Purulent pleurisy is a pathology that was first described by Imhotep, an Egyptian doctor, around 3000 BC [1]. Hippocrates, the most cited and considered the father of medicine, described this pathology around 500 BC [1]. Since the latter, many therapies have been adopted and developed. LIGHT [2], in his book Pleural Diseases, gives us a perfect summary of these therapies, as well as the epidemiology of this pathology (incidence, bacterial ecology, etc.).

Pneumonia is believed to be the primary etiology of pleurisy. In England, 60% of hospitalizations for pneumonia have been reported to be complicated by pleurisy [3]. On the other hand, LIGHT [4] reports a share of 20-40% of hospitalizations for pneumonia in the United States, each year. The British Thoracic Society reports an annual incidence of 80,000 cases, United States and United Kingdom combined [1]. These are all data that reveal the importance and interest of clearly identifying this pathology.

In Africa, few studies have yet been carried out on a large scale, but some work already attracts attention among the most recent:

• In Togo, in 2003, Dagnra et al. [5] showed a clear predominance of gram-positive germs in purulent pleurisy; it was Staphylococcus aureus, followed by Streptococcus pneumoniae

• In Morocco, at the Pneumology department of Marrakech, purulent pleurisy represented 12.6% of pleurisies hospitalized from 2005 to 2009 [6]; Still in the same hospital, Rachidi et al. [7] published in 2019 a sex ratio of 1.31 and also a predominance for gram positive bacteria (Streptococcus pneumoniae followed by Staphylococcus aureus)

• In Cameroon, Perfura et al. [8] reported in 2012 that purulent pleurisy represented 11.6% of pleurisies hospitalized at Jamot hospital; There was a sex ratio of 1.7 and HIV was the most common predisposing factor. In the Ivory Coast, the Pneumology department of the Cocody University Hospital reported in 2013 that gram-negative germs are the most responsible for purulent pleurisy in HIV-positive patients [9]

Taking the case of our country Gabon, Ondo N’dong et al. [10] published in 2007 a work just as remarkable as the others mentioned above. In the Thoracic and Vascular Surgery department of the Jeanne Ebori Foundation. The prevalence of purulent pleurisy was 4.94%, with male predominance. Gram negative cocci were the most common, including Proteus mirabilis and Klebsiella pneumoniae. It should be noted that this was a retrospective study conducted from 1990 to 2005.

Thus, in order to obtain recent data and to make a new inventory, we carried out this retrospective and descriptive study over 3 years and 5 months, in the Surgery Department of the University Hospital of Libreville. It aimed to bring out the epidemiological and therapeutic aspects of purulent pleurisy in the said service.

Patient and Method

We carried out a retrospective, descriptive and analytical study in the Thoracic and Vascular Surgery department of the Center Hospitlier Universitaire de Libreville over a period of three (3) years and five (5) months.

We included in the study all records of patients hospitalized for purulent pleurisy, regardless of etiology. The diagnosis was to be made by pleural puncture, bringing back a cloudy or clearly purulent fluid with a viscous yellowish, greenish or other color.

We did not include all the files concerning tumors and other pleuropulmonary diseases, as well as the incomplete files.

Data was collected using an anonymous collection form. This sheet was based on several levels of information:

• The identity of the patient: Sex, age, profession, month and year of consultation

• History and Comorbidities: Vaccination status, pleuropulmonary infection and/or intervention, existence of underlying pathologies such as HIV, malnutrition, diabetes, etc ...

• Habits and Lifestyle: Alcohol, tobacco, etc ...

• Anamnesis and clinical signs (functional and physical)

• Morphological paraclinical examinations (chest X-ray, CT, pleural ultrasound) and biological (NFS, CRP, bacteriology of pleural puncture fluid, etc.)

• The etiology: pneumonia, lung abscess, etc ...

• The Type of Treatment: Medicinal (antibiotic in particular), surgical (drainage, decortication)

• The evolution during hospitalization, after a gesture of pleural evacuation

We have collected and analyzed clinical, radiological, biological, therapeutic (medical and surgical) and even evolutionary parameters.

The clinical and biological parameters were reported only on the basis of the medical file and the reports of the biological examinations included therein.

The radiological results were reported either on the basis of the radiologist's report or on the surgeon's interpretation.

