Miliary tuberculosis with diagnostic challenges in infancy: a case report from Vietnam Authors: Bao Linh Nguyen, Anh Hao Nguyen, Hong-Nhung Nguyen-Thi Published: 2026 ABSTRACT BackgroundMiliary tuberculosis in infants is a life-threatening condition often misdiagnosed due to non-specific symptoms that mimic common childhood illnesses such as bronchiolitis. Obtaining microbiological confirmation is challenging in paucibacillary pediatric cases, frequently delaying life-saving treatment, especially in high-burden settings such as Vietnam.Case presentationWe describe a case of a 3-month-old Vietnamese male infant who presented with progressive wheezing unresponsive to standard bronchiolitis therapy. Although vaccinated with Bacille Calmette–Guérin at birth, no vaccination scar was observed. Chest computed tomography revealed diffuse miliary nodules and necrotic lymphadenopathy, causing severe airway compression. While gastric aspirate samples were negative by both acid-fast bacilli smear and the Xpert Mycobacterium tuberculosis/rifampicin assay, a diagnosis was confirmed after a non-invasive nasopharyngeal swab tested positive for rifampicin-sensitive Mycobacterium tuberculosis using the same molecular assay. The patient showed rapid clinical improvement following the initiation of a 6-month anti-tuberculosis regimen.Discussion and conclusionsThis case illustrates that miliary tuberculosis can be a critical differential diagnosis in infants with persistent wheezing. The successful use of a nasopharyngeal swab for molecular testing in this instance suggests its potential as a valuable, non-invasive diagnostic tool when traditional specimens are negative or difficult to obtain; however, its overall diagnostic yield requires evaluation in larger studies. Furthermore, the development of disseminated disease despite vaccination in an infant lacking a Bacille Calmette–Guérin scar raises questions about host immune responses and suggests a need for further research into the clinical and immunopathological significance of this finding. Background Tuberculosis (TB) remains one of the leading infectious diseases causing death worldwide. According to the World Health Organization (WHO), an estimated 10.8 million people developed TB in 2023, with 1.25 million deaths, including 161,000 among people living with HIV [1]. Despite being preventable and curable, TB continues to pose a major threat to global health, particularly in low- and middle-income countries. Pediatric tuberculosis, especially miliary tuberculosis, remains a significant cause of morbidity and mortality in high-burden countries such as Vietnam [1]. Transmission typically occurs via airborne droplets from adults with active pulmonary TB, particularly in overcrowded and poorly ventilated living conditions. Due to their immature immune systems, young children are at high risk of progression from TB infection to active disease [2], often developing severe forms such as miliary tuberculosis or tuberculous meningitis. Miliary TB, a hematogenously disseminated form of the disease, progresses rapidly and carries a high fatality rate if not diagnosed and treated promptly [3]. The Bacille Calmette–Guérin (BCG) vaccine, administered shortly after birth, plays a crucial role in reducing the risk of severe TB manifestations in children, particularly miliary and meningeal TB. However, its protective effect is not absolute and may wane over time, especially in immunocompromised individuals [4]. Diagnosis of TB in children remains challenging due to nonspecific clinical symptoms, the paucibacillary nature of the disease, difficulty in obtaining respiratory specimens, and a generally low yield of microbiological confirmation. The Xpert MTB/RIF assay offers improved diagnostic yield, with reported sensitivity of 90% (95% CI 54.1–99.5) on sputum and 68.8% (95% CI 53.6–80.9) on gastric aspirate specimens, and specificity consistently exceeding 98.5–99.3%, using culture as the reference standard [5]. Acid-fast bacilli (AFB) smear microscopy remains useful as an initial screening tool, although its sensitivity in children is limited; a meta-analysis estimated that only 6.8% (95% CI 2.2–12.2%) of pediatric TB cases were sputum smear-positive, compared with 52.0% (95% CI 40.0–64.0%) in adults [6]. Despite requiring a longer turnaround time, Mycobacterial Growth Indicator Tube (MGIT) culture system offers higher sensitivity (88%) and facilitates drug susceptibility testing when needed [7]. In addition to microbiological evidence, radiological imaging plays a critical role in identifying the characteristic features of primary tuberculosis in infants, including diffuse bilateral miliary nodules and mediastinal lymphadenopathy with airway