Article 12 : Who Fails to Complete Tuberculosis Treatment in Haiti ’ s National Prison , and What Innovations Might Improve Success ?
Table 2 : Tuberculosis Treatment Outcomes of Men Incarcerated in Haiti ’ s National Penitentiary from January 2016 to December 2018
Treatment Outcome No . %
Completed Treatment |
556 |
84.1 |
Treated |
405 |
61.3 |
Treatment Terminated * |
151 |
22.8 |
Did Not Complete Treatment |
105 |
15.9 |
Released |
52 |
7.9 |
Died |
32 |
4.8 |
Transferred |
2 |
0.3 |
Treatment Failure |
3 |
0.5 |
Unknown |
16 |
2.4 |
* In accordance with WHO guidelines , presumptive TB cases without a |
positive smear prior to starting treatment are classified as treatment |
terminated after 6 months of treatment |
Univariate logistic regression Failing to complete treatment was significantly associated with TB type and HIV status [ Table 3 ]. EPTB cases were 1.76 times less likely to complete treatment than pulmonary TB cases ( P = 0.0424 ). Individuals living with HIV were 3.19 times less likely to complete treatment than those who were HIVnegative ( P < 0.0001 ). The odds of failing to complete treatment increased with increased age at start of treatment , however , this association was not significant . Retreatment cases were slightly more likely to complete treatment than new cases , but this association was also not significant .
Multivariate logistic regression EPTB patients and patients living with HIV were significantly less likely to complete treatment in prison [ Table 4 ]. EPTB cases were almost twice as likely to not complete treatment as pulmonary cases ( aOR 1.96 , 95 % CI 1.08-3.52 ).
Patients living with HIV were about three times as likely to not complete treatment as HIV-negative patients ( aOR 3.31 , 95 % CI 1.82-6.02 ). The confounding assessment presented evidence that age may be a confounder of the relationship between TB type , HIV status , and not completing treatment in prison , so estimates were adjusted for patient ’ s age at the start of treatment . The adjusted estimates provided more precise estimates ( narrower confidence intervals ).
McNemar chi-squared test An additional analysis explored mortality rates by TB type , TB case category , and HIV status using a McNemar chi-squared test . The test found that mortality rates were significantly lower for pulmonary versus EPTB cases , new versus retreatment cases , and persons without HIV infection versus those living with HIV .
Cox regression analysis Cox regression analysis was conducted to investigate who may take longer to complete treatment in prison . Only TB case category ( new cases versus retreatment cases ) was significantly associated with
181