The medical treatment reported in the data consisted of antibiotic therapy and was adapted according to the clinical course or the result of the antibiogram. Only antibiotic therapy that allowed clinical remission of symptoms was reported in this study for negative cultures. In cases with positive culture, antibiotic therapy according to the antibiogram alone was postponed.

Management by drainage was indicated from the outset in all patients with moderate or abundant effusion. This method was considered to be unsuccessful in the absence of remission of the sepsis and/or absence of drying up of the pleurisy or its reconstitution.

As endoscopic surgery was not available (although ideal in all patients who could benefit from surgery at stage 2/3 of the empyema), decortication thoracotomy was the only surgical technique performed in case of failure of the empyema. drainage.

The patient's progress before discharge was postponed, as well as the course of action to be taken for subsequent follow-up (i.e. a surgical check-up or transfer to another service/structure).

Medium and long term course and particularly after discharge, was not part of this study.

The data were entered using Microsoft Excel 2010 software and were analyzed using Epi info 7 software (version 7.2.0.1). We performed a simple descriptive analysis of the different variables. The quantitative variables were reported either with their mean± standard deviation, or with their number and frequency.

Epidemiological Data

Prevalence of the 3219 patients received at the Surgery Department of the University Hospital of Libreville, During the study period, 41 presented with purulent pleurisy, i.e. a prevalence of 1.3%. Nine (9) files were not selected because they did not meet all the inclusion criteria. Thirty-two cases, or (32) patients were selected for the study.

Gender

Among the 32 patients selected, 18 were female and 14 were male, for an M/F sex ratio of 0.78.

Age

Age ranged from 2 months to 60 years, with a median age of 2 years and an interquartile range [1-3]. The children were under 5 years of age in 25 cases (78.1%), including 21 cases of infants, or 65.6% of the total number. The proportion of the latter was followed by that of adolescents (4 cases, or 12.5%) then adults (3 cases, or 9.4%).

Consultation Period

The months of December to February (9 cases) and from June to August (8 cases) saw the most consultations, for a total of 17 cases (Figure 1). These periods correspond respectively to the small and the large dry season.

Figure 1: Seasonal Distribution of Patients

Figure 2: Breakdown of Patients According to Consultation Time

Background and Field

Twenty-five (25) patients (78.1%) had no specific medical-surgical history. Seven (7) patients had an isolated or associated history: HIV infection (2 cases), malnutrition (1 case), nephrotic syndrome (1 case), hemothorax (1 case) and history of pleural empyema (1 case), sero pleurisy -fibrinous (1 case).

Clinical Data

Consultation time and origin Twenty-seven (27) patients (84.4%) consulted directly from their homes. The rest were referred from a medical facility. The consultation time varied between one (1) day and one (1) month, for an average of 14.5±9.9 J.

Twelve (12) cases among the patients from their home had consulted more than a week from the onset of symptoms, of which 8 (or 66.7% of this group) had already received antibiotic therapy (Figure 2).

Reasons for Consultation and Clinical Signs

Fever was present in all cases and was always one of the reasons for consultation. It was estimated at 38.9±0.8°C. Dyspnea (23 cases) and cough (18 cases) were the two other reasons for consultation (Table 1).

Imagery

All patients underwent a chest x-ray. Only some patients had subsequently undergone an ultrasound (4 patients) or chest CT (6 patients), under the   following  conditions:

• Septic pleurisy

• Estimation of pleural fluid

• Persistent symptoms despite a treatment (antibiotic therapy and/or surgery) well conducted

The patients who received only an x-ray were those whose images directly referred to free pleurisy.

The radiological lesion was that of a completely opaque lung, or 18 cases (56.2%). The septate pleurisy was the clinico-radiological form in 11 cases, i.e. 34.4%. Exceptionally, we found 1 case (3.1%) of lung abscess and 1 case of complete lung atelectasis (Table 2).

The other lesions (free pleurisy, alveolar syndrome and interstitial syndrome) are described in Table 3.

Pleural fluid was abundant in 24 cases, or 75%.

Biology

Pleural Puncture the culture of the liquid came back sterile in 24 cases (75%). Staphylococcus aureus (S. aureus) was present in 6 cases, or 18.7% (Figure 3).

BK (Koch's bacillus) was suspected in 3 cases (9.4%), following a favorable history, sterile culture and favorable clinical outcome under anti-tuberculosis drugs.

