Pediatric drug-sensitive tuberculosis treatment regimen profiles and treatment outcomes in mainland China, 2017–2022: a multi-center study Authors: Xinxin Zeng, Hao Xu, Aole Li, Juanjuan Liu, Yuhang Wu, Xingyu Dong, Xin Yu, Hongmei Xu, Ping Liu, Juan Ma, Qingshan Cai, Haiyan Li, Man Tian, Qing Fang, Guangxu Yang, Fei Wang, Bin Chen, Lisu Huang Published: 2025 ABSTRACT BackgroundDespite the establishment of WHO treatment guidelines for childhood tuberculosis (TB), evaluation of TB treatment regimens for children in real-world settings in China remained limited, particularly regarding treatment duration, utilization of second-line anti-TB drugs (SLDs), and adverse events. This study aimed to describe the treatment protocols for childhood TB in China.MethodsA multi-center retrospective observational study was conducted at ten state-mandated TB referral hospitals across four regions in China during 2017–2022. Children with drug-susceptible TB who were hospitalized and under 15 years-old were enrolled. And their clinical information, including diagnosis, treatment regimen, duration, adverse events and treatment response, were collected by professional pediatricians through the hospital medical records system and telephone follow-ups.ResultsA total of 597 children were included, about 75% of children were pulmonary TB (PTB)-only; the remaining 25% children were diagnosed with extrapulmonary TB (EPTB). In total, 63.5% of all children treated on a standard core regimen (HRZ(E)/HR(E)) while 27.8% were on a regimen that included SLDs and 8.71% were on other regimens. 67.0% of PTB-only patients received a standard core regimen, while 41.7% of EPTB patients were on regimens including at least one SLDs. Duration of regimen varied by form of TB: 72.4% of PTB-only patients and 90.1% of EPTB patients received a regimen longer than six months. Among EPTB patients, extended treatment was most common in those with severe manifestations—such as meningitis and osteoarticular TB—with median durations of 18.2 months and 12.7 months, respectively. The treatment success rates for children with PTB-only and with EPTB were 96.5% and 94.4%, respectively. A total of 133 (22.3%) children experienced adverse reactions. Treatment duration exceeding 12 months (OR: 5.52, P < 0.01) and inclusion of SLDs (OR: 2.67, P < 0.01) were significantly associated with increased odds of adverse reactions.ConclusionIn our study, a relatively proportion of pediatric drug-sensitive TB received extended duration regimens, and SLDs were often incorporated for cases of EPTB. While the treatment regimens resulted in few unfavorable outcomes, they were associated with a relatively high frequency of side effects.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12879-025-12112-w. Introduction Tuberculosis (TB), caused by Mycobacterium tuberculosis (MTB), remained a formidable global infectious disease with substantial health burdens. Pediatric TB accounted for 12% of total cases (1.3 million annually), with an alarming mortality rate of 16% among affected children [1]. The substantial disease burden underscored the critical need for heightened focus on this vulnerable population. Notably, untreated pediatric TB carries a mortality risk as high as 22%, while standardized anti-TB regimens could reduce this rate to 0.9% [2]. Consequently, standardized treatment was crucial for effective pediatric TB management. World Health Organization (WHO) recommended several standard regimens for drug-susceptible TB, including regimens for non-severe pulmonary TB (PTB), other pulmonary or extrapulmonary TB (EPTB, such as TB meningitis), while non-severe EPTB was often overlooked. In reality, a significant proportion of childhood TB cases were extrapulmonary, accounting for approximately three-tenths of all cases [3], and might face greater treatment challenges [4], such as prolonged treatment durations, the use of second-line anti-TB drugs (SLDs), and increased adverse drug reactions. For example, a real-world study in Italy found the use of SLDs exceeding the incidence of drug-resistant TB [5], and the impact of these drugs on treatment outcomes and adverse effects in children is not fully understood. As the third highest TB-burdened nation globally, China faces unique challenges in pediatric TB management. Current WHO treatment guidelines primarily derive from studies in low-resource African and Southeast Asian settings, potentially misaligning with China’s advanced healthcare infrastructure and distinct MTB strain prevalence. The divergent treatment durations recommended by WHO and Chinese domestic guidelines may serve as evidence of this misalignment. WHO guidelines recommended a 4-month duration for non-severe TB, 6-month for PTB and most EPTB case, and for 12-month for TB meningitis (TBM) [6]. Whereas Chinese guidelines extended 12 months for severe PTB (such as miliary TB and bronchial TB), and TBM, 9–12 months for TB pleuritis [7]. Currently, the selection of TB