Tuberculosis is a contagious disease caused by airborne Mycobacterium tuberculosis. This bacterium most often develops in the respiratory tract (pulmonary tuberculosis). It can spread to other organs (extrapulmonary tuberculosis) [1].

In 2020, an estimated 9.9 million people worldwide will be infected with TB, including 1.1 million children. Tuberculosis is the second most common cause of death from infectious diseases after VTEC-19 and the 13th most common cause of death in the world. It is also the leading cause of death among people living with HJV and a major cause of death due to antimicrobial resistance [2].

TB is present in all countries and in all age groups. In 2020, the World Health Organization Region with the highest number of new TB cases was Southeast Asia (43% of all new cases) followed by the African Region (25%) and the Western Pacific Region (18%) [2].

Pediatric tuberculosis remains a major global problem with more than one million new cases each year according to the World Health Organization. Mortality from tuberculosis is proportionally higher in children than in adults: while children represent 10% of all tuberculosis cases, they account for 15% of deaths [3]. Delayed diagnosis and rapidly progressive forms in young children are the main reasons for this high mortality. In a recent systematic evaluation of the factors associated with pediatric TB deaths, it was shown that 80% of deaths occur in children under 5 years of age, and 96% in children who have not had access to anti-TB treatment [3].

In 2020, according to the annual report of the National Tuberculosis Control Program in Mali, 274 cases of all forms of tuberculosis were recorded in children aged 0 - 14 years, i.e. a frequency of 4% [4].

A study of a series of 12 cases was carried out in 2017 in children aged 0 - 15 years in the pediatric department of the Mali Hospital [5]. The aim of this work was to collect cases of tuberculosis in children aged 0 - 15 years in order to study the epidemiological and clinical aspects in the pediatric department of the Mali Hospital during the period 2015-2021.

1) Frequency

From January 1, 2015, to December 31, 2021, out of 9438 children aged 1 month to 15 years hospitalized, we recorded 69 cases of tuberculosis of all forms or a frequency of 0.73%.

2) Socio-demographic characteristics of the children.

The age group [1 - 9 years] was the most represented, with a frequency of 58.0%. The average age was 6.16 years with extremes of 3 months and 15 years. The male sex was slightly in the majority with a sex ratio of 1.1. More than half of our children (43/69) were from an urban area, with a frequency of 62.3% (Table 1).

3) Clinical characteristics

The notion of tuberculosis infection was found in 21 of the 69 children, or a frequency of 30.4%. Almost all children had been vaccinated against tuberculosis (88.4%) (Figure 1).

The most common cause of hospitalization was fever (76.8%), followed by weight loss (73.9%) and cough (69.6%) (Table 2).

Pulmonary tuberculosis was the most frequent clinical form (60.9%) followed by extrapulmonary tuberculosis (39.1%) (Figure 2).

Pulmonary involvement was the most common (60.9%), with or without other involvement, followed by pleural involvement (11.5%) and lymph node involvement (10.0%) (Table 3).

Tuberculosis was associated with severe acute malnutrition in 46 of the 69 children, with a frequency of 66.6% (Table 4).

Tuberculosis was confirmed by bacteriology in 30 of the 69 children (43.5%). The direct examination was positive in 33.3% of the children, culture (18.8%), and PCR (13.0%). Mycobacterium tuberculosis was present in all confirmed cases. We isolated a rifampicin-resistant strain of Mycobacterium tuberculosis in 2/30 children, a multidrug-resistant strain in 1/30 children, and an extensively drug-resistant strain in 1/30 children (Table 5).

For microscopy-negative cases, the diagnosis was made on the basis of characteristic lesions on radiology in 34.8% of children, histology (11.6%), and cytology (10.1%).

The tuberculin intradermal test performed on 37/69 children was positive in 32.4% of them. HIV serology was positive in 5 of the 54 children tested (9.3%) (Table 6 and Table 7).

1^st line anti-tuberculosis drugs; R = Rifampicine, H = Isoniazide, Z = Pyrazinamide, E = Ethambutol; 2^nd line anti-tuberculosis drugs; Am = amykacine, Mfx = Moxifloxacine, Pto = Prothionamide, Cfz = Clofazimine, Bdq = Bédaquiline, Lzd = Linézolide, Cs = Cyclosérine.

