Tuberculin responsiveness as a surrogate prognostic marker in hospitalized HIV-infected patients with tuberculosis in a resource-limited setting

Abstract

Introduction

In resource-limited settings, access to CD4 lymphocyte counts for prognostication in HIV-associated tuberculosis (HIV-TB) remains limited. Tuberculin skin test (TST) reactivity reflects cell-mediated immunity and may serve as a simple surrogate prognostic marker. This study evaluated the association between tuberculin responsiveness and clinical outcomes in hospitalized HIV-infected patients with pulmonary tuberculosis.

Materials and Methods

A hospital-based observational cohort study was conducted among 107 HIV-positive adults admitted with pulmonary tuberculosis. Tuberculin reactivity was assessed using the Mantoux method with 10 TU purified protein derivative. Nutritional status was assessed using body mass index (BMI). The primary outcome was clinical status at hospital discharge. A parallel cohort of 100 HIV-negative pulmonary tuberculosis patients was evaluated using identical methods.

Results

Among HIV-positive patients, 33 (30.8%) demonstrated TST induration >10 mm; all had mild-moderate undernutrition and were discharged alive within six weeks. Twenty-six patients (24.3%) had induration 5–10 mm; all were severely undernourished, with 23.1% mortality. Forty-eight patients (44.9%) were anergic or had induration <5 mm; 62.5% died and 25.0% left moribund. Increasing induration showed a graded inverse association with mortality (p < 0.001). HIV-negative patients demonstrated substantially lower overall mortality (3% vs 33.6%). While preserved TST reactivity (>10 mm) predicted 100% survival in both groups, anergy carried dramatically different implications: 5.7% mortality in HIV-negative versus 62.5% in HIV-positive patients.

Conclusions

Tuberculin reactivity demonstrated a strong, HIV-specific graded association with outcomes. TST may retain adjunctive prognostic value in resource-limited settings where advanced immunological testing is unavailable.

Highlights

• •TST reactivity strongly predicted in-hospital mortality in HIV-TB coinfection.
• •11-fold mortality difference between HIV+ and HIV− TB patients (34% vs 3%).
• •Tuberculin anergy signals irreversible immunosuppression in HIV, not malnutrition.
• •TST >10 mm indicated 100% survival in both HIV+ and HIV− patients.
• •Simple bedside TST provides prognostic value in resource-limited settings.

1. Introduction

Tuberculosis (TB) remains the leading cause of mortality among people living with HIV, with the burden concentrated in low- and middle-income countries such as India [1], [2]. Hospitalized HIV-TB patients often present with advanced disease and experience poor short-term outcomes, highlighting the need for simple bedside prognostic markers in resource-limited settings [3], [4].

Progressive HIV infection impairs CD4+ T-lymphocyte function, resulting in diminished delayed-type hypersensitivity responses [5]. The tuberculin skin test (TST), although limited diagnostically in HIV due to frequent anergy [6], [7], may reflect residual immune competence rather than absence of infection.

We hypothesized that the magnitude of tuberculin responsiveness could serve as a pragmatic surrogate marker of immune reserve and short-term prognosis in hospitalized HIV-TB patients. This study evaluated the association between TST reactivity, nutritional status, and in-hospital outcomes in a cohort from pre-ART-era India. Such historical cohorts provide valuable benchmarks for understanding immune reserve and mortality patterns in late-presenting HIV-TB patients.

2. Methods

2.1. Study design and setting

This observational cohort study was conducted at Rangaraya Medical College and Government General Hospital, Kakinada, Andhra Pradesh, India, between January 1998 and October 1999, during the pre-antiretroviral therapy era when CD4 testing and viral load estimation were not routinely available. The present manuscript represents a secondary analysis of clinical data prospectively collected as part of a postgraduate dissertation.

2.2. Study population

Consecutive HIV-seropositive adult patients admitted with pulmonary tuberculosis were eligible for inclusion. HIV infection was confirmed using the standard two-test sequential serological algorithm recommended in India during the study period, consisting of an initial ELISA screening test followed by a second ELISA employing a different antigen preparation or confirmatory Western blot testing in reactive cases. Formal WHO clinical staging was not routinely documented and therefore could not be reconstructed retrospectively. However, based on the high prevalence of constitutional symptoms, advanced pulmonary tuberculosis requiring hospitalization, and poor short-term outcomes, the majority of patients would be consistent with advanced HIV disease.

