English (pdf)
Article in xml format
Article references
Automatic translation
Send this article by e-mail
Cited by SciELO 
Cited by Google 
Similars in
SciELO 
Similars in Google 
Permalink
Introduction
The risk of tuberculosis (TB) transmission among healthcare workers (HCWs) is a worldwide issue that has been reported by all countries, regardless of their burden of disease. However, it has not received the attention it deserves in low-income countries such as Colombia.1
In the country, most healthcare institutions lack the resources and policies to prevent the transmission of Mycobacterium tuberculosis (the bacterium responsible for most TB cases), not only between the patient-HCW duo, but also between visitors, administrative staff and other collaborators. Moreover, due to the lack of recognition of this problem, the epidemiological link is difficult to demonstrate in most cases. 2,3
The risk that HCWs have of contracting TB depends on many factors, including the time they spend in the health facility, the area where they perform their duties (emergency room, radiology department, clinical laboratory, morgue, etc.), their position (respiratory therapist, pulmonologist, otolaryngologist, diagnostic imaging technician), the use of personal protective equipment, their training in infection control and, of course, their immune status. 4
The World Health Organization (WHO) 5 estimated that in 2018, around the world, there were 10 million (range: 9.0-11.1 million) new cases of all forms of TB -8.6% among individuals with HIV (human immunodeficiency virus)- and that 1.2 million people (range 1.1-1.3 million) died from TB disease (137 deaths per hour), of which 251 000 (range: 223 000-281 000) were HIV positive.
In Colombia, the Instituto Nacional de Salud (National Health Institute) 6 established that 14 480 cases of all forms of TB were reported in 2017, being more frequent in men (64.33%). Likewise, it was established that 20.04% of the cases corresponded to people over 65 years of age; 83.30% were cases of pulmonary TB; 78.99% were admitted as laboratory-confirmed cases, with acid-fast bacillus (AFB) stain in sputum being the most used diagnostic method; incidence was 26.5 cases per 100 000 inhabitants, and HIV co-infection was 11.09%.
Also in 2017, the Health Department of Bogotá7 stated that 1 455 cases of TB in all forms had been reported in the city, also with a predominance in males (64%). Furthermore, it revealed that the overall mortality was 16.66%; that 20.83% of the cases had HIV co-infection, making Bogotá the city with the highest TB/HIV co-infection rate in the country; that 2.40% of the cases occurred in HCWs; and that of the 1 128 cases in which the type of TB could be determined, 68.70% corresponded to pulmonary TB, 22.78% to extra-pulmonary TB, and 8.52% to meningeal TB.
Huaroto & Espinoza8 suggest that the first step that healthcare institutions should take to prevent TB transmission, regardless of their size or level of care, is to assess the risk of the facility in order to recognize the areas with the highest risk of transmission; they should then design and implement appropriate interventions.
In Colombia, research on the real impact of TB on HCWs is limited; moreover, until 2019, national standards on specific measures to control TB transmission in health institutions were virtually non-existent. 3,9
Recently, in order to prevent TB, the Colombian Ministry of Health and Social Protection issued Resolution 227 of 2020, 10 whereby "the technical and operational guidelines of the National Program for the Prevention and Control of Tuberculosis" were adopted. Although two of its chapters state that HCWs should be prioritized in the diagnosis of active and latent TB, the measures described are comparable to the recommendations made for other populations, such as diabetic patients or smokers. In this sense, on the one hand, it is considered that the intentions of government authorities are not yet in line with the relevance of this problem in the country and, on the other hand, it is evident that the recommendations made in this document are far from being a reality, even in health institutions with available resources.
In 2017, Ochoa et al.11 published a study carried out in 1 218 HCWs in Medellín, Colombia, who were administered a latent tuberculosis infection (LTBI) prevalence survey, using the tuberculin skin test (TST), and a Quantiferon test (QFT), in which they found that the marginal estimate of the prevalence P(LTBI+) was 62.1% (95%CI: 53.0-68.2).
Given this scenario, the present study aimed to describe the demographic and clinical characteristics of HCWs with TB treated in a university hospital in Bogotá, D.C., the histological and radiological findings, and the main criteria for diagnosing TB in these patients.
Materials and methods
Case series study conducted in a quaternary care university hospital in Bogotá, D.C., whose users belong to both the contributory and subsidized insurance regimes of the Colombian health system.
