Abstract
Purpose
This study aimed to evaluate the prognosis of the not evaluated (NE) group by comparing it with the lost to follow-up (LTFU) group among patients with multidrug/rifampin-resistant tuberculosis (MDR/RR-TB).
Materials and Methods
This was a retrospective longitudinal follow-up study using an integrated database constructed by data linkage of the three national databases. This database included 7226 cases of MDR/RR-TB notified between 2011 and 2017 in South Korea.
Results
Among the 7226 MDR/RR-TB cases, 730 (10.1%) were classified as LTFU group, and 353 (4.9%) as NE group. When comparing NE group with LTFU group, there were no significant differences in the all-cause mortality rate (18.1% vs. 13.8%, p=0.065), median time to death [404 days (interquartile range, IQR 46–850) vs. 443 days (IQR 185–1157), p=0.140], and retreatment rate (26.9% vs. 22.2%, p=0.090). After adjusting for potential confounders, the adjusted hazard ratio (aHR) for all-cause mortality (aHR 1.11; 95% confidence interval 0.80-1.53; p=0.531) in NE group was not significantly different than that in LTFU group. Among retreated cases, NE group had a higher treatment success rate (57.9% vs 43.8%, p=0.029) and a lower LTFU rate (11.6% vs 38.3%, p<0.001) compared to LTFU group.
Conclusion
NE group had an unfavorable outcome comparable to LTFU group, suggesting undetected cases of LTFU or deaths during the referral process. Establishing an efficient patient referral system would contribute to reducing the incidence of NE cases.
Keywords: Tuberculosis, multidrug-resistant; lost to follow-up; patient transfer; mortality; risk factors; retreatment
Graphical Abstract
INTRODUCTION
Conventional treatment for multidrug/rifampin resistant tuberculosis (MDR/RR-TB) is lengthy, toxic, and complicated. Consequently, completion of MDR/RR-TB treatment is often not achieved.1 Lost to follow-up (LTFU) and not evaluated (NE) are two outcome categories designated for patients who have not completed their treatment as planned. NE is defined as a patient for whom no treatment outcome is assigned, which includes cases transferred out to another treatment unit and whose treatment outcome is unknown.2 The World Health Organization (WHO) globally estimated that 14.3% and 3.1% of MDR/RR-TB patients were LTFU and NE, respectively.3 In South Korea, these rates were estimated to be 11.4% for LTFU and 14.6% for NE among MDR/RR-TB patients notified between 2011 and 2014.4
LTFU group has been categorized as an unfavorable outcome without controversy. Numerous studies have highlighted the high mortality rate of LTFU group, as well as its potential to disseminate drug-resistant bacilli in the community.5,6 However, there was a controversy regarding how to define the prognosis of NE group; some studies have categorized NE group as an unfavorable outcome,7,8 whereas others categorized it as an unknown outcome.9
NE group shares similarities with LTFU group in that both groups have not completed treatment and cannot be traced by the National TB Control Program. Due to these similarities, some studies have merged NE group with LTFU group,10,11 but the supporting evidence for this approach is limited. Tracking what happens to patients assigned to LTFU and NE groups may provide insights into the two groups. While a prospective follow-up study is optimal, it is often impractical due to high costs and time requirements. Data linkage has been used as a complementary tool to track patients with failed follow-up in prospective studies,12 making it a potentially valuable tool for tracking the post-outcomes of both groups.
This study aimed to evaluate the prognosis and predictors of NE group by comparing it with LTFU group. We established an integrated TB database by linking the three national databases13 and reported nationwide treatment outcomes for patients with MDR/RR-TB in South Korea.14 Using this integrated TB database, we conducted a comparative analysis between NE and LTFU groups, focusing on the baseline characteristics, predictors of outcomes, timing of outcomes, and post-outcomes among patients with MDR/RR-TB in South Korea.
MATERIALS AND METHODS
Data sources and collection
The Korean Tuberculosis and Post-Tuberculosis (TB-POST) cohort was constructed by linking the following three national databases: 1) Korean Tuberculosis Surveillance System (KTBS) between 2011 and 2018, 2) National Health Insurance Database between 2006 and 2018, and 3) Causes of Death Statistics database between 2011 and 2018.13
The KTBS is a web-based notification system, which was established in 2000. TB notification is mandatory in South Korea, and its completeness reached 94% in 2014.15 Patient information related to TB is continuously registered with KTBS from notification to the end of treatment. The National Health Insurance Database is a public database on healthcare utilization, health screening, sociodemographic variables, and mortality formed by the National Health Insurance Service. South Korea has a universal healthcare insurance system. As of 2014, the National Health Insurance covered almost 98% of the total population in South Korea.16 The Causes of Death Statistics databases were developed based on death certificates and cover almost the entire population. Causes of death were obtained from the International Statistical Classification of Disease and Related Health Problems, 10th revision (ICD-10).
Study design and population
This was a retrospective, longitudinal follow-up study of patients with MDR/RR-TB. The study included MDR/RR-TB cases notified from January 1, 2011 to December 31, 2017, which were extracted from the Korean TB-POST cohort. We selected patients whose treatment outcomes were assigned as LTFU and NE, and followed their post-treatment outcomes from the date of end-of-treatment to July 30, 2020.
