Factors Associated with Delayed Tuberculosis Diagnosis and Predicted Population Cases generated during the Delays in Isiolo County, Kenya

Abstract

Twenty-three (23%) percent of patients who visit a healthcare provider with respiratory symptoms fail to get diagnosed at the first point of contact and could accelerate the transmission of Mycobacterium tuberculosis to 10-15 people by a single infected individual annually. The study assessed various factors of delayed Tuberculosis (TB) diagnosis among patients attending the Isiolo County Referral Hospital and predicted the cases generated during the delays in Isiolo County in Northern Kenya. The study employed a cross-sectional mixed methods study design and systematically sampled 172 tuberculosis patients. We subsequently collated their data from January 2018 to January 2019 and abstracted epidemiological and clinical characteristics from their records to serve as our independent variables. The outcome variable was delayed diagnosis dichotomized into <21 days or >21 days and treated as a binary outcome. Pre-tested interviewer-administered questionnaires, focused group discussions, and key informant interview guides were used to collect data. All data were analysed using SPSS for Windows version 20 and R . The number of TB cases was predicted by employing a simple mathematical model. The SIRV-type dynamics model was parameterized using the data collected at the Isiolo County Referral Hospital between January 2018 and January 2019, as well as from the existing published literature. Most (n=89, 57.8%) of the TB diagnoses fell in the category of >21 days delay, which constituted a median of 27.6 and a mean of 37.3, with a standard deviation of 57 days (range 0 to 414 days). Delayed diagnosis associated factors included (i) use of dispensary and private health facilities (OR=4.3, 95% CI: 1.44, 13.14; P= 0.009) and (OR= 4.9, 95% CI: 1.64, 14.73; P= 0.004), (ii) self-employed individuals (OR=21.7, 95% CI: 2.47,190.93; P = 0.006) and employed individuals (OR=9.9, 95% CI: 1.14, 85.80; P= 0.038), (iii) secondary level of education (OR= 0.03, 95% CI: 0.01, 0.21; P=0.000), and tertiary education (OR= 0.033, 95% CI: 0.01, 0.23; P=0.001). The predicted number of TB cases generated by the delayed diagnosis of existing TB cases over a period of 10 years was approximately 2316, with numbers of infectious cases oscillating between two and 23 on any given day. Seventy-five (75%) of the rates of recovery were <0.1 corresponding to a delayed diagnosis of <10 days. Only 9% of these rates were ≥0.2 corresponding to a delayed diagnosis of ≥20 days. The predicted effect of increasing vaccination coverage from the current KDHS 2019 BCG coverage of 94.2% up to 99.2% (by 1% unit), under the same delay periods, reduces the number of TB cases among the susceptible population. The findings in the current study showed a significant delay of more than 21 days with median days of 27.6 in the diagnosis of TB. The health facility of diagnosis, occupation, and education levels, were associated with delayed diagnosis of TB in the current study area. The modelling framework has projected the number of TB cases generated by delayed diagnosis for 10 years at 2316, which is a probable determinant of the disease endemicity. Therefore, there is a need to increase health education and promotion in the community, to deliberately strengthen healthcare workers’ knowledge of symptoms and signs of TB in high-burden settings, and to implement dedicated TB-specific public-private health facility linkages. Prospective studies are needed in order to disentangle the factors and their interactions linked to delay in seeking healthcare (patient delay) or delay in receiving a confirmed diagnosis (health system delays) in settings with a high TB burden.