Impact of SARS-CoV-2 infection history on the clinical features and outcomes of patients with recurrent pulmonary tuberculosis: a retrospective case–control study Authors: Wei Tang, Ke Mi, Haiyan Xue, Yating Liang, Haiming Zhang, Lei Xu, Sha Lin, Yunzhu Long, Zhixiong Fang Published: 2026 ABSTRACT AbstractObjectiveThe impact of prior SARS-CoV-2 infection on the clinical course of recurrent pulmonary tuberculosis (TB) remains poorly defined. This study aimed to compare clinical features, radiographic evolution, and outcomes in recurrent TB patients with and without a history of SARS-CoV-2 infection.MethodsA retrospective case-control study was conducted. Patients diagnosed with recurrent pulmonary TB between January 2019 and January 2025 were consecutively enrolled. A history of SARS-CoV-2 infection was defined by laboratory confirmation (RT-PCR or antigen test). The primary outcome was favorable treatment outcome (cure or treatment completion). The secondary outcome was time to radiographic lung recovery (≥50% lesion resolution). Propensity score matching (PSM), false discovery rate (FDR) correction for multiple comparisons, and multiple imputation by chained equations were applied.ResultsOf 355 patients with recurrent pulmonary TB, 139 had a history of SARS-CoV-2 infection. After PSM (105 matched pairs), the infection history group had a significantly higher proportion of patients undergoing invasive procedures (51.4% vs 36.2%) and a shorter median time to radiographic lung recovery (93 days vs 108 days). The difference in dyspnea (35.2% vs 25.7%) became non-significant after FDR correction (q = 0.12). No significant difference was observed in favorable treatment outcomes (62.9% vs 67.6%). Multivariable regression identified extra-pulmonary TB (adjusted OR = 5.30, 95%CI: 1.91–14.72) and drug resistance (adjusted OR = 11.92, 95%CI: 1.48–96.23) as independent risk factors for unfavorable outcomes, but not a history of SARS-CoV-2 infection (aOR = 1.18, 95%CI: 0.68–2.05, p = 0.55).ConclusionA history of SARS-CoV-2 infection is associated with a greater need for invasive procedures and faster radiographic recovery, but is not independently associated with treatment outcomes. Enhanced monitoring is warranted, but established risk factors remain the primary prognostic determinants.Keywords: Recurrent tuberculosis, SARS-CoV-2 infection, clinical characteristics, radiographic recovery, prognosis Introduction Pulmonary tuberculosis (TB) remains one of the world’s deadliest infectious diseases. According to the World Health Organization, an estimated 10.7 million new TB cases occurred globally in 2024, corresponding to an incidence rate of 131 per 100,000 population. Approximately 1.23 million deaths from TB were reported, resulting in a case fatality rate of 11.5% [1]. Recurrent TB is a significant contributor to ongoing disease transmission and poses a major challenge to TB control efforts [1]. Recurrence may arise either from reactivation of latent endogenous infection or from exogenous reinfection, and it is influenced by multiple factors including host immune status, comorbidities, and socioeconomic conditions [2]. In recent years, the global pandemic of Coronavirus Disease 2019 (COVID-19), caused by Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2), has profoundly impacted public health systems. Substantial evidence indicates that the COVID-19 pandemic severely disrupted global TB diagnosis, treatment, and follow-up services, potentially leading to delayed case detection, treatment interruptions, and consequently an increased risk of recurrence and poor outcomes [3–5]. From a pathophysiological perspective, a complex interplay exists between pulmonary TB and COVID-19. Both diseases primarily target the respiratory tract and can trigger significant immune responses. SARS-CoV-2 infection is known to induce broad immune dysregulation, including lymphocyte depletion, cytokine storms, and impaired adaptive immunity [6,7]. Such immune perturbations may compromise the host’s ability to control Mycobacterium tuberculosis, thereby potentially facilitating reactivation of latent infection or