Other Biological Examinations

The CBC of 28 patients (87.5%) demonstrated hyperleukocytosis with an average of 22,154.7±12,134/mm3, as well as anemia (Table 4). CRP averaged 131.6±99.3 mg/L.

Figure 3: Bacteriology of Pleural Fluid

Figure 4: Number of Bacterial Strains Found by Age Group

Table I: Summary of Clinical Signs

Table 2: Distribution of Radiological Signs

Table 3: Distribution of other Biological Anomalies

Exceptionally, K. pneumoniae has been isolated from one patient at ECBU.

Distribution of Germs Found According to Age

Staphylococcus aureus was most isolated in children (6 cases), infants (4 cases) in particular (Figure 4). 

Etiology

Bronchopneumonia was the cause in 20 patients (62.5%). The other causes are shown in Figure 5.

Treatment

Drug treatment was initiated in all patients, according to the rules of conventional antibiotic therapy, then readjusted in cases of positive culture, or in strong suspicions of pulmonary tuberculosis.

Ceftriaxone was the most widely used antibiotic (46.9%), followed by gentamicin (43.7%), metronidazole (18.7%), vancomycin (18.7%) and amoxicillin- clavulanic acid (12.5%). The rest is shown in Table 4.

Pleural drainage was performed in all patients except one (1) only who benefited from iterative punctures.

Surgery (by decortication thoracotomy) was performed in 11 cases (34.4%), concomitantly with pleural drainage. 

Evolution

The majority of patients (23 cases, or 71.9%) had a favorable outcome.

Figure 5: Distribution by Etiology

Figure 6: Distribution by Exit Delay

Table 4: Frequency of use of Antibiotics

Table 5: Evolution According to Instrumental and Surgical Treatment

Among the persistence cases (5 cases), none could benefit from decortication thoracotomy. Three of them were released against medical advice, 1 case (HIV positive) had been transferred to Infectious Disease because his CD4 count (too low) did not allow him to benefit from surgery; the latter had a subsequent appointment for surgery, but was lost to follow-up.

The discharge time was 23.7±8.8 J on average (for extremes of 9 to 50 J), with a greater proportion of patients (12 cases), between 20 and 29 J (Figure 6).

Among the patients who had a favorable outcome, 22 cases (62.5%) had undergone drainage while only one (3.1%) was treated by iterative puncture. Eleven (11) patients among the cases of drainage underwent concomitant pulmonary decortication, but only 10 had a favorable outcome (Table 5).

A proportion of 16.7% (5 cases, all types of treatment combined) had an unknown outcome due to lack of sufficient information in the medical records.

We did not note any cases of death in our study.

We carried out a retrospective and descriptive study in the Surgery Department of the University Hospital of Libreville, in order to determine the therapeutic epidemiological profiles of the management of purulent pleurisy.

Epidemiological Data

Prevalence during our study, we found a total of 3219 hospitalized patients, including 41 cases of purulent pleurisy. Which gives us a frequency of 1.3%.

Many African authors have found variable data, even under more restricted conditions (i.e. for samples of adults only or children). Ondo Ndong et al. [10] in Gabon, found a prevalence of 0.5%, which is close to that of Koueta et al. [11], with 0.7% in Burkina Faso. In other skies, in Niger, Garba et al. [12] report a prevalence of 1.04%. The similarity and variability of these results, however, can be explained by the proximity or proportionality of the sample size to our study, the similar study duration and also the retrospective nature of the majority of these studies.

Gender and Age

Our study reports a sex ratio of 0.78 in favor of women. Other authors show a clear predominance for males [5,13,14]. However, Goyal [15], as well as Guyon [16] highlight data similar to those of our study.

It must be deduced from this that the sex, in particular male, would be strongly linked to a risk of purulent pleurisy. However, in the light of these studies, we noted the following phenomenon: the lower the average age of the study population, the more the female sex would tend to be affected to the detriment of the male sex. This is an observation that remains to be confirmed, however and could be the subject of further study.

Seasonal Distribution

The dry season (and the coldest of the year in particular) is the period with the most cases of purulent pleurisy. In our review of the literature, some authors [11,12,16] find the same trend as ours, while in other skies, the resurgence of cases is observed more in the hot and humid season [17,18].

Each of the observations is just as justifiable as the other. Humidity and heat are known to everyone to promote the proliferation and transmission of most pathogenic bacteria in humans, while cold and dryness weaken most tissues, especially those of the respiratory mucous membranes. However, the length of the seasons, the incubation times of the germs and to a certain extent the overall susceptibility to infections of individuals in the same region, are all elements that could better justify the seasonal profile of a location in the region. globe to another.