compression. While these clinical, radiological, and microbiological manifestations of infantile miliary tuberculosis are well recognized, the potentially novel aspect of our case lies in microbiological confirmation by Xpert MTB/RIF on a nasopharyngeal swab after negative gastric aspirate testing, an approach for which published pediatric data remain limited. Case presentation A 3-month-old Vietnamese male infant with no known history of TB exposure and no family history of TB presented with progressive wheezing and low-grade fever for one week prior to admission. The patient had a history of intermittent wheezing since one month of age, which temporarily improved after nebulization and medication from a private clinic; however, no chest radiography was performed at that time. The symptoms subsequently worsened over time despite continued treatment, leading to respiratory distress and subsequent admission to the Children’s Hospital in Ho Chi Minh City. The patient was born full-term via spontaneous vaginal delivery with a birth weight of 3.4 kg and had received BCG vaccination at birth, though no BCG scar was observed. Our patient lived with both parents. Neither parent had a prior history of TB or known TB risk factors such as malnutrition, HIV, or immunosuppressive therapy. The mother remained asymptomatic throughout pregnancy and delivery, and maternal microbiological testing, chest radiography, and placental examination were not performed at that time. Upon admission to Children’s Hospital, the patient was alert but required supplemental oxygen support, with arterial blood gas analysis performed on FiO₂ 40%, PaO2 106.6 mmHg (Table 1). Physical examination revealed mild subcostal retraction and diffuse wheezing on expiration. No cervical lymphadenopathy, hepatosplenomegaly, or abnormal heart sounds were noted. Diagnostic workup showed elevated white blood cell count (WBC 17.9 K/µL), neutrophils at 28.6%, lymphocytes at 58.5%, and C-reactive protein (CRP) at 27.3 mg/L (Table 1). Blood culture was negative for bacterial growth. Three gastric aspirate specimens were collected during separate fasting periods; AFB smear was performed on two specimens and Xpert MTB/RIF on one, all of which were negative for Mycobacterium tuberculosis. Nasopharyngeal swab Xpert MTB/RIF detected rifampicin-sensitive Mycobacterium tuberculosis. Frontal chest radiography on admission revealed diffuse bilateral micronodular opacities, homogeneous consolidation in the right lower lung zone, mediastinal widening consistent with lymphadenopathy, and narrowing of the left main bronchus (Fig. 1A). Thoracic contrast-enhanced computed tomography (CT) demonstrated three main findings. First, multiple enlarged mediastinal and bilateral hilar lymph nodes showed peripheral rim enhancement with central low attenuation, consistent with caseous necrotic lymphadenopathy (Figs. 2 and 3). Second, the necrotic lymph node conglomerate caused extrinsic compression of the carina, left main bronchus, and bronchus intermedius, providing the anatomical explanation for the patient’s persistent wheezing, which had been refractory to bronchodilator and nebulization therapy (Figs. 2 and 3). Third, the lung parenchyma showed diffuse, bilaterally randomly distributed micronodules and a focal area of necrotizing consolidation in the right lower lobe (Fig. 4). Additional findings included necrotic lymph nodes in the cervical and supraclavicular regions, splenic lesions, and a small right pleural effusion. Echocardiography showed a patent foramen ovale without pericardial effusion. Cranial ultrasound was normal. Table 1Laboratory blood test results on admission at Children’s HospitalParameters (unit)ResultsReferencesLeukocyte (K/µL)17.95.0–13.5Neutrophil (%)28.615–45Neutrophil (K/µL)5.11.0–3.5Lymphocyte (%)58.545–75Lymphocyte (K/µL)10.53.0–8.0Hemoglobin (g/dL)10.29.5–12.8Platelet (K/µL)472150–450D-dimer (µg/mL)3.22< 0.5C-reactive protein (mg/L)27.3< 5Lactate (mmol/L)3.10.7–2.1Aspartate aminotransferase-AST (U/L)49.810–45Alanine aminotransferase-ALT (U/L)19.210–45Urea (mmol/L)1.11.8–5.4Creatinine (µmol/L)27.915–35Glucose (mmol/L)5.43.3–6.1pH7.357.35–7.45Partial pressure of carbon dioxide-PaCO2 (mmHg)35.935–45Partial pressure of oxygen-PaO2 (mmHg)106.680–100Fraction of inspired oxygen-FiO2 (%)4021Bicarbonate-HCO3 (mmol/L)19.222–26 Table 1Laboratory blood test results on admission at Children’s HospitalParameters (unit)ResultsReferencesLeukocyte (K/µL)17.95.0–13.5Neutrophil (%)28.615–45Neutrophil (K/µL)5.11.0–3.5Lymphocyte (%)58.545–75Lymphocyte (K/µL)10.53.0–8.0Hemoglobin (g/dL)10.29.5–12.8Platelet (K/µL)472150–450D-dimer (µg/mL)3.22< 0.5C-reactive protein (mg/L)27.3< 5Lactate (mmol/L)3.10.7–2.1Aspartate aminotransferase-AST (U/L)49.810–45Alanine aminotransferase-ALT (U/L)19.210–45Urea (mmol/L)1.11.8–5.4Creatinine (µmol/L)27.915–35Glucose (mmol/L)5.43.3–6.1pH7.357.35–7.45Partial pressure of carbon dioxide-PaCO2 (mmHg)35.935–45Partial pressure of oxygen-PaO2 (mmHg)106.680–100Fraction of inspired oxygen-FiO2 (%)4021Bicarbonate-HCO3 (mmol/L)19.222–26 Laboratory blood test results on admission at Children’s Hospital Fig. 1Frontal chest radiographs on admission (A) and after completion of anti-tuberculosis therapy (B). Legend: (A) Diffuse bilateral 1–3 mm random micronodules (white arrowheads), homogeneous consolidation in the right lower zone (white arrow), widening of the right paratracheal stripe consistent with mediastinal lymphadenopathy (black arrow), and narrowing of the left main bronchus (black arrowhead). B Marked radiological resolution of miliary nodules, consolidation, and mediastinal widening after 6 months of treatment Frontal chest radiographs on admission (A) and after completion of anti-tuberculosis therapy (B). Legend: (A) Diffuse bilateral 1–3 mm random micronodules (white arrowheads), homogeneous consolidation in the right lower zone (white arrow), widening of the right paratracheal stripe consistent with mediastinal lymphadenopathy (black arrow), and narrowing of the left main bronchus (black arrowhead). B Marked radiological resolution of miliary nodules, consolidation, and mediastinal widening after 6 months of treatment Fig. 2Coronal reconstruction of contrast-enhanced chest CT (lung window). Legend: Marked narrowing of the left main bronchus (arrow) caused by extrinsic compression from an adjacent necrotic mediastinal lymph node conglomerate (asterisk) Coronal reconstruction of contrast-enhanced chest CT (lung window). Legend: Marked narrowing of the left main bronchus (arrow) caused by extrinsic compression from an adjacent necrotic mediastinal lymph node conglomerate (asterisk) Fig. 3Axial contrast-enhanced chest CT (mediastinal window) at the level of the carina. Legend: Multiple enlarged mediastinal and bilateral hilar lymph nodes (white arrows) demonstrate peripheral rim enhancement with central low attenuation, characteristic of caseous necrotic lymphadenopathy. The lymph node conglomerate extrinsically compresses the carina and both main bronchi (black arrows) Axial contrast-enhanced chest CT (mediastinal window) at the level of the carina. Legend: Multiple enlarged mediastinal and bilateral hilar lymph nodes (white arrows) demonstrate peripheral rim enhancement with central low attenuation, characteristic of caseous necrotic lymphadenopathy. The lymph node conglomerate extrinsically compresses the carina and both main bronchi (black arrows) Fig. 4Axial chest CT (lung window). Legend: Diffuse, bilateral, randomly distributed 1–3 mm micronodules (white arrowheads) — the characteristic “miliary” pattern of hematogenous dissemination — with confluent necrotizing consolidation in the right lower lobe (black arrowheads) Axial chest CT (lung window). Legend: Diffuse, bilateral, randomly distributed 1–3 mm micronodules (white arrowheads) — the characteristic “miliary” pattern of hematogenous dissemination — with confluent necrotizing consolidation in the right lower lobe (black arrowheads) Following confirmation of pulmonary tuberculosis, the patient was transferred to Pham Ngoc Thach Hospital, the national tertiary referral hospital for tuberculosis and pulmonary diseases in southern Vietnam. At our hospital, a lumbar puncture was performed. The cerebrospinal fluid (CSF) analysis showed a protein level of 0.66 g/L, glucose at 2.77 mmol/L, and a leukocyte count of 6 cells/µL (100% lymphocytes). CSF AFB smears, bacterial/fungal cultures, and Xpert MTB/RIF were all negative (Table 2). Brain magnetic resonance imaging (MRI) was not performed due to the risk of respiratory failure from sedation, the lack of pediatric anesthesia protocols at our hospital, which primarily serves adult patients, and because this imaging was not covered by health insurance for a neurologically asymptomatic patient. Anti-tuberculosis treatment was initiated; the patient weighed 6 kg and received the standard 6-month regimen 2HRZE/4HR (comprising isoniazid (H), rifampicin (R), pyrazinamide (Z), and ethambutol (E)) according to the Vietnam National Tuberculosis Program (VNTP) pediatric guidelines, consisting of one fixed-dose combination (FDC) dispersible tablet of RHZ 75/50/150 mg plus one tablet of ethambutol 100 mg daily during the intensive phase, followed by one FDC dispersible tablet of RH 75/50 mg daily for 4 months during the maintenance phase. Pyridoxine and adjunctive corticosteroids were not administered. Table 2Laboratory tests and microbiology of cerebrospinal