Of 231 patients included, the age ranged from 31 - 59 months is more representative with 54.7% for men and 50.7% for women; in each of these genders as showed in Table 1 . There is no significant difference among the age groups (P = 0.777).

The [31 - 59 years] age group had more cases in both sexes with 51.9%. However, the difference is not statistically significant at the 5% level.

Table 2 below shows the distribution of patients by age groups and types of effusion fuids. There is any statistically significant difference at the 5% level among the age groups (P = 0.090).

Although the 31 to 59 age group appears to be the most affected, the difference is not statistically significant at the α = 5% (P = 0.090) level.

Considering the distribution of patients according to clinical services, surgical emergencies recorded the most important cases with 25.1% [95% CI: 19.6% - 33.7%] ( Table 3 ).

Following the distribution of patients by department, we have the distribution of samples according to clinical services.

Table 4 below gives the distribution of fluids effusion according the clinical services.

Surgical emergencies, medical emergencies and surgery had 58, 52 and 42 biologics, respectively.

Of 231 samples, 62 cultures were positive and 64 bacterial strains were identified. E. coli was most representative with 25 strains followed by S. aureus with 10 strains and K. pneumoniae with 9 strains. Streptococcus spp was less found with only 2 cases ( Table 5 ).

Legend: UC (Surgical Emergencies); UM (Medical Emergencies); GE (Big Kids); MI (Infectious Diseases); Sheave (Resuscitation).

Out of 231 samples, 62 cultures were positive and 64 bacterial strains were identified (27%). The most common bacteria were E. coli (39.7%) and S. aureus (15.9%).

Of the 106 pleural punctures collected, 20 bacterial strains were isolated by culture. Among the bacterial strains isolated, Staphylococcus aureus was the most represented (25%), followed by E. coli and Pseudomonas spp (20% for each bacterium). Table 6 shows the distribution of strains isolated from pleural fluid.

There were 35 bacteria isolated from 91 peritoneal fluids. Thus, there was a polymicrobial culture with two bacteria. The most common bacterial strains were E. coli (57%) and K. pneumoniae (17%). Table 7 below shows the bacterial strains found in the peritoneal fluid.

Of these samples, 8 were culture-positive, including a two-germ polymicrobial. The most involved bacteria were S. aureus at 33% of the germs and P. aeruginosa at 22%. Table 8 below shows the distribution of the identified bacterial strains in joint fluid.

For the 4 pericardial fluids, no germs were isolated.

For the sample which was positive on culture, bacterial strains underwent antibiotics susceptibility testing.

In the following lines, we described the antibiotics susceptibility profile of bacterial profiles according to the type of effusion fluid.

Out of 20 bacterial strains isolated from pleural fluids, 07 resistance phenotypes were identified, including 03 Gram-negative bacilli (GNB) that were ESBL-producing and 04 Gram-positive cocci (CGP) resistant to penicillin G. All strains of staphylococci were susceptible to cefoxitin. Table 9 shows the antibiotics susceptibility of these strains.

Legend: AmoxiClav: Amoxicillin + Clavulanic Acid; C3G: Third-generation cephalosporins (Ceftriaxone/ceftazidime).

In this work, 35 bacterial strains were isolated from peritoneal puncture fluids. Of these, 09 (33%) strains of Enterobacteriaceae were ESBL-producing, 04 (17%) NGBs were resistant to imipenem, and 20% were MRSA-resistant. Resistance to penicillin G from Staphylococcus spp was 83%. Table 10 below highlights antibiotics susceptibility of strains isolated from peritoneal fluid.

Legend: AmoxiClav = Amoxicillin + Clavulanic Acid; C3G = Ceftriaxone/ceftazidime.

In joint fluid, 09 bacterial strains were isolated. Antibiotics susceptibility showed that isolated staphylococci were sensitive to cefoxitin (4/4) and all PGCs resistant to Penicillin G (3/3). Table 11 presents details on antibiotics susceptibility.

At the end of the presentation of the results of antibiotic susceptibility according to the bacteria involved in fluid infections, we propose a synthesis of active antibiotics by treated biological product.

This distribution is common in paediatrics and has also been used by some authors [12] [19] [23] . The most affected age group is older children and adolescents aged 31 - 59 months, regardless of the effusion fluid involved. This age group accounted for 56.7% of cases and the mean age was 75.8 months. The same trend has been reported by previous studies [15] [24] [25] . For joint and pericardial fluid, no cases were collected in infants.

