DETERMINATION OF PREVALENCE OF DEFAULTING AND ASSOCIATED FACTORS FROM TREATMENT AMONG TUBERCULOSIS PATIENTS IN KILIFI NORTH SUB COUNTY

Abstract

Tuberculosis is a major public health problem and remains one of the deadliest communicable diseases. Understanding the characteristics of patients who default from treatment for tuberculosis will help modify the strategies to reduce such default to the barest minimum and achieve higher levels of adherence. The key objective of the present study was to identify determinants and prevalence of TB defaulting and associated factors from TB treatment in Kilifi North Sub County. The study was a mixed method employing both qualitative and quantitative data. It involved a retrospective analysis of routine data, patients’ survey, group discussions and microbiological culture analysis of sputum. Logistic regression was used to calculate completion rate. The significance level was set at p = 0.05.93.6% of the participant consented to participate while 6.4% did not consent to participate in the study. In the univariate analysis, some of the key factors underlying defaulting were; distance to the health centre and belief that herbal drugs could be used to treat TB. Multivariate regression analysis revealed that travel distance to health facility (p = 0.00), first day upon diagnosis with tuberculosis (p = 0.03), owning a radio at home (p=0.02) and offering family support while on TB treatment (p = 0.01) were significantly associated with defaulting among tuberculosis patients. Of the fifty-five (55) samples of patients’ sputum, 63% of them had normal nonpathogenic bacterial flora. GeneXpert tests on 55 sputum cultures showed that 20 patients had; Haemophilus influenza (3), Streptococcus pneumoniae (6), Escherichia coli (4), staphylococcus aureus (1), Klebsiella pneumonia (2), pseudomonas (3), Proteus mirabilis (1) as the most common associated microbiomes. The findings suggest that a large proportion of the consented participants had concomitant bacteria colonizing their lower airways in addition to T. mycobacterium. It was observed that normal bacterial flora and some staphylococci had the highest microbial load in TB positive sputum xiv isolates. Staphylococcus aureus and Klebsiella pneumonia had the lowest. This study found that prevalence of TB defaulting was 34% and approximately9.09% patients were positive using Ziehl-Neelesn staining. One sample had streptococcus pneumonia and two samples (3.6%) were found to be resistant to rifampicin. The results suggest the need for establishing continued systems for diagnosis of TB coupled with the clinical manifestation and community awareness to help minimize default from treatment. It would also be interesting to investigate in retrospective the history of other bacterial infections resulting from indications of the observed normal bacterial flora in the patients’ samples.