exacerbating active TB [8,9]. Furthermore, acute lung injury and potential long-term pulmonary fibrosis associated with COVID-19 may have additive or synergistic effect with tuberculous lung pathology, leading to more complex clinical presentations and radiographic changes [10,11]. Comorbidities such as diabetes mellitus are established risk factors for both diseases, and their management became more challenging during the pandemic, potentially further increasing the risk of TB recurrence [12,13]. Although clinical observations and epidemiological studies suggest a potential association between SARS-CoV-2 infection and an increased risk of TB recurrence [14], current understanding of how a history of SARS-CoV-2 infection specifically influences the subsequent clinical course of recurrent TB remains very limited. For instance, systematic evidence is lacking on whether such an infection history alters symptom profile, microbiological characteristics, radiographic evolution patterns, treatment requirements, and ultimate outcomes in patients with recurrent TB. Clarifying these issues is crucial for optimizing individualized clinical management of patients with recurrent TB in the “post-pandemic era”. Therefore, this study was designed to systematically compare clinical manifestations, laboratory findings, radiographic features, therapeutic interventions, and prognosis between recurrent pulmonary TB patients with and without a history of SARS-CoV-2 infection, and to explore whether such a history is associated with unfavorable outcomes. Materials and methods Study design and population This single-centre retrospective case-control study consecutively enrolled patients diagnosed with recurrent pulmonary TB between January 2019 and January 2025. Data were retrieved in July 2025 from the Electronic Medical Record (EMR) system using the International Classification of Diseases, Tenth Revision (ICD-10) criteria. Active TB cases included both bacteriologically confirmed and clinically diagnosed cases. Clinical diagnoses were determined through discussion by an in-hospital expert panel comprising TB specialists and radiologists. Patients were divided into two groups based on the presence or absence of a laboratory-confirmed history of SARS-CoV-2 infection (positive RT-PCR or antigen test). Exclusion criteria included concomitant lung cancer and missing key data. This study was approved by the Ethics Committee of the Central Hospital of Xiangtan (Approval No. LY-2025006), which included a waiver of informed consent. The study was conducted in accordance with the principles of the Declaration of Helsinki. Data collection Data collected included baseline demographic information, clinical symptoms, laboratory test results (including complete blood count, inflammatory markers, liver and kidney function, lipid profile, blood glucose, and coagulation function), sputum/bronchoalveolar lavage (BAL) fluid smear results, Mycobacterium tuberculosis DNA, TB drug resistance gene detection, culture and phenotypic drug susceptibility test (DST) results, chest computed tomography (CT) imaging features, and treatment information. Outcomes and definitions The primary outcome of this study focused on the rate of favourable treatment outcomes for pulmonary TB. The secondary outcome included the lung imaging recovery time. A favourable treatment outcome was defined as either cure or treatment completion. Patients categorized as having an unfavourable outcome included those with treatment failure, death, loss to follow-up, or unevaluated outcomes. The Lung imaging recovery time was defined as the time from treatment initiation to the first point at which ≥50% of baseline lung lesions (by extent and density) had resolved or stabilized, as determined by two independent radiologists. And invasive procedures refer to diagnostic or therapeutic interventions that require entry into the body: Bronchoscopy, percutaneous lung biopsy, thoracic drainage, endotracheal intubation/mechanical ventilation, or thoracic surgery (e.g. lobectomy). Radiographic assessment All patients underwent chest CT at baseline, 1, 3, 6, and 12 months