Information on TB cases diagnosed between January 2008 and December 2018 in the hospital was extracted from the Institutional Office of Hospital Epidemiology's databases, where it was found that the institution reported 199 confirmed cases of TB in all forms to the National Public Health Surveillance System (SIVIGILA), of which 24 (12%) were healthcare workers. The medical records of these patients were consulted in the hospital's medical records system in order to collect information on their clinical and sociodemographic characteristics, as well as on radiological and laboratory findings (pathology, microbiology, and molecular biology).
In all cases, it was found that the laboratory received more than one sample for each test (microbiological, histological, molecular) and that the first test reported as positive was the one considered as the diagnostic criterion. Antibiotic sensitivity tests were performed using the law of multiple proportions.
Once collected, data were entered into a Microsoft Excel spreadsheet and a descriptive analysis was performed, calculating median and interquartile ranges for quantitative variables and relative and absolute frequencies for qualitative variables.
The study, which was approved by the Corporate Research Ethics Committee of the Hospital Universitario Fundación Santa Fe in accordance with Minutes No. 3 of February 18, 2019, took into account the ethical principles for medical research involving human subjects established by the Declaration of Helsinki12 and the provisions on health research of Resolution 8430 of 1993 of the Colombian Ministry of Health. 13
Results
Of the 24 HCWs with TB reported to SIVIGILA during the study period, 12 were women (50%). The median age of the patients was 33.5 years (IQR 24-52.7), 4 patients had simultaneous lung and other organ involvement, 4 had diabetes mellitus (DM), 4 had TB/HIV coinfection, 2 suffered from autoimmune diseases requiring immunosup-pressive therapy, and 1 was on immunosuppressive treatment after receiving a liver transplant. Most cases (41.66%) occurred in physicians, followed by medical students (16.66%), and the most frequent form of TB was pulmonary TB (62.5%). All patients were classified as new TB cases (Table 1) and one case of mono-resistance to isoniazid and one to ethambutol were found.
Table 1 General characteristics of the healthcare workers with tuberculosis analyzed.
| Profession | Sex | Age (years) | Form of TB | History | RS | |
|---|---|---|---|---|---|---|
| 1 | Medical student | M | 19 | Pulmonary | - | Yes |
| 2 | Radiology Secretary | F | 23 | Pulmonary | - | Yes |
| 3 | General practitioner | M | 25 | Pulmonary | - | No |
| 4 | Dentist | F | 25 | Pulmonary and pleural | - | Yes |
| 5 | Nurse | M | 29 | Pulmonary and cutaneous | HIV | Yes |
| 6 | General services and housekeeping | F | 38 | Pulmonary | -- | No |
| 7 | Otolaryngologist * | M | 56 | Laryngeal | DM | No |
| 8 | Anesthesiologist | M | 63 | Pulmonary | Liver transplantation | Yes |
| 9 | Anesthesiologist | M | 68 | Pulmonary | DM | Yes |
| 10 | Emergency department assistant | F | 24 | Pleural | Cigarette consumption | No |
| 11 | Radiology Secretary | F | 47 | Pulmonary | Rheumatoid arthritis | Yes |
| 12 | Medical student | F | 22 | Pulmonary | -- | Yes |
| 13 | Occupational health physician | M | 54 | Miliary, peritoneal | Cirrhosis, NAFLD, DM | No |
| 14 | General surgeon | M | 47 | Pulmonary | -- | No |
| 15 | General practitioner | M | 30 | Pulmonary, miliary | HIV, ART: Yes, VL: 292 copies, CD4 lymphocytes: 97/mm3 | Yes |
| 16 | Medical student | F | 24 | Pleural | -- | Yes |
| 17 | Medical student | M | 21 | Pulmonary | -- | No |
| 18 | Dietitian | F | 49 | Pulmonary | -- | ND |
| 19 | General practitioner | F | 48 | Pulmonary | HIV, ART: No, VL -, CD4 | Yes |
| lymphocytes: 7/mm3 | ||||||
| 20 | General practitioner | M | 62 | Pulmonary | DM | Yes |
| 21 | Health visitor | F | 37 | Pulmonary | SLE, systemic sclerosis, | No |
| dermatomyositis | ||||||
| 22 | Gastroenterologist | M | 66 | Pulmonary | HIV, ART: No, VL 5 400 000 copies, | Yes |
| CD4 lymphocytes: 93/mm3 | ||||||
| 23 | Medical student * | F | 23 | Pulmonary, peritoneal | -- | No |
| 24 | Dietician | F | 26 | Pulmonary, lymph node | Yes |
TB: tuberculosis; RS: respiratory symptom; M: male; F: female; HIV: human immunodeficiency virus; DM: diabetes mellitus; NAFLD: nonalcoholic fatty liver disease; ART: antiretroviral therapy; VL: viral load; SLE: systemic lupus erythematosus; ND: no data.
