Prevalence of latent tuberculosis infection among prison officers in Paraná, Brazil: a longitudinal study Authors: Kristiana Giacomini, Carlos Eduardo de-Souza-Matos, Matheus Arengheri-Vicente, Lia Beatriz Henke-de-Azevedo, Leonardo Garcia-Velasquez, Lirane Elize Defante-Ferreto Published: 2026 ABSTRACT AbstractObjectives: To evaluate the baseline prevalence of latent tuberculosis infection (LTBI) and longitudinal changes in interferon-gamma release assay (IGRA) positivity among male and female prison officers in Paraná, Brazil.Material and method: A longitudinal study was conducted with prison officers in direct contact with inmates. LTBI was diagnosed using the QuantiFERON-TB® Gold Plus (IGRA) test.Results: The proportion of IGRA-positive individuals decreased from 22.6% in 2022 to 5.7% in 2024. However, increases were observed in cigarette consumption, alcohol use, and daily consumption of medications.Discussion: The findings highlight the importance of periodic monitoring of latent tuberculosis infection and strengthening occupational health promotion strategies for prison officers. INTRODUCTION Latent tuberculosis infection (LTBI) occurs when an individual is infected with Mycobacterium tuberculosis without symptoms of active disease. Individuals with LTBI are not contagious; however, the infection may progress to active tuberculosis, particularly among immunocompromised individuals1. According to the WHO Global Tuberculosis Report 2023, approximately one-quarter of the global population is infected with Mycobacterium tuberculosis, representing the reservoir of LTBI2. In the Americas, the Pan American Health Organization (PAHO) reports persistent transmission in vulnerable populations, particularly among prison inmates and other high-risk groups3. In Brazil, tuberculosis surveillance and control have been strengthened through national clinical protocols, notification systems such as SINAN/DATASUS, and professional training initiatives4,5. However, adherence to LTBI preventive treatment remains low despite the availability of prevention strategies, including BCG vaccination, early diagnosis, and the use of personal protective equipment (PPE). Prison environments present a particularly high risk of Mycobacterium tuberculosis transmission due to overcrowding, poor ventilation, and limited access to healthcare4,6. In 2023, people deprived of liberty (PDL) represented only 0.3% of Brazil’s population but accounted for 9% of new tuberculosis cases7. This scenario highlights the occupational exposure of correctional officers8-10. Strengthening occupational health policies for correctional officers is essential to reduce transmission within prisons and prevent it from spreading into the community11,12. This study evaluates baseline LTBI prevalence and longitudinal changes in IGRA positivity among officers of both sexes in Paraná, Brazil, in 2022 and 2024, and identifies factors associated with infection. To our knowledge, this represents one of the first longitudinal analyses of LTBI in correctional officers in southern Brazil. MATERIAL AND METHOD Study design and setting This longitudinal study was conducted among prison officers working at the Francisco Beltrão State Prison (Paraná, Brazil), a medium-security facility with high TB-transmission potential caused by enclosed environments and frequent contact with incarcerated individuals. Data collection occurred in May 2022 (baseline) and May 2024 (follow-up). The basis used to define the two-year interval was the Brazilian Ministry of Health recommendations for periodic screening of high-risk occupational groups and to allow (i) observation of potential IGRA conversions/reversions, (ii) completion of LTBI treatment among those indicated, and (iii) sufficient time for institutional educational interventions to be implemented without recall bias. Population and sample The center had 104 eligible staff members with direct contact with inmates. The sample size for the baseline study had been previously calculated and described in an earlier study, assuming an expected LTBI prevalence of 29.1%, 80% power, a 5% alpha level, and an anticipated 10% attrition rate, resulting in 71 staff members included in 202213. Of these, 53 completed follow-up in 2024 (attrition 25.3%), with losses caused solely be transfer to other prison units or dismissal. The study flow is presented in Figure 1. Baseline characteristics of prison