More than half of our children (65.2%) were treated with 1^st line anti-tuberculosis drugs and with a short 6 months regimen. Two children out of 69 were treated with 1^st line anti-tuberculosis drugs with the long 12 months regimen. Two of the 69 children were multidrug-resistant and one was extensively drug-resistant. They were treated with 2^nd line anti-tuberculosis drugs with the 9- and 20-month regimen. Twenty of the 69 children did not receive treatment because they died before bacteriological confirmation.

We observed cures without sequelae in 29/69 (42.0%), cures with sequelae (5.8%), and therapeutic failure in 1 child. The case fatality rate was 39.1% and the dropout rate was 11.6% (Table 8).

Limitations

Despite the limitations of this study related to its retrospective nature, poor record-keeping, the size of our sample, and the unavailability of certain additional examinations essential for diagnosis, the results obtained have made it possible to highlight the problems of tuberculosis in children in our department.

Epidemiological aspects

1) Frequency

From January 1, 2015, to December 31, 2021, 9438 children aged 0 - 15 years were hospitalized among whom we diagnosed 69 cases of tuberculosis of all forms or a hospital frequency of 0.73%. It represents 2.3% of all pediatric tuberculosis cases according to the data of the Sectoral Unit for the Fight against Tuberculosis of Mali [4] during the same period.

Our prevalence is lower than that of Ibrahim et al. [8] who reported a frequency of 2.3% during their study conducted in 2020 in the pediatric department of the Donka National Hospital in Guinea. It is however higher than that of Tamini/Toguyeni et al. [9] who found a frequency of 0.078% during their study conducted in 2016 at the Centre Hospitalier Universitaire Pédiatrique Charles De Gaules in Ouagadougou, Burkina Faso. This low frequency could be explained by the fact that tuberculosis in children often goes unnoticed because of the unspecific symptoms and also because of the difficulty in obtaining bacteriological confirmation.

2) Socio-demographic characteristics

Age:

The study showed that the age group [1 - 9 years] was the most represented (58%). The mean age was 6.16 with extremes of 3 months and 15 years. It is close to that of Tamini/Toguyeni et al. [9] in Burkina Faso and that of Ben Bechir et al. [10] in Tunisia who respectively reported an average age of 6.2 and 6.3 years. It is lower than that of Hamdi et al. [11] and Aw et al. [12] who found an average age of 9 years in their studies conducted respectively in Tunisia and Mauritania. It is higher than that of Soumana et al. [13] in Niamey (4.6 years) and those of the literature (5 years) [14].

Gender

Both sexes were affected with a slight male predominance (53.6%) (sex ratio: 1.1). This male predominance was reported by the Sectoral Unit for Tuberculosis Control in Mali [4] and by most African authors, notably by Tamini/Toguyeni et al. [9] in Burkina Faso and by Djegbeton et al. [15] in Ivory Coast. A female predominance of 52.8% was observed by Zemour et al. [16] in their study carried out in 2016 in Remchi in Algeria. This male predominance in our series could be explained by the male predominance in the general population of Bamako where the majority of our children resided.

Residence

The majority of the children in our series lived in urban areas (62.3%) as in the study by Tamini/Toguyeni et al. [9] in Burkina Faso (74.4%). In the study by Rakotomizao et al. [17], more than half of his patients were from rural areas (67%). Urban areas could be a risk factor because of the high population density which would favor TB transmission.

3) Clinical characteristics

The tuberculosis vaccine was administered to 61/69 children, i.e. a proportion of 88.4%. This result is close to that of Radoui et al. [18] who found a frequency of 90%. This result is different from that of L. Donato [19] who reported in the literature that BCG confers a protection rate of 50% for all forms. It is different from that of Tamini/Toguyeni et al. [9] who reported cases of tuberculosis in children who were mostly unvaccinated (57.5%).

The notion of infection was found in (21/69) or 30.4%. This frequency is lower than that reported by Radoui et al. [18] (65.3%) and that of Ben Bechir et al. [10] (42.6%). However, it is higher than that of Tamini/Toguyeni et al. [9] who found the notion of tuberculosis infection in 19.4% of the children in their study in Burkina Faso.