Definition and measurement
MDR-TB was defined as TB resistant to at least isoniazid (INH) and rifampin (RIF).2 Extensively drug-resistant TB (XDR-TB) was defined as TB resistant to at least INH and RIF plus any fluoroquinolones (FQ) and at least one of the injectable second-line drugs (amikacin, kanamycin, or capreomycin).
Treatment outcomes were assigned by the attending physicians according to the criteria suggested by the WHO.2 NE was defined as a patient for whom no treatment outcome was assigned, which included cases transferred out to another treatment unit and whose treatment outcome was unknown. LTFU was defined as a TB patient whose treatment was interrupted for 2 consecutive months or more.
A treatment episode was defined as a set of consecutive events without treatment interruption for more than 2 months. If a patient experienced multiple treatment episodes, the treatment outcome was defined as that of the first treatment episode. For example, if a patient was transferred and subsequently registered in a new institution within 2 months, it was considered a continuous treatment episode. Conversely, if the transferred patient was not registered with another institution within 2 months, it was designated as NE.
Post-outcomes such as retreatment, death, or survival were traced in the integrated database. The post treatment follow-up period was defined as the period from the end of the first treatment episode to July 30, 2020. The end-of-treatment date for both the LTFU and NE cases was determined by adding 60 days from the date of their last hospital visit. Retreatment was defined as the occurrence of a new treatment episode in a patient who already had a treatment outcome. These included both bacteriologically confirmed and clinically diagnosed cases. Deaths were classified as TB-related and non-TB-related deaths according to the ICD-10 codes in the Cause of Death Statistics database. Household income was classified into the 5th quintile (1=lowest, 5=highest), according to the national health insurance premium. Medical aid beneficiaries were classified into group 0.
Statistical analysis
Continuous variables were presented as mean±standard deviation if the variable was normally distributed, and as median (interquartile range, IQR) otherwise. Categorical variables were expressed as numbers (percentages). The Student’s t-test or Mann–Whitney test was used to compare continuous variables. The chi-square test or Fisher’s exact test was used to compare categorical variables. The Cox proportional hazard model was used to calculate the hazard ratios (HR) of the predictors in LTFU and NE groups, adjusting for covariates that were significantly different in the univariate model. NE and LTFU groups were compared to the combined group for success, failure, and death, respectively. Variables with p values <0.2 on univariate analysis were entered into the multivariate models. The Kaplan–Meier method and log-rank test were used to compare the survival times between two or more groups. All p-values were two-tailed, and a p-value <0.05 was deemed statistically significant. All statistical analyses were performed using STATA/MP version 17 (StataCorp LLC, College Station, TX, USA).
Ethics statement
This study was conducted in accordance with the Declaration of Helsinki. The study protocol was reviewed and approved by the Institutional Review Board of the National Evidence-based Healthcare Collaborating Agency (NECAIRB19-008-1). The requirement for informed consent was waived due to the retrospective nature of the study using public de-identified data.
RESULTS
Baseline characteristics
A total of 7226 cases of MDR/RR-TB were identified in the integrated TB database. Among these cases, 730 (10.1%) were classified as LTFU group and 353 (4.9%) as NE group. The baseline and clinical characteristics of both groups are shown in Table 1. NE group had a higher mean age, larger proportion of female, and greater percentage of XDR-TB cases compared to LTFU group. The annual trend of LTFU and NE rates were slightly different. The LTFU rates decreased from 14.2% to 7.1% between 2011 and 2017, while the NE rates decreased from 11.9% to 1.8% between 2011 and 2015, and then increased again to 3.0% in 2017 among the entire MDR-TB population.
Table 1. Baseline Characteristics of Patients Who Were NE and LTFU.
