Analysis of prognostic factors and construction of a prediction model for patients with initially treated severe pulmonary tuberculosis

Introduction

Tuberculosis is one of the diseases that seriously endanger human health and impose a substantial socio-economic burden worldwide. The most recent Global Tuberculosis Report released by the World Health Organization (WHO) in 2024 states that an estimated 10.8 million new cases of tuberculosis occurred globally in 2023, with 1.25 million people dying from tuberculosis, making it once again a leading cause of death from a single infectious disease globally (1). The number of deaths caused by tuberculosis is almost twice that of human immunodeficiency virus (HIV)/acquired immunodeficiency syndrome (AIDS).

Severe pulmonary tuberculosis (PTB) (2) refers to active tuberculous lesions occurring in the lung tissue, trachea, bronchi, and pleura, manifesting as severe hypoxemia or acute respiratory failure requiring ventilatory support, or presenting with hypotension, shock, and other signs of circulatory failure, as well as dysfunction of other organs. With the emergence of multidrug-resistant Mycobacterium tuberculosis, co-infection with tuberculosis and HIV, and the impact of the coronavirus disease 2019 (COVID-19) pandemic in recent years, the incidence of severe PTB has shown an upward trend (3). Moreover, the onset time has significantly advanced, and a considerable proportion of patients are already in a state of severe PTB at the time of initial treatment (4,5). Studies have found that delayed treatment and treatment regimen selection are associated with mortality (6-8). Severe PTB progresses rapidly, has a high mortality rate, and is difficult to treat. It imposes a huge economic burden on patients’ families and society, and poses a significant challenge to tuberculosis control and prevention in China, the realization of Healthy China 2030, and the United Nations’ goal of reducing tuberculosis mortality by 95% by 2035. Therefore, it is crucial to reasonably assess the condition of these patients.

Given the severe situation of severe PTB, early prognostic risk assessment and timely intervention are particularly critical. This study conducted an in-depth retrospective analysis of clinical data from patients with newly diagnosed severe PTB, aiming to explore the key factors affecting the prognosis of these patients and to construct a corresponding prognostic prediction model. This model will assist clinicians in accurately assessing patients’ risk of mortality at an early stage, thereby providing a simple and effective clinical decision-making tool for prognostic evaluation, with the goal of optimizing the diagnostic and treatment process and improving patients’ prognosis. We present this article in accordance with the TRIPOD reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1059/rc).

Methods

Study population

A retrospective collection of 347 tuberculosis patients who were observed in the emergency department of Beijing Chest Hospital, Capital Medical University, from January 2024 to October 2024 was conducted. According to the inclusion and exclusion criteria, 189 patients with newly diagnosed severe PTB were enrolled and randomly assigned to a modeling group (n=107) and a validation group (n=82) (Figure 1).

Figure 1 Flow chart of participants included in this analysis.

Inclusion criteria

Confirmed severe PTB:

• Patients aged 18 years and older with active PTB, who have severe radiological findings (see grading standards below).
• A PO₂ <80 mmHg without oxygen therapy or an oxygenation index <300 with oxygen therapy (2,9-11).

Newly diagnosed patients: patients who are diagnosed for the first time and have not received any anti-tuberculosis drug treatment.

Exclusion criteria

• Patients whose underlying cause of death is unrelated to tuberculosis, i.e., patients who died from comorbidities.
• Patients with old tuberculosis.
• Patients with concurrent tumors or immune deficiency.
• Patients with missing clinical data that cannot be used for statistical analysis.
• Patients with unknown treatment outcomes.