after treatment initiation. Two chest radiologists, blinded to SARS-CoV-2 infection history, independently reviewed all images. A third radiologist arbitrated any disagreement. Inter-rater agreement was high (kappa coefficient = 0.84, 95% CI: 0.79–0.89). Lung lesion extent was scored using a grid system (total score 0–20). Recovery was defined as a ≥ 50% reduction from baseline score (Supplementary Materials 2). Statistical analysis Continuous variables are presented as mean ± standard deviation or median (interquartile range). Between-group comparisons were performed using independent samples t-tests or Mann-Whitney U tests. Categorical variables are expressed as frequency (percentage) and were compared using chi-square tests or Fisher’s exact test. For missing data, multiple imputation by chained equations (MICE, 5 imputations, 10 iterations) was employed. The Benjamini–Hochberg false discovery rate (FDR) method was applied to all comparisons of clinical symptoms and key laboratory parameters (q < 0.05 considered significant). To reduce confounding, 1:1 nearest-neighbour matching (caliper = 0.2) without replacement was performed, matching on age, sex, diabetes mellitus, COPD, BMI, and smoking history. After matching, 105 patients from the infection history group were successfully matched to 105 patients from the non-infection group (total n = 210). Standardized differences < 0.1 indicated good balance (Supplementary Materials 3 and 4). Univariate and multivariate logistic regression were used to assess risk factors for unfavorable treatment outcomes. Variables with p < 0.10 in univariate analysis or known clinical importance were entered into the multivariate model. Regression was performed on the matched cohort. All analyses were performed using R software (version 4.2.2), p < 0.05 was considered statistically significant. Results Study population From January 2019 to January 2025, screening identified 2,848 hospitalized pulmonary TB patients based on Electronic Medical Record (EMR) coding. Among them, 139 recurrent pulmonary TB patients were selected as the study group (with a history of SARS-CoV-2 infection), while 216 recurrent pulmonary TB patients (without a history of SARS-CoV-2 infection) were selected as the control group. After propensity score matching (1:1, caliper = 0.2) on age, sex, diabetes, COPD, BMI, and smoking history, 105 pairs (n = 210) were successfully matched. Baseline characteristics were well balanced (all standardized differences < 0.1) (Table 1). All subsequent analyses are based on this matched cohort unless otherwise specified. A detailed flowchart is provided in Supplementary Materials 1. Table 1.Baseline characteristics of recurrent pulmonary TB patients after propensity score matching. History of SARS-CoV-2 infection CharacteristicNo (n = 105)Yes (n = 105)SMD¹p-value²Age (years)61 (53–69)60 (52–68)0.060.71Gender 0.040.74Female24 (22.9%)22 (21.0%) Male81 (77.1%)83 (79.0%) Occupation 0.040.77Manual workers35 (33.3%)37 (35.2%) Non-manual workers70 (66.7%)68 (64.8%) Residential location 0.040.76City31 (29.5%)33 (31.4%) Countryside74 (70.5%)72 (68.6%) BMI (kg/m²)21.3 (19.0–23.6)21.5 (19.2–23.8)0.070.68Vaccination history with BCG 0.100.80No62 (59.0%)64 (60.9%) Unknown25 (23.8%)28 (26.7%) Yes18 (17.2%)13 (12.4%) Smoking history 0.040.78No60 (57.1%)58 (55.2%) Yes45 (42.9%)47 (44.8%) Extra-pulmonary TB 0.110.65No75 (71.4%)80 (76.2%) Yes30 (28.6%)25 (23.8%) Diabetes 0.040.75No79 (75.2%)77 (73.3%) Yes26 (24.8%)28 (26.7%) Hepatitis B 0.140.76No102 (97.1%)99 (94.3%) Yes3 (2.9%)6 (5.7%) HIV positive 0.00>0.999No105 (100%)105 (100%) COPD 0.030.85No90 (85.7%)89 (84.8%) Yes15 (14.3%)16 (15.2%) Data are median (IQR) or n (%). ¹ SMD = standardized mean difference (absolute value); SMD < 0.1 indicates good balance. ² p-value from Wilcoxon rank sum test, chi-squared test, or Fisher’s exact test as appropriate.PTB: pulmonary tuberculosis; COPD: chronic obstructive pulmonary disease. Baseline characteristics of recurrent pulmonary TB patients after propensity score matching. Data are median (IQR) or n (%). SMD = standardized mean difference (absolute value); SMD < 0.1 indicates