Before the coronavirus disease pandemic, tuberculosis (TB) was one of the leading causes of mortality due to a single infectious agent. In 2020, there were approximately 10 million cases, nearly 5.8 million new cases of TB and 1.3 million fatalities from TB globally (World Health Organization, 2021), and more than 80% of these were pulmonary TB (PTB) (Cui et al., 2020). In China, according to the Expert Consensus on Diagnosis and Prevention of Inactive Pulmonary Tuberculosis published in China 2021 (Cheng et al., 2021) and the Expert consensus on a standard of activity judgment of pulmonary tuberculosis and its clinical implementation published in China 2020 (Deng and Lu, 2020), the PTB could be divided into active PTB (APTB) and inactive PTB (IPTB). APTB was defined as positive microbiological examination with/without typical symptoms, or clinical diagnosis (Deng and Lu, 2020). And the presence of abnormal stable radiography results in a person with a positive tuberculin skin test (TST) or interferon-release assay (IGRA), negative bacteriologic assay (if performed), and no clinical and/or radiographic evidence of current disease (after excluding other infections) is defined as inactive PTB (IPTB) (Goldberg et al., 1999). The classification is different from the classification provided by Drain PK in 2018, tuberculosis is divided into five categories, named eliminated TB infection, latent TB infection, incipient TB infection, subclinical TB disease and active TB disease, respectively (Drain et al., 2018). For reasons of different treatment regimens and duration, the spectrum of TB disease was divided into APTB and IPTB according to the presence or absence of radiological or microbiological evidence of M. Tuberculosis (Deng and Lu, 2020; Cheng et al., 2021). However, compared with patients with LTBI or incipient TB infection with no pulmonary abnormalities, patients with IPTB have a higher risk of developing APTB (Walter et al., 2014; Gao et al., 2018). China is a country with a high burden of TB. About a quarter of the population is infected with M. tuberculosis, but most of them do not progress to active TB in their lifetime. Therefore, providing TB preventive treatment for IPTB is more cost-effective and targeted, especially in remote areas and economically underdeveloped areas. Additionally, in clinical practice, for IPTB patients, abnormal pulmonary imaging has become the main reason for them to come to the hospital for help. Based on these, the differential diagnosis between APTB and IPTB is crucial for clinical decision-making, and once IPTB is identified, treatment strategies will change significantly. Delayed therapy in patients with APTB due to misdiagnosis results in disease progression, whereas overtreatment in patients with IPTB increases the cost burden, risk of adverse effects, and risk of drug resistance (Goldberg et al., 1999). Hence, it is critical to develop a promising method that can precisely discriminate between APTB and IPTB. Unfortunately, existing studies have focused on discriminating between active TB infection (ATBI) and LTBI, and no studies have aimed to distinguish between APTB and IPTB. Hence, there are still great limitations to distinguishing APTB from IPTB.

Five methods, namely serum biomarker examination, immunological examination, histopathological examination, laboratory microbiological examination, and imaging examination, are valuable for diagnosing APTB. However, these methods have certain limitations. Regarding serum biomarker examination, no single biomarker is effective in determining TB activity; hence, serum biomarker examination produces a low level of evidence in activity judgments. The tuberculin skin test (TST) and interferon-release assay (IGRA) are two major immunological examinations employed and included in the World Health Organization guidelines for the diagnosis of LTBI. However, neither test can distinguish between LTBI and ATBI (Wallis et al., 2013; Lewinsohn et al., 2017; Cohen et al., 2019), let alone APTB and IPTB. Histopathological and laboratory microbiological examinations are the gold standards for diagnosing APTB. However, histopathological examinations, such as invasive biopsies accompanied by trauma, are inconvenient for patients and in clinical practice. In addition, the sensitivity of laboratory microbiological examinations, including smear microscopy (Davies and Pai, 2008), Mycobacterium tuberculosis (MTB) culture (Seifert et al., 2021), and molecular detection by polymerase chain reaction (Noordhoek et al., 1994) or Xpert MTB/RIF (Trébucq et al., 2011; Ponnudurai et al., 2018), is unsatisfactory or limited. Karen et al. demonstrated that 50% of patients had negative results when identifying an etiology (Jacobson, 2017). Moreover, recent studies have demonstrated the TB-specific antigen-to-phytohemagglutinin (TBAg/PHA) ratio for T-SPOT. TB (T-SPOT) assay has potential value in discriminating ATBI from LTBI (Wang et al., 2016; Bosco et al., 2017; Zhou et al., 2017). However, it is difficult to apply this extensively, particularly in low-income countries or areas, because it is an expensive and complex assay. Radiographic examinations play an important role in judging the activity of PTB (Skoura et al., 2015), Chest X-ray is the most used image examination, but it has been gradually replaced by chest computed tomography (CT) due to the low detection rate of small lesions and concealed lesions (Long et al., 1998; Skoura et al., 2015). Chest CT had received more and more attention in diagnosing APTB (Bhalla et al., 2015). Unfortunately, chest CT has some limitations due to its inability to distinguish between approximately 20% of active lesions and 11% of inactive lesions.