Study design

The time of patient admission to the emergency department was considered as the study start point, and the time of patient death was the study endpoint. The observation period for cases was 60 days. Patients in the modeling group were assigned to a survival group (n=73) and a mortality group (n=34) based on their prognosis. Clinical data were collected, including (I) baseline information: age, gender, underlying diseases (diabetes, lung diseases, cardiac insufficiency, cerebral infarction), clinical symptoms (mental status, presence of fever), Mycobacterium tuberculosis etiological results, and rifampicin resistance status; (II) laboratory test results: 25 variables including white blood cell (WBC), hemoglobin (HGB), platelet count (PLT), lymphocyte count (LY), neutrophil-to-lymphocyte ratio (NLR), albumin (ALB), blood urea nitrogen (BUN), creatinine (Cr), sodium (Na), C-reactive protein (CRP), D-dimer, imaging grade, partial pressure of carbon dioxide (PaCO₂), partial pressure of oxygen (PaO₂), respiratory failure, etc. Modeling analysis was used to identify prognostic risk factors in patients with newly diagnosed severe PTB, and a predictive model was established to predict the 60-day prognosis of these patients upon admission. The model was validated using a validation set to assess its predictive value, calibration, and clinical utility. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Beijing Chest Hospital ethics committee (ethical approval No. YJS-2023-10). Written informed consent was acquired from each participant.

Treatment regimens

For sensitive PTB, treatment followed the WHO’s “Consolidated Guidelines on Tuberculosis, Module 4 - Treatment of Drug-sensitive Tuberculosis” (12); for drug-resistant PTB, treatment adhered to the WHO’s “Consolidated Guidelines on Drug-resistant Tuberculosis Treatment” (13). All patients received effective treatment regimens based on drug susceptibility test results and previous treatment history and they do have adhere to the treatment. The anti-tuberculosis treatment plan has been adjusted timely based on the patient’s blood routine, liver and kidney function and gastrointestinal reactions during the whole treatment process. The severe PTB patients may have the problems of hypoalbuminemia, intestinal mucosal edema, intestinal microbiota disorder, which leads to intestinal drug absorption disorders. Therefore, for severe PTB patients with gastrointestinal dysfunction, it is recommended to receive intravenous anti-tuberculosis drug treatment in the early stage. Then gradually transition to oral administration when the condition stabilizes. All patients died in the hospital. They all had received effective anti-tuberculosis treatment until their death and none of them give up treatment.

X-ray or plain chest radiographs were obtained for all recruited patients. There is no universally recognized, accurate and comprehensive definition of severe PTB. So based on a previously published high-quality research in 2023, The radiographic extent of disease was graded into “Mild”, “Moderate” or “Severe” stages according to the extent of lesions and presence of cavitation and pleural effusion. The imaging grading standards (9-11) are presented in Table 1.

Statistical analysis

Statistical analyses were conducted using SPSS version 27. Measurement data following a normal distribution were described as “x¯±s” and analyzed using t-tests. Measurement data not following a normal distribution were described as “M (Q1, Q3)” and analyzed using the Mann-Whitney U test. Categorical data were expressed as counts (percentages) and analyzed using the χ² test. Univariate analysis was initially performed on general data and test indicators within groups, followed by least absolute shrinkage and selection operator (LASSO) regression analysis to identify factors influencing the prognosis of patients with newly diagnosed severe PTB. The factors identified through regression analysis were used to construct a nomogram prediction model with R software. The constructed model was validated using receiver operating characteristic (ROC) curves, decision curve analysis (DCA), and calibration curves. A P value of <0.05 was considered statistically significant.

Results

General data

The modeling group included 107 patients with newly diagnosed severe PTB, comprising 75 males and 32 females, aged 18–92 years, with an average age of 64.45±20.65 years. Among them, 73 patients survived and 34 died, resulting in a mortality rate of 31.78%.

Univariate analysis of prognostic risk factors in patients with newly diagnosed severe PTB

In the modeling group, univariate analysis was performed on the following 25 variables. The results showed that eight indicators, including LY, NLR, ALB, BUN, Na, CRP, respiratory failure, and consciousness disorder, were statistically significant (P<0.05) and were identified as risk factors for mortality in patients with newly diagnosed severe PTB. The remaining 17 factors were not statistically significant. See Table 2 for details. Among the patients, 22.4% had concomitant consciousness disorder, and 61.7% had concomitant respiratory failure. There were no significant differences in underlying diseases between the groups.