officers retained vs. lost to follow-up are presented in Table 1. No meaningful baseline imbalances were observed, suggesting non-differential attrition. Figure 1Flowchart of participant inclusion, IGRA status, and follow-up between 2022 and 2024. Flowchart of participant inclusion, IGRA status, and follow-up between 2022 and 2024. Table 1Sensitivity analyses for LTBI prevalence estimates (2022-2024).Analysis scenario AssumptionsLTBI prevalence 2024Difference vs. 2022 (22.6%)InterpretationMain analyses (complete-case)Only participants assessed in both years (n=53)5.7% (3/53)−16.9 pp†Major decline in LTBI* prevalenceWorst-case (realistic)All 4 lost IGRA+ ‡ participants were positive in 20249.8% (7/71)-12.8 ppDecline persists even in conservative scenarioWorst-case (extreme, implausible)All 18 lost participants were positive in 202429.5% (21/71)+6.9 ppExtreme assumption not supported by baseline dataBest-case scenarioAll lost participants were negative in 20244.2% (3/71)−16.4 ppIdentical to main analysisTipping-point analysisNumber of lost participants needed to remove the difference≥13/18 would need to be positiveImpossibleAttrition alone cannot explain declineIPW§Not performed --Not needed; attrition non-differential MI||Not performed --No missing covariates required MI Note. *LTBI: latent tuberculosis infection; †pp: percentage points; ‡IGRA: interferon gamma release assay; §IPW: Inverse Probability Weighting; ||MI: Multiple Imputation. Table 1Sensitivity analyses for LTBI prevalence estimates (2022-2024).Analysis scenario AssumptionsLTBI prevalence 2024Difference vs. 2022 (22.6%)InterpretationMain analyses (complete-case)Only participants assessed in both years (n=53)5.7% (3/53)−16.9 pp†Major decline in LTBI* prevalenceWorst-case (realistic)All 4 lost IGRA+ ‡ participants were positive in 20249.8% (7/71)-12.8 ppDecline persists even in conservative scenarioWorst-case (extreme, implausible)All 18 lost participants were positive in 202429.5% (21/71)+6.9 ppExtreme assumption not supported by baseline dataBest-case scenarioAll lost participants were negative in 20244.2% (3/71)−16.4 ppIdentical to main analysisTipping-point analysisNumber of lost participants needed to remove the difference≥13/18 would need to be positiveImpossibleAttrition alone cannot explain declineIPW§Not performed --Not needed; attrition non-differential MI||Not performed --No missing covariates required MI Note. *LTBI: latent tuberculosis infection; †pp: percentage points; ‡IGRA: interferon gamma release assay; §IPW: Inverse Probability Weighting; ||MI: Multiple Imputation. Sensitivity analyses for LTBI prevalence estimates (2022-2024). Note. *LTBI: latent tuberculosis infection; †pp: percentage points; ‡IGRA: interferon gamma release assay; §IPW: Inverse Probability Weighting; ||MI: Multiple Imputation. Inclusion and exclusion criteria Inclusion criteria: Permanent or temporary officersAge ≥18 years≥1 year of direct contact with inmates BCG vaccination >6 months before baselineOn active duty at time of data-collectionConsent given to participation Permanent or temporary officers Permanent or temporary officers Age ≥18 years Age ≥18 years ≥1 year of direct contact with inmates ≥1 year of direct contact with inmates BCG vaccination >6 months before baseline BCG vaccination >6 months before baseline On active duty at time of data-collection On active duty at time of data-collection Consent given to participation Consent given to participation Exclusion criteria: Administrative staff without direct contact with inmatesMedical leave or <1 year in positionHistory of active TBHistory of previously treated LTBI (added per reviewer request)Pregnancy Administrative staff without direct contact with inmates Administrative staff without direct contact with inmates Medical leave or <1 year in position Medical leave or <1 year in position History of active TB History of active TB History of previously treated LTBI (added per reviewer request) History of previously treated LTBI (added per reviewer request) Pregnancy Pregnancy Data collection and variables Data were collected at the prison health unit through a structured questionnaire and blood samples for IGRA testing (QuantiFERON-TB® Gold Plus [Qiagen]; sensitivity: 94.09%; specificity: 97.27%). This assay evaluates the cellular immune response to specific Mycobacterium tuberculosis antigens by quantifying interferon-gamma production. Variables related to TB