Pulmonary TB was diagnosed using clinical features, chest radiography, sputum smear microscopy where available, and physician assessment consistent with national guidelines. Culture and drug susceptibility testing were not routinely performed. Chest radiography was performed systematically, with findings interpreted by pulmonary specialists. Clinical diagnosis was established based on constitutional symptoms (fever, weight loss, night sweats), compatible radiological findings, and response to empirical anti-tubercular therapy, reviewed by senior clinicians experienced in HIV-TB management.

Advanced diagnostic procedures including mycobacterial culture, drug-susceptibility testing, bronchoscopy, induced sputum examination for Pneumocystis jirovecii, and molecular diagnostic assays were not routinely available during the study period. Diagnostic classification therefore relied on clinical assessment by experienced chest physicians, chest radiographic findings, sputum smear microscopy where available, and subsequent clinical response to anti-tubercular therapy.

Patients with prior anti-tubercular therapy for the current episode or incomplete clinical records were excluded. A total of 107 patients met inclusion criteria. All consecutive eligible patients admitted during the study period were included to minimize selection bias; no formal a priori sample size calculation was performed.

2.3. HIV-negative comparison cohort

To contextualize the prognostic significance of tuberculin responsiveness specifically in HIV-TB coinfection, a parallel cohort of 100 consecutive HIV-seronegative adults hospitalized with pulmonary tuberculosis during the same period was evaluated using identical diagnostic criteria, tuberculin testing procedures, nutritional assessment methods, and outcome definitions. This comparison cohort was drawn from the same hospital wards and managed by the same clinical team, ensuring procedural consistency.

The HIV-negative cohort had comparable demographic characteristics with similar age distribution and sex ratio. Clinical presentation, radiological findings, and prevalence of severe malnutrition were similar to the HIV-positive cohort within corresponding TST reactivity categories.

2.4. Tuberculin skin testing

Tuberculin responsiveness was assessed using the Mantoux method. Ten tuberculin units (10 TU) of purified protein derivative (PPD) were injected intradermally into the volar aspect of the forearm [8]. Induration was measured in millimeters at 48–72 h by trained clinical staff using the palpation method and transverse diameter measurement. Based on maximum transverse diameter, patients were categorized into three groups: (1) >10 mm; (2) 5–10 mm; and (3) <5 mm (including anergic responses). Tuberculin testing was performed after stabilization of acute illness where clinically feasible. These cutoffs were chosen based on commonly used clinical thresholds in HIV-related tuberculin interpretation. BCG vaccination status was not systematically recorded and could not be reliably reconstructed retrospectively. Two-step tuberculin testing was not performed; a single Mantoux test was used, consistent with routine clinical practice during the study period.

2.5. Nutritional assessment

Nutritional status was assessed at admission using body mass index (BMI), calculated as weight in kilograms divided by height in meters squared (kg/m²). BMI was categorized using World Health Organization criteria [9]: severe undernutrition (BMI <16), moderate undernutrition (BMI 16–17), mild undernutrition (BMI 17–18.5), and normal (BMI ≥18.5). In this hospitalized population, BMI was considered a clinical proxy for overall disease burden and chronicity. Preserved BMI was interpreted as an indirect indicator of relatively maintained immune status, while severe malnutrition in advanced HIV infection is known to contribute to CD4 lymphocyte depletion [10], impaired cytokine signaling, and blunting of delayed-type hypersensitivity responses.

2.6. Outcome measures

The primary outcome was clinical status at hospital discharge, categorized as: (1) survival with discharge within six weeks; (2) survival with prolonged hospitalization (>6 weeks); (3) discharge against medical advice (LAMA) in moribund condition; (4) in-hospital death. Mortality was defined as all-cause in-hospital death occurring before discharge. Duration of hospitalization was recorded in weeks. Outcomes were ascertained from hospital records and treating physician documentation, with vital status verified through ward registers and discharge summaries.

