MODELLING THE USE OF GENEXPERT AND SMEAR MICROSCOPY ON TIME TO DIAGNOSIS, TREATMENT INITIATION AND TREATMENT OUTCOMES AMONG TB PATIENTS AT UNIVERSITY TEACHING HOSPITALS-ADULT HOSPITAL, CHEST CLINIC, LUSAKA ZAMBIA
Title: MODELLING THE USE OF GENEXPERT AND SMEAR MICROSCOPY ON TIME TO DIAGNOSIS, TREATMENT INITIATION AND TREATMENT OUTCOMES AMONG TB PATIENTS AT UNIVERSITY TEACHING HOSPITALS-ADULT HOSPITAL, CHEST CLINIC, LUSAKA ZAMBIA
John Mathias Zulu, MSc1; Sheila Masaku, MSc.1; Moses Mukosha, MSc.1 Chipo Nkwemu, MSc.1 Dr. Gershom Chongwe PhD1,2 Dr. Patrick Kaonga, PhD1; Prof. Patrick Musonda PhD1
1University of Zambia-School of Public Health, 2Tropical Diseases Research Centre

Abstract
Background: Timely diagnosis and treatment initiation are important determinants of TB prevention and treatment outcomes. The study explored the diagnostic performance of the GeneXpert and Smear Microscopy on time to diagnosis, treatment initiation, and improvement of treatment outcomes.
Methods: A retrospective cohort study which followed up TB patients who used the GeneXpert and smear microscopy at UTH-chest clinic from 2015 january-2019 December. Median time was analyzed at two levels, time to diagnosis and time to treatment initiation. Multivariable Cox proportion models were used to analyze predictors of treatment initiation. While lognormal regression model was used to determine predictors of unfavorable outcomes. A p-value of 0.05 was deemed statistically significant.
Results: A total of 417 TB patients were enrolled in the study, 243 (58.3%) used GeneXpert, and 174(42.7%) used the smear microscopy. Of the total patients, 299(70.5%) were diagnosed in <=5days and 125(30%) in >5 days. Of the total participants, 229(54.0%) were treated in <=14 days and 195(46.0%) > 14days. Further, 156(45.0%) had successful and 191(55.0%) unsuccessful outcomes. Compared to smear, in Xpert arm median days of diagnosis were 1 day [IQR] 1–41) versus 4 days [IQR] 1–15), p<0.0001; Compared to Smear Microscopy, in Xpert arm median days for treatment initiation were 7 days [IQR] 1–31 versus 18 days [IQR] 5–72), p<0.0001. Predictors of delayed treatment initiation were, residence outside district but within catchment area vs within 12 km adjusted [HR] 2.133, 95% CI: 1.375,3.310, P=0.001, GeneXpert vs Smear microscopy: adjusted [HR] 0.777, 95% CI: 0.514–1.052, p<0.135). MDR vs Susceptible: adjusted ([HR] 1.608, 95% CI: 1.094–2.362, p=0.016). diagnosed <=5 days’ vs >5 days: adjusted: [HR] 0.762, 95%CI: 0.561,1.034, p=0.031. Predictors of unfavorable outcomes were method of diagnosis GeneXpert vs Smear microscopy: adjusted [TR] 0.216, 95% CI: 0.139–1.321, p<0.0001). outside district vs within 12km, adjusted [TR] 0.391, 95% CI: 0.149–1.028, p=0.050).
Conclusion: Patients who used the microscopy compared to those who used Xpert and who came from outside district compared to those who came from within 12km radius were more likely to meet the unfavorable TB treatment outcomes more quickly. The study recommends more distribution and use of the GeneXpert in all diagnostics centers and areas of rural nature. The national TB programme and its partners should strengthen efforts to monitor the implementation of the use of the GeneXpert in all its TB diagnostics centers. National wide studies are needed to identify and establish implementation strategies and challenges of GeneXpert.

John Mathias Zulu MSc

University of Zambia

Appearances