exposure (e.g., knowing someone with TB or contact with TB cases) referred to self-reported exposure occurring within the previous 12 months prior to each assessment. Sociodemographic variables Age, sex, race/ethnicity, education, marital status, income bracket, years spent in profession, work sector and shift. Behavioral and lifestyle variables Tobacco use, alcohol consumption, illegal drug use, physical activity, number of household contacts. Clinical variables and comorbidities In 2022, clinical comorbidities were collected through a structured questionnaire asking prison officers whether they had a previous diagnosis of HIV, hepatitis B, hepatitis C, syphilis, diabetes mellitus, hypertension, dyslipidemia, thyroid disorders, cancer, prior gastric surgery, osteoporosis, or a history of hepatitis/jaundice. Because the frequency of each condition was low, all comorbidities were analyzed as a grouped binary variable (present/absent). Additional clinical variables included BMI category and presence of a BCG scar. In 2024, IGRA-positive prison officers underwent laboratory testing for HIV, hepatitis B, hepatitis C, and syphilis exclusively as part of the national clinical safety protocol for LTBI treatment. These tests were not used as comorbidity variables but only to assess treatment eligibility, in accordance with Brazilian Ministry of Health guidelines. LTBI diagnosis IGRA results were interpreted according to the manufacturer’s criteria for the QuantiFERON-TB® Gold Plus assay (Qiagen), with values ≥0.35 IU/mL being considered positive. No indeterminate results occurred in either year13; therefore, no retesting was required. Prison officers with positive IGRA results were referred to public tuberculosis reference services of the Brazilian Unified Health System (SUS) for clinical evaluation and possible initiation of LTBI preventive treatment according to national guidelines. Treatment decisions were made by the attending physicians of the public health services and were not part of the study protocol. Educational intervention Between assessments, an educational intervention was implemented targeting TB transmission and protection practices. Contents included: Clinical signs and symptoms of TB.LTBI concept and treatment indications.Use of PPE (N95) and environmental control measures.Symptom surveillance and notification flow inside the prison.Post-test individual feedback and counseling. Clinical signs and symptoms of TB. Clinical signs and symptoms of TB. LTBI concept and treatment indications. LTBI concept and treatment indications. Use of PPE (N95) and environmental control measures. Use of PPE (N95) and environmental control measures. Symptom surveillance and notification flow inside the prison. Symptom surveillance and notification flow inside the prison. Post-test individual feedback and counseling. Post-test individual feedback and counseling. Format and duration Two 1-hour educational sessions were conducted offline. Individual feedback meetings were held after receiving the IGRA results. Printed informational materials and institutional posters were created. Educators Multidisciplinary team of health professionals specialising in infectious diseases and occupational health. Evaluation of learning Pre- and post-intervention checklist on TB signs, preventive practices, and PPE use. The educational intervention was implemented as part of routine institutional health education activities and was not designed as an experimental intervention. Therefore, its impact on IGRA outcomes was not formally evaluated. Recognition of study design limitations This study did not include a control group for the educational intervention. Therefore, while changes in IGRA positivity and related practices are reported descriptively and temporally, causal inference cannot be established. Changes are interpreted as associations within the follow-up period. Statistical analysis Given that four IGRA-positive individuals from 2022 were lost to follow-up, sensitivity analyses were conducted under the assumption (i) that all lost individuals remained IGRA-positive, and (ii) a complete-case approach. These analyses were performed to evaluate the potential impact of attrition on LTBI prevalence estimates. Descriptive statistics were used to characterize the sample. LTBI prevalence with 95% CIs was calculated for each period. Attrition bias was addressed through sensitivity analyses