treatment regimens, adverse event and treatment outcomes for children in China, as well as the actual implementation of WHO-recommended regimens, remained unclear. This study addresses these evidence deficiencies through a multicenter real-world analysis of treatment outcomes from 2017 to 2022 across China’s designated pediatric TB hospitals. Methods Patients and study design This multicenter retrospective observational study was conducted at ten state-mandated TB referral hospitals across four regions in China between January 2017 and December 2022. The inclusion criteria were as follows: diagnosed with TB and without a diagnosis of drug-resistant TB; hospitalized children; ≤ 14 years old at the time of their initial TB diagnosis; and having completed at least the intensive phase of treatment with antituberculosis medications. A trained researcher then independently reviewed each record for completeness and accuracy before its final inclusion in the centralized study database. The study flowchart was shown in Supplementary Figure S1. Criteria for pediatric TB diagnosis and classification Pediatric TB was diagnosed based on the industry standards for TB diagnosis issued by the National Health Commission of China [8]. All included children had either bacteriologically confirmed TB (bacteriological evidence obtained by culture, smear microscopy, or WHO-approved rapid diagnostic tests, such as the GeneXpert MTB/RIF) or clinically diagnosed TB (bacteriological criteria were not met but a diagnosis of active TB was made by a clinician based on clinical manifestations, radiological and immunologic evidence). TB is categorized into PTB and EPTB according to the site of the lesion [9]. PTB was defined as any diagnosed case of TB only involving intrathoracic tissue (e.g. lung parenchyma, the tracheobronchial tree, intrathoracic lymph nodes (mediastinal and/or hilar), and pleura) [10–12]. EPTB was defined as any case of TB that only involved organs other than the lungs (e.g. peripheral lymph nodes, abdomen, genitourinary tract, skin, osteoarticular, meninges) [10, 12, 13]. The specific subtypes of PTB and EPTB included in our study are detailed in Supplementary Appendix S1. Data collection Demographic and clinical characteristics were abstracted by professional pediatricians through the hospital medical records system and telephone interviews, including age, gender, weight, history of close contact with TB patients, and bacillus Calmette-Guérin (BCG) vaccination status, immunologic test results, bacteriological results, symptoms before admission, treatment regimens, duration, adverse events and treatment outcomes, etc. Pediatricians routinely documented treatment outcomes under China’s national TB control policy [14]. The Chinese Medical Association defines low body weight as being below the third percentile for Chinese children aged 0–14 years [11]. Pediatric TB treatment regimens The current first-line antituberculosis drugs recommended by Chinese guidelines for children include isoniazid (INH), rifampicin (RFP), pyrazinamide (PZA), and/or ethambutol (EMB), the detail regimens and duration was presented in Supplementary Table 1S [7]. Treatment regimens containing only the first three drugs or all four drugs are defined as the standard core regimen (HRZ(E)/HR(E)). Patient adding SLDs (e.g., linezolid, levofloxacin, moxifloxacin, amikacin, etc.) to the standard core regimen were classified as the SLDs-containing regimen. The absent usage of INH, RFP, or PZA during the intensive or continuation phase was defined as other regimens. Adverse events and treatment outcomes Adverse events were monitored, focusing on liver and renal impairment, hematological events, gastrointestinal reactions, rashes, ototoxicity, visual changes, central nervous system reactions, arthralgia, peripheral neuropathy, and other discomforts. The WHO guidelines for TB surveillance define the following treatment outcomes: cure, treatment completed, treatment failed, died, and lost to follow-up (Appendix S2) [15, 16]. Treatment outcomes were based on the standard national TB programme outcomes documented on the TB register. Unfavorable outcome has two definitions: one includes death, treatment failure, or loss to follow-up, while the other as death or treatment failure. Statistical analyses The information on pediatric patients’ basic and clinical characteristics, antituberculosis treatment medication regimens and duration, adverse events, and treatment outcomes was presented as frequencies (percentage, %) and medians (P25, P75). The differences in medication regimens, and treatment duration among pediatric TB subtypes were tested by Chi-square test, Fisher’s exact test, or non-parametric Kruskal-Wallis test as appropriate. Logistic regression models examined the associations of antituberculosis medication regimens and duration with unfavorable outcomes and adverse events and were expressed as unadjusted odds ratios (ORs) and 95% confidence