2.7. Statistical analysis

Categorical variables were summarized as frequencies and percentages. Associations between ordered TST categories and mortality were assessed using χ² test for trend, with p < 0.05 considered significant. Analyses were exploratory without multivariable adjustment. BMI values were missing in fewer than 5% of cases and handled by complete-case analysis. This study is reported in accordance with STROBE guidelines. Data were double-entered into Microsoft Excel and verified against source records.

2.8. Ethical considerations

The study was conducted in accordance with ethical standards applicable at the time of data collection (1998–1999). The current manuscript represents a secondary analysis of anonymized archival dissertation data with no identifiable patient information. All patients provided verbal informed consent for HIV testing, tuberculosis treatment, and inclusion of their clinical data in educational and research activities at hospital admission, consistent with institutional practices during that period. As the analysis involved routinely collected clinical records and posed minimal risk, formal informed consent for this specific analysis was not required. Patient confidentiality was maintained throughout, consistent with the Declaration of Helsinki. No consent to publish individual patient data was required, as the manuscript contains only aggregate, de-identified results.

3. Results

3.1. Study population

Between January 1998 and October 1999, 134 HIV-seropositive adults with pulmonary tuberculosis were admitted to the participating wards. Of these, 27 patients were excluded: 18 had received prior anti-tubercular therapy for the current episode, 6 had incomplete clinical documentation, and 3 were transferred before tuberculin testing. The final analytical cohort comprised 107 patients with complete baseline and outcome data.

Median age was 34 years (range: 21–58 years), and 68 patients (63.6%) were male. Urban residence was documented in 41 patients (38.3%). Constitutional symptoms dominated clinical presentation: fever in 103 patients (96.3%), weight loss >10% in 89 patients (83.2%), and chronic cough in 98 patients (91.6%). Sputum smear microscopy was positive for acid-fast bacilli in 52 patients (48.6%), negative in 37 (34.6%), and 18 (16.8%) could not produce sputum. Chest radiography demonstrated bilateral infiltrates in 64 patients (59.8%), cavitary lesions in 31 (29.0%), and miliary pattern in 12 (11.2%). All patients were hospitalized at presentation, reflecting advanced disease severity with high prevalence of undernutrition.

3.2. Tuberculin skin test reactivity

Based on Mantoux testing with 10 TU purified protein derivative, 33 patients (30.8%) demonstrated induration >10 mm, 26 patients (24.3%) had induration 5–10 mm, and 48 patients (44.9%) were anergic or had induration <5 mm (Table 1).

Table 1.

Tuberculin skin test reactivity and nutritional status in hospitalized HIV-TB patients (N = 107).

Tuberculin skin test induration	Number of patients, n (%)	Nutritional status (BMI)
>10 mm	33 (30.8)	Mild–moderate undernutrition (BMI 16–18.5)
5–10 mm	26 (24.3)	Severe undernutrition (BMI <16)
<5 mm / anergic	48 (44.9)	Severe undernutrition (n = 42); Normal BMI (n = 6)
Total	107 (100)

BMI = body mass index; undernutrition categories defined according to WHO criteria.

3.3. Nutritional status

All patients with induration >10 mm had BMI 16–18.5 kg/m², indicating mild to moderate undernutrition (Table 1). All patients in the 5–10 mm induration group were severely undernourished (BMI <16). Among the anergic or < 5 mm group, severe undernutrition was present in the majority; six patients had BMI within the normal range.

3.4. Clinical outcomes

Clinical outcome data were available for all included patients; BMI values were missing in fewer than 5% of cases.

All 33 patients (100%) with TST induration >10 mm survived and were discharged within six weeks. Among the 26 patients with induration 5–10 mm, six (23.1%) died during hospitalization, while the remaining 20 (76.9%) survived but required prolonged hospitalization exceeding six weeks. In the anergic or < 5 mm group (n = 48), 30 patients (62.5%) died during hospitalization, 12 (25.0%) were discharged against medical advice in moribund condition, and only 6 (12.5%) survived to discharge after prolonged hospitalization (Table 2). Notably, all six survivors in the anergic group had normal BMI and required prolonged hospitalization exceeding six weeks before discharge.