In recent years, with the development of analytical approaches, the construction of mathematical models based on multiple markers has been increasingly applied in the field of medicine. A nomogram was used for developing a scoring system to identify malignant pleural effusion, and the scoring system exhibited good diagnostic performance (Wang et al., 2020). Similarly, A promising diagnostic model to discriminate between active tuberculosis and latent tuberculosis infection was constructed based on machine learning (Luo et al., 2022). All approaches combine a series of significant parameters to generate a predictive model to achieve better diagnostic performance.

In a variety of previous studies, routine blood examination indices and routine biochemical indicators commonly used in clinical practice have been proven to be of minor or moderate value in the diagnosis of TB (Luo et al., 2020; Luo et al., 2021a; Luo et al., 2021b). However, a single diagnostic indicator is insufficient owing to its poor sensitivity and specificity. This study attempted to construct a predictive model based on routine laboratory parameters to differentiate between APTB and IPTB. The predictive model was shown to be a nomogram, which was then transformed into a scoring system for use in clinical practice and easy clinical application.

APTB and IPTB represent two PTB states. However, existing laboratory evaluations have limitations, which limit differential diagnosis using conventional methods. Moreover, the clinical treatment and outcomes were dissimilar. Therefore, it is essential to discriminate between these two PTB states. Thus far, research has mainly focused on identifying ATBI and LTBI. Luo (Luo et al., 2020) established a diagnostic model based on the TBAg/PHA ratio and iron metabolism indices for the differential diagnosis of ATBI and LTBI, with a sensitivity and specificity of 88.80% and 91.09%, respectively. Similarly, a predictive model based on routine laboratory indicators to differentiate between ATBI and LTBI, which has high diagnostic ability, exhibited an AUC, sensitivity, and specificity of 0.9880, 92.72%, and 95.99%, respectively (Luo et al., 2022). Although the predictive model has high diagnostic efficiency, it involves 15 indices and is not convenient for clinical practice. Unfortunately, there is no specific method to distinguish between APTB and IPTB. Hence, our study is the first to develop a scoring system based on routine blood examination indices and routine biochemical indicators (MCV, ESR, ALB, ADA, MHR, and HSCLR) to discriminate APTB from IPTB. Moreover, other advantages of the present study were that the indices used in the scoring system were easily obtained, had less time cost than those of conventional laboratory microbiological approaches, and had less cost because the indices involved in the scoring system were routine examinations.

Although there are many differences between APTB and IPTB in terms of clinical characteristics, a single feature has limitations in distinguishing between the two different states of PTB because of its low sensitivity and specificity. Recently, with the growth of clinical data and advanced machine learning methodologies, an increasing number of studies have developed mathematical models based on several markers to improve the diagnostic performance for similar diseases (Wang et al., 2020). In the present study, a scoring system for differentiating APTB from IPTB was developed to enhance the diagnostic possibility of APTB or IPTB more easily for clinicians. Hence, the indices in the scoring system established in this study are easy to access, particularly in economically underdeveloped areas or primary hospitals. In the present study, 52 indices were intergraded, including primary indices and informative ratios commonly used in clinical practice, such as MHR and HSCLR. In recent years, MHR and HSCLR have been newly discovered inflammatory markers. Compared with other inflammatory markers, such as interleukins, MHR and HSCLR are simple, easy to measure, and relatively stable. It is worth noting that previous studies mainly used the C-reactive protein-to-lymphocyte ratio (CLR) as the research focus. The present study considered HSCLR due to the higher sensitivity of high-sensitivity C-reactive protein than that of C-reactive protein. Emerging data suggest that higher MHR and CLR values are associated with various diseases and organ dysfunctions. Recent research has demonstrated that MHR can indicate the prevalence of diabetic retinopathy in patients with type 2 diabetes mellitus (Tang et al., 2021). Higher MHR values represent the highest predictive value for the risk of atherosclerosis (Zhou et al., 2021). In addition, the clinical value of CLR has been researched in previous studies on decompensated cirrhosis, colorectal liver metastases (Taniai et al., 2021), and pneumonia (Cillóniz et al., 2021). Interestingly, there is no literature on MHR and HSCLR in TB patients. Our investigation revealed that MHR and HSCLR had greater diagnostic significance in distinguishing between APTB and IPTB, as evidenced by AUC values of 0.721 and 0.799, respectively.