* Previously reported cases: references 3 and 9, respectively.
Source: Own elaboration.
The histopathological study, which was performed on biopsies from 19 patients (16 pulmonary specimens, 2 mediastinal lymph node specimens, and 1 bone marrow specimen), was useful in the diagnosis of 66.66% of cases. Necrotizing granulomas were the most frequent finding in this study (14 cases), followed by acid-alcoholic resistant bacilli using the Ziehl Neelsen staining technique (10 cases). Diagnostic imaging scans were available for all cases, and findings, in both plain radiography and tomography, were diverse; however, the result of the plain radiography was interpreted as normal in 2 patients (Table 2).
Table 2 Histological and radiological characteristics of the analyzed healthcare workers with tuberculosis.
| Main diagnostic criteria | Histological findings | Radiological findings | Drug sensitivity (RHZE) | ||
|---|---|---|---|---|---|
| Rx | CT | ||||
| 1 | Histological, microbiological | EGC | Left apical opacity | - | SSSS |
| 2 | Microbiological | - | Right basilar opacity | - | SRSS |
| 3 | AFB+, histological, microbiological | NCGI | Left basilar opacity | - | SSSS |
| 4 | Microbiological | - | Right basilar opacity and PE | - | SSSS |
| 5 | Histological, microbiological | NCGI | - | Miliary pattern | SSSR |
| 6 | Histological, microbiological | NCGI | Right basilar opacity | - | SSSS |
| 7 | Histological, microbiological | NCGI | Normal | - | SSSS |
| 8 | Histological, microbiological | NCGD | Normal | - | SSSS |
| 9 | Histological, microbiological | NCGD | Left apical opacity | - | SSSS |
| 10 | Histological, molecular | NCGI | - | PE | SSSS |
| 11 | AFB+, histological, molecular, microbiological | NCGI | - | Nodules, cavitation, tree-in-bud sign | SSSS |
| 12 | Microbiological, molecular | -- | - | Nodules, tree-in-bud sign, PE | SSSS |
| 13 | Histological, microbiological | NCGD (BM) | - | Miliary pattern | SSSS |
| 14 | Microbiological (AAFB in BAL) | -- | - | Consolidation in RUL and LLL, tree-in-bud sign | - |
| 15 | Histological, microbiological | NCGI | - | Miliary pattern, nodules, generalized adenopathies | SSSS |
| 16 | Histological, molecular | NCGI | PE, atelectasis | - | - |
| 17 | Molecular | Acute inflammation | - | Tree-in-bud sign | SSSS |
| 18 | Molecular, microbiological (AAFB in BAL) | -- | Opacity in RUL | SSSS | |
| 19 | Clinical, imaging | Cellular augmentation, PMN | - | Consolidation in LML and LLL | - |
| 20 | Molecular, microbiological (AAFB in BAL) | NCGD | Opacity in lingula | - | SSSS |
| 21 | Microbiological | Chronic inflammation | - | Consolidation of the anterior segment of the RUL | - |
| 22 | Histological | CGD (mediastinal lymph nodes) | Frosted glass | - | - |
| 23 | Microbiological, molecular | NCGI | - | Cavitating nodule in RUL, pleural thickening | SSSS |
| 24 | Histological | NCGI (mediastinal lymph nodes) | Consolidation in RUL, cavitation, tree-in-bud sign | - | |
CGD: chronic granulomatous disease; AFB, acid-fast bacilli smear (R: rifampicin, H: isoniazid, Z: pyrazinamide, E: ethambutol, S: sensitive, R: resistant); NCGI: necrotizing chronic granulomatous inflammation caused by mycobacterium; PE: pleural effusion; NCGD: necrotizing chronic granulomatous disease; BM: bone marrow; AAFB: acid-alcohol-fast-bacilli; BAL: bronchoalveolar lavage; RUL: right upper lobe; LLL: left lower lobe; PMN: polymorphonuclear cells; LML: left middle lobe.
Source: Own elaboration.