The reasons for hospitalization were dominated by fever (76.6%), followed by deterioration of general condition (73.9%) and cough (69.6%). Our symptoms are similar to those reported in the literature [20] [21] and those observed by Hamdi et al. [11]: fever (65%), cough, and alteration of general condition (56%). They are different from those reported by Tamini/Toguyeni et al. [9] in Burkina Faso who observed in their study that pallor was the most frequent reason for hospitalization in children (87.23%) followed by cough (53.19%). They are different from symptoms found by Ben Bechir et al. [10]: weight loss (50%) and cervical adenopathy (46%).

In our series, the pulmonary form was more frequent than the extrapulmonary form (60.9% versus 39.1%). Our data are similar to those of Tamini/Toguyeni et al. [9] who found more pulmonary than extrapulmonary forms (68.08% versus 25.53%) in Burkina Faso as did Radoui et al. [18] who observed 92 cases of pulmonary tuberculosis and 42 cases of extrapulmonary tuberculosis in their study. our data are different from those of Soumana et al. [13] who found in their study more extrapulmonary forms than pulmonary forms (62.1% versus 37.9%) like Zemour et al. [16] who reported more extrapulmonary forms than pulmonary forms (77.8% versus 20.2%) in their study.

In our sample, tuberculous pleurisy was the most frequent extrapulmonary form (11.5%) followed by peripheral adenitis (10.0%) as reported in the literature [20] [22] but with higher proportions (pleurisy: 40%, adenitis: 30%). Radoui et al. [18] also showed in their series that tuberculous pleurisy was the most frequent extrapulmonary involvement (36.5%) followed by lymph node involvement (23.1%).

Co-morbidity was found in 54 of the 69 children, or a frequency of 78.1%. It was mostly associated with severe acute malnutrition (66.7%) followed by HIV (7.2%), acute myeloblastic leukemia (1.4%), nephrotic syndrome (1.4%), and pulmonary hypertension (1.4%). Our data are close to those of Soumana et al. and Djegbeton et al. [15] who respectively reported that tuberculosis was predominantly associated with severe acute malnutrition in 55% and 43% of children in their studies.

4) Paraclinical characteristics

Intradermal tuberculin testing was performed in 37/69 children (53.6%). It came back positive in 32.4% of them. This rate is lower than that of most African authors, notably Soumana [13] (85.7%), Radoui [18] (69.3%), Ibrahim [8] (62.2%), Hamdi [11] (58%) and Ben Bechir [10] (57%). However, it is higher than that of Djimbélé et al. [23] who found a frequency of 9.35% in their study conducted at the Pediatric Center of Bangui. This low rate of positivity of the tuberculin intradermal test could be explained by the immune immaturity of our children due to their young age and the high prevalence of severe acute malnutrition in our series.

Tuberculosis was confirmed by bacteriology in 30 of the 69 children (43.5%). This proportion represents 1% of all confirmed pediatric cases according to data from the Sectoral Unit for Tuberculosis Control in Mali [4]. It is lower than that of Ben Bechir et al. [10] who obtained bacteriological confirmation in 65% of children with tuberculosis in their series. However, it is higher than those reported by most African authors, notably those of Djegbeton et al. [15] (45.8%), Tamini/Toguyeni et al. [9] (23.4%), Ibrahim et al. [8] (18.75%) and of Radoui et al. [18] (14.1%).

The direct examination was positive in 33.3% of our children. This positivity rate is higher than that reported in the literature (20%) [22] [24] and that of Hamdi et al. [11] (10%). However, it is lower than that of Zemour et al. [16] who reported that 53% of children were positive on direct examination in their study. This increase in the positivity rate in our series could be explained by the fact that we send our samples to 2 different bacteriology laboratories for direct examination.

The culture came back positive in 18.8% of our children. This frequency is lower than that reported in the literature (50%) [22] [24]. However, it is higher than that obtained by Hamdi et al. [11] (10%).

If the sensitivity of detection of Mycobacterium tuberculosis in children by the PCR method is of the order of 45-100% [21] [24], it came back positive in 13.0% of our children. This proportion is lower than those obtained by most African authors, notably Djegbeton et al. [15] (66.7%) and by Ben Bechir et al. [10] (54%). However, it is higher than the detection rate of tuberculosis of all forms reported by the Sectoral Cell for the Fight against Tuberculosis in Mali (8.2%) [4].