assuming: (i) lost prison officers remained LTBI-positive, and (ii) complete-case analysis. Analyses were processed using SPSS v23 (SPSS v23). Ethics The study was approved by the UNIOESTE Research Ethics Committee (CAAE: 6.745.006). All prison officers signed informed consent per the Declaration of Helsinki and Brazilian Resolution 466/2012. RESULTS A total of 18 prison officers (25.3%) were lost to follow-up between 2022 and 2024 due to institutional transfers to other prison units (n = 2) and dismissals (n = 16). Most baseline characteristics were similar between prison officers who completed follow-up and those lost to follow-up (Table 2), except for time in service, which showed a significant difference. This finding suggests that attrition was largely non-differential. Table 2Baseline characteristics of participants retained vs. lost to follow-up (2022-2024).Variable Retained (N = 53)Lost (N = 18)P-valueSMD*Age, mean ± SD† (years)42.4 ± 5.744.4 ± 7.10.2190.32Time in service (years)9.23 ± 2.446.67 ± 3.760.0010.81Sex (male), n (%)47 (88.7)16 (88.9)0.9810.006Smoker, n (%)4 (7.5)4 (22.2)0.0890.41Comorbidities, n (%)8 (15.1)2 (11.1)0.7030.12Work assignment (shift-based), n (%)50 (94.3)18 (100)0.3020.34Note. SMD < 0.20 is typically interpreted as a negligible imbalance between groups; values of 0.20-0.50 indicate a small imbalance, 0.50-0.80 moderate imbalance, and >0.80 large imbalance. This table compares sociodemographic and clinical characteristics at baseline for participants who completed both study phases (retained) and those who did not return in 2024 (lost to follow-up). Standardized Mean Differences (SMD) are presented as a measure of covariate balance; values of <0.20 indicate negligible imbalance between groups. *SMD: Standardized Mean Difference; †SD: standard deviation. Table 2Baseline characteristics of participants retained vs. lost to follow-up (2022-2024).Variable Retained (N = 53)Lost (N = 18)P-valueSMD*Age, mean ± SD† (years)42.4 ± 5.744.4 ± 7.10.2190.32Time in service (years)9.23 ± 2.446.67 ± 3.760.0010.81Sex (male), n (%)47 (88.7)16 (88.9)0.9810.006Smoker, n (%)4 (7.5)4 (22.2)0.0890.41Comorbidities, n (%)8 (15.1)2 (11.1)0.7030.12Work assignment (shift-based), n (%)50 (94.3)18 (100)0.3020.34Note. SMD < 0.20 is typically interpreted as a negligible imbalance between groups; values of 0.20-0.50 indicate a small imbalance, 0.50-0.80 moderate imbalance, and >0.80 large imbalance. This table compares sociodemographic and clinical characteristics at baseline for participants who completed both study phases (retained) and those who did not return in 2024 (lost to follow-up). Standardized Mean Differences (SMD) are presented as a measure of covariate balance; values of <0.20 indicate negligible imbalance between groups. *SMD: Standardized Mean Difference; †SD: standard deviation. Baseline characteristics of participants retained vs. lost to follow-up (2022-2024). Note. SMD < 0.20 is typically interpreted as a negligible imbalance between groups; values of 0.20-0.50 indicate a small imbalance, 0.50-0.80 moderate imbalance, and >0.80 large imbalance. This table compares sociodemographic and clinical characteristics at baseline for participants who completed both study phases (retained) and those who did not return in 2024 (lost to follow-up). Standardized Mean Differences (SMD) are presented as a measure of covariate balance; values of <0.20 indicate negligible imbalance between groups. *SMD: Standardized Mean Difference; †SD: standard deviation. Of the 16 prison officers who were IGRA-positive at baseline, 4 (25) did not return for the 2024 assessment. Among the 12 prison officers with positive IGRA results at baseline, three initiated and completed LTBI preventive treatment. Among the remaining IGRA-positive prison officers, preventive therapy was not initiated due to clinical contraindications, participant refusal, or operational factors related to referral and follow-up in the public health system. No indeterminate IGRA results were identified in either evaluation year, and all samples were valid for analysis. Sensitivity analyses showed that the reduction in IGRA positivity remained evident even under conservative assumptions. These findings indicate that the decline in IGRA positivity cannot be explained solely by attrition. In addition, because baseline characteristics were similar between retained and lost prison officers, attrition was unlikely