intervals (CIs). Potential confounders included data on children BCG vaccination status, body weight and bacteriologic confirmation status. All analyses were performed using Empower® (www.empowerstats.com; X&Y Solutions, Inc., Boston, MA, USA) and R software (http://www.T-project.org). P ≤ 0.05 was considered statistically significant. Results A total of 597 hospitalized children met the inclusion criteria, and their characteristics are summarized in Table 1. About half of them were boys (54%), with median age of 7.58 years (3.00, 12.0). Approximately 11% were low body weight. Table 1Characteristics of pediatric TB in China stratified by antituberculosis medication regimens (N = 597)Characteristicn (%)Medication RegimensP-ValueStandard Core RegimenaSLDs-containing RegimenbOther Regimensc(n = 379)(n = 166)(n = 52) Basic Characteristics Age, years, n (%) < 146 (7.71)24 (6.33)20 (12.0)2 (3.85)< 0.01 1–4160 (26.8)117 (30.9)26 (15.7)17 (32.7) 5–9168 (28.1)118 (31.3)38 (22.9)12 (23.1) 10–14223 (37.4)120 (31.7)82 (49.4)21 (40.4) Gender, n (%) Male320 (53.6)198 (52.2)87 (52.4)35 (67.3)0.12 Female277 (46.4)181 (47.8)79 (47.6)17 (32.7) Low body weight, n (%)65 (10.9)39 (10.3)21 (12.7)5 (9.62)0.69 Clinical Characteristics Contact history, n (%)217 (36.4)149 (39.3)55 (33.1)13 (25.0)0.08 BCG vaccination status, n (%)530 (91.2)332 (90.7)150 (91.5)48 (94.1)0.72 Immunologic test, n (%) PPD Positive311 (80.4)216 (81.8)70 (75.3)25 (83.3)0.36 IGRAs Positive473 (86.2)309 (87.3)123 (83.1)41 (87.2)0.45 Bacteriologic test Result, n (%) Negative271 (48.3)189 (53.5)60 (37.7)22 (44.9)< 0.01 Positive290 (51.7)164 (46.5)99 (62.3)27 (55.1) Bacteriologic Test Type, n (%) AFB stain153/521 (29.4)88/335 (26.3)53/149 (35.6)12/37 (32.4)0.11 MTB culture153/433 (35.3)88/257 (33.5)52/137 (38.0)15/39 (38.5)0.62 MTB nucleic acid test177/384 (46.1)93/236 (39.4)64/118(54.2)20/30 (66.7)0.02 Site of disease, n (%) PTB-only446 (74.7)299 (78.9)103 (62.0)44 (84.6)< 0.01 EPTB (not TBM and OTB)54 (9.05)32 (8.44)19 (11.4)3 (5.77) EPTB (TBM and/or OTB)97 (16.3)48 (12.7)44 (26.5)5 (9.62) Total treatment duration, months, n (%) ≤ 451 (8.56)13 (3.43)14 (8.43)24 (47.1)< 0.01 4–687 (14.6)76 (20.1)6 (3.61)5 (9.80) 6–12228 (38.3)160 (42.2)51 (30.7)17 (33.3) > 12230 (38.6)130 (34.3)95(57.2)5 (9.80)Abbreviations: SLDs: Second line antituberculosis drugs; BCG: Bacillus Calmette-Guérin; PPD: Purified protein derivative; IGRAs: Interferon-gamma release assays; AFB: Acid fast bacteria; MTB: Mycobacterium tuberculosis; PTB: Pulmonary tuberculosis; EPTB: Extrapulmonary tuberculosis; TBM, Tuberculous meningitis; OTB, Osteoarticular tuberculosis Table 1Characteristics of pediatric TB in China stratified by antituberculosis medication regimens (N = 597)Characteristicn (%)Medication RegimensP-ValueStandard Core RegimenaSLDs-containing RegimenbOther Regimensc(n = 379)(n = 166)(n = 52) Basic Characteristics Age, years, n (%) < 146 (7.71)24 (6.33)20 (12.0)2 (3.85)< 0.01 1–4160 (26.8)117 (30.9)26 (15.7)17 (32.7) 5–9168 (28.1)118 (31.3)38 (22.9)12 (23.1) 10–14223 (37.4)120 (31.7)82 (49.4)21 (40.4) Gender, n (%) Male320 (53.6)198 (52.2)87 (52.4)35 (67.3)0.12 Female277 (46.4)181 (47.8)79 (47.6)17 (32.7) Low body weight, n (%)65 (10.9)39 (10.3)21 (12.7)5 (9.62)0.69 Clinical Characteristics Contact history, n (%)217 (36.4)149 (39.3)55 (33.1)13 (25.0)0.08 BCG vaccination status, n (%)530 (91.2)332 (90.7)150 (91.5)48 (94.1)0.72 Immunologic test, n (%) PPD Positive311 (80.4)216 (81.8)70 (75.3)25 (83.3)0.36 IGRAs Positive473 (86.2)309 (87.3)123 (83.1)41 (87.2)0.45 Bacteriologic test Result, n (%) Negative271 (48.3)189 (53.5)60 (37.7)22 (44.9)< 0.01 Positive290 (51.7)164 (46.5)99 (62.3)27 (55.1) Bacteriologic Test Type, n (%) AFB stain153/521 (29.4)88/335 (26.3)53/149 (35.6)12/37 (32.4)0.11 MTB culture153/433 (35.3)88/257 (33.5)52/137 (38.0)15/39 (38.5)0.62 MTB nucleic acid test177/384 (46.1)93/236 (39.4)64/118(54.2)20/30 (66.7)0.02 Site of disease, n (%) PTB-only446 (74.7)299 (78.9)103 (62.0)44 (84.6)< 0.01 EPTB (not TBM and OTB)54 (9.05)32 (8.44)19 (11.4)3 (5.77) EPTB (TBM and/or OTB)97 (16.3)48 (12.7)44 (26.5)5 (9.62) Total treatment duration, months, n (%) ≤ 451 (8.56)13 (3.43)14 (8.43)24 (47.1)< 0.01 4–687 (14.6)76 (20.1)6 (3.61)5 (9.80) 6–12228 (38.3)160 (42.2)51 (30.7)17 (33.3) > 12230 (38.6)130 (34.3)95(57.2)5 (9.80)Abbreviations: SLDs: Second line antituberculosis drugs; BCG: Bacillus Calmette-Guérin; PPD: Purified protein derivative; IGRAs: Interferon-gamma release assays; AFB: Acid fast bacteria; MTB: Mycobacterium tuberculosis; PTB: Pulmonary tuberculosis; EPTB: Extrapulmonary tuberculosis; TBM, Tuberculous meningitis; OTB, Osteoarticular tuberculosis Characteristics of pediatric TB in China stratified by antituberculosis medication regimens (N = 597) Abbreviations: SLDs: Second line antituberculosis drugs; BCG: Bacillus Calmette-Guérin; PPD: Purified protein derivative; IGRAs: Interferon-gamma release assays; AFB: Acid fast bacteria; MTB: Mycobacterium tuberculosis; PTB: Pulmonary tuberculosis; EPTB: Extrapulmonary tuberculosis; TBM, Tuberculous meningitis; OTB, Osteoarticular tuberculosis Over 90% of children were vaccinated with BCG. Of all children, 74.7% were diagnosed with PTB only and 25.3% with EPTB (including 97 cases (64.2% of EPTB) diagnosed with TBM or OTB), 51.7% of them with positive results in pathogen testing (Table 1). Regarding specific TB subtypes, TB mediastinal lymphadenitis (TBML) (39.6%) was the dominant PTB subtypes, and was more prevalent in children PTB children < 5 years old, accounting for 45.8% of those under 1 year and 60.9% of those aged 1–4 years; TBM (42.4%) was the most notably EPTB subtypes, especially children1-4 years old (Fig. 2S, Supplementary Table 2S). Antituberculosis treatment medication regimens and duration Among all patients treated for drug-susceptible TB, 63.5% were on a standard core regimen, 27.8% received a SLDs-containing regimen, and 8.71% were on other regimens (Supplementary Table 3S). Of TB children received a standard core regimen, approximately 30.6% received the standard intensive phase duration of two months while 67.5% had extended treatment during the intensive phase. During the continuation phase, 19.8% of these children received the standard duration of four months while 74.2% experienced extended treatment durations (Fig. 1 and Supplementary Table 3S). Among children with SLDs-containing regimen, about 79.5% experienced an intensive phase lasting longer than two months, and 42.0% of extended continuation phase lasted beyond ten months (Fig. 1 and Supplementary Table 3S). Fig. 1Distribution of treatment duration for different medication regimens in pediatric TB. SLDs-containing regimen as those adding second-line antituberculosis drugs (e.g., linezolid, levofloxacin, moxifloxacin, amikacin, etc.) to the standard core regimen. a: In 30.6% of these patients, the intensive phase drug regimen was 2 months Distribution of treatment duration for different medication regimens in pediatric TB. SLDs-containing regimen as those adding second-line antituberculosis drugs (e.g., linezolid, levofloxacin, moxifloxacin, amikacin, etc.) to the standard core regimen. a: In 30.6% of these patients, the intensive phase drug regimen was 2 months The majority of children with PTB-only (67.0%) treated with a standard core regimen, with the median duration of the intensive phase was 3.00 (2.00, 4.60) months and the median duration of the continuation phase was 6.40 (4.00, 9.00) months. The Kruskal-Wallis test showed significant difference in continuation phase duration between PTB subtypes (P = 0.01). Besides, 23.1% of children with PTB-only were treated with a SLDs-containing regimen, for which the median duration of the intensive phase was 3.00 (2.15, 5.25) months and the median duration of the continuation phase was 9.00 (6.15, 11.6) months. Miliary PTB children was predominantly therapized with adding SLDs (55.6%), with median intensive and continuation periods of 4.60 (3.00, 4.90) and 10.7 (9.00, 12.4) months, respectively (Fig. 2, Supplementary Table 4S and Table 5S). Fig. 2Composition and length of medication regimens for pediatric PTB and EPTB subtypes. (A1, A2) Composition of pediatric PTB and EPTB subtypes in different medication regimens. 446 PTB-only patients exclusive of PTB with EPTB (A1), and 151 individuals with EPTB or both PTB + EPTB (A2). (B1, B2) Treatment duration for standard core Regimen in pediatric TB subtypes. (C1, C2) Treatment duration for SLDs-containing regimen in pediatric TB subtypes. The expected duration of the intensive phase is indicated by a red dashed line (B1, C1) while the expected duration of the continuation phase for uncomplicated TB (4 months) and severe TB (10 months) was indicated by a gray solid line and gray dashed line, respectively (B2, C2). The Kruskal-Wallis test showed significant difference in continuation phase duration between PTB subtypes (P = 0.01). Abbreviations: PTB: Pulmonary tuberculosis; TBML: Tuberculous mediastinal lymphadenitis; EPTB: Extrapulmonary tuberculosis Composition and length of medication regimens for pediatric PTB and EPTB subtypes. (A1, A2) Composition of pediatric PTB and EPTB subtypes in different medication regimens. 446 PTB-only patients exclusive of PTB with EPTB (A1), and 151 individuals with EPTB or both PTB + EPTB (A2). (B1, B2) Treatment duration for standard core Regimen in pediatric TB subtypes. (C1, C2) Treatment duration for SLDs-containing regimen in pediatric TB subtypes. The expected duration of the intensive phase is indicated by a red dashed line (B1, C1) while the expected duration of the continuation phase for uncomplicated TB (4 months) and severe TB (10 months) was indicated by a gray solid line and gray dashed line, respectively (B2, C2). The Kruskal-Wallis test showed significant difference in continuation phase duration between PTB subtypes (P = 0.01). Abbreviations: PTB: Pulmonary tuberculosis; TBML: Tuberculous mediastinal