Regarding the prevalence of Tuberculosis among children aged 6 months to 59 months admitted with Severe Acute Malnutrition (SAM) at Jinja Regional Referral Hospital and we found the prevalence to be 23.4%. This prevalence was high. The high prevalence can be attributed to the big proportion of participants whose immunization status was not up to date, since the BCG vaccine offers protection against TB [6, 16]. Also, the fact that majority of the participants had a clinical diagnosis (25/32), makes it possible that some of the participants considered to have TB could have been falsely classified. This is because we did not do any investigations to confirm bacterial pneumonia and therefore some of the patients clinically diagnosed with TB could have had bacterial pneumonia.
Our prevalence was higher than that reported in India (13%) [17], in Nepal (4.67%) [18], in Ethiopia (10.39%) [13] and in Congo (8.4%) [19], but slightly lower than that reported in South Africa (25.6%) [20]. The differences in the prevalence of tuberculosis can be attributed to the study population differences, the burden of tuberculosis varies geographically, and the criteria used for diagnosis of tuberculosis in these studies. For example, in our study, the clinical criteria, TB-LAM and Gene x-pert were used to diagnose TB which could have resulted in a higher prevalence. The study in South Africa that reported a prevalence of 25.6% (almost similar to ours but slightly higher) also used the clinical criteria for diagnosis.
The radiological manifestations of pulmonary tuberculosis in children exhibit variations across different populations and studies. Common findings include consolidation, hilar/mediastinal adenopathy, and pleural effusion. However, there are notable differences in the distribution of these abnormalities, as highlighted by the findings from this study compared to others [21,22,23].
The distribution of consolidation and hilar/mediastinal adenopathy aligns with the typical presentation of pulmonary tuberculosis in children [21, 23]. These findings serve as important diagnostic markers, aiding in the identification and management of the disease. However, the absence of cavitary lesions in the current study suggests a potentially less severe disease course compared to findings from other studies where cavitary lesions were more prevalent [24].
The predominance of right unilateral lesions in the abnormal radiographs observed in the present study is consistent with previous reports, indicating a propensity for TB to affect the right lung more [23]. However, the distribution of lesions may vary among populations, as evidenced by differences in lesion laterality across studies.
The distribution of radiological findings observed in this study contrasts with those reported by Hassanzad et al. [24] and findings from Brieflands [25]. While our study found no cavitary lesions (0%), both Hassanzad et al. and Brieflands reported the presence of cavitary lesions in a subset of children with pulmonary tuberculosis.
According to Brieflands, cavitary pulmonary tuberculosis typically seen in adults, is rare in children, which aligns with our findings of no cavitary lesions observed in the studied population. However, the presence of cavitary lesions in other studies indicates a higher mycobacterium load, serving as a potential source of infection [25]. This discrepancy underscores the variability in disease presentation across different populations and highlights the importance of considering regional and demographic factors when interpreting radiological findings in pediatric tuberculosis.
Understanding the variations in radiological findings of pulmonary tuberculosis among children has significant implications for clinical practice. Recognizing common patterns such as consolidation and adenopathy facilitates accurate diagnosis and timely initiation of treatment. The absence of cavitary lesions in the current study may suggest a less advanced disease stage, influencing treatment decisions and prognostic assessments.
Regarding the presenting signs and symptoms, the common symptoms seen among patients with TB were fever seen in 75% of the TB patients and cough in 78.1%. Our findings were in agreement with a cross sectional hospital based study done in Kassala Teaching Hospital and Kwaiti Paediatric Hospital in Sudan which reported that the commonest symptoms among children with TB and SAM were fever in 95.5% and cough in 79.8% [26]. The presence of fever and cough as the commonest symptoms is likely because pulmonary tuberculosis affects the lungs causing the irritation of the airways that results in Cough. In addition, the inflammatory reaction to the infection results in the release of markers of inflammation in the blood resulting in fever.
Regarding the factors associated with Tuberculosis among children aged 6 months to 59 months admitted with SAM at JRRH. We observed that a child coming from a rural area had the odds of having TB increased by 1.205 times compared to the one from the urban areas. The social-economic status has been noted to differ between the rural and the urban areas. The poor social-economic status in the rural areas could be responsible for the increased odds of TB in the children from rural areas. A case in point is a study which included 9 rural eastern Uganda communities that reported a high prevalence of tuberculosis among the children [27]. In addition to the above, the use of traditional treatments in rural communities may also be contributing to the high burden of TB in the rural areas. These treatments could worsen the clinical presentation of the disease and even delay the parents from seeking medical attention on time and thereby increasing the number of TB contacts.
Though some studies reported the history of TB contact to be significantly associated with TB, this was not the case in this study. This is possibly because only 5 participants reported a history of TB contact. Therefore, the number of participants was small, making it not possible to make statistically relevant comparisons regarding this variable. Having a small number of participants reporting TB contact yet we had a big number diagnosed with TB could be because: we did not do contact tracing nor did we screen the household members of these children.
A child who was HIV positive was 1.619 times more likely to have TB compared to one who was HIV negative. HIV infection attacks the white blood cells resulting in the impairment of the child’s immune system [28]. This impairment in the immune system is even made worse by the presence of malnutrition in these children [13]. This makes it difficult for the body to contain the mycobacterium tuberculosis organisms resulting in clinical tuberculosis. This was in agreement with two studies in Ethiopia where HIV infection increased odds of having TB [13, 28].
The altered platelet count was also found to be associated with tuberculosis, with thrombocytopenia increasing the odds of having TB by 1.407 times and thrombocytosis increasing the odds by 1.202 times. Disseminated tuberculosis, which is more common in immune compromised children such as these with severe acute malnutrition results in platelet activation which may result in thrombocytosis [29]. However, a number of studies have also reported the occurrence of immune thrombocytopenia in patients diagnosed with tuberculosis [30]. This could explain why these derangements in platelet count were found to be significantly associated with the occurrence of tuberculosis.
Study limitations
Not all patients had chest X-Ray done, yet even those who did not have respiratory distress could have had radiological findings consistent with TB. This could have introduced bias in the study. We did not do any investigation to rule out bacterial pneumonia, which could have resulted in some bacterial pneumonia being clinically diagnosed as TB. The BCG vaccination confirmation was not done in this study; therefore, it was not able to ascertain its association with PTB in SAM. This study was cross-sectional, so the associations found do not necessarily infer a causal relationship. Therefore, future studies on the subject matter should consider the use of other study designs like case control. The study took place during a 3-month period at a single location in a semi-urban setting, which limits the generalizability of our findings; however, these findings can be used as a basis to design multicentre studies over a longer period of time.