fluidParameters (unit)ResultsReferencesProtein (g/L)0.660.55–1.05Glucose (mmol/L)2.772.0–5.6AFB stain and cultureNegativeNegativeXpert MTB/RIFNegativeNegativeBacterial cultureNegativeNegativeFungus cultureNegativeNegativeLeukocyte – cells/µL (%lymphocyte)6 (100%)< 9 (100%) Table 2Laboratory tests and microbiology of cerebrospinal fluidParameters (unit)ResultsReferencesProtein (g/L)0.660.55–1.05Glucose (mmol/L)2.772.0–5.6AFB stain and cultureNegativeNegativeXpert MTB/RIFNegativeNegativeBacterial cultureNegativeNegativeFungus cultureNegativeNegativeLeukocyte – cells/µL (%lymphocyte)6 (100%)< 9 (100%) Laboratory tests and microbiology of cerebrospinal fluid Within 72 h of treatment initiation, the patient showed significant clinical improvement, becoming alert, feeding well, and afebrile, with resolution of wheezing, marked reduction in cough, and complete cessation of oxygen support. The patient was discharged after one week of clinical stabilization. At this time, the mother also underwent TB screening, including symptom evaluation, chest radiography, and sputum Xpert MTB/RIF, all of which were negative. She was subsequently monitored for respiratory symptoms during the child’s monthly follow-up visits and remained free of TB-related symptoms throughout the 6-month treatment course. He was subsequently followed monthly at the local TB clinic according to VNTP guidelines, including anthropometric assessment, clinical examination, complete blood count, renal and liver function tests, and screening for visual adverse effects. The patient completed the full 6-month regimen without reported adverse events, with steady weight gain and progressive clinical improvement. Follow-up frontal chest radiographs obtained at 2 and 6 months demonstrated progressive radiological improvement. The end-of-treatment radiograph at 6 months (Fig. 4B) confirmed marked resolution of miliary nodules, consolidation, and mediastinal widening. Discussion and conclusions Miliary tuberculosis is a rare but severe form of disseminated TB resulting from hematogenous spread of Mycobacterium tuberculosis. It occurs more frequently in infants and young children due to their immature immune systems and higher risk of progression from infection to active disease. Clinical presentation in pediatric miliary TB is typically nonspecific, with symptoms such as low-grade fever, respiratory distress, or failure to thrive, often leading to delayed diagnosis. Despite extensive pulmonary and mediastinal disease, our patient exhibited only subtle clinical manifestations that persisted for several weeks before diagnosis, highlighting the diagnostic challenges of infantile miliary tuberculosis, particularly in primary care settings where wheezing is often attributed to more common respiratory conditions. Bronchiolitis is the most common cause of wheezing in young infants and was initially presumed in our patient, resulting in treatment with nebulized bronchodilators without chest imaging. However, persistent wheezing despite standard therapy should prompt consideration of tuberculosis and other alternative diagnoses, particularly in endemic settings. In our patient, chest CT demonstrated diffuse miliary nodules, consolidation, and necrotic mediastinal lymphadenopathy with airway compression, findings strongly suggestive of severe pulmonary tuberculosis. In conjunction with the clinical presentation and absence of features suggestive of aspiration or gastroesophageal reflux, these findings made alternative causes of infant wheeze, including bronchiolitis, pertussis, aspiration, and congenital airway abnormalities, less likely. Viral and pertussis-specific testing were not performed due to limited availability and the strong radiologic-clinical concordance supporting tuberculosis as the leading diagnosis. Microbiological confirmation of tuberculosis is particularly challenging in young infants, in whom history-taking is limited and obtaining lower respiratory tract or gastric aspirate specimens is technically difficult, distressing, and often of low yield. In our patient, three gastric aspirates were negative, and confirmation was ultimately achieved through a nasopharyngeal swab Xpert MTB/RIF. To our knowledge, published pediatric data on Xpert MTB/RIF performance on nasopharyngeal swab specimens are lacking. As an indirect reference, one study reported Xpert sensitivity of 22% (95% CI, 15%–32%) and specificity of 100% (95% CI, 91%–100%) on oral swab specimens [8], which sample a different anatomical site from nasopharyngeal swabs. The nasopharyngeal swab collection technique was identical to that used for COVID-19 real-time Polymerase Chain Reaction (PCR) testing in Vietnam — a faster, less invasive procedure than gastric