The overall bacteriological profile of the effusion fluids showed Enterobacteriaceae as E. coli, K. pneumoniae and Staphylococci. All these microorganisms are commensal to the skin or mucous membranes. This could mean that the infection of the effusion fluids would come from a local commensal flora. According to François Dénis, infections of effusion fluids could be the consequence of the extension of a local infection or bacterial translocation from the digestive flora or bacterial dissemination by the blood-borne route from a primary focus [21] .

Pleural fluid: In pleural fluid, we had 35% Staphylococcus spp whose S. aureus (25%) and 20% strains of Pseudomonas spp. Some studies in Burkina Faso had yielded Pseudomonas spp. (19.2%) and S. aureus (30.8%) [19] . Likewise in Democratic Republic of Congo and Benin, it also meant a predominance of strains of S. aureus respectively 30% and 40% [12] [13] . In addition, 40% of Enterobacteriaceae were also isolated. In the literature, Enterobacteriaceae ranging from 7% to 30% had also been reported in the sub-Saharan zone [12] [13] [19] . However, only 5% of Streptococcus spp have been isolated in our work, while strains of S. pneumoniae are described. A prevalence of S. pneumoniae who goes 10% in Lubumbashi at 39.8% in Dakar [11] [12] [13] . The absence of S. pneumoniae in this study could be explained by the introduction of PCV13 in the Expanded Programme on Immunization on the one hand and sampling fluctuations on the other hand. In addition, there is evidence that S. aureus would be predominant in developing countries, while the S. pneumoniae predominates in developed countries [13] .

Peritoneal fluid: The bacteria responsible for community peritonitis is mainly from the intestinal flora with a predominance of E. coli [23] [26] . The bacteriological profile of bacterial strains isolated from peritoneal fluids showed: E. coli 57%, K. pneumoniae 17%, P. aeruginosa 6%. Bhat’s and Dumont’s teams have reported similar cases in India and France [24] . However, in Bamako, one team reported 29% of E. coli and 36% other Enterobacteriaceae [16] . Some teams had drawn up the bacterial profile of European paediatric peritonitis by showing a similar profile to ours with 65.1% E. coli, 10% Klebsiella spp and P. aeruginosa respectively [7] . Indeed, the difference in proportion with this study could be explained either in the size of the sample since this study concerned more than 1259 germs isolated from 16 countries, or by the fact that Gram positive cocci were not taken into account in their study. In addition, 17% of Staphylococci were isolated. Dumont and Castagnola teams had reported 11% and 5% of CGP successively in their series [24] [27] .

Strains isolated from Joint fluid: In joint infections, S. aureus et K. kingae are the most common bacteria; better, K. kingae had even become the first pathogen responsible for septic arthritis in infants. Our data showed a bacteriological profile dominated by CGP (66%). Indeed, S. aureus accounted for one-third of the strains. In Tunis, some studies found a similar profile with 31.8% of S. aureus and Gram-negative bacilli in 45% of cases [17] . In Morocco, according to research conducted in 2012, the authors reported a clear predominance of S. aureus 73% [15] . In 2005, in Australia, it was concluded that S. aureus was the most isolated strain in 76% of cases of hematogenous osteomyelitis and 39% of septic arthritis [25] . In addition, according to the results of work using classical culture, the etiological diagnosis is missing in 50% to 80% of cases [10] . Likewise, K. kingae is a pathogen emerging from very difficult crops. This could justify its absence among the isolated pathogens.

Pericardial fluid: Although, any pathogens were isolated from the pericardial fluid. Cytology revealed that half of the samples had a lymphocytic cytological profile. This could refer to a viral, tuberculous or idiopathic etiology. The macroscopic and microscopic appearance reveals that the other half of the effusions were haemorrhagic. Earlier studies had already made similar observations [28] . This most often refers to a tuberculous etiology. In some underdeveloped geographic areas, massive pericardial effusion and cardiac tamponade are frequently secondary to tuberculosis. At the end of this presentation of bacteria by type of effusion fluid, it’s important to look about their antibiotics susceptibility profile.