Table 2.

Clinical outcomes by tuberculin skin test reactivity among hospitalized HIV-positive patients with pulmonary tuberculosis (N = 107).

Tuberculin skin test induration	Discharged ≤6 weeks	Prolonged hospitalization >6 weeks	Moribund discharge (LAMA)	In-hospital death
>10 mm (n = 33)	33 (100%)	0	0	0
5–10 mm (n = 26)	0	20 (76.9%)	0	6 (23.1%)
<5 mm / anergic (n = 48)	0	6 (12.5%)	12 (25.0%)	30 (62.5%)

LAMA = left against medical advice.

A contemporaneous cohort of 100 HIV-negative patients hospitalized with pulmonary tuberculosis was also evaluated. Among these patients, 24 (24.0%) demonstrated TST induration >10 mm, 41 (41.0%) had induration 5–10 mm, and 35 (35.0%) were anergic or had induration <5 mm. All patients with induration >10 mm were discharged alive. All HIV-negative patients with TST induration ≤10 mm had severe undernutrition (BMI <16); nevertheless, in-hospital mortality remained low, with only one death (2.4%) among patients with induration 5–10 mm and two deaths (5.7%) among anergic patients, contrasting sharply with the corresponding HIV-positive groups (Table 3, Table 4).

Table 3.

Tuberculin skin test reactivity, nutritional status, and clinical outcomes in hospitalized HIV-negative pulmonary tuberculosis patients (N = 100).

TST Induration	N (%)	Nutritional Status (BMI)	Discharged	In-Hospital Death
>10 mm	24 (24.0%)	Mild malnutrition (BMI 16–18.5)	24 (100%)	0 (0%)
5–10 mm	41 (41.0%)	Severe malnutrition (BMI <16)	40 (97.6%)	1 (2.4%)
<5 mm / anergic	35 (35.0%)	Severe malnutrition (BMI <16)	33 (94.3%)	2 (5.7%)
Total	100 (100%)	–	97 (97.0%)	3 (3.0%)

Note: BMI = body mass index. Undernutrition categories defined per WHO criteria; This table presents the distribution of tuberculin reactivity, nutritional status, and clinical outcomes in the HIV-negative comparison cohort using the same outcome definitions as the HIV-positive cohort.

Table 4.

Comparative outcomes by HIV status and tuberculin reactivity.

TST Induration	HIV-Negative (n = 100)	HIV-Positive (n = 107)	Relative Mortality (HIV-positive/HIV-negative)
>10 mm	0/24 (0%)	0/33 (0%)	–
5–10 mm	1/41 (2.4%)	6/26 (23.1%)	9.5×
<5 mm/anergic	2/35 (5.7%)	30/48 (62.5%)	11.0×
Overall	3/100 (3%)	36/107 (33.6%)	11.2×

All percentages are expressed to one decimal place.

Note: TST = tuberculin skin test. Both cohorts were hospitalized with pulmonary tuberculosis and evaluated during the same period (1998–1999) using identical tuberculin testing procedures and outcome definitions. Mortality ratios calculated as HIV-positive mortality rate divided by HIV-negative mortality rate within each TST category. The dramatic mortality gradient observed across TST categories in HIV-positive patients (0% to 62.5%) contrasts with the minimal gradient in HIV-negative patients (0% to 5.7%), demonstrating HIV-specific prognostic significance of tuberculin responsiveness. χ² test for trend in HIV-positive cohort: p < 0.001.

3.5. Association between tuberculin reactivity and outcome

A graded relationship was observed between increasing tuberculin induration size and more favorable clinical outcomes. Higher TST reactivity was associated with shorter hospitalization and survival to discharge, whereas minimal or absent reactivity was associated with increased mortality, prolonged hospitalization, and moribund discharge (Fig. 1). The association between decreasing TST reactivity and increasing mortality demonstrated a highly significant trend (χ² test for trend: p < 0.001). Pairwise comparisons revealed significant differences in mortality between all TST categories (>10 mm vs 5–10 mm: p = 0.004; 5–10 mm vs <5 mm: p < 0.001; >10 mm vs <5 mm: p < 0.001).

Fig. 1.

