failure. CONCLUSION: Using fractal geometry relationships to drug pharmacokinetics, low weight has profound influence on failure of anti-tuberculosis treatment among patients at risk for diabetes mellitus.

Prognostic factors among TB and TB/DM comorbidity among patients on short course regimen within Nairobi and Kiambu counties in Kenya.

BACKGROUND: The double burden of diabetes mellitus (DM) and pulmonary tuberculosis (TB) is one of the global health challenges. Studies done in different parts of the world indicate that 12%-44% of TB disease is associated with DM. In Kenya TB-DM co-morbidity data is scarce and is not readily available. In this study we set to determine the difference in treatment outcomes among TB and TB/DM comorbidity patients and their respective clinical and socio-demographic characteristics. OBJECTIVE: To determine prognostic factors among TB and TB/DM comorbidity among patients on short course regimen within Nairobi and Kiambu counties in Kenya. METHODS: We carried out a prospective cohort study of non-pregnant patients aged 15 years and above that tested positive for TB in two peri­urban counties in Kenya between February 2014 and August 2015. Clinical and socio demographic data were obtained from a questionnaire and medical records of the National TB program patient data base at two, three, five and six months. The data consisted of TB status, HIV status, TB lineage, County, (Glucose, %HbA1c, creatinine) weight, height, BMI, regimen, sex, level of education, employment status, distance from health facility, number of cigarettes smoked, home size, and diet. Univariate analysis was then used to compare each potential risk factor in the TB and TB/DM patients by the Pearson x2 test of proportions or fisher exact test, as appropriate. RESULTS: DM prevalence (HbA1c > 6%) among TB infected patients was 37.2%. Regimen, employment status, alcohol intake, smoking, age and household size were some of the factors associated with DM among TB patients at p-value < 0.05. The number of cigarettes smoked per day and the value of the BUN were significant risk factors of developing DM among TB patients (p values = 0.045). Mean time to conversion from positive to negative was slightly higher for the TB-DM patients compared to the TB patents, though not statistically significant (p = 0.365). CONCLUSION: Patients regimen, employment status, alcohol intake, smoking, age and are associated with DM among TB patients.

Molecular clustering of patients with Mycobacterium tuberculosis strains cultured from the diabetic and non-diabetic newly diagnosed TB positive cases.

BACKGROUND: Social determinants of health, biological, and individual variants have been associated with Pulmonary TB (PTB) case clustering. None of the studies have focused on diabetes mellitus (DM) despite it being one of the co-morbidity affecting TB patients. Minimal data is available and it is not clear whether patients with DM and TB are more likely than TB patients without DM to be grouped into similar molecular clusters thus indicating a bias in transmission among TB/DM co-morbidity patients. OBJECTIVE: To determine proportion of TB strains within TB and TB/DM cases that were clustered with their corresponding clinical outcomes and hence could be attributable to active TB transmission in the two urban counties of Nairobi, Kenya. METHODS: We carried out a prospective cohort study of non-pregnant patients aged 15 years and above that tested positive for TB in two peri­urban counties in Kenya between February 2014 and August 2015. Clinical and socio-demographic data were obtained from a questionnaire and medical records of the National TB program patient data base at two, three, five and six months. Spoligotyping data was then obtained and compared from previously identified strains in a data bank from the spolDB4. RESULTS: We identified 7 different TB strains out of which East Asia Beijing, Euro America and Indo oceanic being the most dominant strain within the two counties accounting for 92.4% of the infections. DM was not a significant factor in increasing the likelihood of PTB patients to cluster according to the genotype of the infecting Mycobacterium tuberculosis bacillus. TB lineages, DM and County of the patient were found to be independent of the clinical outcomes that were observed in the study. CONCLUSION: Diabetes mellitus is not a significant factor in increasing the molecular clustering among PTB patients.

Kenya tuberculosis prevalence survey 2016: Challenges and opportunities of ending TB in Kenya.

BACKGROUND: We aimed to determine the prevalence of pulmonary TB amongst the adult population (≥15 years) in 2016 in Kenya. METHOD: A 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. RESULT: The 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. CONCLUSION: This 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.