LASSO regression analysis results for prognosis of patients with newly diagnosed severe PTB

After identifying statistically significant indicators through univariate analysis, multivariate analysis was conducted. LASSO regression was used as the primary analytical tool due to its advantages in effectively handling high-dimensional data, preventing overfitting, and automatically selecting the most influential features for the predictive variables. The indicators with statistical significance from the univariate analysis were subjected to LASSO regression, using ten-fold cross-validation to select the optimal tuning parameter λ, which was λ=0.0142346121972067. At this point, the model’s mean squared error was minimized, ensuring high predictive accuracy and stability. The results indicated that eight variables, all with non-zero coefficients, were included. The identified prognostic factors influencing mortality in patients with newly diagnosed severe PTB were LY, NLR, ALB, BUN, Na, CRP, respiratory failure, and consciousness disorder. The corresponding non-zero coefficients were −0.23526199, 0.09096384, −0.29879209, 0.08217159, 0.56438757, 0.79732634, 0.62780684, and 1.14600626, respectively (see Figures 2,3).

Figure 2 Dynamic process diagram of variable selection using LASSO. LASSO, least absolute shrinkage and selection operator.

Figure 3 Process diagram of selecting different combinations of the cross-validation parameter λ for included variables.

Establishment of a prognostic prediction model for newly diagnosed patients with severe PTB

Based on the eight influential factors screened by LASSO regression analysis, consciousness state, CRP, respiratory failure, Na, ALB, LY, NLR, BUN are independent influencing factors of patients with initially treated severe PTB. a nomogram model (Figure 4) for predicting the prognosis of newly diagnosed patients with severe PTB was constructed using R software. It includes the points axis, variable axis, total points axis, and mortality risk axis. In the nomogram, each variable is converted to a standardized score ranging from 0 to 100 based on its regression coefficient. The score of continuous variables are changes based on numerical ranges. Categorical variables such as respiratory failure and consciousness state are assigned different scores in a binary classification form. 0 refers to the patient has no respiratory failure or consciousness disorder, while 1 indicates either respiratory failure or consciousness disorder existed. The total score is obtained by adding up the scores of all variables, which is then mapped to the value on the bottom axis to indicate the patient’s risk of mortality.

Figure 4 Nomogram for prognostic prediction model in newly diagnosed patients with severe pulmonary tuberculosis. 0 refers to the patient has no respiratory failure or consciousness disorder, while 1 indicates either respiratory failure or consciousness disorder existed. ALB, albumin; BUN, blood urea nitrogen; CRP, C-reactive protein; LY, lymphocyte count, Na, sodium; NLR, neutrophil to lymphocyte ratio.

Validation of the prognostic prediction model for newly diagnosed patients with severe PTB

The baseline characteristics of patients in the modeling and validation groups are presented in Table 3. The results showed no statistically significant differences between the two groups in terms of age, gender, underlying diseases, clinical manifestations, tuberculosis etiology, and drug resistance, indicating that they are comparable.

The results of the ROC curve analysis showed that the area under the curve (AUC) for the predictive value of the model in the modeling group was 0.8774, with a sensitivity of 88.24% and a specificity of 70.42% (Figure 5A). In the validation group, the AUC was 0.8341, with a sensitivity of 92.86% and a specificity of 72.88% (Figure 5B). These results suggest that the model has good predictive value for the prognosis of newly diagnosed patients with severe PTB in both the modeling and validation groups. The results of the DCA curve analysis indicated that the clinical benefit of the model in both the modeling and validation groups was superior to the “all” (assuming all patients die) or “none” (assuming no patients die) curves (Figure 6). The results of the calibration curve (Figure 7) suggested that the actual and corrected curves of the model fit well and were close to the ideal curve.

Figure 5 ROC curve of nomogram model for predicting the prognosis of newly diagnosed patients with severe pulmonary tuberculosis. (A) Training set; (B) validation set. AUC, area under the curve; ROC, receiver operating characteristic.