to introduce systematic bias. The sociodemographic characteristics of the study population (2022-2024) indicate a predominantly male profile, with 88.7% (n=47) of prison officers being men and 11.3% (n=6) women. Most prison officers identified as White (71.7%, n=38), followed by Black (26.4%, n=14) and Mixed-race individuals (1.9%, n=1). The figures for educational qualifications showed that the majority held a higher education degree (88.7%, n=47), while smaller proportions had completed high school (7.5%, n=4) or had not completed high school (3.8%, n=2). As regards marital status, 73.6% (n=39) were married, 15.1% (n=8) were single, and 11.3% (n=6) were separated and/or widowed. Overall, the study population consisted predominantly of male, white individuals who were highly educated, and married. The data presented in Table 3 indicate that when considering factors associated with latent tuberculosis infection (LTBI), the most relevant lifestyle-related findings were the marked reduction in drug use, which decreased from 49.1% in 2022 to 7.5% in 2024, and the improvement in hepatitis B vaccination coverage, rising from 47.2% to 67.9%. From an LTBI risk perspective, indicators of exposure to tuberculosis showed a clear decline. The proportion of prison officers who reported knowing someone with TB decreased from 34.0% in 2022 to 17.0% in 2024, and reported contact with individuals with active TB also dropped substantially. In 2024, 84.9% reported no contact with TB cases, compared to 66.0% in 2022, suggesting reduced occupational or community exposure during the study period. Table 3Distribution of lifestyle-related characteristics of the study participants at two time points. 2022 2024 2022 2024 n%n%n%n%Fixed sexual partner Alcohol consumption Yes611.3611.3 Yes2649.13158.5No 4788.74788.7No 2750.92241.5Daily consumption of medication BCG scar* Yes1018.91834 Yes4279.24177.4No 4381.13566No 1120.81222.6Illegal drug use Knows someone with tuberculosis Yes2649.147.5 Yes1834917.0No 2750.94992.5No 35664279.2Smoker Contact with a person with tuberculosis† Yes611.31120.8 Less than once a week1222.611.9No 4788.74279.21 to 2 times a week 47.523.8Tattoo in the last year More than 3 times a week 11.923.8Yes35.711.9 Every day 11.911.9No 5094.35298.1No contact35664584.9Hepatitis B vaccination Comorbidities‡ Yes2547.23667.9 Yes1018.91018.9No 2852.81732.1No 4381.14381.1Note. *BCG: Calmette-Guérin bacillus. †Two participants did not respond (3,8%). ‡Comorbidities include HIV, diabetes mellitus, chronic kidney disease, immunosuppressive therapy, chronic liver disease, and hepatitis B/C. Due to the low frequency of individual conditions, all comorbidities were analyzed as a single dichotomous variable (present/absent). Table 3Distribution of lifestyle-related characteristics of the study participants at two time points. 2022 2024 2022 2024 n%n%n%n%Fixed sexual partner Alcohol consumption Yes611.3611.3 Yes2649.13158.5No 4788.74788.7No 2750.92241.5Daily consumption of medication BCG scar* Yes1018.91834 Yes4279.24177.4No 4381.13566No 1120.81222.6Illegal drug use Knows someone with tuberculosis Yes2649.147.5 Yes1834917.0No 2750.94992.5No 35664279.2Smoker Contact with a person with tuberculosis† Yes611.31120.8 Less than once a week1222.611.9No 4788.74279.21 to 2 times a week 47.523.8Tattoo in the last year More than 3 times a week 11.923.8Yes35.711.9 Every day 11.911.9No 5094.35298.1No contact35664584.9Hepatitis B vaccination Comorbidities‡ Yes2547.23667.9 Yes1018.91018.9No 2852.81732.1No 4381.14381.1Note. *BCG: Calmette-Guérin bacillus. †Two participants did not respond (3,8%). ‡Comorbidities include HIV, diabetes mellitus, chronic kidney disease, immunosuppressive therapy, chronic liver disease, and hepatitis B/C. Due to the low frequency of individual conditions, all comorbidities were analyzed as a single dichotomous variable (present/absent). Distribution of lifestyle-related characteristics of the study participants at two time points. Note. *BCG: Calmette-Guérin bacillus. †Two participants did not respond (3,8%). ‡Comorbidities include HIV, diabetes mellitus, chronic kidney disease, immunosuppressive therapy, chronic liver disease, and hepatitis B/C. Due to the low frequency of individual conditions, all comorbidities were analyzed as a single dichotomous variable (present/absent). Most prison officers reported having a BCG scar in both years (79.2% in 2022 and 77.4% in 2024), indicating