good balance. p-value from Wilcoxon rank sum test, chi-squared test, or Fisher’s exact test as appropriate. PTB: pulmonary tuberculosis; COPD: chronic obstructive pulmonary disease. Baseline characteristics As shown in Table 1, the two matched groups were comparable across all demographic and clinical characteristics. The median age was 60 years (IQR 52–68) in the infection history group and 61 years (IQR 53–69) in the non-infection group (p = 0.71). Male patients constituted 79.0% of the infection history group and 77.1% of the non-infection group (p = 0.74). Body mass index (BMI) was similar between groups (median 21.5 vs 21.3, p = 0.68). The prevalence of diabetes mellitus (26.7% vs 24.8%, p = 0.75), COPD (15.2% vs 14.3%, p = 0.85), smoking history (44.8% vs 42.9%, p = 0.78), and extra-pulmonary TB (23.8% vs 28.6%, p = 0.65) did not differ significantly between the two groups. All standardized differences were below 0.1, indicating excellent balance. Symptomatology Table 2 presents the symptom profiles of the matched cohort. The most common symptom was cough, reported in 74.3% of the non-infection history group and 76.2% of the infection group (p = 0.75). Expectoration occurred in 61.9% vs 60.0% (p = 0.78), hemoptysis in 21.0% vs 19.0% (p = 0.73), and chest pain in 29.5% vs. 26.7% (p = 0.65). Dyspnoea was more frequent in the infection history group (35.2% vs 25.7% p = 0.048). However, after FDR correction, the difference was no longer statistically significant (q = 0.12). Systemic symptoms—fever (41.9% vs 43.8%, p = 0.78), night sweats (26.7% vs 28.6%, p = 0.76), and weight loss (18.1% vs 16.2%, p = 0.71)—showed no significant differences between groups. Table 2.Symptoms of patients with recurrent pulmonary TB after propensity score matching. History of SARS-CoV-2 infection SymptomNo (N = 105)Yes (N = 105)p-value¹FDR q-value²Cough78 (74.3%)80 (76.2%)0.750.75Expectoration65 (61.9%)63 (60.0%)0.780.78Hemoptysis22 (21.0%)20 (19.0%)0.730.75Chest pain31 (29.5%)28 (26.7%)0.650.74Dyspnea27 (25.7%)37 (35.2%)0.0480.12Fever44 (41.9%)46 (43.8%)0.780.78Night sweats28 (26.7%)30 (28.6%)0.760.78Weight loss19 (18.1%)17 (16.2%)0.710.75Data are n (%). ¹ p-value from Pearson’s chi-squared test. ² False discovery rate (Benjamini–Hochberg) correction for multiple comparisons applied to all symptoms; q < 0.05 considered significant.Bold p-values indicate statistical significance (p < 0.05). Symptoms of patients with recurrent pulmonary TB after propensity score matching. Data are n (%). p-value from Pearson’s chi-squared test. False discovery rate (Benjamini–Hochberg) correction for multiple comparisons applied to all symptoms; q < 0.05 considered significant. Bold p-values indicate statistical significance (p < 0.05). Imaging and laboratory findings Radiographic features and laboratory test results are presented in Tables 3 and 4. Chest CT imaging revealed whole-lung involvement in 52.4% of patients with a history of SARS-CoV-2 infection compared to 46.7% of those without (p = 0.29). Cavity formation was observed in 36.2% vs 27.6% (p = 0.18). Tuberculin skin test positivity rates were similar (61.9% vs 63.8%, p = 0.77). Acid-fast bacilli (AFB) smear positivity was 52.4% vs 45.7% (p = 0.33). Mycobacterium tuberculosis DNA detection was positive in 51.4% vs 61.0% (p = 0.16). Culture positivity trended higher in the infection history group (48.6% vs 38.1%), but the difference did not reach statistical significance (p = 0.12). Drug resistance testing was positive in 37.6% vs 48.1% (p = 0.55). Table 3.Imaging, pathological, and immunological test results of patients with recurrent pulmonary TB after propensity score matching. History of SARS-CoV-2 infection CharacteristicNo (n = 105)Yes (n = 105)p-value [1]Whole-lung involvement 0.29No56 (53.3%)50 (47.6%) Yes49 (46.7%)55 (52.4%) Lung Cavity 0.18No76 (72.4%)67 (63.8%) Yes29 (27.6%)38 (36.2%) Tuberculin skin test 0.77Negative38 (36.2%)40 (38.1%) Positive67 (63.8%)65 (61.9%) Acid-fast staining 0.33Negative57 (54.3%)50 (47.6%) Positive48 (45.7%)55 (52.4%) TB DNA testing 0.16Negative41 (39.0%)51 (48.6%) Positive64 (61.0%)54 (51.4%) TB culture 0.12Negative65 (61.9%)54 (51.4%) Positive40 (38.1%)51 (48.6%) TB drug resistance results 0.55No42/81 (51.9%)*53/85 (62.4%)* Yes39/81 (48.1%)*32/85 (37.6%)* *Denominator varies due to missing results. Data are n (%). 