In-hospital mortality by tuberculin skin test (TST) reactivity category.

Bar chart showing in-hospital mortality rates stratified by Mantoux induration size. Mortality demonstrated a strong inverse gradient with tuberculin responsiveness: 0% mortality in patients with induration >10 mm (n = 33), 23.1% mortality in those with induration 5–10 mm (n = 26), and 62.5% mortality among patients with induration <5 mm or anergy (n = 48).

3.6. Comparison with HIV-negative tuberculosis patients

Among 100 HIV-negative pulmonary tuberculosis patients, tuberculin reactivity showed a markedly different prognostic profile. Twenty-four patients (24%) demonstrated induration >10 mm, all with mild malnutrition; all were discharged alive after brief hospitalization. Forty-one patients (41%) had induration 5–10 mm, all severely malnourished; 40 (97.6%) were discharged alive and one (2.4%) died. Thirty-five patients (35%) were anergic or had induration <5 mm; 33 (94.3%) were discharged alive and two (5.7%) died (Table 3).

Overall mortality in HIV-negative TB patients was 3% (3/100) – an 11-fold lower mortality compared to the HIV-positive cohort (33.6%, 36/107). While preserved tuberculin reactivity (>10 mm) predicted 100% survival in both groups, diminished reactivity showed dramatically different prognostic implications by HIV status. Among patients with induration 5–10 mm, HIV-positive patients experienced 9.5-fold higher mortality (23.1% vs 2.4%). Among anergic or minimally reactive patients (<5 mm), HIV-positive patients experienced 11-fold higher mortality (62.5% vs 5.7%) (Table 4). These findings demonstrate that the graded mortality gradient associated with decreasing tuberculin responsiveness is HIV-specific rather than a general feature of tuberculosis or severe malnutrition.

4. Discussion

This hospital-based cohort study demonstrates a strong and graded association between tuberculin skin test (TST) reactivity and short-term clinical outcomes among hospitalized HIV-infected patients with pulmonary tuberculosis. The inclusion of a parallel HIV-negative tuberculosis cohort provides critical context: while preserved TST reactivity (>10 mm) predicted favorable outcomes in both groups, tuberculin anergy carried profoundly different prognostic implications depending on HIV status 5.7% mortality in HIV-negative versus 62.5% in HIV-positive patients.

4.1. Diagnostic limitations versus prognostic utility

While TST has limited diagnostic utility in this setting, our findings show it conveys meaningful prognostic information. Patients with preserved TST reactivity (>10 mm) demonstrated 100% survival in both HIV-positive and HIV-negative cohorts, marking it as a reliable indicator of physiological reserve regardless of HIV status. Conversely, tuberculin anergy showed dramatically different implications: among HIV-negative patients, anergy was associated with only 5.7% mortality, likely reflecting transient immune suppression from acute illness with preserved recovery potential. In stark contrast, HIV-positive anergic patients experienced 62.5% mortality (an 11-fold difference), suggesting that anergy in HIV-TB may reflect advanced impairment of cell-mediated immunity and systemic inflammatory burden [11], [12] rather than acute illness severity alone.

Most patients did not present with severe breathlessness, a clinical feature commonly associated with moderate-to-severe Pneumocystis jirovecii pneumonia. Instead, the predominance of productive cough (91.6%), constitutional symptoms, and chest radiographic findings, including apical, miliary and cavitary patterns, was considered more consistent with pulmonary tuberculosis. Nevertheless, given the limited diagnostic resources available during the study period, the contribution of unrecognized opportunistic co-infections, including Pneumocystis jirovecii pneumonia, cannot be excluded. Consequently, the observed association between TST anergy and mortality may reflect profound immune dysfunction with increased susceptibility to multiple opportunistic infections rather than tuberculosis severity alone. These findings suggest that tuberculin skin test responsiveness should be interpreted as an integrated surrogate marker of residual cell-mediated immunity rather than a disease-specific marker of tuberculosis severity. Accordingly, TST anergy may identify patients with advanced immune dysfunction who are at increased risk of adverse clinical outcomes, including death, resulting from the cumulative burden of opportunistic infections, irrespective of the precise pathogen responsible.