Antibiotic susceptibility profile: The antibiotic susceptibility profile of the 64 isolated bacterial strains showed 30 resistance phenotypes. Of these, 33% of Enterobacteriaceae strains are ESBL producing. According to the results of a study carried out, the lack of routine surveillance in most countries and in the West African region does not allow for an accurate estimation of the proportions of ESBL among strains isolated during infectious processes [22] . Nevertheless, in Guinea-Bissau, 32.6% ESBL carriage in children under 5 years of age has been described [29] . Similarly, among the hospital population, a prevalence of ESBL carriage of 31% among children hospitalized for malnutrition in Niger was reported [30] . In Ghana, authors reported that 49.4% of Enterobacteriaceae isolated at Korle-Bu Hospital were ESBL-producing [31] . Then, the prevalence of ESBLs in our work is consistent with that previously described in West African Region. In addition, we also had 13% carbapenem-resistant strains. Carbapenemase-producing strains have also been described in Sierra Leone and Senegal [32] [33] . In Nigeria, varying prevalences have been reported depending on the level of care: for example, in regional hospitals, the prevalence is about 10% compared to 36% in reference [34] [35] [36] . All these different studies show that we are facing an emergence of resistance to carbapenems.

We report 10% resistance to methicillin. Some authors had isolated 37% of MRSA from pleural fluid [19] . This finding is consistent with the prevalences of MRSA in the sub-Saharan African region with 10% in Yaoundé, 13% in Niamey, 34% in Dakar and 36% in Benin [37] [38] .

We note that strains isolated from peritoneal fluids include the majority of resistance phenotypes (75% of ESBLs, 100% of carbapenemases and 100% of MRSAs). Of 35 pathogens isolated from peritoneal puncture fluids, the most common bacterial strains were Enterobacteriaceae with 33.7% ESBL 33.7% of K. pneumoniae and 6.8% E. coli [7] . Isolated staphylococci were resistant to penicillins G. Among them, 17% is MRSA. As reported on clinical samples from five African cities, there is a prevalence of more than 15% of MRSA on the continent [37] .

Susceptibility Profile of Bacterial Strains Isolated from Joint Fluid: The most common strains in joint fluid were methicillin-sensitive staphylococci. No MRSA has been isolated from this pathology. In literature, Trifa et al. reported in Tunis, 59% of S. aureus methicillin-susceptible during osteoarticular infections in children. However, the same team reported 12% resistance to methicillin [15] .

Antibiotics active on isolated bacterial strains: Some studies suggested the standardization of an effective antibiotic therapy on the main pathogens involved in pleurisy. Indeed, in their work, the probabilistic treatment was based on the combination of a beta-lactam associated with an aminoglycoside or a macrolide [11] [13] . We believe that in this combination, beta-lactam must be a 3rd generation carbapenem or cephalosporin due to the low sensitivity of other cephalosporins. However, empiric antibiotic therapy in paediatric peritonitis is often penicillin + inhibitor-based monotherapy (ampicillin + sullbactam/piperacillin + tazobactam) or ertapenem. Unfortunately, more than 38% resistance to ampicillin + sullbactam has been reported [27] . According to the work of Lob et al., only imipenem and amikacin had greater than 90% sensitivity in peritonitis [7] . However, the recommendation of monotherapy with penicillin plus inhibitors or carbapenems remains unlikely because of ESBL-producing strains..

The initial therapeutic choice is empirical in osteoarticular infections. Despite decades of experience with numerous and diverse protocols, the analysis of clinical studies does not determine the best antibiotics for osteoarticular infections. An association is required. Primary treatment may be a combination of penicillin-resistant penicillinase and a third-generation cephalosporin. An alternative therapy is the combination of vancomycin, fosfomycin, rifampicin or clindamycin with a third-generation cephalosporin [39] .

We can notice that effusion fluids make the patient uncomfortable. The best way to avoid them is to eliminate the etiologies of the effusion. This could be done towards the treatment of the diseases. Similarly, we don’t have any information about the influence of local weather on the fluids effusion infection. Nevertheless, these infections are commonly found in diseases as cancers particularly in T‑lymphoblastic lymphoma even it is a rare subtype of non‑Hodgkin lymphoma [40] . Elsewhere, many cases of fluids effusion are observed in some inflammatory diseases like polyserositis. Then, it can be observed some cases of neoplasm with an increase level of pleural lactate deshydogenase [41] . Moreover, some studies had shown that the abnormal effusion fluid were observed in auto-immune diseases [42] . Finally, some unknown etiologies cases are described with effusion fluids [41] . Despite these limitations, we were able to study the bacteriological profile of effusion fluids.