| Total (n=1083) | NE (n=353) | LTFU (n=730) | p value | ||
|---|---|---|---|---|---|
| Age (yr) | 48.5±16.0 | 50.5±16.3 | 47.5±15.8 | <0.001 | |
| Age group | 0.003 | ||||
| ≤24 years | 70 (6.5) | 15 (4.3) | 55 (7.5) | ||
| 25–34 years | 158 (14.6) | 44 (12.5) | 114 (15.6) | ||
| 35–44 years | 215 (19.9) | 74 (21.0) | 141 (19.3) | ||
| 45–54 years | 287 (26.5) | 85 (24.1) | 202 (22.7) | ||
| 55–64 years | 176 (16.3) | 56 (15.9) | 120 (16.4) | ||
| 65–74 years | 88 (8.1) | 42 (11.9) | 46 (6.3) | ||
| ≥75 years | 89 (8.2) | 37 (10.5) | 52 (7.1) | ||
| Gender | 0.007 | ||||
| Female | 261 (34.1) | 103 (29.2) | 158 (21.6) | ||
| Male | 822 (75.9) | 250 (70.8) | 572 (78.4) | ||
| Resident region | 0.002 | ||||
| Metropolitan | 538 (49.7) | 151 (42.8) | 387 (53.0) | ||
| Others | 545 (50.3) | 202 (57.2) | 343 (47.0) | ||
| Nationality | <0.001 | ||||
| Korean | 929 (85.8) | 326 (92.4) | 603 (82.6) | ||
| Immigrant | 154 (14.2) | 27 (7.7) | 127 (17.4) | ||
| Household incomes | 0.001 | ||||
| 0 (lowest) | 162 (15.0) | 53 (15.0) | 109 (14.9) | ||
| 1 | 216 (19.9) | 64 (18.1) | 152 (20.8) | ||
| 2 | 243 (22.4) | 60 (17.0) | 183 (25.1) | ||
| 3 | 220 (20.3) | 71 (20.1) | 149 (20.4) | ||
| 4 | 135 (12.5) | 59 (16.7) | 76 (10.4) | ||
| 5 (highest) | 107 (9.9) | 46 (13.0) | 61 (8.4) | ||
| Treatment history | 0.089 | ||||
| New | 491 (45.3) | 147 (41.6) | 344 (47.1) | ||
| Previously treated | 592 (54.7) | 206 (58.4) | 386 (52.9) | ||
| Lesion site | 0.090 | ||||
| Pulmonary | 1063 (98.2) | 350 (99.2) | 713 (97.7) | ||
| Extra-pulmonary | 20 (1.8) | 3 (0.9) | 17 (2.3) | ||
| Notifying institution 1 | 0.923 | ||||
| PPM | 315 (29.1) | 102 (28.9) | 213 (29.2) | ||
| Non-PPM | 768 (70.9) | 251 (71.1) | 517 (70.8) | ||
| Notifying institution 2 | 0.172 | ||||
| Health center | 115 (10.6) | 31 (8.8) | 84 (11.5) | ||
| Private institution | 968 (89.4) | 322 (91.2) | 646 (88.5) | ||
| Smear | 0.507 | ||||
| Positive | 591 (54.6) | 184 (52.1) | 407 (55.8) | ||
| Negative | 431 (39.8) | 147 (41.6) | 284 (38.9) | ||
| ND/unknown | 61 (5.6) | 22 (6.2) | 39 (5.3) | ||
| Culture | <0.001 | ||||
| Positive | 737 (68.1) | 213 (60.3) | 524 (71.8) | ||
| Negative | 120 (11.1) | 43 (12.2) | 77 (10.6) | ||
| ND/unknown | 226 (20.9) | 97 (27.5) | 129 (17.7) | ||
| Resistance pattern | <0.001 | ||||
| RR | 75 (6.9) | 17 (4.8) | 58 (8.0) | ||
| MDR | 816 (75.3) | 240 (68.0) | 576 (78.9) | ||
| Pre-XDR (SLID) | 22 (2.0) | 2 (0.6) | 20 (2.7) | ||
| Pre-XDR (FQ) | 43 (4.0) | 20 (5.7) | 23 (3.2) | ||
| XDR | 127 (11.7) | 74 (21.0) | 53 (7.3) | ||
| Comorbidity | |||||
| Diabetes mellitus | 217 (20.0) | 81 (23.0) | 136 (18.6) | 0.096 | |
| Cancer | 19 (1.8) | 5 (1.4) | 14 (1.9) | 0.556 | |
| Time to outcome (days) | 274 [106–471] | 99 [13–227] | 319 [188–518] | <0.001 | |
NE, not evaluated; LTFU, lost to follow-up; PPM, public-private mix; ND, not done; RR, rifampin resistance; MDR, multidrug resistance; SLID, second-line injectable drug; FQ, fluoroquinolone; XDR, extensively drug-resistant; IQR, interquartile range.
Data are presented as mean±standard deviation, n (%), or median [IQR].
Timing of outcomes
For NE group, the outcome occurred with a median of 99 days [IQR 13–227] from the initiation of MDR-TB treatment, which was significantly earlier than the median of 319 days [IQR 188–518] in LTFU group (p<0.001). The cumulative proportion of incident cases for NE group was 46.7% at 90 days and 61.4% at 180 days after treatment initiation. In contrast, for LTFU group, these proportions were 9.7% at 90 days and 23.3% at 180 days (Fig. 1).
Fig. 1. Number of incident cases of NE group and LTFU group. NE, not evaluated; LTFU, lost to follow-up.
Predictors of outcomes
Table 2 shows the results of the univariate and multivariate analyses of predictors in LTFU and NE group. Age >75 years, the lowest income class, previously treated cases, and XDR-TB were identified as common independent predictors for both groups. Interestingly, while XDR reduced the risk of LTFU, it increased the risk of NE. Moreover, being male, a non-metropolitan resident, and an immigrant were identified as independent predictors of LTFU group but not of NE group.