4.2. The HIV-specific nature of the mortality gradient

The comparison between cohorts reveals that the dramatic mortality gradient across TST categories is not a general feature of tuberculosis or malnutrition, but a specific marker of HIV-induced immune dysfunction. The overall 11-fold mortality disparity (33.6% vs 3%) was entirely concentrated among patients with diminished tuberculin responses; those maintaining robust reactivity (>10 mm) showed identical survival regardless of HIV status.

Crucially, this gradient persisted despite comparable severe malnutrition in the corresponding HIV-negative TST groups. Among patients with induration 5–10 mm, HIV-positive patients experienced 23.1% mortality compared to only 2.4% in HIV-negative patients, a 9.5-fold difference, despite similar rates of severe malnutrition (BMI <16) in both groups. These findings suggest that tuberculin responsiveness reflects HIV-related immune dysfunction beyond nutritional status alone. However, because patient-level analyses adjusting simultaneously for BMI and TST responsiveness were not feasible within this archival dataset, the independent prognostic contribution of each variable cannot be determined. BMI and TST responsiveness are biologically interrelated through their shared association with immune competence, and residual confounding cannot be excluded.

Indeed, an unadjusted analysis of nutritional status showed that the pooled mortality rate among severely undernourished HIV-positive patients was 52.9% (36/68), comprising the 6 deaths in the 5–10 mm stratum and the 30 deaths in the anergic stratum. While this underscores the profound mortality risk associated with severe undernutrition, the higher mortality observed among completely anergic patients (62.5%) compared with partially reactive patients (23.1%) is consistent with the possibility that TST responsiveness provides additional prognostic information even within a uniformly malnourished cohort.

The contemporaneous HIV-negative cohort provides additional evidence supporting this interpretation. Severe malnutrition was present in all HIV-negative patients with TST induration ≤10 mm, yet mortality remained low (2.4–5.7%). In contrast, mortality among HIV-positive patients with comparable TST categories ranged from 23.1% to 62.5%. These findings suggest that the adverse prognostic significance of tuberculin anergy in HIV-TB cannot be explained by nutritional status alone and more likely reflects profound HIV-related impairment of cell-mediated immunity.

The presence of six HIV-positive patients with normal BMI despite tuberculin anergy further supports this interpretation. All six survived, although each required prolonged hospitalization exceeding six weeks. This observation suggests that preserved nutritional status does not necessarily indicate preserved cell-mediated immunity, while potentially conferring greater physiological reserve for recovery. These findings further support the concept that tuberculin skin test responsiveness should be interpreted as an integrated surrogate marker of residual cell-mediated immunity rather than a disease-specific marker of tuberculosis severity. Accordingly, TST anergy may identify patients with advanced immune dysfunction who are at increased risk of adverse clinical outcomes, irrespective of nutritional status or the precise opportunistic pathogen responsible. Early CD4 depletion following recent HIV infection may permit TB reactivation before weight loss develops, making anergy a marker of acute immunodeficiency.

Conversely, profoundly malnourished and anergic HIV-positive patients may manifest reduced inflammatory responses, resulting in atypical clinical features and poor short-term outcomes.

4.3. Comparison with existing literature

Prior studies have documented associations between tuberculin anergy and adverse outcomes in HIV-infected populations. Selwyn et al. (1992) reported that HIV-infected individuals with cutaneous anergy had significantly higher risk of developing active tuberculosis [11], while Moreno et al. (1993) demonstrated that tuberculin anergy was strongly associated with advanced HIV disease and lower CD4 lymphocyte counts, supporting the concept that anergy reflects progressive impairment of cell-mediated immunity [12]. Our findings extend these observations by demonstrating a graded relationship between TST reactivity magnitude and short-term clinical outcomes and by documenting through direct comparison that this mortality gradient is HIV-specific rather than a general feature of tuberculosis or malnutrition.