lymphadenitis; EPTB: Extrapulmonary tuberculosis About 59.3% of children with EPTB (excluding TBM and OTB) received a standard core regimen, while nearly half (45.4%) of children with TBM or OTB were treated with SLD-containing regimens (Table 1). The standard core regimens were most frequently utilized in patients with TB lymphadenitis and skin TB. For children with TBM children receiving the SLDs-containing regimen had a median duration of 6.70 (4.10, 8.40) months during the intensive phase, and 11.7 (8.30, 15.1) months during the continuation phases. Overall, the treatment duration for intestinal TB, skin TB, and TB lymphadenitis was relatively short, while the treatment duration for TBM, OTB, and peritoneal TB was significantly longer (Fig. 2, Supplementary Table 4S and Table 5S). Adverse effects A total of 133 (22.3%) children experienced any adverse reactions (Table 2). In the standard core regimen, 62 (16.4%) children experienced adverse reactions, 56 (33.7%) patients in SLDs-containing regimen, and 15 (28.8%) in other regimens. The most common adverse reactions were liver impairment (10.1%), hematotoxicity (6.04%), and gastrointestinal impairment (3.85%). The incidence of liver impairment, renal impairment and hematological toxic reactions in the SLDs-containing regimen group was significantly higher than that in the standard core regimen (P < 0.01). Other regimens were more likely to experience gastrointestinal impairment (P = 0.02) (Table 2). Considering the therapy duration, the incidence of adverse reactions remained around 10% within the 12 months. However, when the treatment duration exceeded 12 months, the adverse reaction rate increased significantly to 40.4%. A similar trend was observed across all specific types of adverse reactions (Table 2). Table 2Adverse events stratified by antituberculosis medication regimens and total durationVariablesn (%)Medication RegimensP-ValueStandard Core RegimenSLDs-containing RegimenOther Regimens(n = 379)(n = 166)(n = 52)Adverse effects, n (%)133 (22.3)62 (16.4)56 (33.7)15 (28.8)< 0.01Liver impairment60 (10.1)24 (6.37)27 (16.3)9 (17.3)< 0.01Renal impairment16 (2.68)10 (2.65)6 (3.61)0 (0.00)0.37Hematotoxicity36 (6.04)15 (3.97)18 (10.8)3 (5.77)< 0.01Ototoxicity2 (0.34)0 (0.00)2 (1.20)0 (0.00)0.07Joint Pain1 (0.17)1 (0.26)0 (0.00)0 (0.00)0.75Peripheral neuropathy2 (0.34)1 (0.26)0 (0.00)1 (1.92)0.10Mental Alterations3 (0.50)1 (0.26)1 (0.60)1 (1.92)0.28Gastrointestinal impairment23 (3.85)8 (2.12)2 (1.20)4 (7.69)0.02Rash16 (2.68)6 (1.59)7 (4.22)3 (5.77)0.77Vision changes8 (1.34)3 (0.79)3 (1.81)2 (3.85)0.17 Variables n (%) Total Duration, months P -Value ≤ 4 4–6 6–12 > 12 ( n = 51) ( n = 87) ( n = 228) ( n = 230) Adverse effects, n (%)133 (22.3)6 (11.8)10 (11.5)23 (10.1)93 (40.4)< 0.01Liver impairment60 (10.1)3 (5.88)8 (9.30)9 (3.95)39 (17.0)< 0.01Renal impairment16 (2.68)0 (0.00)1 (1.15)2 (0.88)13 (5.65)< 0.01Hematotoxicity36 (6.04)0 (0.00)3 (3.45)6 (2.63)26 (11.4)< 0.01Ototoxicity2 (0.34)0 (0.00)0 (0.00)0 (0.00)2 (0.87)0.36Joint Pain1 (0.17)0 (0.00)0 (0.00)0 (0.00)1 (0.43)0.66Peripheral neuropathy2 (0.34)0 (0.00)0 (0.00)2 (0.88)0 (0.00)0.36Mental Alterations3 (0.50)0 (0.00)0 (0.00)2 (0.88)1 (0.43)0.72Gastrointestinal impairment23 (3.86)1 (1.96)1 (1.15)2 (0.88)19 (8.26)< 0.01Rash16 (2.68)2 (3.92)0 (0.00)3 (1.32)11 (4.78)0.04Vision changes8 (1.34)1 (1.96)0 (0.00)3 (1.32)4 (1.74)0.66Abbreviations: SLDs: Second line antituberculosis drugs Table 2Adverse events stratified by antituberculosis medication regimens and total durationVariablesn (%)Medication RegimensP-ValueStandard Core RegimenSLDs-containing RegimenOther Regimens(n = 379)(n = 166)(n = 52)Adverse effects, n (%)133 (22.3)62 (16.4)56 (33.7)15 (28.8)< 0.01Liver impairment60 (10.1)24 (6.37)27 (16.3)9 (17.3)< 0.01Renal impairment16 (2.68)10 (2.65)6 (3.61)0 (0.00)0.37Hematotoxicity36 (6.04)15 (3.97)18 (10.8)3 (5.77)< 0.01Ototoxicity2 (0.34)0 (0.00)2 (1.20)0 (0.00)0.07Joint Pain1 (0.17)1 (0.26)0 (0.00)0 (0.00)0.75Peripheral neuropathy2 (0.34)1 (0.26)0 (0.00)1 (1.92)0.10Mental Alterations3 (0.50)1 (0.26)1 (0.60)1 (1.92)0.28Gastrointestinal impairment23 (3.85)8 (2.12)2 (1.20)4 (7.69)0.02Rash16 (2.68)6 (1.59)7 (4.22)3 (5.77)0.77Vision changes8 (1.34)3 (0.79)3 (1.81)2 (3.85)0.17 Variables n (%) Total Duration, months P -Value ≤ 4 4–6 6–12 > 12 ( n = 51) ( n = 87) ( n = 228) ( n = 230) Adverse effects, n (%)133 (22.3)6 (11.8)10 (11.5)23 (10.1)93 (40.4)< 0.01Liver impairment60 (10.1)3 (5.88)8 (9.30)9 (3.95)39 (17.0)< 0.01Renal impairment16 (2.68)0 (0.00)1 (1.15)2 (0.88)13 (5.65)< 0.01Hematotoxicity36 (6.04)0 (0.00)3 (3.45)6 (2.63)26 (11.4)< 0.01Ototoxicity2 (0.34)0 (0.00)0 (0.00)0 (0.00)2 (0.87)0.36Joint Pain1 (0.17)0 (0.00)0 (0.00)0 (0.00)1 (0.43)0.66Peripheral neuropathy2 (0.34)0 (0.00)0 (0.00)2 (0.88)0 (0.00)0.36Mental Alterations3 (0.50)0 (0.00)0 (0.00)2 (0.88)1 (0.43)0.72Gastrointestinal impairment23 (3.86)1 (1.96)1 (1.15)2 (0.88)19 (8.26)< 0.01Rash16 (2.68)2 (3.92)0 (0.00)3 (1.32)11 (4.78)0.04Vision changes8 (1.34)1 (1.96)0 (0.00)3 (1.32)4 (1.74)0.66Abbreviations: SLDs: Second line antituberculosis drugs Adverse events stratified by antituberculosis medication regimens and total duration Abbreviations: SLDs: Second line antituberculosis drugs Adjustments based on child characteristics stratified by adverse outcomes (Supplementary Table 6S), and multivariate analysis showed that using the SLDs-containing regimen and other regimens significantly increased the occurrence of adverse reactions, with OR values of 2.67 (95% CI: 1.70, 4.20) and 2.00 (0.99, 4.04) respectively. In multivariate analysis, treatment duration exceeding 12 months was identified as an independent risk factor for adverse reactions [OR = 5.52, 95% CI: 2.63, 11.6)], whereas neither age nor disease site demonstrated statistically significant associations (Table 3). Table 3Associations between treatment characteristics and adverse effects in pediatric TB patients (n = 597)VariablesEvent number/Total number (%)UnivariableMultivariableOR (95% CI)OR (95% CI)Age, years ≥ 1117/551 (21.2)ReferenceReference < 116/46 (34.8)1.98 (1.04, 3.75)1.22 (0.56, 2.65)Site of diseasePTB-only85/446 (19.1)ReferenceReferenceWith EPTB (not TBM and OTB)8/54 (14.8)0.74 (0.34, 1.62)0.41 (0.16, 1.07)With TBM or/and OTB40/97 (41.2)2.98 (1.87, 4.76)1.39 (0.80, 2.43)Medication regimensStandard Core regimen62/379(16.4)ReferenceReferenceSLDs-containing regimen56/166 (33.7)2.60 (1.71, 3.97)2.67 (1.70, 4.20)Other regimens15/52 (28.8)2.07 (1.07, 4.01)2.00 (0.99, 4.04)Total duration, months ≤ 616/138 (11.6)ReferenceReference 6–1223/228 (10.1)0.86 (0.44, 1.68)1.13 (0.53, 2.43) > 1293/230 (40.4)5.18 (2.89, 9.28)5.52 (2.63, 11.6)The logistic regression models were expressed as ORs and 95% CI, and multivariable regression adjusted for BCG vaccination status, body weight and bacteriologic confirmation statusAbbreviations: OR, Odd ratio; CI, Confidence interval; PTB: Pulmonary tuberculosis; EPTB: Extrapulmonary tuberculosis; TBM, Tuberculous meningitis; OTB, Osteoarticular tuberculosis; SLDs: Second line antituberculosis drugs; BCG: Bacillus Calmette-Guérin Table 3Associations between treatment characteristics and adverse effects in pediatric TB patients (n = 597)VariablesEvent number/Total number (%)UnivariableMultivariableOR (95% CI)OR (95% CI)Age, years ≥ 1117/551 (21.2)ReferenceReference < 116/46 (34.8)1.98 (1.04, 3.75)1.22 (0.56, 2.65)Site of diseasePTB-only85/446 (19.1)ReferenceReferenceWith EPTB (not TBM and OTB)8/54 (14.8)0.74 (0.34, 1.62)0.41 (0.16, 1.07)With TBM or/and OTB40/97 (41.2)2.98 (1.87, 4.76)1.39 (0.80, 2.43)Medication regimensStandard Core regimen62/379(16.4)ReferenceReferenceSLDs-containing regimen56/166 (33.7)2.60 (1.71, 3.97)2.67 (1.70, 4.20)Other regimens15/52 (28.8)2.07 (1.07, 4.01)2.00 (0.99, 4.04)Total duration, months ≤ 616/138 (11.6)ReferenceReference 6–1223/228 (10.1)0.86 (0.44, 1.68)1.13 (0.53, 2.43) > 1293/230 (40.4)5.18 (2.89, 9.28)5.52 (2.63, 11.6)The logistic regression models were expressed as ORs and 95% CI, and multivariable regression adjusted for BCG vaccination status, body weight and bacteriologic confirmation statusAbbreviations: OR, Odd ratio; CI, Confidence interval; PTB: Pulmonary tuberculosis; EPTB: Extrapulmonary tuberculosis; TBM, Tuberculous meningitis; OTB, Osteoarticular tuberculosis; SLDs: Second line antituberculosis drugs; BCG: Bacillus Calmette-Guérin Associations between treatment characteristics and adverse effects in pediatric TB patients (n = 597) The logistic regression models were expressed as ORs and 95% CI, and multivariable regression adjusted for BCG vaccination status, body weight and bacteriologic confirmation status Abbreviations: OR, Odd ratio; CI, Confidence interval; PTB: Pulmonary tuberculosis; EPTB: Extrapulmonary tuberculosis; TBM, Tuberculous meningitis; OTB, Osteoarticular tuberculosis; SLDs: Second line antituberculosis drugs; BCG: Bacillus Calmette-Guérin Outcome The cohort comprised 527 (88.3%) children demonstrating successful therapeutic outcomes, with 70 cases developing unfavorable treatment. After excluding children with durations ≤ 4 months, the overall treatment effectiveness rate was 96.0%, specifically, 96.5% for PTB, 96.1% for EPTB, and 93.5% for TBM and/or OTB (Fig. 3 and Supplementary Table 7S). In the outcome of PTB-only cases, success rates were 96.9% for standard core regimen, 96.7% for SLDs-containing regimen, and 52.3% for other regimens. PTB-only case with a treatment duration exceeding 6 months achieved a treatment success rate of > 95% (Fig. 3, Supplementary Table 8S and Table 9S). Among EPTB (non-TBM/ OTB) cases, success rates were 97.9% for standard regimen versus 100% for SLDs-containing regimen. Success rates reached 96% in children completing 6–12 months of therapy, whereas extended treatment beyond 12 months demonstrated complete success (Fig. 3, Supplementary Table 8S and Table 9S). In the TBM or/and OTB cohort, standard core regimen achieved 95.8% success versus 93.0% for SLDs-containing