Table 2. Predictors of NE Group and LTFU Group.
| LTFU | NE | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Univariate HR (95% CI) | p value | Multivariate aHR (95% CI) | p value | Univariate HR (95% CI) | p value | Multivariate aHR (95% CI) | p value | ||
| Age group | |||||||||
| ≤24 years | Reference | Reference | Reference | Reference | |||||
| 25–34 years | 1.11 (0.80–1.52) | 0.543 | 1.07 (0.77–1.48) | 0.693 | 1.53 (0.85–2.75) | 0.156 | 1.40 (0.78–2.52) | 0.262 | |
| 35–44 years | 1.31 (0.96–1.79) | 0.088 | 1.34 (0.98–1.85) | 0.068 | 2.46 (1.41–4.29) | 0.001 | 2.09 (1.19–3.67) | 0.010 | |
| 45–54 years | 1.52 (1.12–2.04) | 0.006 | 1.31 (0.97–1.79) | 0.082 | 2.29 (1.32–3.97) | 0.003 | 1.81 (1.03–3.17) | 0.039 | |
| 55–64 years | 1.13 (0.82–1.55) | 0.459 | 1.01 (0.72–1.40) | 0.976 | 1.85 (1.05–3.27) | 0.034 | 1.52 (0.85–2.73) | 0.158 | |
| 65–74 years | 0.81 (0.55–1.20) | 0.288 | 0.83 (0.56–1.24) | 0.371 | 2.32 (1.28–4.18) | 0.005 | 1.92 (1.05–3.51) | 0.033 | |
| ≥75 years | 1.14 (0.78–1.67) | 0.500 | 1.50 (1.01–2.22) | 0.044 | 2.26 (1.24–4.11) | 0.008 | 2.35 (1.27–4.34) | 0.006 | |
| Gender | |||||||||
| Female | Reference | Reference | Reference | Reference | |||||
| Male | 1.78 (1.50–2.13) | <0.001 | 1.72 (1.43–2.07) | <0.001 | 1.24 (0.98–1.55) | 0.071 | 1.10 (0.87–1.40) | 0.419 | |
| Region | |||||||||
| Metropolitan | Reference | Reference | Reference | ||||||
| Others | 0.84 (0.72–0.97) | 0.016 | 0.79 (0.68–0.91) | 0.001 | 1.23 (0.99–1.52) | 0.057 | 1.22 (0.98–1.51) | 0.072 | |
| Nationality | |||||||||
| Korean | Reference | Reference | Reference | Reference | |||||
| Immigrant | 2.72 (2.25–3.30) | <0.001 | 3.02 (2.46–3.71) | <0.001 | 1.22 (0.83–1.81) | 0.317 | 1.48 (0.99–2.23) | 0.058 | |
| Income | |||||||||
| 0 (lowest) | 2.78 (2.03–3.80) | <0.001 | 2.64 (1.92–3.63) | <0.001 | 1.94 (1.31–2.88) | 0.001 | 1.60 (1.07–2.40) | 0.021 | |
| 1 | 2.09 (1.55–2.81) | <0.001 | 1.79 (1.32–2.42) | <0.001 | 1.28 (0.88–1.87) | 0.201 | 1.20 (0.81–1.76) | 0.364 | |
| 2 | 2.29 (1.71–3.06) | <0.001 | 2.04 (1.52–2.74) | <0.001 | 1.09 (0.74–1.60) | 0.661 | 1.15 (0.78–1.70) | 0.488 | |
| 3 | 1.82 (1.35–2.45) | <0.001 | 1.51 (1.11–2.05) | 0.009 | 1.24 (0.86–1.80) | 0.251 | 1.18 (0.81–1.73) | 0.392 | |
| 4 | 1.12 (0.80–1.58) | 0.494 | 1.09 (0.78–1.53) | 0.619 | 1.20 (0.82–1.77) | 0.344 | 1.19 (0.81–1.76) | 0.372 | |
| 5 (highest) | Reference | Reference | Reference | Reference | |||||
| Treatment history | |||||||||
| New | Reference | Reference | Reference | Reference | |||||
| Previously treated | 1.42 (1.23–1.64) | <0.001 | 1.41 (1.22–1.54) | <0.001 | 1.87 (1.51–2.31) | <0.001 | 1.62 (1.30–2.02) | <0.001 | |
| Lesion site | |||||||||
| Pulmonary | Reference | Reference | Reference | ||||||
| Extra-pulmonary | 0.83 (0.51–1.35) | 0.454 | - | - | 0.28 (0.09–0.86) | 0.027 | 0.19 (0.06–0.61) | 0.005 | |
| Notifying institution 1 | |||||||||
| PPM | Reference | Reference | Reference | Reference | |||||
| Non-PPM | 0.83 (0.71–0.97) | 0.023 | 1.09 (0.89–1.34) | 0.413 | 0.80 (0.63–1.00) | 0.055 | 1.05 (0.82–1.35) | 0.693 | |
| Notifying institution 2 | |||||||||
| Health center | Reference | Reference | Reference | Reference | |||||
| Private institution | 0.79 (0.63–0.99) | 0.044 | 0.85 (0.64–1.13) | 0.273 | 0.99 (0.68–1.43) | 0.940 | - | - | |
| Smear | |||||||||
| Positive | Reference | Reference | |||||||
| Negative | 0.93 (0.80–1.09) | 0.379 | - | - | 1.01 (0.81–1.25) | 0.960 | - | - | |
| ND/unknown | 1.16 (0.83–1.60) | 0.390 | - | - | 1.33 (0.85–2.07) | 0.206 | - | - | |
| Culture | |||||||||
| Positive | Reference | Reference | Reference | Reference | |||||
| Negative | 1.14 (0.90–1.45) | 0.290 | 1.12 (0.88–1.42) | 0.374 | 1.42 (1.03–1.98) | 0.035 | 1.52 (1.09–2.12) | 0.013 | |