Figure 6 DCA curve evaluation of the prognostic nomogram prediction model for newly diagnosed patients with severe pulmonary tuberculosis. (A) Training set; (B) validation set. “All” indicates that all patients died within 60 days of admission, and “None” indicates that no patients died. When the risk threshold is within the range of 0 to 0.6, the net benefit of this model is the highest. DCA, decision curve analysis.

Figure 7 Calibration curve evaluation of the prognostic nomogram prediction model for newly diagnosed patients with severe pulmonary tuberculosis. (A) Training set; (B) validation set.

Discussion

Patients with severe PTB typically require hospitalization or intensive care unit support, yet their mortality rate remains high. Among them, the mortality rate for those with severe PTB complicated by respiratory failure is as high as 69% (11,14,15). In this study, the mortality rate of 107 patients with newly diagnosed severe PTB was still as high as 31.78%, indicating that the problems faced by newly diagnosed severe patients are similarly grave. Additionally, we found that infection markers and other indicators play a significant role in assessing the condition and prognosis of these patients. This study further explores the observed phenomena.

Clinical characteristics related to condition and prognosis

Through a retrospective analysis of patients with newly diagnosed severe PTB, this study identified three core factors influencing short-term survival rates: inflammatory immune response status, homeostasis, and vital organ function.

Firstly, the inflammatory immune response status is a critical factor in determining patient prognosis (11). This is specifically reflected in CRP, LY, and the NLR. As a non-specific acute phase protein, CRP significantly increases in response to infection or inflammation, with its levels closely related to the severity of tuberculosis. This elevation not only indicates bacterial infection but also reflects the body’s strong immune response to pathogens. Such a strong immune response can exacerbate lung tissue damage, leading to alveolar structural impairment, inflammatory cell infiltration, and exudate accumulation, thereby affecting ventilation function. The LY is a key indicator of immune function; low lymphocyte levels suggest immune suppression, which is detrimental to anti-tuberculosis treatment, especially in elderly patients (16,17). Reduced lymphocyte levels can also lead to immune dysregulation, increasing the risk of complications. NLR, an emerging inflammatory marker, indicates a stronger inflammatory response when elevated, reflecting an imbalance between increased neutrophils and decreased lymphocytes. This imbalance is closely related to the systemic inflammatory response caused by Mycobacterium tuberculosis (18-20). Abnormalities in these inflammatory immune response statuses can influence the body’s immune function and the intensity of the inflammatory response, thereby affecting the progression and prognosis of PTB.

Secondly, homeostasis is crucial for maintaining the patient’s life status. Among these, the balance of ALB, BUN, and Na is particularly key. Low ALB levels not only reflect malnutrition but are also closely related to poor prognosis in patients with severe PTB. It affects the transport of nutrients and the maintenance of plasma colloid osmotic pressure, thereby worsening the condition (21-24). Increased BUN concentration may indicate renal impairment, affecting the elimination of toxins in the body and further aggravating the condition (25). Particularly, the balance of Na is vital for fluid stability. PTB patients may experience increased antidiuretic hormone (ADH) secretion due to lung inflammation and infection, reducing free water production and increasing free water reabsorption, leading to water retention in the body and decreased extracellular Na concentration, thus causing hyponatremia. Additionally, PTB patients may experience reduced blood volume due to night sweats and hemoptysis, further stimulating ADH secretion and exacerbating water-Na metabolic disorders. The presence of hyponatremia not only affects normal cellular function but can also lead to a series of severe complications, such as cerebral edema, posing a life-threatening risk (26). Furthermore, hyponatremia may also lead to further immune function decline, making tuberculosis more difficult to control and forming a vicious cycle. Abnormalities in homeostasis, by affecting the body’s metabolic and excretory functions, thus influence the progression and prognosis of PTB.

