Prevalence of latent tuberculosis and treatment adherence among patients with chronic kidney disease in Campo Grande , State of Mato Grosso do Sul Aucely

Introduction: The primary strategy for tuberculosis control involves identifying individuals with latent tuberculosis. This study aimed to estimate the prevalence of latent tuberculosis in chronic kidney disease (CKD) patients who were undergoing hemodialysis in Campo Grande, State of Mato Grosso do Sul, Brazil, to characterize the sociodemographic and clinical profi les of patients with latent tuberculosis, to verify the association between sociodemographic and clinical characteristics and the occurrence of latent tuberculosis, and to monitor patient adherence to latent tuberculosis treatment. Methods: This epidemiological study involved 418 CKD patients who were undergoing hemodialysis and who underwent a tuberculin skin test. Results: The prevalence of latent tuberculosis was 10.3%. The mean patient age was 53.43±14.97 years, and the patients were predominantly men (63.9%). The population was primarily Caucasian (58.6%); half (50%) were married, and 49.8% had incomplete primary educations. Previous contact with tuberculosis patients was reported by 80% of the participants. Treatment adherence was 97.7%. Conclusions: We conclude that the prevalence of latent tuberculosis in our study population was low. Previous contact with patients with active tuberculosis increased the occurrence of latent infection. Although treatment adherence was high in this study, it is crucial to monitor tuberculosis treatment administered to patients in health services to maintain this high rate.

Tuberculosis (TB) is a serious public health problem.It is the second most frequent cause of death from infectious disease worldwide.According to the World Health Organization 1 , Brazil ranks 17 th in a group of 22 countries that account for 82% of new TB cases worldwide.
In 2012, 70,047 new cases emerged in Brazil, with an incidence rate of 36.1 cases per 100,000 inhabitants.In 2011, 4,563 deaths from TB occurred, resulting in a mortality rate of 2.4 deaths per 100,000 individuals.The cases were concentrated in 315 of the 5,564 municipalities in the country; these 315 municipalities accounted for 70% of all cases 2 .The disease often affects people of reproductive age and frequently leads to hospitalization, thus signifi cantly increasing National Health Service costs 3,4 .
In 2012, the incidence of TB in the State of Mato Grosso do Sul was 37.7 cases per 100,000 inhabitants and resulted in 65 deaths 2 .In the capital, Campo Grande, 316 cases were reported; however, latent infection with Mycobacterium tuberculosis was not documented 5 .Therefore, one of the aims of this study was to investigate the prevalence of latent tuberculosis (LTB) in patients with chronic kidney disease (CKD), the 12 th and 17 th leading causes of death and disability worldwide, respectively 3,4 .
The morbidity and mortality rates of TB are high, particularly in patients of reproductive age, which results in increased treatment costs for CKD patients and renders TB a serious public health problem [6][7][8] .
According to the 2011 Brazilian dialysis census report, the estimated number of patients participating in a chronic dialysis program in Brazil exceeded 91,314.The cost of the dialysis program combined with renal transplants in Brazil is approximately 1.4 billion reals per year 9 .
Tuberculosis diagnosis in dialysis patie nts can be complicated and diffi cult because of the increased frequency of extrapulmonary involvement, which may result in atypical clinical manifestations