regimen. A marked duration-response relationship emerged, with 66.7% vs. 96.0% success rates for ≤ 12-month and > 12-month treatments, respectively (Fig. 3 and Supplementary Table 9S). Fig. 3Proportions of treatment outcomes by different disease sites, medication regimens and total treatment duration. (A) In total patients, treatment success rates approximate 90% across disease sites. Standard core regimen and SLDs-containing regimen demonstrated > 90% success rates, significantly outperforming other regimens. Treatments ≤ 4 months exhibit elevated loss to follow-up rates, whereas extended duration maintain > 90% success rates. (B) In PTB-only patients, success rates across regimens and duration parallel those in the overall patients. (C) In patients with EPTB (not TBM/OTB), no treatment failures or deaths observed. SLDs-containing regimen yields the highest success rate, followed by standard core regimen and other regimens. Duration > 6 months correlate with enhanced success rates. (D) In patients with TBM or/and OTB, standard core regimen achieves 100% success rate, contrasting with other regimens (20%). Treatments > 12 months attain > 95% success rate, substantially exceeding rates for ≤ 12-month duration Proportions of treatment outcomes by different disease sites, medication regimens and total treatment duration. (A) In total patients, treatment success rates approximate 90% across disease sites. Standard core regimen and SLDs-containing regimen demonstrated > 90% success rates, significantly outperforming other regimens. Treatments ≤ 4 months exhibit elevated loss to follow-up rates, whereas extended duration maintain > 90% success rates. (B) In PTB-only patients, success rates across regimens and duration parallel those in the overall patients. (C) In patients with EPTB (not TBM/OTB), no treatment failures or deaths observed. SLDs-containing regimen yields the highest success rate, followed by standard core regimen and other regimens. Duration > 6 months correlate with enhanced success rates. (D) In patients with TBM or/and OTB, standard core regimen achieves 100% success rate, contrasting with other regimens (20%). Treatments > 12 months attain > 95% success rate, substantially exceeding rates for ≤ 12-month duration Discussion In this multi-center retrospective observational study in mainland China, we present the current situation of TB treatment for children, including therapeutic regimen, the occurrence of adverse reactions and the treatment outcomes. We found that PTB-only children were primarily treated with the standard WHO medication regimen, while nearly half of children with EPTB were treated with SLD-containing regimens, especially in TBM or/and OTB. Treatment durations were generally extended. Adverse reactions tend to occur more frequently in children who have received treatment for more than one year, but the therapeutic effect is generally good. Our results suggested that pediatric TB patients in China achieved favorable treatment outcomes. Currently, the treatment of pediatric TB internationally primarily followed WHO guidelines, which recommend a six-month regimen for PTB and TBL, twelve-month regimen for TBM and OTB [6]. However, our research found that treatment durations tend to be prolonged in TB patients. The discrepancies of treatment duration might be related to different guidelines. In contrast, domestic guidelines recommended a treatment duration of 12 months for severe pulmonary tuberculosis (such as miliary TB and bronchial TB), 9 to 12 months for TB pleuritis, 18–24 months for TBM/OTB, and 9 months for superficial lymph node TB [7]. Potential explanations for this observation might involve pathogen strain characteristics; for instance, reports suggest that Beijing strains were associated with high virulence and accelerated disease progression [17]. Another possibility was delayed diagnosis in the pediatric population in China [18], resulting in advanced disease requiring prolonged treatment. Nevertheless, while TBM was characterized by distinct severity grades, lymph node and peritoneal TB didn’t have well-defined staging classifications. Therefore, physicians faced challenges in determining which guideline to follow due to the lack of clear reference, which reflected a deficiency in evidence, particularly the absence of randomized controlled trials for severe PTB and EPTB. Although a 4-month treatment regimen for non-severe PTB had been proposed by the WHO [18], and a 4-month HPZM regimen had been investigated for PTB in individuals ≥ 12 years old [19], the applicability of these shorter regimens in China warranted further investigation through dedicated clinical studies. The selection of SLDs might be influenced by factors such as intolerance to first-line medications, the severity of the disease, or the extent of the lesions. However, the specific reasons required further investigation. Our study found that the incidence of adverse events among children was 22.3%, with