Background and Favoring Factors

The majority of patients (78.1%) in our study had no particular history, even if some patients had a particular situation (HIV, malnutrition, nephrotic syndrome, etc.). However, some authors [8,19] show an influence of HIV infection, smoking, alcoholism and malnutrition.

One would be tempted to say that in our context at the CHUL, there is no terrain or favorable factor. However, the size of the sample, the retrospective nature of the study and the difference in methodology compared to the aforementioned work largely explain why our study does not find any favorable factor.

Clinical Data

Consultation Deadline: A period of 14.5±9.9 J elapsed between the onset of symptoms and the admission of patients to our study. The review of the literature reports a similar delay, even sometimes greater than ours [8,18]. However, there are works which reduce the time to 7-9 days shorter [20, 21].

The relatively long delay is justified by the fact that many people resort to self-medication before consulting. Also, in many African countries, as in our study, many patients have difficult access to care, due to their geographical location, or the cost of examinations which is sometimes high. On the other hand, the short delay is seen in developed countries with more developed health systems, better access to care, better-off and more educated populations (from a health point of view).

 Reasons for Consultation and Clinical Signs

Fever, followed by cough, then dyspnea are the most recurring reasons for consultation. The same tripod of symptoms is found in many series [11,22,23], without any particular order of succession. In other series [8,10,21], chest pain takes precedence over dyspnea, but the other two symptoms remain predominant.

These differences are easily explained by simply observing the age of the study population. We note that in populations made mostly or only children, the first tripod of symptoms is recurrent; whereas in an adult population, fever and cough are followed directly by chest pain.

Location of the Effusion

The effusion is located on the right in the majority of cases (53.1%). Data from the literature for the last 10 or even 20 years [11,15,17,24,25] show this right predominance.

The anatomy of the lower respiratory tract could justify such a distribution; the right main bronchus and lung being naturally larger than those on the left. This physiologically leads to a greater exposure to airborne germs and therefore a greater risk of developing right pleuropneumonia.

Radiological Data

The radiologic lesions were dominated by a "white lung" appearance, followed by radiologic alveolar syndrome and interstitial syndrome.

Many authors [11,18,22,26] report the same results in varying proportions, with a predominance of either a "white lung" appearance or reticulo-interstitial lesions.

This variability would be the subject of further study, but could be explained by the nature of the germs encountered in these studies, taking prior antibiotics can modify the behavior of the germs, as well as the formation of lesions. Also, it should be added that age would be an additional factor: the total opacities of a hemi thorax would tend to appear much more in children, in particular infants.

Biological Aspects

Bacteriology of Pleural Fluid: Staphylococcus aureus (18.7%) is the first germ the most found in our study. Many studies [5,11,18,23,26] find this germ, followed by pneumococcus. Other series [8,12,28,29] differ in the predominance of the latter. It is also necessary to note the high rate of sterile cultures which was 75%.

Such results in our study can be explained by many intertwined factors:

• The effect of pneumococcal vaccination (EPI and national vaccination programs) which produces a natural selection of bacterial ecology and decreases the risk of pneumococcal infection [21,30]

• The use of antibiotics before admission (especially at home), especially beta-lactams, which are generally more active on pneumococcus than staphylococci (secretors of beta-lactamases). This observation influences both our ecology and the crop positivity rate

• The good conditions for collecting and transporting biological fluids to the laboratory, which are not always correct

• The limited resources (insufficiency, lack of appropriate materials) and the difficulty of cultivating anaerobic germs in our context in particular

In addition, the studies underlined above with a predominance of pneumococcus, all have the first to be prospective and to have been carried out with a better technical platform (Kit for research of soluble antigens, PCR ...) and a better follow-up. patients. Another peculiarity and observation made in these studies is the presence of at least one immunosuppressive factor (alcohol, tobacco, HIV infection, malnutrition, etc.) in relatively large proportions, favorable to the development of pneumococcus. We must add, despite the vaccination programs established around the world, the possibility of the emergence of pneumococcal serotypes not included in the vaccines used for these studies.

Blood Disturbances

Hyperleukocytosis (87.5%), anemia (87.5%) and increased CRP (75%) were the three blood abnormalities found. Many authors [12,18,26,27] find the same anomalies, with a few similarities.