| ND/unknown | 1.61 (1.33–1.96) | <0.001 | 1.70 (1.38–2.09) | <0.001 | 2.86 (2.25–3.64) | <0.001 | 3.05 (2.35–3.95) | <0.001 | |
| Resistance pattern | |||||||||
| RR | Reference | Reference | Reference | Reference | |||||
| MDR | 1.10 (0.84–1.44) | 0.509 | 1.07 (0.82–1.41) | 0.614 | 1.93 (1.18–3.16) | 0.009 | 1.60 (0.97–2.62) | 0.064 | |
| Pre-XDR (SLID) | 0.92 (0.55–1.53) | 0.738 | 0.99 (0.60–1.66) | 0.981 | 0.42 (0.10–1.83) | 0.249 | 0.43 (0.10–1.85) | 0.255 | |
| Pre-XDR (FQ) | 0.52 (0.32–0.85) | 0.009 | 0.49 (0.30–0.79) | 0.004 | 2.10 (1.10–4.02) | 0.025 | 1.77 (0.92–3.42) | ||
| XDR | 0.59 (0.40–0.86) | 0.006 | 0.52 (0.35–0.76) | 0.001 | 3.63 (2.14–6.18) | <0.001 | 2.74 (1.60–4.69) | ||
| Comorbidity | |||||||||
| Diabetes mellitus | 1.20 (1.00–1.45) | 0.053 | 1.21 (1.00–1.48) | 0.055 | 1.50 (1.17–1.92) | 0.001 | 1.29 (0.99–1.68) | ||
| Cancer | 1.32 (0.78–2.24) | 0.302 | - | - | 0.80 (0.33–1.93) | 0.618 | - | ||
NE, not evaluated; LTFU, lost to follow-up; HR, hazard ratio; aHR, adjusted hazard ratio; CI, confidencd interval; PPM, public-private mix; ND, not done; RR, rifampin resistance; MDR, multidrug resistance; SLID, second-line injectable drug; FQ, fluoroquinolone; XDR, extensively drug-resistant.
Post-treatment mortality
Table 3 shows the post-treatment outcomes for both groups during the median follow-up period of 4.2 years [IQR 2.2–7.1]. Of the 353 NE cases, 64 (18.1%) died at a median of 404 days [IQR 46–850] after from the end-of-treatment date; 27 (7.6%) deaths were TB-related and 37 (10.5%) were non-TB-related. The all-cause mortality rate, non-TB-related mortality rate, and time to all-cause death did not significantly different between NE group and LTFU group. However, TB-related mortality rate was significantly higher in NE group.
Table 3. Post-Treatment Outcomes of Patients Who Were NE and LTFU.
| Total (n=1083) | NE (n=353) | LTFU (n=730) | p value | ||
|---|---|---|---|---|---|
| Retreatment | 257 (23.7) | 95 (26.9) | 162 (22.2) | 0.090 | |
| Death, all cause | 165 (15.2) | 64 (18.1) | 101 (13.8) | 0.065 | |
| TB-related | 52 (4.8) | 27 (7.6) | 25 (3.4) | 0.002 | |
| Non-TB-related | 113 (10.4) | 37 (10.5) | 76 (10.4) | 0.960 | |
| Alive | 918 (84.8) | 289 (81.9) | 629 (86.2) | 0.065 | |
| Time to retreatment (days) | 312 [152–775] | 203 [22–602] | 398 [195–872] | <0.001 | |
| Time to death (days) | 431 [143–1003] | 404 [46–850] | 443 [185–1157] | 0.140 | |
NE, not evaluated; LTFU, lost to follow-up; TB, tuberculosis; IQR, interquartile range.
Data are presented as n (%) or median [IQR].
Fig. 2 shows Kaplan–Meier survival curve comparing NE group with LTFU group as a reference. This was adjusted for potential confounders and other variables (age, gender, nationality, income, TB treatment history, lesion site, health institution, sputum smear result, drug susceptibility pattern, diabetes, and cancer). In NE group, the adjusted hazard ratio (Ahr) of all-cause mortality [Ahr 1.11; 95% confidence interval (CI) 0.80–1.53; p=0.531] and non-TB-related mortality (Ahr 0.85; 95% CI 0.57–1.26, p=0.415) were not significantly different, but TB-related mortality (Ahr 1.86; 95% CI 1.05–3.31, p=0.034) was significantly higher than that in LTFU group.
Fig. 2. Comparison of survival between NE group and LTFU group. (A) All-cause mortality. (B) TB-related mortality. (C) Non-TB-related mortality. Adjusted for age, gender, nationality, income, tuberculosis treatment history, lesion site, health institution, sputum smear result, drug susceptibility pattern, diabetes, and cancer. NE, not evaluated; LTFU, lost to follow-up; aHR, adjusted hazard ratio; TB, tuberculosis; CI, confodence interval.