All confirmed cases were Mycobacterium tuberculosis. PCR detected rifampicin resistance in 2 children (6.7%) and culture revealed a multidrug-resistant strain of Mycobacterium tuberculosis in 1 child and an extensively drug-resistant strain in 1 child. Our data are different from those reported by the authors. Our rate of resistance to rifampicin is higher than that of Ben Bechir et al. [10] (3.2%) and that of Diatta et al. [25] (0.91%). These authors did not observe any multidrug-resistant or extensively drug-resistant cases in their studies.

For microscopy-negative cases, the diagnosis was made on the basis of characteristic signs of tuberculosis on radiology in 34.8% of children, on histology in 11.6%, and cytology in 10.1%.

5) Therapeutic and evolutionary aspects

More than half of our children (65.2%) were treated with 1^st line anti-tuberculosis drugs and with the short 6-month treatment regimen (2RH/4RH).

Two out of 69 children with tuberculosis meningitis were treated with 1^st line anti-tuberculosis drugs with the long 12-month regimen (2RH/10RH).

Two of the 69 children had multidrug-resistant peritoneal TB and extensively drug-resistant urinary TB, respectively. The 2 children were treated with 2^nd line anti-tuberculosis drugs with the 9- and 20-month regimen.

Twenty of the 69 children did not receive treatment because they died before bacteriological confirmation.

Forty-five of the 69 children were treated with the (2RH/4RH) or (2RHE/4RH) regimen, or 68% as in the literature [19] [21] [24]. This regimen is used by most African authors through their national TB control programs, notably Tamini/Toguyeni et al. [9] used it in 72.34% of children and Hamdi et al. [11] in 100% of children. Second-line anti-tuberculosis drugs have not been used by other authors because they have not reported cases of multidrug resistance or ultra resistance in their series.

During the course of the disease, we observed an overall cure rate of 47.8%, of which 42.0% were cured without sequelae and 5.8% with sequelae (vertebral forms). Our cure rate is lower than the overall rate recorded in 2020 by the Malian Tuberculosis Control Sector Unit (82%) [4] and the objective of 85% therapeutic success. It is also lower than that of Ibrahim et al. [8] and Moh et al. [26] who respectively obtained a cure rate of 84.54% and 71% in their hospital studies.

In our series, we recorded a case fatality rate of 39.1%. According to the data of the Sectoral Cell for the Fight against Tuberculosis in Mali, the case fatality rate among children is 46.5% [4]. Our rate represents 1% of the overall pediatric mortality rate in Mali.

It is also higher than that of Ibrahim et al. [8] who recorded a case fatality rate of 5.4% and Moh et al. [26] who found a case fatality rate of 16% in their study conducted in 2015 in 2 University Hospitals of Abidjan. This high mortality rate in our series could be explained by the young age of the majority of our children, the delay in diagnosis, the existence of severe forms (miliary, meningitis, multidrug resistance, ultra drug resistance), and comorbidities.

Eight children out of 69 were lost to follow-up, or a proportion of 11.6%, slightly higher than that observed by Ibrahim et al. [8] (10.05%) and by Moh et al. [26] (8.5%). This could be explained by the lack of communication and the low level of education of the parents of our children who abandon the treatment to resort to traditional medicine.

Tuberculosis is common in the pediatric ward of the Mali Hospital despite good BCG vaccination coverage. Children under 9 years of age are the most affected. The notion of tuberculosis infection was found in a small proportion of our children. Fever and alteration of the general condition are the most frequent symptoms. Bacteriological confirmation remains difficult despite the popularization of molecular diagnostic methods (Xpert system). The case fatality rate is high due to the young age of the children, HIV co-infection, and the emergence of resistant strains of bacilli. In order to assess the extent of tuberculosis in children in Mali and to study its clinical and paraclinical characteristics, all children hospitalized for acute and severe malnutrition should be tested for Koch’s bacillus in their gastric fluid. Investigations should be carried out around an index case in the family, including chest radiography, skin tests, and bacteriological samples from all contacts.

The authors declare no conflicts of interest regarding the publication of this paper.

Kané, B., Diallo, K.W., Sangaré, A., Dembélé, Y., Dabo, G., Traoré, M., Traoré, M.M., Berthé, M. and Maiga, M. (2022) Study of Tuberculosis in Children Aged 1 Month to 15 Years in the Pediatric Ward of the Hospital of Mali 2015-2021. Open Journal of Pediatrics, 12, 433-448. https://doi.org/10.4236/ojped.2022.122047