METHODS
and nonspecifi c symptoms 10 .Extrapulmonary TB was more commonly identifi ed than pulmonary TB, with tuberculous lymphadenitis being the most frequent form 11 .
Latent tuberculosis was characterized by the presence of bacilli in host tissues without clinical or radiological evidence of disease 12 .LTB affects approximately one-third of the world's population.It is estimated that 50 million Brazilians are infected with M. tuberculosis.Approximately 5% of the individuals carrying the tuberculosis bacillus will develop the active form of the disease, and 95% will present with LTB 12 .
The tuberculin skin test (TST) is widely used to detect individuals with latent infection 13 .However, some TST limitations have been reported, including anergy as a consequence of chronic impaired renal immunity, possible false-negative results, and low test sensitivity 14,15 .
Impaired immunity in people with CKD is the main risk factor for the development of TB.In recent years, the incidence rate of TB has been increasing in this population because of immunosuppression, particularly in individuals who have undergone renal transplantation 9 .
Immune system weakness in CKD occurs as a result of chronic infl ammatory conditions, renal replacement therapies, and chronic renal failure or other comorbidities 15 .Moreover, nutritional status, vitamin D defi ciency, and hyperparathyroidism contribute to impaired immunity in CKD patients 16 .
An increased risk for TB in dialysis patients was fi rst reported by Pradhan et al. 17 in 1974.Decades later, another study showed a signifi cantly increased risk of TB in patients with chronic renal failure who were undergoing dialysis 18 .In cohort studies, the relative risk of TB among such patients was 3.4 to 25.3 times greater than that among the general population [19][20][21][22] .
Tuberculosis is the second leading cause of death from infectious disease, and its control is one of the millennium development goals, which aim to halve the numbers of cases and deaths from TB by 2015 23 .Therefore, the ability to identify individuals at increased risk for developing TB has important implications for public health policy and patient care 24 .
Considering this goal, the aims of this study were to estimate the prevalence of LTB in CKD patients who were undergoing hemodialysis, to characterize the sociodemographic and clinical profi les of patients with latent tuberculosis, to verify the association between sociodemographic and clinical characteristics and the occurrence of latent tuberculosis, and to monitor the treatment adherence of LTB patients in Campo Grande, State of Mato Grosso do Sul (MS), Brazil.
In this epidemiological study, we conducted a descriptive survey of the six existing dialysis services in Campo Grande, MS, Brazil, from July 2011 to December 2013.
A nonprobability convenience sample was used.The six dialysis centers in Campo Grande served 772 CKD patients, 418 (54.1%) of whom participated in this study.
Patients older than 18 years, CKD patients, and patients undergoing hemodialysis were included in the study.Indigenous people, pregnant women, prisoners, individuals being treated for TB, and patients who refused to sign the informed consent form were excluded.
We used a standardized sociodemographic data questionnaire (i.e., sex, age, race, education, marital status, provenance, origin, family income, and occupation prior to CKD) and gathered information concerning other variables, such as the duration of hemodialysis-associated disease, use of medication, presence of a bacillus Calmette-Guérin (BCG) vaccination scar, and history of contact with TB or an active TB patient.
The participants underwent TSTs using the Mantoux method, which consists of the application of intradermal tuberculin purifi ed protein derivative (PPD) RT23 (State Serum Institute, Copenhagen, Denmark) in the middle third of the anterior surface of the forearm at a dose of 0.1mL, equivalent to 2 tuberculin units.Forty-eight hours after the application, the reading was performed by a trained professional who measured the induration at its largest transverse axis with a ruler.The result was expressed in millimeters.
In the TST, an induration cutoff value of 10mm was used for patients with chronic renal failure.This value aligned with the recommendations of the Ministry of Health and III Guidelines for Tuberculosis of the Brazilian Society of Pneumology and Tisiology 2,26 .Patients with TST values ≥10mm underwent chest radiography and were evaluated by a pulmonologist to exclude active disease.
All patients who were considered positive for LTB but with no radiological evidence of active TB were directed to undergo a 6-month treatment with isoniazid at the maximum allowable dose of 300mg/day after hemodialysis, as recommended by the Ministry of Health and III Guidelines for Tuberculosis of the Brazilian Society of Pneumology and Tisiology 2,26 .
A 95% confi dence interval (CI) for the prevalence of TB was estimated assuming a fi nite population size and using an F distribution 25 .
For statistical analysis of categorical or nominal variables, frequency distribution tables were compiled.To test the associations among categorical variables, we used Pearson's chi-squared test or, when the expected values in the tables fell below fi ve, Fisher's exact test.All analyses were performed using the Statistical Analyses System (SAS) 9.1 program (World Headquarters, SAS Institute, USA) 27 .The prevalence ratios were calculated by point and 95%CI, with the estimates obtained via Poisson regression, and robust variance using the GENMOD procedure in SAS.