Lastly, the functional status of vital organs, particularly respiratory failure and consciousness disorder, are direct factors affecting patient prognosis. Respiratory failure is primarily caused by pulmonary parenchymal destruction, inflammatory response, and fibrosis (27). In this study, the incidence of respiratory failure in patients with newly diagnosed severe PTB was high and was closely associated with the risk of death. Consciousness disorders may be caused by central nervous system tuberculosis (28,29), respiratory failure, shock, electrolyte disturbances, among other reasons, making it an independent risk factor for mortality. Abnormalities in these vital organ functions directly threaten patient safety and are crucial indicators for evaluating patient prognosis.

In summary, through an in-depth study of inflammatory immune response status, homeostasis, and vital organ function, we have gained a more comprehensive understanding of the prognostic factors for patients with newly diagnosed severe PTB. This provides valuable treatment decision-making information for clinicians. The close relationship between hyponatremia and tuberculosis prognosis, as well as the impact of inflammatory immune response status and homeostasis on prognosis, provide important references for clinicians to promptly identify and adjust relevant indicators during treatment to improve patient prognosis. Future research should further explore how to integrate these biomarkers and clinical indicators into individualized treatment plans to improve patient survival rates and quality of life and verify the predictive value of these factors in different tuberculosis populations.

Prediction model construction and performance evaluation

In exploring the prognostic factors for patients with newly diagnosed severe PTB, this study skillfully combined multifactorial LASSO regression analysis, nomogram model construction, and rigorous statistical validation methods to construct an accurate and practical prediction model for clinical decision-making.

Firstly, we used multifactorial LASSO regression analysis for feature selection. This method can automatically select the most predictive variables, effectively avoiding overfitting issues that may arise in traditional regression analysis, thus ensuring the model’s stability and generalizability. Based on this, we constructed a nomogram prediction model, translating complex statistical results into an intuitive and understandable scoring system, allowing doctors to quickly quantify the patient’s risk of death in two months based on the scores of eight variables, which is assisting clinical decision-making.

To comprehensively evaluate the model’s effectiveness and reliability, we conducted rigorous statistical validation. ROC curve analysis showed that the AUC of the nomogram model was 0.8774 in the modeling group and 0.8341 in the validation group. This result demonstrates the model’s excellent discriminatory ability, accurately distinguishing between high-risk and low-risk patients. Additionally, the AUC values approaching 0.9 indicate a high level of predictive accuracy, providing strong support for clinical decision-making.

Furthermore, DCA curve analysis revealed the model’s practical value in clinical applications. Compared to the extreme assumptions of “all patients die” and “no patients die”, our prediction model showed higher clinical net benefit in both the modeling and validation groups. It should be noted that although the ideal effect range of the Model curve is relatively narrow, suggesting limited predictive efficacy in patients with extreme risk values. But this does not hinder the model’s outstanding predictive performance within the critical threshold range. Particularly in patients with moderate risk (high-risk threshold 0.2–0.6), its predictive capability remains reliable. These results confirm the scientific validity and effectiveness of our approach. Building on this foundation, we will focus on further optimizing the model and exploring additional potential predictive variables to broaden its application scenarios and enhance its predictive accuracy and generalizability. Additionally, calibration curve results further validated the model’s calibration performance. The close fit between the actual and corrected curves and their proximity to the ideal curve indicate a good consistency between the predicted probabilities and the observed outcomes. This finding not only enhances our confidence in the model’s predictive results but also lays a solid foundation for its widespread application in clinical practice.

Limitations

This study has several limitations. (I) Limited sample size: this study only included 189 patients with newly diagnosed severe PTB, which may limit the generalizability of the model. (II) Single-center design: all data were sourced from a single hospital, which may introduce selection bias and affect the external validity of the study results. (III) Retrospective nature: this study was based on existing data, which may be subject to incomplete information or bias.

Future directions

In the future, we aim to expand the sample size and conduct multicenter studies to enhance the reliability and generalizability of the research results on severe PTB. Additionally, we will adopt a prospective design to collect data, ensuring completeness and accuracy of information. We also plan to validate the model in more external datasets and actively explore its clinical application value, providing scientific evidence and decision support for personalized treatment of severe PTB.

