Risk Factors for Tuberculosis Going Beyond HIV:A Case Control Study in Kisii County, Kenya

*George Kasera1, 2, S. Karanja3, C. Mwachari4, E. Masini1, J. Kioko1

1Ministry of Health, Kenya 2Field Epidemiology and Laboratory Training Program, Kenya , 3Jomo Kenyatta University of Agriculture & Technology, Kenya4Kenya Medical Research Institute, Kenya

Corresponding Author: George Kasera| kadondikasera@gmail.com | +254 (0) 722 994573

• Kenya is not on track to attain one or more of the three targets for reducing tuberculosis (TB) incidence, prevalence and mortality.

• While HIV remains an important driver of TB in Kenya, 67% of cases are not HIV related suggesting the need to identify and address other drivers of the epidemic.

• Growing evidence describing the links between TB and a number of NCDs and their risk factors, such as diabetes mellitus, smoking- and alcohol-related conditions, chronic obstructive pulmonary disease (COPD), mental illness and malnutrition.

• The prevalence of diabetes mellitus in Kenya is 1.9%, while 11.6% and 19.3% adults smoke cigarettes and consume alcohol respectively.

Background

1. Kenya Stepwise Survey for Non Communicable Diseases Risk Factor Report. Nairobi: Ministry of Health, 2015

2. Guidelines for Management of Tuberculosis and Leprosy in Kenya. Nairobi: National Tuberculosis, Leprosy and Lung Disease Program, 2013

References

The study was conducted across �ve health facilities (Kisii Teaching and Referral Hospital, Gucha, Iyabe, Keumbu Sub-County Hospitals and Oresi Health Centre) in Kisii County, South Western Kenya.

Kisii County

Study Site

• To determine the association of TB with cigarette smoking, alcohol consumption, diabetes mellitus and malnutrition in Kisii County, Kenya.

Objective

• This study showed that diabetes is positively associated with a higher chance of getting TB and that malnutrition was associated with the highest likelihood of having TB among all the exposures assessed. While smoking and alcohol showed increased odds for TB that was not signi�cant, their role cannot be ignored since we had a small sample size that was further thinned out by strati�cation.

• A total of 268 study participants enrolled

Conclusions

Results

Study Design• Unmatched case control study

Eligibility Criteria• Cases: Current TB patients registered for treatment at the

aforementioned health facilities• Controls: Persons from the catchment areas of the facilities

and did not have TB.

Variables’ De�nition and Assessment• Outcome variable: TB• Exposure variables: Alcohol consumption, cigarette

smoking, diabetes mellitus (RBS > 200 mg/ dL or FBS > 126 mg/ dL) and malnutrition (BMI<18.5 kg/ m2).

• Co-variates: Age, sex, household size and education level.

Data Sources/ Measurement• Pretested structured questionnaire: Demographic, alcohol

consumption and cigarette smoking data.• Physical examination: Height (cm) & weight (kg)• Cardiochek PA ® machines (PTS Diagnostics Inc., USA):

Blood glucose and cholesterol levels• Data collected from 09/16/2015 to 01/25/2016

Sample Size• 268 (67 cases & 201 controls), Kelsey formula

Data analysis• STATA software version 12 (StataCorp LP. Texas, USA) used

Materials and Methods

• There is need to recon�gure our health systems to focus on tackling NDCs and the challenges they pose to public health. Such a realignment of services could be guided by four key principles of integration of services, innovation in service delivery, inclusion of communities and information and communication for better care

Recommendations

Figure 1: Participant Recuitment, Kisii, 2016

Table 1: Participant Characteristics, Kisii, 2016

Table 2: Odds Ratios (95% CI) of having TB in relation to selected exposures, Kisii County

National Tuberculosis, Leprosy and Lung Disease Program1st Floor, Afya Annex, Kenyatta National Hospital Grounds

P. O. Box 20781- 00200, Nairobi, KenyaTel: +254 (0) 773 977 440

Email: info@nltp.co.kewww.nltp.co.ke

Ethical Considerations• Both ethical approval for the study was obtained and local

authorization was granted.• Informed consent was obtained from all successful

participants.• Prior plans to link those who tested HIV positive for HIV

care, screened TB positive for TB diagnostic services and those who were diabetic with diabetes services.

• No incentives were given to induce participation.