1 Pearson’s Chi-squared test based on matched cohort (n = 210). All q > 0.05 after FDR correction. Imaging, pathological, and immunological test results of patients with recurrent pulmonary TB after propensity score matching. Denominator varies due to missing results. Data are n (%). Pearson’s Chi-squared test based on matched cohort (n = 210). All q > 0.05 after FDR correction. Table 4.Laboratory examination findings of patients with recurrent pulmonary TB after propensity score matching. History of SARS-CoV-2 infection CharacteristicNo (n = 105)Yes (n = 105)p-value [1]RBC (×10¹²/L)4.80 (4.50–5.20)4.84 (4.50–5.40)0.42HB (g/L)145 (135–158)144 (136–156)0.80WBC (×10⁹/L)7.80 (5.90–9.95)7.70 (5.20–9.70)0.31LYM (×10⁹/L)2.60 (2.00–3.80)2.30 (1.90–3.60)0.31NEU (×10⁹/L)4.20 (3.05–5.45)4.20 (2.80–5.30)0.42PLT (×10⁹/L)200 (147–245)201 (160–247)0.59CRP (mg/L)34 (7–66)32 (8–65)0.58ESR (mm/h)54 (25–77)50 (27–76)0.91PCT (μg/L)0.04 (0.02–0.06)0.03 (0.02–0.05)0.16ALT (U/L)33 (21–47)37 (23–47)0.27AST (U/L)35 (22–61)35 (25–61)0.35T-BIL (μmol/L)10.9 (6.7–16.4)10.8 (6.9–15.3)0.53D-BIL (μmol/L)3.40 (2.20–4.85)3.20 (2.20–4.50)0.34TP (g/L)73.3 (69.1–79.1)74.4 (70.5–79.7)0.29ALB (g/L)42.7 ± 4.443.4 ± 4.30.39BUN (mmol/L)5.60 (4.30–7.20)6.10 (4.60–7.30)0.20Scr (μmol/L)73 (63–93)74 (65–88)0.92UA (μmol/L)298 (242–367)305 (256–350)0.89TG (mmol/L)1.46 (0.96–2.26)1.50 (0.97–2.42)0.69TC (mmol/L)4.70 (3.84–5.43)4.84 (4.10–5.69)0.10LDL (mmol/L)2.88 (2.35–3.47)2.94 (2.50–3.35)0.62FPG (mmol/L)5.54 (4.68–6.46)5.65 (4.74–6.65)0.35D-dimer (mg/L)0.47 (0.28–1.54)0.59 (0.31–1.25)0.85PT (s)12.40 (11.80–13.10)12.50 (11.80–13.10)0.62APTT (s)37.1 (34.7–39.6)36.5 (33.9–39.3)0.16FIB (g/L)3.77 (2.98–5.03)3.63 (2.87–4.82)0.56Data are median (Q1, Q3) or mean ± SD. 1 Wilcoxon rank sum test or Welch two-sample t-test based on matched cohort. All q > 0.05 after FDR correction. Laboratory examination findings of patients with recurrent pulmonary TB after propensity score matching. Data are median (Q1, Q3) or mean ± SD. Wilcoxon rank sum test or Welch two-sample t-test based on matched cohort. All q > 0.05 after FDR correction. Routine blood, inflammatory, and biochemical markers were comparable between the two groups. The median white blood cell count was 7.70 × 109/L (IQR 5.20–9.70) in the infection history group vs 7.80 × 109/L (IQR 5.90–9.95) in the non-infection group (p = 0.31). Lymphocyte counts were similar (2.3 vs 2.6 × 109/L, p = 0.31). C-reactive protein levels showed no significant difference (median 32 vs 34 mg/L, p = 0.58). Albumin levels (mean ± SD: 43.4 ± 4.3 vs 42.7 ± 4.4g/L, p = 0.39) and alanine aminotransferase (ALT) (median 37 vs 33 U/L, p = 0.27) were also similar. Treatment interventions and outcomes Table 5 summarizes the treatment interventions and clinical outcomes. The proportion of patients undergoing invasive procedures was significantly higher in the infection history group compared to the non-infection group (51.4% vs 36.2%, p = 0.003). The median hospital stay was 16 days (IQR 11–24) in the infection history group and 14 days (IQR 10–21) in the non-infection group (p = 0.09). Table 5.Treatment interventions and outcomes of patients with recurrent pulmonary TB after propensity score matching. History of SARS-CoV-2 infection CharacteristicNo (n = 105)Yes (n = 105)p-value [1]Invasive procedure 0.003No67 (63.8%)51 (48.6%) Yes38 (36.2%)54 (51.4%) Length of hospital stay (days)14 (10–21)16 (11–24)0.09Lung imaging recovery time (days)108 (86–138)93 (78–122)0.024Treatment outcome 0.47Favorable71 (67.6%)66 (62.9%) Unfavorable34 (32.4%)39 (37.1%) Data are n (%) or median (IQR). 