Previous studies have suggested that MCV is a marker of pulmonary inflammation (Abba, 2009) and that MCV levels are negatively correlated with bacillary load in the lungs (Baruch Baluku et al., 2020). ESR is an inflammatory index, which can reflect the state of TB. The levels of ESR in active TB patients were significantly higher than those in the healthy population. Similarly, ESR levels were significantly higher in the ATB group than in LTBI (Luo et al., 2022). The role of ALB in PTB has attracted increased attention. Hypoalbuminemia is positively associated with the severity of PTB clinical manifestations of PTB (Xu et al., 2021). In addition, ALB levels were significantly lower in the ATB group than in LTBI (Luo et al., 2022). Additionally, a previous study suggested that ADA level has a 98% positive predictive value in high TB prevalence areas (Ding and Zhang, 2018). In addition, ADA activity in ATB patients was significantly higher than that in LTBI patients, and the level of ADA activity was significantly decreased after the completion of anti-TB prophylaxis treatment (Tozkoparan et al., 2007). In addition to the indicators involved in this scoring system, other parameters like hemoglobin, and white blood cell (WBC) have shown a significant difference in our study. These results are consistent with the previous studies. Haemoglobin could be used as a chronic consumption index for active TB (Rathish and Siribaddana, 2018). Additionally, the WBC count in TB patients is significantly higher than in healthy individuals, and the absolute WBC counts decreased significantly during Anti-tuberculosis treatment (Chedid et al., 2020). Neutrophil and lymphocyte are subtypes leukocyte. Previous studies have shown that neutrophil count is an independent predictor of the radiologic severity of PTB at the end of the treatment period (Jones et al., 2021) and lymphocytopenia represented critical inflammatory states (Li et al., 2021).

However, our study had some limitations. First, this study was based on retrospective data, and a selection bias may have existed. Therefore, forward-looking external validation is required. Second, because the participants enrolled in this study were aged>18 years, the performance of the diagnostic model in individuals aged <18 years was unclear. Third, changes in MCV, ALB, and ADA levels after Suspicious were not investigated. Finally, although the scoring system showed relatively excellent performance, the reference range of indicators involved in this model may not be extensive because of confounding factors, such as testing instruments, specimen collection, and testing personnel. Further multicenter validation of this scoring system is required.

In conclusion, MCV, ESR, ALB, ADA, MHR, and HSCLR were significant in differentiating between APTB and IPTB. This scoring system achieved good diagnostic performance and calibration for discriminating APTB from IPTB.

The original contributions presented in the study are included in the article/ Supplementary Material . Further inquiries can be directed to the corresponding authors.

The studies involving human participants were reviewed and approved by the ethics committee of Wuhan Jinyintan Hospital, Tongji Medical College of Huazhong University of Science and Technology. Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.

FG and XW conceived the study. QY, JY, HL, GL, GW, and JM collected patients’ clinical features and radiological and laboratory examination results. QY, WW, and ST were responsible for statistical analysis. QY, JY, and FG interpreted the data and wrote the manuscript. All authors contributed to the article and approved the submitted version.

This work was supported by the National Science and Technology Major Special Foundation of China (2020ZX9201001), Scientific Research Fund of Wuhan in Hubei Province (WX16C33), and Scientific Research Fund of Wuhan in Hubei Province (WX12C02).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.