Ethical considerations
The ethics committee of the Federal University of Mato Grosso do Sul approved this study under protocol number 2042/0138.0.049000.
The study sample consisted of 418 patients who voluntarily agreed to participate.The prevalence of LTB was 10.3% (95%CI, 8.40-12.54).The hemodialysis services, patient population undergoing hemodialysis, sample size, and TST results that were considered positive for LTB are shown in Table 1.
Regarding their provenance, the patients were divided into two groups, namely those who came from the interior and those who came from the capital.These groups were further divided according to city region and organized into health districts.Of the study population, 81.1% came from the following districts in the capital: east, north, west, and south.In Campo Grande in 2012 (Table 2), we observed that 331 active TB cases were reported; this result was proportional to the number of LTB cases found in this study.
The results of the evaluation of the association between the TST results and the epidemiological characteristics are described in Table 3.No signifi cant association was found (p>0.05).
The clinical characteristics were also evaluated to assess the existence of an association with the TST results (Table 4).The only variable analyzed that resulted in a signifi cant association was contact with a TB patient (p≤0.05), as evidenced by the greater proportion of positive results among patients who reported contact with a person with TB.The risk ratio for having LTB among those who were previously exposed to TB was 8.44 (95%CI=4.97-14.35).Therefore, among the individuals who had contact with persons with active TB, LTB was 8.5 times more prevalent than in those who did not have contact.
More deaths were observed among the patients with positive TST results than among those with negative results; however, the causes of death were not related to the development of active TB.
The etiologies of CKD and the evaluated comorbidities, which are presented in Table 5, showed that patients with hypertension and glomerulonephritis, whether separately or together, did not have positive TST results.In contrast, patients with uropathies (18.8%) did have positive TST results.
Upon chest radiography, 39 patients (90.7%) showed no changes, whereas 4 patients presented with radiographic changes and underwent thorax tomography and bronchoscopy.None of the patients were diagnosed with active TB after TST, and none presented with respiratory symptoms.One patient, who was in the fourth month of treatment for LTB, underwent kidney transplantation and developed symptoms of weight loss, low-grade fever, and pleural effusion, which were diagnosed by the medical team as extrapulmonary TB.After beginning TB treatment, the patient's clinical condition signifi cantly improved.
Adherence to LTB treatment with isoniazid was 97.7%.After 4 days of treatment, two patients developed psychosis, making it necessary to discontinue the medication and replace isoniazid with another drug, rifampin.Three patients had adverse reactions to epigastralgia, although these reactions did not prevent the use of isoniazid.