aspiration. BCG vaccination provides substantial protection against severe forms of childhood TB, including miliary tuberculosis and tuberculous meningitis, but this protection is incomplete [9]. Despite receiving BCG vaccination at birth, our patient developed disseminated tuberculosis. Notably, no BCG scar was observed at admission, although BCG scars typically develop within 6–8 weeks after vaccination. In a retrospective study, Liao et al. reported that vaccinated children without a BCG scar remained at lower risk of miliary tuberculosis or tuberculous meningitis than unvaccinated children [10], although direct comparisons between scar-positive and scar-negative vaccinated children were not performed. The clinical significance of BCG scar absence remains uncertain, and scar formation may be influenced by factors unrelated to host immunity, including injection technique, vaccine strain, and timing of vaccination. A causal relationship between scar absence and the development of disseminated tuberculosis therefore cannot be established in this case. Nevertheless, two non-mutually exclusive hypotheses may be considered. First, reduced immunogenicity following BCG vaccination may have contributed to susceptibility to severe tuberculosis. Second, congenital or early postnatal infection with Mycobacterium tuberculosis may have interfered with the development of a local vaccine response, thereby preventing scar formation. Consistent with the first hypothesis, ineffective host immunity can lead to poorly formed granulomas and inadequate containment of Mycobacterium tuberculosis, resulting in hematogenous dissemination — a mechanism most extensively documented in HIV-associated tuberculosis [11, 12]. The chest CT findings in our patient — extensive miliary nodules and necrotic mediastinal lymphadenopathy — are radiologically consistent with this disseminated pattern [13, 14]. Further studies are needed to clarify the prevalence, determinants, and clinical significance of absent BCG scars in infants with early tuberculosis exposure. Once tuberculosis is suspected or confirmed in infants, assessing for central nervous system (CNS) involvement is an essential step. Contrast-enhanced brain MRI is more sensitive than CSF examination for detecting CNS lesions in miliary tuberculosis [15]. However, brain MRI was not feasible in our patient due to three factors: the sedation risk in an infant with severe airway compression, the lack of pediatric anesthesia protocols at our predominantly adult-focused hospital, and limited insurance coverage for brain MRI in patients without neurological manifestations. CNS involvement was therefore assessed through a normal neurological examination, normal CSF analysis, and normal cranial ultrasound performed through the anterior fontanelle. Cranial ultrasound is less sensitive than MRI for meningeal enhancement, tuberculomas, and small infarcts that may characterize CNS tuberculosis. To our knowledge, no well-designed study has determined the sensitivity of cranial ultrasound, or of cranial ultrasound combined with CSF analysis, for the diagnosis of CNS tuberculosis in infants. Although these concordantly normal findings reduced the likelihood of CNS involvement, they cannot fully exclude it; close clinical follow-up was therefore maintained throughout treatment. Immune-based diagnostic assays such as interferon-gamma release assay (IGRA) were not performed in this case due to limited resources. In young infants, both IGRA and tuberculin skin testing are known to have suboptimal performance due to immature immune responses, limiting their diagnostic utility in this age group. Sensitivity of IGRA in children under 12 months has been reported to be approximately 80% [16], although both IGRA and TST are affected by immunologic immaturity. In high tuberculosis burden settings, clinical judgment, supported by radiological and microbiological findings, remains central to diagnosis. The 6-month regimen (2HRZE/4HR) was selected in accordance with both the WHO 2022 pediatric TB guideline [17] and the VNTP 2024 guideline for severe pediatric pulmonary tuberculosis, as the patient presented with miliary disease, extensive intrathoracic lymphadenopathy with airway compression, extensive parenchymal involvement, and clinical severity requiring hospitalization. The shorter 4-month WHO regimen is restricted to non-severe drug-susceptible pulmonary tuberculosis and is therefore not applicable in this case. Given the extensive intrathoracic lymphadenopathy and airway compression, adjunctive corticosteroid therapy was considered. Both the WHO 2022 and VNTP 2024 pediatric TB guidelines recommend corticosteroids for tuberculous meningitis and pericarditis; the