Losses in the Care Cascade for Latent Tuberculosis Infection in the Tracing Contact Studies

Background: The control of latent tuberculosis infection (LTBI) encompasses multiple stages. The objective was to calculate the losses in the LTBI care cascade for pulmonary TB contacts in Catalonia (Spain). Methods: The LTBI care cascade was studied for pulmonary TB contacts reported from 1 January 2019 to 30 June 2021, considering three dependent variables: non-performance of testing; non-receipt of a treatment prescription; and non-adherence to treatment. Variables associated with the cascade were analysed using adjusted OR (aOR) and 95% confidence intervals (CI). Results: Identified from 847 cases of pulmonary TB were 7087 contacts, of whom 6537 (92.2%) could be screened for LTBI. LTBI prevalence was 25.5% (1670/6537); 69.4% of persons with LTBI (1159/1670) received a treatment prescription and 71.3% (827/1159) completed it. Treatment prescription was associated with age ≥65 years (aOR = 0.3; 95%CI: 0.2–0.6) and a daily exposure of ≥6 h to the TB index case (aOR = 3.6; 95%CI: 2.6–5.0). Treatment adherence was lower in men (aOR = 0.7; 95%CI: 0.5–1.0) and immigrants (aOR = 0.7; 95%CI: 0.5–0.9). Conclusions: Under 50% of contacts make it to the end of the LTBI cascade. Losses need to be reduced through education of both healthcare providers and patients and through treatment monitoring. The greater involvement of primary care physicians could help in monitoring and controling LTBI.


Introduction
The United Nations (UN) and the World Health Organization (WHO), having set ambitious targets to reduce the global burden of tuberculosis (TB) by 2030, recognize the essential role of tackling latent tuberculosis infection (LTBI) as a strategy for controlling and eliminating TB [1].
The control of LTBI encompasses multiple stages, starting with identifying a population for testing-using the tuberculin skin test (TST) and/or the interferon gamma release assay (IGRA)-and ending with treatment completion [2][3][4].Several studies have indicated Pathogens 2023, 12, 1403 2 of 9 that significant losses occur at all stages of the LTBI cascade, but especially in the early stages and with treatment prescription and adherence to treatment [5].
In the early stages of the TB contact cascade, public health teams need to ensure that all pulmonary TB cases are reported and registered and also that their contacts are identified and registered for screening purposes [2,6].In addition, protocols need to clearly establish the priority of prescribing treatment to LTBI-positive contacts, while healthcare providers, especially those in primary care, need to be aware of and correctly apply the recommendations of TB elimination guidelines [7,8].Finally, public health services need to monitor treatment adherence to reduce the risk of new TB cases that could hinder the goal of elimination [9].
Mathematical modelling studies have demonstrated that diagnosing and treating LTBI in people at a high risk of developing active TB, such as contacts, accelerates TB elimination [10,11].However, the obstacles to the practical implementation of an elimination policy need to be understood.Important information for TB management can be obtained by evaluating the LTBI care cascade and the factors associated with losses at different stages [12].
The aim of the study was to estimate LTBI care cascade losses at three stages, namely, the identification and testing of contacts, treatment prescription, and treatment adherence, in contacts of pulmonary TB cases in Catalonia (Spain).