DISCUSSION
This study demonstrated that the incidence of LTB was low in chronic renal failure patients who were undergoing hemodialysis.Previous contact with individuals with active TB was signifi cantly associated with latent infection.
Note that one-third of the world's population has an M. tuberculosis infection 23 .For effective TB control, it is important to identify and to treat active and latent TB cases 9,28 .
Chronic renal failure patients who are undergoing hemodialysis are approximately 10 times more likely to develop TB than the general population 30 .The greatest risk for developing the disease occurs two years after being infected with M. tuberculosis 31,32 .
In Brazil, TST is the classic diagnostic tool for LTB.It is essential to perform the TST on an admitted chronic renal failure patient attending a regular hemodialysis program 33 .
The prevalence of LTB in this study was 10.3%, which is similar to that observed in a cohort study conducted in India (10.5%) 34 .These results, however, disagree with the prevalences of 15% from studies in Belgium 35 and 20% from studies in Berlin 36 .
The LTB prevalence detected in this study also differs from that obtained in Minas Gerais, in which the value was 28.5% 37 , and in Turkey, in a study by Ates et al., in which the prevalence was 37.5% 38 .We emphasize that this variation may be related to the burden of TB in each region.Although Brazil is classifi ed as a country with a high TB burden, the State of Mato Grosso do Sul only has a moderate burden of the disease, with an incidence rate of 37.7 cases per 100,000 inhabitants 1,39 .
Concerning the demographic characterization of the sample, the results regarding sex and marital status were similar to those obtained in the studies conducted by Fonseca et al. 37 and Rao et al. 40 .
There was no signifi cant association between the presence of LTB and the measured variables (i.e., age, gender, ethnicity, origin marital education class, BCG vaccination scar, and hemodialysis duration).These results align with those in the study performed in Turkey by Ates et al. 38 and with the study of Ribeiro et al. 41 , which took place in the interior of São Paulo and included study subjects that had been on hemodialysis for at least three years.
Among the underlying causes of chronic renal failure, hypertension, diabetes mellitus, and glomerulonephritis were dominant.In the case of the latter, when assessed separately in this study, none of the patients had a positive TST result.The prevalence of these diseases was also detailed in the studies performed in Brazil by Cherchiglia et al. 8 and Sesso et al. 42 .
We observed that the most common medications used were vitamins and minerals, antianginals and antiplatelets, antihypertensives, and cardiotonic supplements.These fi ndings differ from those in a study that found erythropoietin, intravenous iron, vitamin D, and sevelamer were the predominant medications used 42 .We believe that these differences may be explained by the different pathological and clinical conditions presented by the patients.
In this study, previous contact with patients with active TB was signifi cantly associated with latent infection.This result was supported by studies that reported that prior contact with an active TB patient is required for the development of latent infection 12,36 .A signifi cant percentage of cases of active TB is usually due to the reactivation of LTB, which reinforces the need to limit contacts with patients with active TB in the community.The frequency of previous contact with TB patients was a key fi nding in this study; 80% of the patients with LTB reported such contact.This result differs from those obtained by Fonseca et al. 37 , whose research showed a rate of 20.6%.This rate may be biased by the patients' lack of clarity regarding what constituted contact.
The treatment of TBL is a preventive measure aimed at reducing the risk of progression among the large group of infected individuals for TB ativa 2,43,44 .Other studies have reported that supervised treatment is one of the main tools for TB treatment adherence 45,46 .
Treatment adherence in this study reached 97.7%, which was greater than the adherence rates reported in other studies (e.g., 87.7% and 44%) 47,48 .
Positive adherence was obtained in the present study.We built on the strategies recommended by the Tuberculosis Control Program regarding the employment of capable human resources and the effective distribution of medications.These strategies include the choice of medications, regimen, and treatment duration, as well as determining patient needs and motivations to stay in treatment 2 .
Many strategies have been proposed 2 .The treatment in this study was self-administered, with monthly delivery of medication by the researcher.The participants were followed up in their respective dialysis clinics and counseled regarding the importance of treatment adherence and adverse reactions.Many of these patients spent much time on dialysis and, thus, experienced fi nancial and mobility diffi culties.We believe that these measures and systematic visits contributed to effective treatment adherence and may have hindered further shifting to and searching for other services to treat LTB.
A study of treatment cessation in users of health services in Recife highlighted the importance of having a health provider with whom the patient has close ties.This relationship is essential to the success of treatment adherence, as it facilitates the negotiation of the proposed actions with the user and the consequent care management 49 .
In Mato Grosso do Sul, the present study was the fi rst to investigate the prevalence of LTB in chronic renal failure patients who were undergoing hemodialysis.To reduce the risk of developing active TB and, therefore, favor treatment that reduces the risk from 90% to 60%, it is essential that health services adopt protocols to detect LTB, particularly for patients who are awaiting transplantation 43,50 .
In this study, we demonstrated that the prevalence of LTB in chronic renal failure patients who were undergoing hemodialysis was 10.29% and that previous contact with individuals with active TB was associated with the occurrence of latent infection.Treatment adherence was high, and the strategy of having patients self-administer drugs combined with systematic patient monitoring by a researcher in the hemodialysis unit might have contributed to the high treatment adherence in this study.Therefore, we suggest the adoption of this strategy in other health service areas that require preventive LTB treatment.
The authors declare that there is no confl ict of interest.

TABLE 1 -
Results regarding the number of patients with CKD and positive TST results obtained from the hemodialysis services of Campo Grande, State of Mato Grosso do Sul, Brazil, in 2013.

TABLE 2 -
Prevalence of individuals with TB in the districts of Campo Grande, according to latent tuberculosis in hemodialysis services, of Campo Grande, State of Mato Grosso do Sul, Brazil, 2013.

TABLE 3 -
Epidemiological characteristics of the CKD patients, according to the TST results obtained from the hemodialysis services of Campo Grande, State of Mato Grosso do Sul, Brazil, in 2013.

TABLE 5 -
Disease distribution among the CKD patients, according to the TST results obtained from the hemodialysis services of Campo Grande, State of Mato Grosso do Sul, Brazil, in 2013.

TABLE 4 -
Clinical characteristics of the CKD according to the TST results obtained from the hemodialysis services of Campo Grande, State of Mato Grosso do Sul, Brazil, in 2013.