Retreatment outcomes
In NE group, 95 patients (26.9%) returned to treatment at a median of 203 days [IQR 22–602] from the end-of-treatment date (Table 3). Retreatment rate was not significantly different but the time to retreatment was significantly earlier in NE group, compared to LTFU group. Of the 95 NE cases who resumed treatment, 55 (57.9%) were successfully treated and 16 (16.8%) died during retreatment (Table 4). NE group had higher treatment success rate and all-cause mortality rate, while the LTFU rate was lower compared to LTFU group. Among the 162 LTFU cases who resumed treatment, 62 (38.3%) were LTFU again.
Table 4. Retreatment Outcomes of Patients Who Were NE and LTFU.
| NE (n=95) | LTFU (n=162) | p value | ||
|---|---|---|---|---|
| Success | 55 (57.9) | 71 (43.8) | 0.029 | |
| Failure | 0 | 5 (3.1) | 0.084 | |
| LTFU | 11 (11.6) | 62 (38.3) | <0.001 | |
| Death | 16 (16.8) | 9 (5.6) | 0.004 | |
| TB-related | 7 (7.4) | 6 (3.7) | 0.193 | |
| Non-TB-related | 9 (9.5) | 3 (1.9) | 0.006 | |
| On treatment | 13 (13.7) | 15 (9.3) | 0.276 | |
NE, not evaluated; LTFU, lost to follow-up; TB, tuberculosis.
Data are presented as number (%).
DISCUSSION
This study revealed that NE group had an unfavorable outcome, with similar all-cause mortality rate and higher TB-related mortality rate compared to LTFU group. Moreover, the outcome occurred significantly earlier in NE group than in LTFU group, suggesting a potentially higher risk of spreading resistant strains in the community due to insufficient treatment. Conversely, NE group demonstrated higher treatment success rate and lower rate of loss to follow-up on retreatment, compared to LTFU group. These findings suggest that NE group is expected to achieve greater improvements in outcomes through active tracing and resumption of treatment.
The completeness of reporting outcome is one of the indicators for the performance of the National TB Control Program. Therefore, the proportion of NE group reflected a weakness in the National TB Control Program. The majority of NE cases were patients who had been transferred to another treatment unit and their treatment outcomes were unknown. During the referral process, TB patients faced a high risk of treatment delay, loss to follow-up, or death.17,18,19,20,21 In this study, NE group had earlier timing of the outcome, higher TB-related death rate, and higher proportion of XDR-TB compared to LTFU group. These findings suggest that NE group was more likely to include patients who were difficult-to-treat, hence they had been transferred earlier, and would eventually reach loss to follow-up or TB-related death.
Our study revealed that NE group and LTFU group were similar with comparable long-term prognosis and shared common predictors of each outcome. There was no significant difference in all-cause mortality rate, time to all-cause death, and retreatment rate between the two groups. Furthermore, the two groups shared important predictors of outcomes, which were age >75 years, the lowest income class, and a previously treated case.
One of the interesting findings was that XDR-TB had opposite effects on the risk of LTFU and NE groups. The impact of XDR-TB on the increased risk of NE group is presumed to result from the frequent referral process, as mentioned earlier, where the risk of LTFU increases. In contrast, XDR-TB was associated with a decreased risk of LTFU. This might be partially explained by the strengthened patient management following the introduction of new drugs in South Korea. In South Korea, since 2014, bedaquiline and delamanid have been used for MDR-TB patients with limited treatment options, particularly those with FQ-resistant MDR-TB. To ensure the proper use of the new drugs, the National Tuberculosis Expert Committee was established. This committee has been responsible for approving and providing partial supervision for all cases of new drug use.22
Despite the similarities, there were some differences between the two groups in terms of the predictors of outcomes, timing of outcome, and cause of death. These findings can be clues to understanding the difference between the two groups. Several demographic factors such as being male, an immigrant, and a non-metropolitan resident were independent predictors of LTFU group, consistent with previous studies,11,23 but these were not independent predictors of NE group. In terms of socioeconomic factors, household income was more strongly associated with the risk of LTFU than that of NE. For disease-related factors, XDR-TB was inversely related to the risk of LTFU, but proportionally related to the risk of NE. These findings suggest that disease-related factors, rather than demographic and socioeconomic factors, were more strongly associated with the risk of NE group, compared to LTFU group.
In this study, the cumulative proportion of incident cases for combining the two groups over a 6-month period was 45.7%. This was closely aligned with the 48% reported in a prior study, where NE group was merged with LTFU group.10 In terms of post-outcome, the higher TB-related mortality in NE group might be attributed to the older age and higher proportion of XDR-TB. Given the earlier timing of outcomes in NE group, uncontrolled TB contributed more significantly to death.
The global presence of NE group suggests that many countries have encountered challenges in establishing efficient patient referral systems for TB patients. Li, et al.24 reported the efficacy of a web-based transfer system in estimating the burden and tracking final outcomes of transferred patients. eHealth solutions, such as mobile phones or SMS, would be a promising tool to enhance the management of transferred patients.25 In South Korea, the Personal Information Protection Act has prohibited the real-time tracking of transferred patients using the TB registration system. Establishing an efficient patient referral system can contribute to the reduction of cases with unknown outcomes, loss to follow-up, or deaths among transferred patients, thereby lowering the incidence of NE cases.
