ew
Insights of Bioinformatcs
Open Access | Page 9 |
Vol 2 | Issue 1 | Pages 9-20
Copyright: © 2022 Mutabari D, et al. This is an open-access artcle distributed under the terms
of the Creatve Commons Atributon License, which permits unrestricted use, distributon, and
reproducton in any medium, provided the original author and source are credited.
SchOlArly PAgeS SCHOLARS. DIRECT
DOI: 10.36959/317/681
Survival Analysis of Sputum Conversion Time among
Patients with Drug Resistant Tuberculosis in Kenya
David Mutabari
1*
, Idah Orowe
2
, Anthony Karanja
3
and Hillary Kipruto
4
1
Medical Statstcs, University of Nairobi, Kenya
2
Senior Lecturer, School of Mathematcs, University of Nairobi, Kenya
3
Senior Lecturer, Masai Mara University, Kenya
4
Senior Statstcian, World Health Organizaton, Kenya
Research Article
Abstract
Background: Drug-resistant tuberculosis (DR-TB) is a form of antmicrobial resistance Mycobacterium tuberculosis that is
difcult and costly to treat. It is caused by TB bacteria resistance to at least one of the frst-line existng TB medicatons,
resultng in fewer treatment optons and increasing mortality rates. In 2019, close to 500,000 people developed DR-TB,
and 182,000 died globally. Less than 60% of DR-TB patents who started on medicaton were completed, mostly due
to high mortality and lost follow-up. This study aimed to determine when people who have been infected with drug-
resistant tuberculosis and started on treatment convert from being sputum positve to sputum negatve, which is a
critcal point to curtail the infectvity of the disease to the general populaton.
Methods: The analysis was through a retrospectve review of patents captured through electronic medical records (TIBU)
of Kenya Natonal Tuberculosis and Leprosy Program from 2014 to 2019, having a two-year follow-up for every subject
enrolled. The total number of patents included in the study was 2674, entailing all patents who tested positve for
tuberculosis drug resistance. We had age, gender, country, sector, registraton, group, the resistance patern, intensive
phase regimen, modifcaton of intensive phase, year, BMI, and a quarter as our variables for analysis. We considered
Gender, County, Resistance Patern, Age category, and BMI at enrollment for the conversion tme. The Log Rank test
was used to compare the survival distributon of the included variables. The Cox Proportonal Hazard model was used
to analyze Year, Quarter, Sector Model of care, Intensive regimen, Modifcaton of Intensive regimen, and County. The
proportonal hazard assumptons and the overall model ft were assessed. Parametric survival models were included in
three levels of factors associated with patents, factors associated with resistance, and insttutonal factors. We evaluated
the conversion variatons where we used the exponental and Weibull distributons and best-ftng model through the
Akaike’s informaton criterion.
Results: Of the patents enrolled for treatment, 50% took eight months to have their sputum conversions, whereas only
25% took four months to record conversions. Gender had no infuence on conversion tme, where 50% of both genders
had a median tme of eight months to get converted. The resistance patern registered a median tme of eight months,
with Rifampicin resistance taking the least tme of seven months. Mono-Resistance and Pre-XDR not having converted
by standard, intensive observaton period - (9-Months). Most patents were between age 20 and 49 years in the age
category, with a median conversion tme of 9 months. BMI, the underweight took nine months to convert, normal was
eight months, overweight was 7 Months, and Obese was 6 Months.
Conclusion: In general, the median tme for the most patents to convert from being sputum positve to sputum negatve
was eight months, with the earliest converters being recorded as being four months but was being infuenced by diferent
factors like the type of resistance and drugs that were given as the frst line ant-TB with also a modifcaton of his frst-line
medicaton. More adherences should be emphasized on the young and middle-aged populaton who are taking more
tme to have their sputum convert hence posing a more signifcant risk to the enormous populaton by spreading the
resistant disease to them.
Keywords
Mycobacterium tuberculosis, Pre XDR, TIBU, DR TB, Log rank test, Kaplan meier estmator, Cox proportonal hazard
model, Residual analysis
Check for
updates