1 Wilcoxon rank sum test or Pearson’s Chi-squared test based on matched cohort. Treatment interventions and outcomes of patients with recurrent pulmonary TB after propensity score matching. Data are n (%) or median (IQR). Wilcoxon rank sum test or Pearson’s Chi-squared test based on matched cohort. The median time to radiographic lung recovery was significantly shorter in the infection history group (93 days, IQR 78–122) than in the non-infection group (108 days, IQR 86–138), with a p-value of 0.024. Despite faster radiographic resolution, the favorable treatment outcome rate (cure or treatment completion) was numerically lower in the infection history group (62.9% vs 67.6%), but this difference was not statistically significant (p = 0.47). Risk factor analysis for unfavourable treatment outcomes Multivariable logistic regression was performed to identify independent risk factors for unfavourable treatment outcomes (Table 6). After adjusting for potential confounders, the presence of extra-pulmonary TB was associated with a significantly increased risk of unfavourable outcomes (adjusted OR = 5.30, 95% CI: 1.91–14.72, p = 0.001). A positive drug resistance test was the strongest risk factor identified (adjusted OR = 11.92, 95% CI: 1.48–96.23, p = 0.020). Elevated aspartate aminotransferase (AST) level, per unit increase, was also modestly associated with worse outcomes (adjusted OR = 1.06, 95% CI: 1.01–1.11, p = 0.018). Table 6.Univariate and multivariate logistic analysis of adverse outcomes after propensity score matching. UnivariateMultivariateCharacteristicOR (95% CI)p-valueaOR (95% CI)p-valueHistory of SARS-CoV-2 infection1.18 (0.68–2.05)0.551.18 (0.68–2.05)0.55Extra-pulmonary TB1.64 (1.02–2.65)0.0425.30 (1.91–14.72)0.001TB drug resistance5.19 (2.79–9.65)<0.00111.92 (1.48–96.23)0.020AST (per U/L increase)1.01 (1.00–1.01)0.0201.06 (1.01–1.11)0.018Age (per year)1.02 (1.00–1.04)0.0231.01 (0.99–1.03)0.34BMI (per kg/m²)0.88 (0.78–0.98)0.0210.92 (0.81–1.04)0.20Diabetes (yes vs no)1.24 (0.73–2.10)0.421.32 (0.77–2.26)0.31Lymphocyte count (per 10⁹/L)0.83 (0.70–0.99)0.0380.87 (0.72–1.05)0.15Hemoptysis (yes vs no)1.79 (1.07–2.98)0.0261.41 (0.82–2.42)0.21Multivariable model adjusted for age, BMI, diabetes, lymphocyte count, hemoptysis, extra-pulmonary TB, drug resistance, and AST.aOR = adjusted odds ratio.Bold p-values indicate statistical significance (p < 0.05). Univariate and multivariate logistic analysis of adverse outcomes after propensity score matching. Multivariable model adjusted for age, BMI, diabetes, lymphocyte count, hemoptysis, extra-pulmonary TB, drug resistance, and AST. aOR = adjusted odds ratio. Bold p-values indicate statistical significance (p < 0.05). Notably, a history of SARS-CoV-2 infection was not independently associated with unfavourable treatment outcomes (adjusted OR = 1.18, 95% CI: 0.68–2.05, p = 0.55). Other variables, including age (adjusted OR = 1.01, p = 0.34) and diabetes (adjusted OR = 1.32, p = 0.31), also did not reach statistical significance. Discussion This retrospective case-control study of 355 patients with recurrent pulmonary TB systematically evaluated the impact of a history of SARS-CoV-2 infection on clinical features, radiographic resolution, and treatment outcomes. After rigorous adjustment for confounding using propensity score matching and multiple imputation, the main findings were as follows: (1) a history of SARS-CoV-2 infection was associated with a significantly higher need for invasive procedures (51.4% vs 36.2%, p = 0.003) and a shorter time to radiographic lung recovery (median 93 vs 108 days, p = 0.024); (2) this history was not independently associated with favorable treatment outcomes (aOR = 1.18, p = 0.55); and (3) established risk factors such as extra-pulmonary TB and drug resistance remained the dominant prognostic determinants. This study found that a history of SARS-CoV-2 infection was associated with a faster time to radiographic resolution of lung lesions. This finding aligns with reports from some studies describing accelerated resolution of pulmonary inflammation following SARS-CoV-2 infection [15]. Potential mechanisms may include the immune activation and cytokine storm triggered by SARS-CoV-2 infection. This state of immune activation might accelerate the subsequent inflammatory resolution and tissue repair of tuberculous lesions. Alternatively, more proactive early clinical interventions for COVID-19 patients could have contributed to the faster absorption of