Our study had several limitations, primarily attributed to its retrospective nature and reliance on routinely collected health data. First, the follow-up duration varied among the included patients, which may potentially introduce significant bias when estimating long-term mortality. Second, the determination of cause of death was reliant on data from the Cause of Death Statistics, which were extracted from death certificates. The accuracy of this data has not been verified through autopsies or alternate methods, possibly resulting in reporting biases in cases of incorrect death certificates. Third, regimen-related factors were not incorporated in the assessment of predictors as they were largely missing in the KTBS. Fourth, since this study was conducted in a country with low HIV prevalence, this should be considered when interpreting the study results.
In conclusion, NE group had an unfavorable outcome comparable to LTFU group, demonstrated by a comparable all-cause mortality rate and time to death. NE group likely represents undetected cases of LTFU or death occurring during the referral process. NE group is expected to achieve greater improvements in outcomes through active tracing and retreatment, compared to LTFU group. Establishing an efficient patient referral system would contribute to the reduction of incidence and improvement of outcomes for NE group.
ACKNOWLEDGEMENTS
This study used the National Health Information Database (NHIS-2019-1-662) of the National Health Insurance Service (NHIS).
This study was financially supported by the National Evidence-based Healthcare Collaborating Agency, funded by the Ministry of Health and Welfare (grant No. NC19-002, NC20-003, and NC21-001).
Footnotes
The authors have no potential conflicts of interest to disclose.
- Conceptualization: Doosoo Jeon and Hee Jin Kim.
- Data curation: Jeongha Mok, Young Ae Kang, and Dawoon Jeong.
- Formal analysis: Hongjo Choi and Hee-Sun Kim.
- Funding acquisition: Hongjo Choi.
- Investigation: Jeongha Mok, Young Ae Kang, and Hee-Yeon Kang.
- Methodology: Jeongha Mok, Young Ae Kang, and Dawoon Jeong.
- Project administration: Hongjo Choi and Doosoo Jeon.
- Resources: Hongjo Choi and Hee-Sun Kim.
- Software: Dawoon Jeong.
- Supervision: Doosoo Jeon and Hee Jin Kim.
- Validation: Doosoo Jeon and Hee Jin Kim.
- Visualization: Dawoon Jeong and Hee-Sun Kim.
- Writing—original draft: Hongjo Choi.
- Writing—review & editing: Jeongha Mok, Young Ae Kang, Dawoon Jeong, Hee-Yeon Kang, Hee Jin Kim, Hee-Sun Kim, and Doosoo Jeon.
- Approval of final manuscript: all authors.
References
- 1.Dheda K, Gumbo T, Maartens G, Dooley KE, Murray M, Furin J, et al. The Lancet Respiratory Medicine Commission: 2019 update: epidemiology, pathogenesis, transmission, diagnosis, and management of multidrug-resistant and incurable tuberculosis. Lancet Respir Med. 2019;7:820–826. doi: 10.1016/S2213-2600(19)30263-2. [DOI] [PubMed] [Google Scholar]
- 2.World Health Organization. Definitions and reporting framework for tuberculosis . 2013 revision: updated December 2014 and January 2020. [accessed on 2024 February 20]. Available at: https://iris.who.int/handle/10665/79199 .
- 3.World Health Organization. Global tuberculosis report 2022. [accessed on 2024 February 20]. Available at: https://www.who.int/publications/i/item/9789240061729 .
- 4.Lee M, Han J, Kim YR, Kwak N, Kim JH, Park O, et al. Multidrug-resistant tuberculosis in South Korea: a retrospective analysis of national registry data in 2011-2015. Int J Tuberc Lung Dis. 2019;23:850–857. doi: 10.5588/ijtld.18.0658. [DOI] [PubMed] [Google Scholar]
- 5.Franke MF, Appleton SC, Bayona J, Arteaga F, Palacios E, Llaro K, et al. Risk factors and mortality associated with default from multidrug-resistant tuberculosis treatment. Clin Infect Dis. 2008;46:1844–1851. doi: 10.1086/588292. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Kuchukhidze G, Baliashvili D, Adamashvili N, Kasradze A, Kempker RR, Magee MJ. Long-term mortality and active tuberculosis disease among patients who were lost to follow-up during second-line tuberculosis treatment in 2011-2014: population-based study in the country of Georgia. Open Forum Infect Dis. 2021;8:ofab127. doi: 10.1093/ofid/ofab127. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Johnston JC, Shahidi NC, Sadatsafavi M, Fitzgerald JM. Treatment outcomes of multidrug-resistant tuberculosis: a systematic review and meta-analysis. PLoS One. 2009;4:e6914. doi: 10.1371/journal.pone.0006914. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Eshetie S, Alebel A, Wagnew F, Geremew D, Fasil A, Sack U. Current treatment of multidrug resistant tuberculosis in Ethiopia: an aggregated and individual patients’ data analysis for outcome and effectiveness of the current regimens. BMC Infect Dis. 2018;18:486. doi: 10.1186/s12879-018-3401-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Orenstein EW, Basu S, Shah NS, Andrews JR, Friedland GH, Moll AP, et al. Treatment outcomes among patients with multidrug-resistant tuberculosis: systematic review and meta-analysis. Lancet Infect Dis. 2009;9:153–161. doi: 10.1016/S1473-3099(09)70041-6. [DOI] [PubMed] [Google Scholar]
