As for anaemia, it is estimated that about a quarter of the world's population has lower than normal haemoglobin levels. Its occurrence during certain infectious diseases such as HIV infection would promote their development or increase mortality (Lee et al., 2006; Isanaka et al., 2012). This multifactorial problem includes inhibiting erythropoiesis by pro-inflammatory cytokines, nutritional deficiencies, iron use defects, and bone marrow aplasia (Metanat et al., 2020).

The anaemia seen in tuberculosis can manifest itself in several forms. Although it is moderate in the vast majority of cases, sometimes it can resemble pernicious anaemia (Jonsson., 2009). In addition, cases of hemolytic anaemia are reported, which may be infectious, toxic, or autoimmune (Eschapasse et al., 2017).

In a 2018 study in India, the authors reported a 71.83% prevalence of anaemia in tuberculosis patients (Mukherjee et al., 2019). In Brazil, this prevalence was of the order of 85.6% and is believed to be due to the association of an HIV infection, particularly those who had a weight loss, a collapsed CD4 count, a high viral load, and ARV treatment. pre-TB (Demitto et al., 2020). In Guinea, it is assumed that the prevalence of anaemia is also high during tuberculosis infection, whether or not associated with HIV. However, to our knowledge, few studies have been interested in the question, which motivated the realization of this present work. The objective was to determine the prevalence of anaemia in tuberculosis patients in the service and identify its factors.

Study setting and population

Surveillance data were analyzed from 218 patients with drug-resistant TB recruited between 2016 and 2018 in a multicentre longitudinal cohort study conducted at Guinea's main multidrug-resistant tuberculosis treatment centers. According to the WHO standardized regimen, all patients were seen initially and followed monthly for 9 months (WHO treatment guidelines for drug-resistant tuberculosis, 2016 update, n.d.). Patients under 18 years were excluded from the analysis.

The authors collected concerned socio-demographic and clinical characteristics, laboratory test results (sputum smear and blood count) and x-rays for all registered patients in treatment centers. Additional details were gathered from records and results reports. The following clinical, paraclinical, and demographic data were extracted: age, sex, place of residence, comorbidity, HIV status, history of tuberculosis due to previously treated tuberculosis, presence of cavities on the chest x-ray, determined by the principal radiologist, data reference on weight, height for the calculation of the BMI.

Definitions of anaemia

The following World Health Organization (WHO) criteria were used to rank the severity of anaemia:

1. No anaemia (haemoglobin [Hb]> 13.0 g / dL for men, 12.0 g / dL for women),

2. Mild anaemia (11.0-12.9 g / dL for men, 11.0-11.9 g / dL for women),

3. Moderate anaemia (8.0-10.9 g / dL for men and women)

4. Severe anaemia (<8.0 g / dL for men and women).

Potential predictors of anaemia are socio-demographic (sex, age); clinical (cough, dyspnoea, chest pain, night sweats, nausea, vomiting, HIV status, history of antituberculosis treatment, depression, adherence, results of treatment) and paraclinical (number of colonies per initial smear, the initial number of colonies count smear, initial culture, radiography).

Statistical analysis

Frequencies (percent) or means (standard deviation; SD) describe categorical and continuous variables. Univariate and multivariate logistic regression was used to identify predictors of anaemia among DR-TB patients; all variables with a p-value less than 0.1 were considered for multivariate regression. Odds ratios (OR) with 95% intervals (95% CI) were used as association parameters, and the level of significance was p less than 5%. The hypothesis of proportionality and log-linearity risk has been verified.

A total of 218 patients were included. The patients were mainly men (68.3%) and women (31.6%). The mean age of the patients was 33.7± 11 years; most lived in urban areas (77.5%). Fifty-one patients (23%) were HIV positive, 36 of whom (72%) had a BMI <18.5 (Table 1).

According to the patient category, 38 patients (17.4%) were new cases of DR-TB, and 180 patients (82.6%) had already been infected. New cases of DR-TB and 180 patients (82.6%) had already received first-line antituberculosis treatment. Among the authors’ patients, they noted 71% anaemia (Table 2).

On univariate analysis (Table 3), the occurrence of anemia was associated with serological status (p_value <0.001), history of treatment for TB (p_value = 0.027); biologically, leukocyte count (p_value = 0.04), platelet count (p_value = 0.008, lymphocyte count (p_value = 0.009), and creatinemia (p_value = 0.040) were associated with the occurrence of 'anemia. On multivariate analysis (Table 3), seropositive patients were less likely to have anemia (p_value <0.001) as well as vomiting (p_value = 0.049). The results of the biology showed an influence of the leukocyte level (p_value = 0.006) and the platelet level (p_value = 0.043).

The emergence of MDR-TB is a serious global threat to TB  control.  Treatment  of  this  disease   is   a  challenge because it typically requires longer-treatment regimens using second-line anti-TB drugs that are more toxic than drug-sensitive  tuberculosis.  The   results    showed   that around three-quarters of patients followed for multidrug-resistant tuberculosis were anaemic. In an earlier study carried out in Iran (Metanat et al., 2020),  the authors had reported a frequency of anaemia of the order of 54.5% in tuberculosis patients; the result is close to that reported in Brazil, 89.1% of anaemia (Oliveira et al., 2014). Moreover, in the study by Gil-Santana L et al. (Gil-Santana et al., 2019), Hemoglobin levels did not change substantially with increasing AFB smear grade when all patients were considered. This frequency could be explained by the evolution of certain multiresistant forms that have benefited from a long treatment before developing resistance.

The patients studied had a variable profile. In another study, young adults were the most affected age category (average age of the patients was 33.7 years), consistent with other African series, particularly in Gabon (Kombila et al., 2019), In Congo Brazzaville (RRDDI et al., 2020). For retroviral serology, nearly one in four patients (23%) was seropositive, TBM-R and HIV coinfection was reported in several series. It was 42.3% in that Kombila et al. (Kombila et al., 2019)in Gabon, 18% in Togo (Adambounou et al., 2020). Based on the BMI, 72% were underweight (<18.5); this observation is close to Oliviera et al. (Oliveira et al., 2014), which reported 68.7% underweight patients. Depending on the category of patients, our result is close to that of Régis et al.(RRDDI et al., 2020)in Congo; Brazzaville reported previous tuberculosis in 88.6% of MDR-TB patients moreover in Adambounou et al. (Adambounou et al., 2020)study, 5.2% of retreatment cases and 2% among new cases.

Among the factors associated with anaemia, serological status alone has been documented in the medical literature as a provider of anaemia; it results from a direct action of the virus, opportunistic conditions and/or antiretroviral treatment (Attinsounon et al., 2017; Fall et al., 2017). However, apart from HIV status, vomiting, increased leukocyte, platelet, and lymphocyte levels were significantly associated with the occurrence of anaemia in patients with multidrug-resistant tuberculosis (P-Value <0.005).  Unlike  ours,   the   series   from  Oliveira  et  al. (Oliveira et al., 2014)reports an association of anaemia with the male gender (p = 0.03) and BMI (p = 0.0004).