We found that the HIV prevalence among those first diagnosed with TB in Guangxi, China in 2011 was 5.6% with a cross-registration-system check and imputation method. This rate was higher than in previous studies based on surveys in hospitalization and routine TB surveillance [9,12], There were 24,849 (58.8%) TB cases that had a documented HIV test in 2011 and the proportion of TB patients testing HIV-positive was 9.5% (2,351/24,849), the percentage of those co-infected with HIV and TB in this study is consistent with testing results globally in 2017, but the proportion of TB patients testing HIV-positive in this study was less than the WHO estimate of 15% positive in 2017 [1].
Over 850 patients in this study died over the 5 year follow-up period. The survival of HIV/MTB patients might govern by the degree and timing of appropriate ART and anti-tubercular therapy. As previous studies showed that ART improves the survival rate by at least 50% in individuals with HIV/MTB [13,14], in 2012 the WHO endorsed a strategy recommending that ART should initiate as early as possible with HIV-positive TB patients. Insufficient anti-tubercular therapy in HIV/MTB caused the inability to control mycobacterial infection and led to high mortality [15]. Our study found that inadequate TB therapy (TBFMA) has led to a 2.97 (95% CI: 2.45–3.61) times higher mortality than those who cured and a 2.91 (95% CI: 2.46–3.43) times higher than those who completed their TB regimen during the 60-month follow-up, HIV and TB coinfection individuals with insufficient TB treatment were likely to die in an earlier stage of followed-up. Even after removing the deaths during the 6-month duration of anti-tubercular therapy, a significant difference in mortality remained. After adjusting with ART, the OR was still 2.84 (95% CI: 1.99–4.05) times higher among those who TB treatment failed, missing, or experienced adverse events than in those who cured or completed TB treatment. This result demonstrates that inadequate anti-tubercular therapy plays an important role in death for those with HIV/MTB coinfection. Similar results have observed in South Africa and Cote d’Ivoire [16,17]. The reasons for high mortality in those whose TB treatment failed, missing, or experienced adverse events may due to MTB dissemination, overwhelming infections, or failure to achieve rapid immunological recovery [18,19], but we did not get access to such details in this study. Future research should focus on the relationship between immune recovery and treatment outcomes concerning HIV/MTB co-infection.
Risk factors for mortality among those HIV/MTB co-infected in China have addressed elsewhere [20,21]. The majority of factors identified by hierarchical analysis in our study are consistent with the findings of previous studies. Adjusted by these risk factors, the HR for those whose treatment failed was still significantly higher than those who completed treatment and those who cured. These results underline that uncontrolled TB with insufficient treatment should consider as a leading cause of mortality among those with HIV/MTB coinfection. According to the WHO TB treatment guidelines [22], the standard treatment duration for cases of drug-susceptible TB is six months, and treatment for drug-resistant TB is longer, however, longer treatment duration may lead to low adherence to therapy and poor compliance has been associated with failure and drug-resistant TB [23–25], and lead to a high risk of medium to long term mortality. Methods to improve adherence such as shorten treatment duration, monitoring, and treatment of adverse events, and improving treatment success rates should be considered as a priority in clinical practice to reduce mortality risk of HIV/TB coinfection.
In our study, the median CD4 count found in the TBFMA group was l05 cells/µL (IQR: 25.5–268.5). This more advanced HIV/MTB stage leads to high mortality due to the effects of immune reconstitution inflammatory syndrome (IRIS). The enhanced immune function associated with ART initiation to treat HIV/MTB can lead to worsened clinical outcomes due to the increased severity of IRIS. TB-IRIS can categorize as “paradoxical” or “unmasking” in the mechanism. In paradoxical TB-IRIS, ART initiated in a patient with known TB, and clinical TB symptoms worsen after ART initiation. In unmasking TB-IRIS, patients with previously undiagnosed and untreated TB, such as latent TB, present with inflammatory features of TB after ART initiation [26]. Studies have shown that those with an advanced stage of HIV/MTB at the time of ART initiation, and the initiation of ART closer to the time of TB treatment initiated were consistent risk factors for TB IRIS [27,28]; other research has found that the development of TB IRIS is associated with the rapid expansion of pathogen-specific CD4 cells following ART initiation [29,30]. However, the results of other research are not consistent [31]. Our study cannot provide more details on TB IRIS because of the manner of data collection; thus, more research on pathophysiology and immunology are needed to define more precisely the mechanisms of TB-IRIS.
A strength of our study is that our results provide direct evidence for policymakers and stakeholders to highlight practical therapeutic developments and to review clinical outcomes of TB treatment for HIV/MTB co-infection. However, there were limitations to this study. First, we could not stratify data into pulmonary and extra-pulmonary TB to determine more precisely the impact of different TB types. Second, we could not analyze the role of TB IRIS on the mortality of HIV/MTB co-infection because of the method of data collection.