Materials and Methods
A prevalence study was conducted on pulmonary TB contacts reported in the 30-month period from 1 January 2019 to 30 June 2021.The inclusion criterion for pulmonary TB cases was having at least one contact who could be identified and screened.The study population consisted of all registered and studied contacts.
Pulmonary TB cases that met the inclusion criteria were epidemiologically surveyed by epidemiology service technicians.All registered contacts were tested (TST and/or IGRA), and also completed a questionnaire on the context and time of exposure to the index case, cohabitation, smoking status, and alcohol-related medical risk (daily intake of >40 g (men) and 24 g (women) or a medical record indicating alcohol abuse).
Contacts who tested positive for IGRA or TST (≥5 mm) were considered to be infected and underwent a posterior-anterior chest X-ray to rule out TB. Sputum samples were obtained from individuals with lesions suggestive of TB to identify acid-fast bacilli and prepare cultures.
We used an LTBI care cascade model, as previously published (12), consisting of the following stages: (1) the identification of close pulmonary TB contacts (reference population); (2) initial screening for LTBI; (3) medical examination including IGRA, TST, and/or X-ray evaluation; (4) recommendation to receive treatment; (5) acceptance and initiation of treatment; and (6) treatment compliance (defined as 80% of the prescribed medication).
Definition of LTBI cascade loss.For the purposes of this study, we considered LTBI cascade losses among the registered participants considered as candidates for initial LTBI screening and also: (1) contacts who failed to show up for testing, test results, or the second TST or IGRA when indicated, or (2) contacts with LTBI who were not prescribed LTBI treatment, or (3) contacts with LTBI who had received a prescription but who failed to treatment compliance.
The LTBI care cascade was studied considering three dependent variables: (1) contacts who were not screened for LTBI, (2) contacts with LTBI who were not prescribed treatment, and (3) contacts with LTBI who were prescribed but failed treatment compliance.Independent variables were age, sex, immigrant status, cohabitation with the index case, the duration of exposure to the index case, tobacco use, alcohol use, exposure to a smear-positive index case, and/or a chest X-ray showing cavernous lesions.
Prevalence rates were calculated as follows: contacts not tested for LTBI; tested contacts who were LTBI-positive; LTBI-positive contacts who received a treatment prescription; and LTBI-positive contacts who adhered to the prescribed treatment.To calculate the cumulative losses at each cascade stage, the proportion remaining in a stage was multiplied by the proportion remaining after the previous stage [13].
Factors associated with treatment prescription and adherence were determined by calculating the odds ratio (OR) and 95% confidence interval (CI).Variables associated with treatment prescription and adherence were calculated as adjusted OR (aOR) and 95% CI values using multivariate logistic regression models developed using the backward stepwise method.
The study was approved by the Ethics Committee of the Arnau Vilanova University Hospital (CEIC-2049) and was conducted according to Declaration of Helsinki principles.All subjects included in the study received detailed information on the study aims before inclusion.
Data collection and cleaning were performed using the Access 12.0 database manager of the MS Office 2013 software package (Microsoft, Redmond, Washington, DC, USA).The analysis was performed using the SPSS v.24 statistical package and the EpiInfo software v7.2.
Prevalence rates were calculated as follows: contacts not tested for LTBI; tested contacts who were LTBI-positive; LTBI-positive contacts who received a treatment prescription; and LTBI-positive contacts who adhered to the prescribed treatment.To calculate the cumulative losses at each cascade stage, the proportion remaining in a stage was multiplied by the proportion remaining after the previous stage [13].
Factors associated with treatment prescription and adherence were determined by calculating the odds ratio (OR) and 95% confidence interval (CI).Variables associated with treatment prescription and adherence were calculated as adjusted OR (aOR) and 95% CI values using multivariate logistic regression models developed using the backward stepwise method.
The study was approved by the Ethics Commi ee of the Arnau Vilanova University Hospital (CEIC-2049) and was conducted according to Declaration of Helsinki principles.All subjects included in the study received detailed information on the study aims before inclusion.
Data collection and cleaning were performed using the Access 12.0 database manager of the MS Office 2013 software package (Microsoft, Redmond, Washington, USA).The analysis was performed using the SPSS v.24 statistical package and the EpiInfo software v7.2.
Another variable associated with treatment prescription was exposure time to the index case.Since exposure time is associated with the risk of developing LTBI and TB [18], increasing the prescription rate would help to reduce this risk.However, briefer or sporadic but intense exposure (in terms of the volume of air shared with a pulmonary TB case) also entails a high risk [18], so any prescription failures in these cases also represent cascade losses.
Our rate of 71.3% for treatment adherence overall is similar to that reported in other studies [13,[19][20][21][22], and was even higher for those aged <18 years, a group at high risk of developing TB [14,15].The main factors associated with a lower adherence and greater cascade losses were immigrant status and risky alcohol use.Social problems in some groups of immigrants make them more susceptible to treatment non-adherence, as reported by a number of studies [23][24][25][26].Immigrants thus need to be especially targeted by TB programmes [3,8,27,28], as health provider proactiveness and the direct monitoring of treatment adherence could reduce the corresponding losses in the LTBI care cascade [9].Although not statistically significant in our logistic regression analysis, risky alcohol use entails an enhanced risk of hepatotoxicity and thus may partially explain treatment nonadherence [2,28,29].
Our results are consistent with the findings of other studies regarding care cascade losses [13,22].The lower level of prescriptions to older people, associated with insufficient guideline knowledge by healthcare providers, coincides with systematic review observations [13], while other documented reasons for losses include contacts not interested in knowing if they have been infected and contacts perceiving a low risk of infection and disease severity [22].
Public health services need to address the serious health risks associated with TB and the potential sequelae for the respiratory system [23].Public health teams and community health agents are crucial in order to, first of all, register all cases of pulmonary TB and any contacts that may be candidates for LTBI screening, and then to ensure that LTBI testing is exhaustive [3,8,9].A key aspect is to identify vulnerable groups at high risk of not contacting the health system.Facilitating access to screening tests, avoiding delays in results, and ensuring social support through community health agents could reduce losses in the early stages of the cascade [30].
Treatment prescription is also crucial to the goal of ultimately eliminating TB, so public health services need to have clear and explicit protocols in place [23,31] and health providers need to scrupulously apply protocol recommendations.Finally, LTBI-positive contacts receiving treatment need to be followed up in TB control programmes to identify and treat possible side effects and ensure adherence.
This study has some limitations.We only included TB cases in which at least one contact was identified.Thus, given that TB cases without contacts could also account for unscreened contacts, losses may be underestimated, although we did not observe this effect.Another possible underestimation arises from the fact that prescription-and treatmentassociated factors in the logistic regression models were estimated only for contacts for whom the relevant data were available.

Conclusions
Our results, pointing to important losses in the LTBI care cascade, suggest a need for patient education, incentives, home visits, and treatment monitoring.The control of LTBI, as key to achieving the goal of ultimately eliminating TB, also requires the implication of the entire health system, but especially of TB clinics [9] and primary care physicians [5].The inclusion of LTBI data in computerized medical record systems, with primary care protocols and alerts already existing for the control of chronic pathologies, could greatly assist in LTBI monitoring and control.

Figure 1 .
Figure 1.Latent tuberculosis infection care cascade for contacts of pulmonary tuberculosis index cases in a low-incidence region.# number of cases

Figure 1 .
Figure 1.Latent tuberculosis infection care cascade for contacts of pulmonary tuberculosis index cases in a low-incidence region.# number of cases.

Table 1 .
Factors associated with latent tuberculosis infection (LTBI) screening of contacts of pulmonary tuberculosis index cases in Catalonia (Spain) (N = 7087).

Table 2 .
Factors associated with treatment prescription for latent tuberculosis infection (N = 1670).