pulmonary lesions [15,16]. Several considerations temper this interpretation. First, faster resolution of opacities does not equal better clinical outcome, as the favourable treatment outcome rate was numerically lower (though not significantly) in the infection history group (62.9% vs 67.6%, p = 0.47). Second, the mechanism underlying faster radiographic resolution remains speculative. One hypothesis is that SARS-CoV-2-induced immune activation—particularly a transient surge in pro-inflammatory cytokines such as IL-6, TNF-α, and IFN-γ—might inadvertently accelerate the clearance of necrotic debris and inflammatory exudates from tuberculous lesions [6,7]. However, such immune perturbations are typically accompanied by T-cell dysregulation and impaired adaptive immunity, which could compromise long-term control of Mycobacterium tuberculosis [7,8]. Starshinova et al. reported that dual infection with SARS-CoV-2 and TB dually affects Th17 cell immune responses, potentially altering the balance between protective immunity and tissue damage [8]. Li et al. further demonstrated that SARS-CoV-2 co-infection downregulates HLA-DR expression on monocytes and reduces Mycobacterium tuberculosis specific CD4+ T-cell proliferation, which may facilitate mycobacterial replication [9]. These immune changes could explain why radiographic recovery appears accelerated while treatment success rates do not improve. Our data did not show a significant difference in cavity formation or whole-lung involvement, leaving the explanation speculative. We therefore recommend that clinicians do not equate faster radiographic improvement with superior prognosis; instead, microbiological and clinical endpoints should guide treatment decisions. Furthermore, although the proportion of pulmonary cavities in the infected group did not reach statistical significance, it showed an increasing trend (36.2% vs. 27.6%). This suggests that SARS-CoV-2 infection might influence the structural healing of TB by exacerbating local tissue destruction or delaying cavity closure. Regarding laboratory testing, the group with a history of SARS-CoV-2 infection exhibited a higher positive rate in Mycobacterium tuberculosis culture, suggesting a potentially higher bacterial load or more active pathogen status. This may stem from the transient immunosuppression induced by SARS-CoV-2 infection, particularly the impairment of T-cell function, which weakens immune control over Mycobacterium tuberculosis. This observation aligns with recent studies proposing that the immunosuppressive state and lymphocyte dysfunction induced by COVID-19 may facilitate mycobacterial replication [17,18]. It is noteworthy that no significant differences were found between the two groups in the positive rates of tuberculin skin tests, sputum smears, or molecular detection, suggesting that SARS-CoV-2 infection might more directly influence the microbial activity reflected by culture, rather than affecting early diagnostic markers. Furthermore, the proportion of patients undergoing invasive procedures was significantly higher in the group with a history of SARS-CoV-2 infection, which aligns with the more complex clinical presentation observed in this group. Residual lung inflammation, fibrosis, or structural alterations following SARS-CoV-2 infection may lead to a suboptimal response to conventional anti-TB therapy or increase the risk of complications (such as hemoptysis or secondary infections), thereby elevating the need for interventions like bronchoscopy, thoracic drainage, or surgery [19,20]. It is noteworthy that despite a faster radiographic resolution in the group with an infection history, the rate of favorable treatment outcomes was numerically lower than in the non-infected group. Although this difference did not reach statistical significance, it suggests that the quality of clinical recovery may not be superior. While there was no significant difference in the length of hospital stay, the increased complexity during the treatment process likely corresponds to higher consumption of medical resources. Multivariate logistic regression analysis revealed that