WHO guideline additionally considers their use in selected cases of severe pulmonary TB with airway compression or extensive intrathoracic lymphadenopathy, leaving decisions in other manifestations to clinical judgment. In our case, the decision not to initiate corticosteroids was based on the patient’s rapid clinical improvement within the first 72 h of anti-tuberculosis therapy, before being transferred to our team’s care. The patient subsequently completed the full treatment course with normal growth and marked radiological resolution. The WHO 2022 pediatric TB guideline recommends pyridoxine supplementation in exclusively breastfed infants receiving isoniazid. The VNTP 2024 guideline primarily recommends pyridoxine for children receiving high-dose isoniazid or with established risk factors for peripheral neuropathy (e.g., malnutrition, HIV co-infection), leaving use in other settings to clinical judgment. Our patient was exclusively breastfed but received standard-dose isoniazid without additional risk factors. Following the VNTP guidance, pyridoxine was not administered. Biochemical assessment of pyridoxine status was not available in our setting; however, the patient completed the full treatment course without clinical manifestations of pyridoxine deficiency, with normal growth throughout follow-up. Post-discharge management followed the VNTP community-based monitoring framework, with monthly follow-up at local TB clinics. Referral back to the tertiary center was indicated for clinical deterioration, laboratory abnormalities, drug-related adverse events, or at the caregiver’s preference. In our case, the caregiver elected to bring the patient for re-evaluation at our institution at the end of the intensive phase and end of treatment. For our patient’s mother, antenatal tuberculosis screening was conducted using the standardized symptom-based approach recommended by both the WHO and the Vietnam National Tuberculosis Program (VNTP) for asymptomatic pregnant women in high-burden settings. Neither guideline recommends routine microbiological or radiographic testing during pregnancy in the absence of symptoms or recognized risk factors. However, symptom-based screening has limited sensitivity: a retrospective study in Sweden found that 67% (6/9) of pregnant women with bacteriologically confirmed tuberculosis were asymptomatic [18]. Placental histopathology was not performed because tuberculosis was not suspected at delivery. Following the infant’s diagnosis, the mother—who remained asymptomatic—underwent tuberculosis symptom screening, chest radiography, and sputum Xpert MTB/RIF testing, all of which were negative. Under the VNTP framework, household contact investigation is conducted by local healthcare services after discharge; in this case, only the mother underwent formal evaluation, while the father and other household contacts were not formally investigated. Taken together, the reliance on symptom-based antenatal screening, the absence of placental histopathology, and the incomplete household contact investigation represent important limitations of this case. Several limitations should be acknowledged, as discussed above. Granular Xpert MTB/RIF data (semi-quantitative load and cycle-threshold values) and mycobacterial culture from the nasopharyngeal swab were unavailable. In addition, repeat Xpert testing, brain MRI, and biochemical assessment of pyridoxine status were not feasible. Comprehensive evaluation of maternal tuberculosis status at delivery and placental histopathology was not performed, nor were the father and other household contacts formally investigated. This case highlights the diagnostic and therapeutic challenges of miliary tuberculosis in infants, particularly in the absence of specific symptoms, known TB exposure, and the lack of early chest radiography. Despite BCG vaccination, ineffective immune protection may still result in disseminated disease. Advanced imaging, especially chest CT, plays an adjunctive role in identifying characteristic patterns of miliary TB and associated complications such as airway compression or other organ lesions. In our patient, nasopharyngeal swab Xpert MTB/RIF provided timely microbiological confirmation; this non-invasive and accessible approach may potentially be valuable in settings where traditional specimens are negative. Early clinical suspicion, integration of radiologic and molecular tools, and prompt initiation of therapy are essential for improving outcomes in pediatric disseminated TB. This case also underscores the need for further research into the clinical significance of absent BCG scarring and the diagnostic performance of nasopharyngeal swab Xpert MTB/RIF testing in young children.