Evaluation of the tuberculosis control programme in prisons and the post-release continuation of tuberculosis treatment in Malaysia

Abstract

Background: Poorly managed tuberculosis (TB) control programmes in prisons have detrimental health consequences, placing prisoners at an increased risk for TB morbidity and mortality. This situation could further fuel the TB epidemic in the general population when prisoners are released with uncompleted treatment in prisons and default from treatment in the community. Despite the recognised risk, limited information exists about the burden of TB, the performance of TB control programmes, and the continuity of TB care after release from prisons in Malaysia, a country with an increasing TB burden over the past two decades.

Objectives: This PhD project was designed to investigate the prevalence and correlates of active TB among new prison entrants, to assess gaps in the performance of the prison’s TB programme using standardised parameters, to investigate the proportion of released prisoners who continue treatment in the community, and to evaluate factors influencing the continuation of TB treatment after release from prisons in Malaysia.

Methods: In the first study, we screened prisoners entering the largest prison in Malaysia to determine those who needed further TB assessment. All HIV-infected and symptomatic non-HIV infected prisoners were asked to submit sputum specimens to be examined using GeneXpert MTB/RIF (Xpert) or culture. Factors associated with TB disease, define as Xpert- or culture-positive tests, were assessed using regression analyses. In the second study, we developed parameters and assessable indicators to evaluate gaps in the performance of the TB control programme in the same prison. The parameters include policies and human resources; screening, case detection and notification; treatment initiation, follow-up, and outcome; TB care for HIV-infected prisoners; and knowledge about TB. Data gathering tools and data sources (local and international TB guidelines and TB system assessment publications) were utilised to measure the performance indicators under these parameters and determine system performance gaps. In the third study, prisoners who were due to be released from five prisons were recruited and followed up to identify the proportion of former prisoners who continued treatment in the community. Factors associated with failure to register at a TB clinic within 30 days of release were assessed in regression analyses. In the fourth study, factors influencing the continuation of TB treatment in the community were evaluated in a group of prisoners with previous TB episodes using in-depth interviews. We utilised a thematic framework analysis to identify relevant themes.

Results: In the first study, 10,335 participants were recruited. Among HIV-infected prisoners (N=214), 12.6% had TB disease compared with 0.29% of non-HIV-infected prisoners. Among non-HIV infected prisoners, prevalent TB disease was independently associated with older age, current drug use, a previous TB episode, and being underweight. In the second study, we found that the national TB guidelines did not include a section on TB in prisons and that there was an average of 2.19 healthcare workers for every 1,000 prisoners. Furthermore, only 54.2% of new entrants were screened for TB, there was a 37.6% case detection ratio, and only 45.5% of TB cases were notified to the national TB programme. While treatment initiation was high (91%), only half (50.7%) were followed up after two months inside the prison, the treatment success rate was 72.8%, the mortality ratio was 125 per 100,000 prisoners, and only 73.3% were offered TB documentation before release. TB care for HIV-infected prisoners was similarly suboptimal with 22.1% screened for TB disease at entry, only 1.6% were provided with preventive therapy, and 12.9% were prescribed HIV treatment while on TB treatment. Knowledge about TB was very limited, particularly among prisoners compared to prison officers (6.8% and 67.2% correctly answered TB questions, respectively). In the third study, 106 participants recruited. Of these, 47 (44.3%) did not register at a TB clinic to continue treatment after release, and this was independently associated with younger age, pre-incarceration unstable housing and employment, failure to provide contact details, a previous TB episode, and not being supplied with TB documentation at the time of the release from prisons. In the fourth study, we recruited five prisoners who continued, and seven who discontinued treatment of their TB after release from prisons. Key themes related to the continuation of TB treatment after release were the prison environment and attention to prisoner care, prisoner perception and attitude, the presence of a supportive environment during the transition to the community, social support, and welcoming community healthcare services.

Conclusion There is a high prevalence of TB disease among new entrants to prison in Malaysia, likely representing cases missed by the community health services. There are several gaps in the performance of the TB control programme in prisons; a situation that may promote TB transmission in prisons and the larger general population. Almost half of the prisoners who are released while still on TB treatment abandons treatment after release and that several factors influence whether they continue treatment in the community. These findings warrant the establishment of an effective TB control programme in prisons supported with policy changes, proper funding, trained healthcare workers and adequate communication between the prisons and the public health department.