Patients maintaining robust delayed-type hypersensitivity responses despite HIV infection and active tuberculosis likely have relatively preserved CD4+ T-cell numbers and function, enabling more effective immune responses and better physiological reserve. The parallel HIV-negative cohort confirms that preserved TST reactivity predicts favorable outcomes regardless of HIV status, suggesting a general protective role for intact cell-mediated immunity. Conversely, anergic HIV-positive patients represent those with profound immunosuppression and limited capacity for immune reconstitution. This interpretation is supported by Moreno et al., who demonstrated a direct association between cutaneous anergy and advanced CD4 lymphocyte depletion in HIV-infected individuals [12], highlighting the biological basis for the adverse outcomes observed in the present study.

4.4. Contemporary relevance

Although studied in the pre-ART era, these findings retain significant relevance. Estimates suggest that ∼30–40% of patients in India still present with advanced HIV (CD4 < 200 cells/μL) or as “late presenters” [13]. Such patients may resemble the historical cohort with respect to advanced immunosuppression and disease severity. Therefore, although derived from a pre-ART cohort, these findings remain clinically relevant for contemporary patients presenting with advanced HIV disease. In settings where flow cytometry remains inaccessible, TST may continue to provide useful, low-cost prognostic insight.

The HIV-negative comparison provides a critical historical benchmark: the 3% mortality rate likely represents the baseline achievable with TB treatment alone, whereas the 33.6% HIV-positive mortality reflects the additive burden of untreated advanced HIV disease. This 11-fold difference underscores the survival benefits that contemporary ART initiation can achieve in late presenters, while highlighting that a subset with preserved immune function (robust TST reactivity) may survive even without immediate ART access.

Detailed information regarding concomitant antibacterial therapy, empirical co-trimoxazole, antifungal therapy, corticosteroid use, nutritional supplementation, and other supportive interventions was not consistently available in the archived dataset and therefore could not be incorporated into the analysis. Differences in these therapies may have influenced clinical outcomes and contributed to the observed mortality gradient across TST categories. Consequently, residual treatment-related confounding cannot be excluded.

Recent studies consistently demonstrate that advanced immunosuppression and undernutrition remain major determinants of mortality in HIV-associated tuberculosis, even with expanded ART access. Large contemporary cohorts and meta-analyses identify low CD4 counts and poor nutritional status as dominant predictors of adverse outcomes [14], [15], underscoring the continued importance of early risk stratification. In the current “Test and Treat” era, this pre-ART dataset, particularly the HIV-negative comparison, serves as a crucial historical baseline against which survival benefits of modern interventions in late presenters can be measured.

The combined assessment of BMI and tuberculin responsiveness may offer a pragmatic bedside framework: severe undernutrition represents a marker of advanced disease burden and adverse prognosis. Although undernutrition itself may also increase susceptibility to progression following tuberculosis exposure, while preserved tuberculin reactivity may indicate residual immune competence despite active tuberculosis. The HIV-negative data validate that this framework specifically captures HIV-related immune reserve rather than general TB severity.

4.5. Clinical and policy implications

The graded association between TST reactivity and outcomes suggests a potential role for TST as an adjunctive prognostic tool in hospitalized HIV-TB patients. Preserved reactivity (>10 mm) was uniformly associated with survival in both cohorts, whereas anergy carried HIV-specific prognostic significance with markedly elevated mortality only in HIV-positive patients. In contemporary resource-limited settings, TST reactivity may help identify HIV-positive patients warranting closer monitoring and enhanced supportive care, including nutritional rehabilitation and timely ART initiation. Such use should complement, not replace, established clinical assessment and laboratory evaluation.

Notably, even patients with robust tuberculin reactivity (>10 mm) developed active tuberculosis in both cohorts, underscoring the intense transmission force and high prevalence of latent tuberculosis infection in endemic communities, where even individuals with measurable immune responses remain at substantial risk.

4.6. Strengths and limitations

Study strengths include consecutive patient enrollment minimizing selection bias, the contemporaneous HIV-negative comparison cohort evaluated using identical procedures, standardized tuberculin testing and systematic outcome ascertainment, and pragmatic design reflecting real-world conditions in resource-limited settings. The clear dose-response relationship between TST reactivity and outcomes in HIV-positive patients, contrasted with minimal gradient in HIV-negative patients, strengthens causal inference regarding HIV-specific immune dysfunction.