- 10.Walker IF, Shi O, Hicks JP, Elsey H, Wei X, Menzies D, et al. Analysis of loss to follow-up in 4099 multidrug-resistant pulmonary tuberculosis patients. Eur Respir J. 2019;54:1800353. doi: 10.1183/13993003.00353-2018. [DOI] [PubMed] [Google Scholar]
- 11.Kim HW, Park S, Min J, Sun J, Shin AY, Ha JH, et al. Hidden loss to follow-up among tuberculosis patients managed by public-private mix institutions in South Korea. Sci Rep. 2022;12:12362. doi: 10.1038/s41598-022-16441-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Harron K. Data linkage in medical research. BMJ Med. 2022;1:e000087. doi: 10.1136/bmjmed-2021-000087. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Jeong D, Kang HY, Kim J, Lee H, Yoo BN, Kim HS, et al. Cohort profile: Korean tuberculosis and post-tuberculosis cohort constructed by linking the Korean national tuberculosis surveillance system and national health information database. J Prev Med Public Health. 2022;55:253–262. doi: 10.3961/jpmph.21.635. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Choi H, Mok J, Kang YA, Jeong D, Kang HY, Kim HJ, et al. Nationwide treatment outcomes of patients with multidrug/rifampinresistant tuberculosis in Korea, 2011-2017: a retrospective cohort study (Korean TB-POST) J Korean Med Sci. 2023;38:e33. doi: 10.3346/jkms.2023.38.e33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Kang HY, Yoo H, Park W, Go U, Jeong E, Jung KS, et al. Tuberculosis notification completeness and timeliness in the Republic of Korea during 2012-2014. Osong Public Health Res Perspect. 2016;7:320–326. doi: 10.1016/j.phrp.2016.08.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Kim JA, Yoon S, Kim LY, Kim DS. Towards actualizing the value potential of Korea Health Insurance Review and Assessment (HIRA) data as a resource for health research: strengths, limitations, applications, and strategies for optimal use of HIRA data. J Korean Med Sci. 2017;32:718–728. doi: 10.3346/jkms.2017.32.5.718. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Vree M, Huong NT, Duong BD, Sy DN, Van le N, Co NV, et al. Mortality and failure among tuberculosis patients who did not complete treatment in Vietnam: a cohort study. BMC Public Health. 2007;7:134. doi: 10.1186/1471-2458-7-134. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Meijnen S, Weismuller MM, Claessens NJ, Kwanjana JH, Salaniponi FM, Harries AD. Outcome of patients with tuberculosis who transfer between reporting units in Malawi. Int J Tuberc Lung Dis. 2002;6:666–671. [PubMed] [Google Scholar]
- 19.Belayneh T, Kassu A, Tigabu D, Asmare G, Tilaye S, Klinkenberg E. Characteristics and treatment outcomes of “transfer-out” pulmonary tuberculosis patients in Gondar, Ethiopia. Tuberc Res Treat. 2016;2016:1294876. doi: 10.1155/2016/1294876. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Lopez E, Yoshimatsu S, Querri A, Coprada L, Bermejo J, Paulino MR, et al. The referral pathway of presumptive drug-resistant tuberculosis in an urban poor setting, the Philippines. Public Health Action. 2016;6:237–241. doi: 10.5588/pha.16.0046. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Li T, Zhang H, Shewade HD, Soe KT, Wang L, Du X. Patient and health system delays before registration among migrant patients with tuberculosis who were transferred out in China. BMC Health Serv Res. 2018;18:786. doi: 10.1186/s12913-018-3583-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Kwon YS, Shin JE, Kang H, Jeon D, Yim JJ, Shim TS. Treatment outcomes of multidrug-resistant TB with selective use of new drugs. Int J Tuberc Lung Dis. 2023;27:55–60. doi: 10.5588/ijtld.22.0343. [DOI] [PubMed] [Google Scholar]
- 23.Soedarsono S, Mertaniasih NM, Kusmiati T, Permatasari A, Juliasih NN, Hadi C, et al. Determinant factors for loss to follow-up in drug-resistant tuberculosis patients: the importance of psycho-social and economic aspects. BMC Pulm Med. 2021;21:360. doi: 10.1186/s12890-021-01735-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Li T, Du X, Shewade HD, Soe KT, Zhang H. What happens to migrant tuberculosis patients who are transferred out using a web-based system in China? PLoS One. 2018;13:e0206580. doi: 10.1371/journal.pone.0206580. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Margineanu I, Louka C, Akkerman O, Stienstra Y, Alffenaar JW. eHealth in TB clinical management. Int J Tuberc Lung Dis. 2022;26:1151–1161. doi: 10.5588/ijtld.21.0602. [DOI] [PMC free article] [PubMed] [Google Scholar]