Discussion
TB is an infectious disease to which HCWs are highly exposed. Joshi et al1 state that the risk of contracting the disease in this population may be up to 40 times higher than in the general population. It has also been established that the risk of contagion depends on many factors such as the local prevalence of TB, the area where professionals work, the specific work they carry out, the correct use of personal protective equipment, their training on infection control and, of course, their immune status.14,15
Studies such as Cascante & Hueto14 have categorized some healthcare-related activities as high, intermediate, and low risk of TB infection; the first group includes cough induction procedures, bronchoscopy, and emergency room tasks; the second refers to work in units where patients with TB are typically treated, such as outpatient clinics, and surgical and diagnostic imaging rooms; and the third involves work activities in areas such as archives, neonatology, obstetrics, etc. In the series described here, most cases (54.2%) occurred in the HCWs group who had recently carried out high-risk activities, followed by those that carried out intermediate-risk activities (45.8%). It should be noted that none of the patients performed or had performed low-risk activities and that the difference between the other two groups was not greater (Table 1).
Bonifacio et al.16 conducted a study involving 54 in-house physicians and 45 residents who began their training in April 2000 at the Hospital Carrion in Lima, Peru, and found that physicians exposed to a large number of TB cases had an annual infection rate of 17%, which is slightly lower than that described in the present study, where it was 20.83% (n=5). Moreover, in a study conducted in 17 Canadian hospitals, Menzies et al.17 reported that the risk of TB transmission was associated with nursing work (adjusted risk ratio: 4.3; 95%CI: 2.7-6.9), respiratory therapy (adjusted risk ratio: 6.1; 95%CI: 3.1-12), physiotherapy (adjusted risk ratio: 3.3; 95%CI: 1.5-7.2), and the cleaning service (adjusted risk ratio: 4.2; 95%CI: 2.3-7.6).
An analysis of the most affected occupations in the studies carried out in Colombia shows that this variable has a low consistency. Llerena & Zabaleta18 conducted a study involving 128 HCWs with TB and reported that most cases (21.4%) occurred in physicians, which is similar to the findings of this series. In turn, Castillo et al.19 studied the behavior of TB among HCWs and established that of the 532 TB cases reported in this population between 2008 and 2012, most (26.12%) were nurses and dental assistants, followed by physicians (14.09%). This is in line with the findings of Vargas-Restrepo 20 and described in his master's thesis, in which he found that of the 715 cases of TB in HCWs reported in Colombia between 2011 and 2017, 36.4% were nurses and dental assistants and 20.1% were physicians.
The present series included the cases of two women (case 6 and 21, Table 1) who were not healthcare workers but worked at the hospital. This shows that the risk of infection/ disease not only affects healthcare workers and their collaborators directly but extends to all those who work in a facility where TB patients are cared for.
Regarding the distribution by sex, it has been established in the world literature that TB affects mainly men, 5,21 which contradicts the results of the present investigation, where there was no predominant sex. It is particularly notable that three studies carried out in Colombia18-20 show that most of the people involved are women (more than 60% of the cases). This could be explained from a biological perspective because hormonal factors may condition a differential immune response; 22,23 however, from an environmental point of view, it would be more appropriate to consider a higher exposure of technical and professional nurses, who are mostly female, 14,16 to smear-positive patients. 19,20
In the general population, it has been demonstrated that pulmonary TB is observed in the majority of cases (≈80%),3,5 as reported in in the present series (79.16%) and by Llerena-Zabaleta18 (78.9%). However, Castillo-Rico et al.19 and Vargas-Restrepo, 20 who analyzed the totality of TB cases reported in HCWs in Colombia during specific periods of time, reported slightly higher rates of extrapulmonary TB (33.1% and 33.6%, respectively). This difference could be explained by the combination of biological and environmental factors that may put a person with a predisposing condition, in this case HCW, at a higher risk of contracting the disease and developing an extrapulmonary or disseminated form.
This scenario was identified in 4 of the 6 patients with extrapulmonary or disseminated TB in the present series.
Medical history, particularly immune response factors, are risk factors for developing TB, with HIV infection being the most important. However, it is known that diseases such as DM and conditions such as smoking and the use of immunosuppressive drugs, which are common in the general population, may have a greater impact on the epidemiology of TB. 23 In this regard, studies such as that of Sepkowitz & Eisenberg24 have described that up to one third of cases of active TB in HCWs are associated with immunosuppressive entities such as HIV.
In the present study, half of the patients had immune response determinants and, therefore, risk factors for developing TB; 4 of these 12 patients had HIV/TB co-infection, and in 3 of them, histological and laboratory findings confirmed that they already had AIDS. The Llerena & Zabaleta18 and Vargas-Restrepo20 studies reported TB/HIV co-infection of 7.9% and 14%, respectively.
