Challenges of tuberculosis control in South west Ethiopia: treatment delays, cost, and outcomes Abyot Asres Shetano
No Thumbnail Available
Date
2018-04
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Addis Ababa Universty
Abstract
Background: Tuberculosis (TB) is among the major public health problems over the world.
Thus, global efforts have been designed to combat three distinct, but overlapping humanitarian,
public health, and economic burdens posed by the TB illness. Timely case detection and
treatment of cases have been a focus and priority in the prevention and control of TB. However,
long delays to initiate anti-TB treatment have been reported for which evidences on predictors of
the delay and impact of the delayed treatment on outcomes are limited. Elimination of
catastrophic costs posed by TB illness to patients and households has taken attention in the latest
end TB strategy. On the other hand, evidences on magnitude and drivers of patient cost across
continuum of TB care are scarce in Ethiopian setting. Treatment regimens play a vital role in
reducing time delays to treatment and cost spent across continuum of TB care, and improving
outcomes. Nonetheless, evidences on the interdependence among delays, patient cost of care and
outcomes in the era of shortened treatment regimen is scanty in Ethiopian setting.
Objectives: The aims of the dissertation were 1) to compare outcomes of six and eight-month
TB treatment regimen, 2) to determine time delays to initiate anti-TB treatment and its
predictors, 3) to assess pre-and post-diagnosis patient costs for TB care and 4) to examine
association between delayed anti-TB treatment initiation and treatment outcome.
Methods: A blend of cross-sectional and longitudinal studies were conducted among 735 TB
cases on treatment and 790 patient records from 14 public health facilities of Bench Maji, Kaffa
and Sheka zones in Southwest Ethiopia. The cases were selected using multistage cluster
sampling technique. Both primary and secondary data were gathered and/or extracted using
structured questionnaire from January 2015 through June 2016. For comparison of outcomes
across the six and eight month regimen, patient clinical profiles and outcomes were extracted
from unit TB register of cases registered during 2008 through 2014. Data for the cases on
treatment were collected at two points; 1) at enrollment when patients‟ sociodemographics, careseeking
practices, direct and indirect costs of TB care seeking were collected and 2) at the end of
treatment when treatment practices, patient cost of TB treatment and outcomes were inquired.
The data were entered in to Epi-Data and processed on SPSS version 21 and STATA version 13.
Since the cost data were right skewed, analysis was made on natural logarithm and reported in
corresponding antilog. Bivariate and multiple logistic, linear, and log-binomial regression
models were fitted to identify predictors of delays, cost, and outcome. In all the statistical tests,
necessary assumptions were checked and significance judged at p<0.05.
x
Results: The overall treatment success among cases registered during 2008 through 2014 was 88
% ( 85.3% vs 90.6%, p=0.02 among those treated for eight months with 2ERHZ/6HE and six
months with 2ERHZ/4RH regimens, respectively). Thus, 4RH continuation phase treatment
adjusted Odds Ratio [aOR=0.55,95% CI;(0.34,0.89)], weight gain at the end of second month
treatment [aOR=0.28, 95% CI; (0.11, 0.72)] predicted lower odds of unsuccessful outcome. On
the other hand, age [aOR=1.02,95%CI; (1.001,1.022)], rural residence
[aOR=2.1,95%CI;(1.18,3.75)] and HIV co-infection [aOR=2.39,95%CI;(1.12,5.07)]
independently predicted higher odds of unsuccessful outcome.
TB patients had spent a median [inter-quartile range (IQR)] of 55(32-100) days to initiate anti-
TB treatment since onset of illness (total delay). Similarly a median (IQR) of 25(15-36) and
22(9-48) days had been elapsed respectively to initiate care seeking (patient delay) and anti-TB
treatment since first consultation (provider delay). Thus 54.6% of the total delay was attributed
to provider (health system) and the rest to the patient. Prior self-treatment (aOR: 1.72, 95%
confidence interval [CI]:1.07-2.75), HIV co-infection (aOR: 1.80, 95% CI: 1.05-3.10) and extra
pulmonary TB (aOR: 1.54, 95% CI: 1.03-2.29) independently predicted higher odds of patient
delay. On the other hand, initial visits to health posts or private clinics (aOR: 1.42, 95% CI: 1.01,
2.0) and delayed to seek care (aOR: 1.81, 95% CI: 1.33-2.50) significantly predicted higher odds
of provider delay.
Since onset of illness, TB patients totally incurred mean [(standard deviation (+SD)] of
US$244.71(+0.1) for care seeking and treatment. Thus mean (+SD) US$108.0(+0.1) and
US$117.0(+0.1) were respectively incurred during pre-diagnosis and post-diagnosis periods.
Mean (+SD) out of pocket patient expenditures during pre-and post-diagnosis were
US$21.46(0.16) and US$43.80(+0.1) respectively. Total indirect and pre-diagnosis costs
constitute 70.6% and 53.6% of the total cost respectively. Patient delay (adjusted coefficient
(βadj)= 0.004, p<0.001), provider delay (βadj =0.004,p<0.001), number of visited healthcare
facilities (βadj =0.17,p<0.001) and diagnosis at private facilities (βadj=0.16,p=0.02) independently
predicted increased pre-diagnosis cost. Similarly, rural residence (βadj=0.27,p<0.001),
hospitalization (βadj=0.91,p<0.001), patient delay (βadj=0.002,p<0.001) and provider delay (βadj
=0.002,p<0.001) predicted increased post-diagnosis costs.
The overall treatment success among the prospectively enrolled cases was 89.7% (86.7% vs.
92.6%, p=0.01) respectively among those initiated anti-TB treatment beyond and within 55days
xi
of onset of illness). Accordingly, treatment initiation beyond 55days of onset [Adjusted Relative
Risk (aRR)=1.92, 95%CI:1.30, 2.81),treatment center being hospital (aRR=3.73, 95%CI:2.23,
6.25), and HIV co-infection (aRR=2.18, 95%CI: 1.47, 3.25) independently predicted higher risk
of unsuccessful treatment outcome. In contrast, weight gain at the end of second month treatment
(aRR=0.40, 95%CI: 0.19, 0.83) predicted lower risk of unsuccessful outcome.
Conclusions:
The switch of continuation phase TB treatment regimen from 6EH to 4RH has brought
significantly higher treatment success that verified applicability of the regimen change in
resource constrained and high burden countries. TB cases in the study area elapsed too long time
to initiate care seeking and treatment. The delays are attributed to the patient, disease and health
system related factors. Throughout the care seeking and treatment pathways, TB cases incurred
substantial direct and indirect cost for TB care despite the “free TB service”. The delay to initiate
anti-TB treatment was significantly associated with increased patient costs and risk of
unsuccessful outcome. Patient and health system attributes predicted both costs incurred across
continuum of TB care and treatment outcome.
Recommendations
Promotion of early care seeking for TB through community level awareness creation; involving
both formal and informal providers can minimize patient delays. Moreover, improving
diagnostic and case-holding efficiencies of both private and public healthcare facilities can
reduce delays to treatment and risk of unfavorable outcomes. On the other hand, adoption of
patient centered TB care, reimbursement mechanisms of costs and scale up of the national
community and social insurance schemes to the study area can reduce the financial burden on
patients. Finally, further studies are required to explore reasons for patient and provider delays
using qualitative designs, costs of TB care, and its impacts on household and health system.
Description
Keywords
TB, Patient delay, provider delay, direct cost, indirect cost, pre-diagnosis cost, postdiagnosis cost, treatment outcome, longitudinal, Ethiopia