rates ranging between 10.0% and 12.0% for those under 12 months of treatment duration. This aligns with the 9% adverse reaction rate reported in Canada and was comparable to the 9% rate observed in recurrent TB populations [19, 20]. However, the adverse event rates increased significantly to 33.7% and 40.4% when SLDs were included or when treatment exceeded 12 months, respectively. These figures were higher than those reported in a recent meta-analysis, which indicated a 26% adverse event rate for WHO-recommended first-line antituberculosis regimens [21]. It suggested that extending treatment beyond 12 months or adding SLDs in children requires careful consideration, weighing the therapeutic benefits against the medication burden. Our research identified liver injury (10.1%), hematologic disorders (6.04%), and gastrointestinal reactions (3.85%) as the three most common adverse events, consistent with other studies reporting liver injury (9%) and gastrointestinal reactions (8%) as frequent adverse effect [21]. However, hematologic disorders are relatively rare with first-line drugs and are primarily associated with SLDs [22], likely due to the use of linezolid. Linezolid was one of the few drugs highly effective against MTB without clear contraindications in children. Nevertheless, it could impair mitochondrial protein synthesis, disrupt the oxidative respiratory chain, and lead to hematologic toxicity. Our effectiveness rate reached 88.3%, surpassing the WHO’s target (85.0%) [23]. The rates in developed countries such as the Netherlands (92%), Australia (89.4%), and Italy (89.9%) ranged from 89% to 92% [5, 24, 25], while those in African countries like Nigeria (79.2%) and Uganda (82%) stood below 85% [26, 27]. Excluding cases with treatment duration of ≤ 4 months, the overall treatment effectiveness rate for pediatric TB from ten TB referral hospitals was 96%; extending treatment beyond 6 months achieved >97% efficacy for PTB and 96% for EPTB (excluding TBM and OTB). These findings suggested that extended treatment duration may be associated with improved outcomes in our cohort. However, as this was an observational study, we cannot rule out confounding by indication, as more severe cases were likely prescribed longer regimens. In China, due to factors such as the social environment, our tolerance for treatment failure and recurrence was relatively low, which might explain our preference for longer treatment durations. However, in this study, prescribing SLDs did not yield superior outcomes compared to standard regimens, after accounting for the potential for more severe disease in the SLDs group. This might be influenced by factors related to the doctors. Generally, children with severe TB were the ones most often prescribed SLDs. The treatment efficacy rate for TBM or/and OTB cases was 93.5% in our present study, reaching 96% for children receiving a ≥ 12 month treatment plan, compared to 90.4% reported in Europe [28]. The treatment results in the other regimen groups were notably poor, predominantly because a significant number of children were lost to follow-up due to the absence of maintenance phase data. It is probable that these children sought medical attention at different hospitals. In our multi-center retrospective study across four regions in China has several limitations. First, as our cases were drawn from inpatients at ten TB referral hospitals, the enrolled patients likely had more severe disease than those treated as outpatients. This introduces selection bias, and limits the generalizability of our findings to less severe outpatient cases. Second, some patients were excluded due to missing case data on intensive medication regimens or treatment duration, which may affect the representativeness of our cohort. Third, the absence of systematic ophthalmic examination during follow-up means that optic neuritis associated with EMB use was likely underreported. Finally, we did not systematically assess hyperuricemia, a known side effect of PZA, and thus could not include it in our analysis of adverse reaction. Conclusion In present multi-center retrospective study, children with drug-sensitive TB were generally on extended duration treatment. Those with PTB-only primarily received a standard core regimen, while EPTB cases were mainly treated with SLDs-containing regimen. The results showed that these regimens achieved favorable treatment outcomes, with success rates for EPTB meeting the WHO target of 85%. Notably, while extended duration and the use of SLDs demonstrated high treatment success rate, they were also associated with an increased risk of adverse effects, particularly when treatment duration exceeded 12 months. This underscores the need to carefully balance efficacy and safety in clinical decision-making. Nonetheless, these findings need to be verified via further high-quality research, especially for EPTB, remains limited. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1