Diagnostic Delay of Tuberculosis in Isiolo Referral Hospital in Kenya: A comparison of associated factors between binary and survival analyses

Abstract

Background: Tuberculosis (TB) is one of the leading causes of mortality worldwide. Early identification of factors associated with delayed TB diagnosis forms an effective strategy for minimising the transmission of the disease in the community. This study aimed to investigate factors associated with diagnostic delay of tuberculosis.
Materials and Methods: The study employed a cross-section cohort design from 154 TB patients between January 2018 and January 2019 at Isiolo County referral hospital, Northern Kenya. Cox regression analysis was performed to determine the association between predictive variables and survival outcome at Hazard Ratios (95% CI). Further analysis was conducted to determine patients’  epidemiological characteristics, health-seeking behaviour and clinical factors associated with diagnostic delay. We used Kaplan–Meier plots and logrank tests to evaluate the survival pattern, and used median to describe delays of TB diagnosis.
Results: After adjustment for covariates, HR (95% CI) aged 16-30 was (HR = 0.24, CI: 0.09, 0.63; P = 0.00), >45 years old (HR = 1.68, CI: 1.06, 2.67; P = 0.03), male gender (HR = 1.64, CI: 0.99, 2.72; P = 0.05), diagnosis at dispensary (HR = 2.00, CI: 1.12, 3.58; P = 0.02), productive cough (HR = 5.67, CI: 1.54, 20.85; P = 0.01), and chest pain (HR = 0.39, CI: 0.19, 0.81 P = 0.01).
Conclusion: Our binary results demonstrated that TB patients’ health-seeking behaviour, health facility of diagnosis, occupation, and education levels were associated with delayed TB diagnosis. Meanwhile, survival analysis indicated age, gender, health facility of diagnosis, productive cough, and chest pain to be associated with delayed TB diagnosis. This affirms the need to strengthen health
education in the community.

Keywords: Factors, Survival Analysis, Binary, Comparison, Isiolo, Northern Kenya