Likewise, the present study found that four patients had DM, a disease that modifies the presentation, course, and prognosis of TB, and is also a risk factor for TB treatment failure, 25 as animal models have shown that DM is related to a decrease in cellular immunity and the response mediated by interferon y, a fundamental molecule of the humoral immune response against M. tuberculosis.26
In Colombia, the diagnosis of pulmonary TB is mainly made by AFB smear, which has a low sensitivity (30-70%) and requires 5 000 to 10 000 bacilli per cubic centimeter of sample to yield a positive result. 27 Another, less commonly used test is culture, which is more sensitive, requires only 10 to 100 viable bacilli per cubic centimeter of sample for yielding a positive result, allows identification of M. tuberculosis in more than 80% of cases, and has a specificity of >98%.28 In the present series, 75% of the cases had positive cultures in one of their samples, which was lower than the 100% reported by Llerena & Zabaleta, 18 while 8.33% of the patients had positive AFB test, which was considerably lower than 57% reported by Vargas-Restrepo. 20
The low number of cases with a positive AFB test is of special relevance, as it is believed that this phenomenon occurs because most of the patients treated in the hospital sought medical assistance in the early stages of the disease. It could be said that they consulted in a timely manner, as evidenced by the fact that lung damage was insufficient to allow bacilli from the parenchymal lesion to appear in the sputum accompanying cough, which is often seen when the disease is at an advanced stage. 9
Histopathology has traditionally been regarded as a reliable method for diagnosing tuberculosis, especially when the presentation is extrapulmonary. 3,9,29 In this case series, the histopathological study, which was performed on 19 biopsies, showed that necrotizing granulomas were the most common finding (14 cases), followed by acid-alcohol resistant bacilli using the Ziehl Neelsen staining technique (10 cases). Thus, the histopathological diagnosis was chronic necrotizing granulomatous disease caused by Mycobacteria in 10 cases. However, in 16 (84%) of these cases, the pathologist's opinion, supported by the histological findings and patient's clinical history, suggested that TB was the cause of the observed histological pattern.
The ineffectiveness of AFB smear for the diagnosis of TB in the series reported here made it necessary to resort to special procedures such as bronchoalvelar lavage and lung biopsy, which significantly favored culture positivity, as well as histological analysis of the affected tissue.
It is worth noting that all of the TB cases in this case series were new, with only two cases of mono-resistance detected (Table 2). Nonetheless, it should be noted that drug-resistant tuberculosis is currently one of the most pressing concerns in tuberculosis control. 5
Several surveillance studies of drug-resistant tuberculosis in untreated patients have been conducted in Colombia. Leon et at. 30 carried out a study between 1999 and 2000 with 1 087 patients and reported that totally drug-resistant (TDR) tuberculosis was observed in 15.6% and multidrug-resistant (MDR) tuberculosis occurred in 1.5% of cases. In turn, Garzon et al.?1 in a study conducted between 2004 and 2005, established that TDR TB was found in 11.8% and MDR TB in 2.3% of 925 untreated patients, while TDR TB was found in 44.3% and MDR TB in 31.4% of264 previously treated TB patients. Finally, Llerena et at., 32 in a study with 128 patients younger than 15 years with TB who underwent susceptibility testing of M. tuberculosis between 2001 and 2009, found that 123 of them had not been treated and that among these, TDR TB was found in 21.1% and MDR TB in 6.5%. According to Llerena et al,18 the prevalence of TDR TB and MDR TB in new TB cases was 7.9% and 4%, respectively, while the prevalence of these two types of resistance was 12.5% in previously treated cases.
One of the strengths of the present study is that it is a good approximation to the TB situation among HCWs in Bogotá between 2008 and 2018 due to the detailed analysis of the demographic, clinical and paraclinical characteristics of the patients. On the other hand, limitations include the fact that only patients from one institution were enrolled, and advanced epidemiological calculations were not possible.
Conclusions
Compared to the literature, this study found a higher proportion of women (1:1), and the histopathological study allowed for the rapid diagnosis of TB in the majority of cases. Moreover, most patients were physicians and medical students.
Due to the fact that the risk of contracting tuberculosis among HCWs in comparison to the general population is one of the indicators proposed by WHO for evaluating the effectiveness of preventive and control measures for disease transmission in health services, the country's health institutions should pay more attention to the biosecurity practices of this population. In addition, the relevant entities should facilitate the development of occupational epidemiological surveillance programs that promote better transmission control.