Conclusions

The patients included in this study were all severe PTB patients who had not started regular anti-tuberculosis treatment, with severe conditions and high mortality rates. This study is the first to conduct a preliminary analysis of the prognostic factors for patients with newly diagnosed severe PTB from multiple dimensions and to construct a prognostic prediction model. This model has good sensitivity and specificity, with several indicators that are simple and easy to obtain, providing a preliminary basis for clinical assessment of mortality risk in severe PTB patients. Early identification of risk factors helps improve clinicians’ understanding of severe PTB, enhances monitoring, and allows for early intervention, thereby reducing mortality rates.

In the future, we need to increase the sample size, conduct multi-center and prospective studies to further verify the results, identify patients with a high risk of death as early as possible, improve patient prognosis and optimize resource allocation.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the TRIPOD reporting checklist. Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1059/rc

Data Sharing Statement: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1059/dss

Peer Review File: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1059/prf

Funding: This work was supported by grants from Beijing Public Health Experts Project (No. 2022-3-040), and Beijing Research Ward Excellence Program (No. BRWEP2024W042160119)

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1059/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Beijing Chest Hospital ethics committee (ethical approval No. YJS-2023-10). Written informed consent was acquired from each participant.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. World Health Organization. Global tuberculosis report 2024. Geneva: World Health Organization; 2024.
2. Tuberculosis Professional Committee of Chinese Research Hospital Association. Guidelines for the Guidelines for definition and diagnosis of severe pulmonary tuberculosis in adults in China(2023). Chinese Journal of Evidence-based Medicine 2024;24:1365-75.
3. Duro RP, Figueiredo Dias P, Ferreira AA, et al. Severe Tuberculosis Requiring Intensive Care: A Descriptive Analysis. Crit Care Res Pract 2017;2017:9535463. [Crossref] [PubMed]
4. Liu Q, Li R, Li Q, et al. High levels of plasma S100A9 at admission indicate an increased risk of death in severe tuberculosis patients. J Clin Tuberc Other Mycobact Dis 2021;25:100270. [Crossref] [PubMed]
5. Wu G. Functional amino acids in nutrition and health. Amino Acids 2013;45:407-11. [Crossref] [PubMed]
6. So C, Ling L, Wong WT, et al. Population study on diagnosis, treatment and outcomes of critically ill patients with tuberculosis (2008-2018). Thorax 2023;78:674-81. [Crossref] [PubMed]
7. Erbes R, Oettel K, Raffenberg M, et al. Characteristics and outcome of patients with active pulmonary tuberculosis requiring intensive care. Eur Respir J 2006;27:1223-8. [Crossref] [PubMed]
8. Zahar JR, Azoulay E, Klement E, et al. Delayed treatment contributes to mortality in ICU patients with severe active pulmonary tuberculosis and acute respiratory failure. Intensive Care Med 2001;27:513-20. [Crossref] [PubMed]
9. Berry MP, Graham CM, McNab FW, et al. An interferon-inducible neutrophil-driven blood transcriptional signature in human tuberculosis. Nature 2010;466:973-7. [Crossref] [PubMed]
10. Sousa J, Cá B, Maceiras AR, et al. Mycobacterium tuberculosis associated with severe tuberculosis evades cytosolic surveillance systems and modulates IL-1β production. Nat Commun 2020;11:1949. [Crossref] [PubMed]