prior vaccination. Table 4A total of 18 prison officers (25.3%) were lost to follow-up between 2022 and 2024 due to institutional transfers to other prison units (n = 2) and dismissals (n = 16) (n=53).Result2022 2024 n%IC* 95%n%IC* 95%Positive1222.60.12-0.3235.70.01-0.16Negative 4177.40.64-0.885094.30.84-0.98Note. *IC: confidence interval. Table 4A total of 18 prison officers (25.3%) were lost to follow-up between 2022 and 2024 due to institutional transfers to other prison units (n = 2) and dismissals (n = 16) (n=53).Result2022 2024 n%IC* 95%n%IC* 95%Positive1222.60.12-0.3235.70.01-0.16Negative 4177.40.64-0.885094.30.84-0.98Note. *IC: confidence interval. A total of 18 prison officers (25.3%) were lost to follow-up between 2022 and 2024 due to institutional transfers to other prison units (n = 2) and dismissals (n = 16) (n=53). Note. *IC: confidence interval. Table 4 shows the results of the IGRA test, used as an immunological marker suggestive of latent tuberculosis infection (LTBI). The findings indicate a substantial reduction in IGRA positivity over the study period. In 2022, 22.6% of prison officers tested positive, with a 95% confidence interval (95% CI: 12.8-32.2%)13 whereas in 2024 the proportion decreased to 5.7% (95% CI: 0.01-0.16). Consequently, the proportion of individuals with a negative test result increased from 77.4% in 2022 (95%CI: 0.64-0.88) to 94.3% in 2024 (95% CI: 0.84-0.98). DISCUSSION The prison environment favors transmission of Mycobacterium tuberculosis due to overcrowding, poor ventilation, and frequent exposure to infected individuals14,15. Such conditions increase the risk of infection and may subsequently result in latent tuberculosis infection (LTBI). In this study, the proportion of IGRA-positive officers decreased from 22.6% in 2022 to 5.7% in 2024. It is important to emphasize that LTBI is conceptually a cumulative condition. Individuals with a previous positive IGRA result remain classified as having latent tuberculosis infection for clinical and public health purposes, regardless of subsequent IGRA reversion. Therefore, the reduction observed in this study refers to a decrease in IGRA positivity over time rather than definitive evidence of infection clearance at the population level. IGRA reversion has been documented in longitudinal studies and may reflect biological variability in IFN-γ responses, values near the assay cutoff, or inherent analytical variability of the test, rather than true elimination of infection. The proportion of IGRA positivity observed in 2022 was similar to international estimates (25.6%)16, but lower than reported in São Paulo (57%)17, Franco da Rocha (33.7%)18, and Colombia (39-56%)19. These differences may be related to contextual factors such as surveillance practices, environmental conditions, or prevention strategies. However, these factors were not directly measured in the present study20. Although educational activities may have contributed to increased awareness, causal inference is not possible. The reduction in reported contact with TB cases may reflect contextual changes within the facility. In addition, given the possibility of false-negative IGRA results, continued monitoring remains essential21. The sociodemographic profile of the prison officers, predominantly male, well-educated, and married, may influence adherence to preventive measures. This profile may also reflect the higher vulnerability of men to tuberculosis due to behavioral and occupational factors7,22. Drug use decreased substantially during the study period (49.1% to 7.5%), which may indicate behavioral changes with potential protective implications. In contrast, smoking increased from 11.3% to 20.8%, a relevant risk factor for tuberculosis23. Increased alcohol use aligns with evidence linking prison work to stress and common mental disorders24. Attrition represented an important methodological consideration in this study (25.3%), including four prison officers who were IGRA-positive at baseline. Among prison officers who completed both assessments, the number of IGRA-positive individuals decreased from 12 in 2022 to 3 in 2024. Several factors may explain this change: First, three IGRA-positive prison officers completed LTBI preventive treatment and subsequently tested negative. Second, four IGRA-positive individuals were lost to follow-up and did not attend the 2024 evaluation. Third, spontaneous IGRA reversion