Predictors of poor outcome in patients diagnosed with drug-resistant tuberculosis in the Torres Strait / Papua New Guinea border region

Drug-resistant tuberculosis (DR-TB) is an ongoing challenge in the Torres Strait / Papua New Guinea border region. Treatment success rates have historically been poor for patients diagnosed with DR-TB, leading to increased transmission and resistance multiplication risk. This study aimed to identify predictors of poor outcome in patients diagnosed with DR-TB to inform programmatic improvements. A retrospective study of all DR-TB cases who presented to Australian health facilities in the Torres Strait between 1 March 2000 and 31 March 2020 was performed. This time period covers four distinct TB programmatic approaches. Univariate and multivariate predictors of poor outcome were analysed. Poor outcome was defined as treatment default, treatment failure and death (versus cure or completion). In total, 133 patients with resistance to at least one TB drug was identified. The vast majority (123/133; 92%) of DR-TB patients had pulmonary involvement; and of these, 41% (50/123) had both pulmonary and extrapulmonary TB. Poor outcomes were observed in 29% (39/133) of patients. Patients living with human immunodeficiency virus, renal disease or diabetes (4/133; 4/133; 3/133) had an increased frequency of poor outcome (p <0.05), but numbers were very small. Among all 133 DR-TB patients, 41% had a low lymphocyte count, which was significantly associated with poor outcome (p <0.05). Overall, outcome improved in recent years with a 50% increase in the chance of a good outcome per year group over the study period; on binary logistic regression analysis. Being a close contact of a known TB case was associated with improved outcome. While DR-TB treatment outcomes have improved over time, it remains important to prevent DR-TB spread and resistance multiplication resulting from suboptimal treatment. Enhanced surveillance for DR-TB, better cross border collaboration and consistent diagnosis and management of comorbidities and other risk factors should further improve patient care and outcomes.

is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted March 22, 2022. ;https://doi.org/10.1101https://doi.org/10. /2022 1 Management of DR-TB is complex and is often further complicated by comorbidities with other 2 communicable and non-communicable diseases (17). Patients with TB and coinfection / 3 comorbidities such as renal impairment, diabetes and HIV are more likely to have poor outcomes 4 (18)(19)(20). Many studies have reported high rates of mortality in TB patients with comorbidities (21-24) 5 and in patients with low levels of haemoglobin, albumin and lymphocytes (25) 6 While the rise in DR-TB is a global concern, TB programs must consider specific local risk factors and 7 programmatic gaps to ensure better outcomes for patients. These considerations may present 8 important opportunities for TB programmes to meet the challenging targets of the WHO to End TB, 9 aiming to reduce the incidence of TB by 90% before 2035 (26) 10 Little is known about predictors of treatment outcomes in patients diagnosed with DR-TB in the 11 Torres Strait / PNG border region. The aim of this study was to evaluate the association of 12 comorbidities / coinfection, the levels of a suite of serum biomarkers and the impact of 13 programmatic changes to models of TB care over time with treatment outcomes in DR-TB patients 14 diagnosed in the Torres Strait / PNG border region. It is the intention that evidence derived from this 15 study will be carefully considered at a programmatic level to further improve outcomes for patients 16 diagnosed with DR-TB in this context. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted March 22, 2022. ;https://doi.org/10.1101https://doi.org/10. /2022 were included in the study. Patients were excluded from the study if they were residents of PNG 26 villages external to the Western Province of PNG who did not enter the Australian health system via 27 a health facility in the Torres Strait Protected Zone (Fig 1).

28
Models of TB Care -Diagnosis and Treatment 29 There were four models of TB care provided in the region over different time periods between 2000 30 to 2020 (2000-2005; 2006-2012; 2013-2015; 2016-2020 is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint Thursday Island 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 1 (100) 1 . CC-BY 4.0 International license It is made available under a perpetuity.
is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted March 22, 2022. ;https://doi.org/10.1101https://doi.org/10. /2022  is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint 146 Table 3 shows that 86% of patients with haemoglobin recorded had anaemia, and of those, 38% had 147 severe anaemia. Low albumin was detected in 76% of patients that had levels recorded. In patients 148 with low lymphocyte levels, 44% of patients had a poor outcome.
149 is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint 151 *Mild anaemia, at least 2 standard deviations (SD) away from the mean; severe anaemia, at least 152 5SD away from the mean; # Low lymphocyte and albumin, at least 2SD away from the mean.

153
In Table 4, patients with comorbidities / coinfection, low lymphocyte levels and AFB positivity were 154 significantly more likely to have poor outcomes. Being a close contact of a known TB case was a 155 protective factor and reduced the odds of a poor outcome occurring (p .008; OR .31). Although the p 156 value for anaemia in Table 3  is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted March 22, 2022. ;https://doi.org/10.1101https://doi.org/10. /2022 166 This is the first study, to the best of our knowledge, to identify predictors of poor treatment 167 outcomes in patients diagnosed with DR-TB and examine the impact of four different models of TB

185
Collegial relationships between TB programs in the Torres Strait and Daru Island were also 186 strengthened through this period with the joint development of procedural documents and 187 processes (36) which enabled each TB program to define data requirements, streamline the 188 exchange of shared patient information, and enhance surveillance capability. These initiatives are 189 consistent with the Australian National Tuberculosis Advisory Committee recommendations to 190 engage in regional and bilateral collaborations in order to improve TB services in high-risk areas (37).
. CC-BY 4.0 International license It is made available under a perpetuity.
is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint  198 Despite the increased transmission risk, close contacts had more favourable outcomes in this study.

199
A possible explanation is that these close contact patients were linking in with healthcare services

204
A limitation in this study is that it did not identify the type of close contact (i.e. household contact), 205 nor ascertain the type of TB that each close contact was exposed to. The recent upgrade of the is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted March 22, 2022. ;https://doi.org/10.1101https://doi.org/10. /2022 Another potential limitation of this study is that only people with TB isolates with at least one 216 frontline drug resistance was included. Hence the full impact of MDR treatment (compared with fully 217 susceptible TB) could not be evaluated. We were however able to compare those with and without 218 rifampicin resistance, finding a trend to worse outcome.

219
Patients with pre-existing comorbidities/coinfection in this study were significantly more likely to 220 have poor treatment outcomes. This finding is consistent with that of other studies that report 221 worse treatment outcomes for patients with comorbidities/coinfection (6,41,42 is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted March 22, 2022. ;https://doi.org/10.1101https://doi.org/10. /2022