11. Wang Y, Sun Q, Zhang Y, et al. Systemic immune dysregulation in severe tuberculosis patients revealed by a single-cell transcriptome atlas. J Infect 2023;86:421-38. [Crossref] [PubMed]
12. WHO consolidated guidelines on tuberculosis: Module 4: Treatment - Drug-susceptible tuberculosis treatment. World Health Organization; 2022.
13. WHO consolidated guidelines on tuberculosis: Module 4: treatment - drug-resistant tuberculosis treatment, 2022 update. World Health Organization; 2022.
14. Götzinger F, Caprile AA, Noguera-Julian A, et al. Clinical presentation, diagnostics, and outcomes of infants with congenital and postnatal tuberculosis: a multicentre cohort study of the Paediatric Tuberculosis Network European Trials Group (ptbnet). Lancet Reg Health Eur 2025;53:101303. [Crossref] [PubMed]
15. Wang L, Hua Y, Wang L, et al. The effects of early mobilization in mechanically ventilated adult ICU patients: systematic review and meta-analysis. Front Med (Lausanne) 2023;10:1202754. [Crossref] [PubMed]
16. Wang W, Wang LF, Liu YY, et al. Value of the Ratio of Monocytes to Lymphocytes for Monitoring Tuberculosis Therapy. Can J Infect Dis Med Microbiol 2019;2019:3270393. [Crossref] [PubMed]
17. La Manna MP, Orlando V, Dieli F, et al. Quantitative and qualitative profiles of circulating monocytes may help identifying tuberculosis infection and disease stages. PLoS One 2017;12:e0171358. [Crossref] [PubMed]
18. Kaushik R, Gupta M, Sharma M, et al. Diagnostic and Prognostic Role of Neutrophil-to-Lymphocyte Ratio in Early and Late Phase of Sepsis. Indian J Crit Care Med 2018;22:660-3. [Crossref] [PubMed]
19. Kartal O, Kartal AT. Value of neutrophil to lymphocyte and platelet to lymphocyte ratios in pneumonia. Bratisl Lek Listy 2017;118:513-6. [Crossref] [PubMed]
20. Kolber W, Kuśnierz-Cabala B, Maraj M, et al. Neutrophil to lymphocyte ratio at the early phase of acute pancreatitis correlates with serum urokinase-type plasminogen activator receptor and interleukin 6 and predicts organ failure. Folia Med Cracov 2018;58:57-74.
21. Tuberculosis Severe Disease Professional Committee of the Chinese Medical Association Tuberculosis Branch. Expert Consensus on Nutritional Therapy for Tuberculosis. Chinese Journal of Tuberculosis and Respiratory Diseases 2020;17-26.
22. Loh WJ, Yu Y, Loo CM, et al. Factors associated with mortality among patients with active pulmonary tuberculosis requiring intensive care. Singapore Med J 2017;58:656-9. [Crossref] [PubMed]
23. Gayaf M, Ayik Türk M, Özdemir Ö, et al. Sociodemographic and clinical risk factors associated with in-hospital tuberculosis mortality in Türkiye, 2008-2018. Tuberk Toraks 2024;72:59-70. [Crossref] [PubMed]
24. Okamura K, Nagata N, Wakamatsu K, et al. Hypoalbuminemia and lymphocytopenia are predictive risk factors for in-hospital mortality in patients with tuberculosis. Intern Med 2013;52:439-44. [Crossref] [PubMed]
25. Chinese Medical Association. Guidelines for the Primary Care Management of Adult Community-Acquired Pneumonia (2018). Chinese Journal of General Practitioners 2019;117-26.
26. Lee P, Ho KK. Hyponatremia in pulmonary TB: evidence of ectopic antidiuretic hormone production. Chest 2010;137:207-8. [Crossref] [PubMed]
27. Tan DTM, See KC. Diagnosis and management of severe pulmonary and extrapulmonary tuberculosis in critically ill patients: A mini review for clinicians. World J Crit Care Med 2024;13:91435. [Crossref] [PubMed]
28. Xiao X, Li Q, Ju Y. Giant central nervous system tuberculoma in pediatric patients: surgical case series. Childs Nerv Syst 2021;37:2935-41. [Crossref] [PubMed]
29. Meregildo Rodriguez ED. Central nervous system tuberculosis following delayed and initially missed lung miliary tuberculosis: a case report. Infez Med 2018;26:270-5.