occurred in three prison officers, a phenomenon frequently reported in longitudinal studies, particularly when baseline interferon-γ values are close to the assay cutoff. First, three IGRA-positive prison officers completed LTBI preventive treatment and subsequently tested negative. First, three IGRA-positive prison officers completed LTBI preventive treatment and subsequently tested negative. Second, four IGRA-positive individuals were lost to follow-up and did not attend the 2024 evaluation. Second, four IGRA-positive individuals were lost to follow-up and did not attend the 2024 evaluation. Third, spontaneous IGRA reversion occurred in three prison officers, a phenomenon frequently reported in longitudinal studies, particularly when baseline interferon-γ values are close to the assay cutoff. Third, spontaneous IGRA reversion occurred in three prison officers, a phenomenon frequently reported in longitudinal studies, particularly when baseline interferon-γ values are close to the assay cutoff. Sensitivity analyses were carried out to assess whether the reduction in IGRA positivity could be explained entirely by attrition. In a conservative worst-case scenario, assuming that all four lost IGRA-positive prison officers had remained positive, the prevalence in 2024 would still be lower than at baseline (12 vs. 7). In the complete-case analysis, which included only prison officers reassessed in both waves, the reduction from 12 to 3 persisted. Furthermore, baseline characteristics did not differ significantly between retained and lost prison officers, suggesting that attrition was unlikely to be systematically associated with LTBI risk. Sensitivity analyses, including worst-case, best-case, complete-case, and tipping-point scenarios, are detailed in Table 1 and demonstrate that the observed reduction in IGRA positivity remains robust even under conservative assumptions. Therefore, IGRA reversion in this cohort should not be interpreted as definitive evidence of infection clearance but rather as a phenomenon potentially influenced by immunological variability and assay-related factors. Although few longitudinal studies have specifically evaluated IGRA dynamics among prison officers, serial testing studies in other occupational groups with increased exposure to Mycobacterium tuberculosis, particularly healthcare workers, have reported similar patterns of IGRA conversion and reversion. Reported reversion rates range from approximately 10% to 40%, and these changes are often attributed to biological variability in interferon-γ responses, fluctuations around the assay cutoff, or analytical variability of the assay25-27. These findings are consistent with systematic reviews demonstrating that IGRA conversion and reversion are relatively common in serial testing programs28. Healthcare workers provide a relevant comparison group because both populations experience occupational exposure to Mycobacterium tuberculosis and are often monitored through serial screening programs. Finally, the absence of a control group precludes attributing the observed reduction in IGRA positivity or behavioral changes to the educational intervention. Although the intervention may have improved knowledge and awareness, the lack of a comparator means that the findings should be interpreted as temporal associations rather than causal effects. In addition, the attrition rate, particularly the loss of four IGRA-positive prison officers, may have partially contributed to the observed decline. Another limitation is that, although clinical comorbidities were assessed, information on some causes of immunosuppression, such as the use of biologic agents or other immunosuppressive therapies, was not specifically collected. These conditions may influence IGRA responses. This study also has several methodological strengths, including its longitudinal design and the use of IGRA testing to monitor LTBI among prison officers. Despite the limitations described above, sensitivity analyses demonstrated that the findings remained robust even under conservative assumptions. These results highlight the importance of systematic screening, preventive strategies, and psychosocial support for prison officers. Strengthening occupational health policies, improving working conditions, and implementing mental health interventions are essential to reduce tuberculosis transmission and mitigate the impact of occupational stress.