extra-pulmonary TB, positive drug resistance, and elevated AST levels were independent risk factors for unfavorable treatment outcomes, consistent with previous studies on TB prognosis [1,21,22]. It is noteworthy that a history of SARS-CoV-2 infection itself was not identified as an independent risk factor in the adjusted model. This suggests that its influence may operate primarily by exacerbating clinical symptoms and increasing treatment complexity, thereby indirectly affecting patient management, rather than directly determining the final treatment failure. This finding implies that in the clinical management of such patients, greater attention should be paid to the presence of extra-pulmonary TB, drug resistance, and liver function status, rather than overemphasizing the history of SARS-CoV-2 infection itself. Limitations This study has several important limitations. First, SARS-CoV-2 infection history was defined solely by laboratory records without serological confirmation; consequently, asymptomatic infections that were never tested could not be identified, which may introduce exposure misclassification and bias the effect estimates toward the null. Second, due to the retrospective single-center design, data on several key confounders, including COVID-19 severity, specific antiviral or corticosteroid treatments, vaccination status, and the time interval between COVID-19 and TB recurrence, were incomplete. The residual confounding cannot be entirely excluded. Third, we applied false discovery rate correction for multiple comparisons to reduce the risk of type I errors, but the possibility of chance findings cannot be completely ruled out. Fourth, while propensity score matching and multiple imputation were used to address confounding and missing data, the assumptions of no unmeasured confounding and missing at random may not fully hold. Finally, the single-center nature of the study may limit generalizability to other populations or healthcare settings. Future multicenter prospective studies with serological confirmation of past infection, detailed COVID-19 severity grading, standardized imaging protocols, and longer follow-up are needed to validate and extend our findings. Conclusion In summary, a history of SARS-CoV-2 infection in patients with recurrent pulmonary TB is associated with a more resource-intensive clinical course, characterized by a higher need for invasive procedures and faster radiographic lung recovery, but it is not independently associated with unfavorable treatment outcomes. Clinicians should prioritize established risk factors (extra-pulmonary TB, drug resistance, elevated liver enzymes) while maintaining heightened awareness for complex presentations in those with prior COVID-19. Enhanced monitoring and individualized management are warranted, but a history of SARS-CoV-2 infection itself should not be over-interpreted as a direct prognostic determinant. Supplementary Material Supplementary Materials 4 Density plot of propensity score distribution.tif Supplementary Materials 2.docx Supplementary Materials 1.tif Supplementary Materials 3 Love Plot.tif Acknowledgments We gratefully acknowledge the Xiangtan Medical Association for its support of this study. Funding Statement This work was supported by the Scientific Research Project Plan of Xiangtan Medical Association (No. 2024-xtyx-10). Ethics approval This study was approved by the Ethics Committee of the Central Hospital of Xiangtan(Approval No. LY-2025006), which included a waiver of informed consent. Disclosure statement No potential conflict of interest was reported by the authors. Availability of data and materials The datasets used and/or analysed during this study are available from the corresponding author upon reasonable request. References Associated Data Supplementary Materials Supplementary Materials 4 Density plot of propensity score distribution.tif Supplementary Materials 2.docx Supplementary Materials 1.tif Supplementary Materials 3 Love Plot.tif Data Availability Statement The datasets used and/or analysed during this study are available from the corresponding author upon reasonable request.