The present study has several limitations. These include the pre-ART design (late 1990s), during which HIV-positive mortality exceeded contemporary rates by approximately two- to three-fold; the unavailability of CD4 lymphocyte counts and HIV viral load measurements; and the inability to directly compare the independent prognostic contributions of BMI and TST responsiveness because patient-level linkage between BMI category and mortality within individual TST strata was not available in the archival dataset. Consequently, although TST responsiveness appeared to show a stronger association with mortality than nutritional status in this cohort, residual confounding by BMI cannot be excluded. Additional limitations include the lack of routine mycobacterial culture and drug-susceptibility testing, with some patients diagnosed on clinical and radiological grounds without microbiological confirmation, the single-center design (n = 107 HIV-positive, n = 100 HIV-negative) limiting generalizability, and descriptive analyses precluding multivariable adjustment for potential confounders including TB disease severity, comorbidities, and supportive care differences.

Formal WHO clinical staging was not available for analysis. The absence of bronchoscopy, induced sputum examination, and Pneumocystis-specific diagnostic testing means that diagnostic overlap with Pneumocystis jirovecii pneumonia cannot be entirely excluded, particularly among the 64 patients presenting with bilateral pulmonary infiltrates.

Future prospective studies incorporating CD4 lymphocyte counts, HIV viral load, standardized assessment of nutritional status, systematic evaluation for opportunistic co-infections, documentation of concomitant therapies, and multivariable analyses are needed to determine the independent prognostic contribution of tuberculin responsiveness.

Mycobacterial culture, drug-susceptibility testing, bronchoscopy, induced sputum examination for Pneumocystis jirovecii, and molecular diagnostic assays were not routinely available during the study period. Consequently, diagnostic misclassification and some degree of overlap between pulmonary tuberculosis and other opportunistic pulmonary infections cannot be completely excluded. These findings may be most applicable to hospitalized late-presenting HIV-TB patients in similar resource-limited inpatient settings rather than ambulatory ART-treated populations. Interferon-gamma release assays (IGRAs) were not available during the study period. Consequently, direct comparison of TST responsiveness with IGRA-based measures of cellular immune response was not possible.

5. Conclusion

Tuberculin skin testing has limited diagnostic value for active tuberculosis in people living with HIV. However, the magnitude of tuberculin responsiveness offers useful prognostic insight specific to HIV-related immunosuppression in hospitalized HIV-TB patients, particularly where advanced immunological testing is unavailable. A robust TST reaction (>10 mm) serves as a reassuring indicator of physiological reserve regardless of HIV status, whereas anergy in HIV-positive patients, but not HIV-negative patients, should prompt heightened clinical vigilance for adverse outcomes.

The dramatic mortality gradient across TST categories in HIV-positive patients (0% to 62.5%), contrasted with the minimal gradient in HIV-negative patients (0% to 5.7%), supports the interpretation that tuberculin responsiveness primarily reflects HIV-related immune competence rather than tuberculosis severity alone, although residual confounding by nutritional status and other clinical factors cannot be excluded. These findings reinforce the value of simple bedside immune markers for risk stratification in resource-limited settings.

CRediT authorship contribution statement

Murali K. Yanamadala: Writing – review & editing, Visualization, Resources, Methodology, Formal analysis, Data curation, Conceptualization.

Ethics statement

The study was conducted in accordance with ethical standards applicable at the time of data collection (1998–1999). HIV testing and TB treatment were performed under routine institutional practices prevailing at that time. The current manuscript represents a secondary analysis of anonymized archival dissertation data with no identifiable patient information. All patients provided verbal informed consent for HIV testing, tuberculosis treatment, and inclusion of their clinical data in educational and research activities at the time of hospital admission, consistent with institutional practices during that period. As the analysis involved routinely collected clinical records and posed minimal risk to participants, formal informed consent for this specific analysis was not required. Patient confidentiality was maintained throughout, consistent with the principles of the Declaration of Helsinki. No consent to publish individual patient data was required, as the manuscript contains only aggregate, de-identified results.

Funding

No external funding was received.

Declaration of competing interest

The author has declared that no competing interests exist. This study was conducted independently, and the author has no relevant financial or non-financial interests to disclose that could be perceived as influencing the research.