Concerning these biological parameters, the literature always reports these same disturbances. In the majority of cases, leukocytes and inflammatory markers (VS or CRP) always come back high, while the Hb level is low. Bacterial etiology, as well as the infectious and inflammatory processes that accompany it, are naturally responsible for these abnormalities.

Etiology

Purulent pleurisy was post-pneumonic in most (62.5%) of the cases in our series.

This data is consistent with that of Ondo N’dong et al. [10], Hicham et al. [24], as well as Malhotraa et al. [31]. This global observation is justified by the natural and physiological position of the lungs. These organs are the last place of exchange with the external environment after the passage of air through the upper airways. The contiguity of the lungs with the pleura is the main factor and mechanism of inoculation of bacteria to the pleural space. However, by mechanisms still poorly understood, there are possibilities of direct inoculation, without passage through the lung [1, 3].

Therapeutic and Evolving Aspects

Medical Treatment: All of our patients received antibiotic therapy according to the latest recommendations [1]. It was a beta-lactam (amoxicillin-clavulanic acid or ceftriaxone), sometimes combined with an aminoglycoside. The treatment was then readjusted according to the antibiogram for the cases of positive culture. No case has been completely cured with antibiotic therapy alone, although some authors such as Hernandez-Bou et al. [39], as well as Carter et al. [32] in particular, show that this is possible.

Instrumental Treatment

All the patients in this series underwent pleural drainage (31 cases), or iterative puncture (1 case). Thirteen patients (40.6%), including 12 (37.5%) cases of drainage and a single case (3.1%) of iterative punctures, responded favorably, without resorting to surgery.

The results of the literature on the success of first-line drainage are very variable in Africa [8,11,12,33] and elsewhere [34-36]. Many parameters have to be taken into account. First, there is accessibility to the chest tube. From a financial point of view, the drainage kit is not always within the reach of all patients. This naturally results in a delay in laying the drain, which is a limiting factor in the effectiveness of drainage. Another factor that could influence the effectiveness of drainage depending on whether it is long or short is the consultation time. The delay caused by the cost of the drain and the long consultation time have a direct relationship with the pleurisy stage. The longer this delay, the greater the risk of encystment, which is linked to a greater risk of failure. We can easily make this observation in the series of Almaramhy and Allama [27] and that of Mandal et al. [37]. Another fact that can influence the result of drainage in a population is the tuberculous or non-tuberculous origin of the empyema. Kundu et al. [38] show in a prospective study that drainage tends to have better results in pleurisy when the origin is non-tuberculous. Our sample is however too small to make such a comparison.

Surgical Treatment

Decortication thoracotomy, indicated as a last resort according to the latest recommendations [15], is the only surgical option in our series. Thoracoscopy, which is the first-line method [15], is not yet available in our department. Eleven patients (34.3%) had undergone decortication thoracotomy and among them 10 (31.2%) had a favorable outcome. These figures show the failure rate of any other measure of pleural evacuation (drainage and thoracentesis in particular) and the success rate of decortication in this series.

Many series report varying results regarding their surgical method. Kundu et al. [38] in their series analyze two types of populations: tuberculous and non-tuberculous etiologies; 55% had required dehulling in the 1st sample, compared to 10.9% in the other. Sakran et al. [35] in their study obtained 64% of their patients who were eligible for surgery (by thoracoscopy in particular). Semenkovich et al. [39] obtained 44% (including 24% by thoracoscopy and 20% by thoracotomy) in a population that initially benefited from drainage. All these observed disparities largely depend on the success rate of the thoracic drainage. It should be added, however, that the rate of patients who benefit from surgery is influenced by age and clinical status; Young adults cope better with heavy surgeries compared to those at the extreme ages of life. In addition, certain clinical situations with comorbidities do not always bode well for good results.

Duration of Hospitalization

We obtained a mean hospital stay of 23.7±8.8 J with extremes of 9 and 50 J. A greater proportion (37.2%) of patients was between 20-29 J.

More or less similar results are found. Massika et al. [18] found a delay of 22.67±10.97 J. Perfura et al. [8] found 26.8±15.6 J. Bremont et al. [32] found an average of 24 J with extremes de 15-45 J. In addition, some authors report different results. Meier et al. [40] found 14.6 J. Maffey et al. [41] found an identical mean of 14±6 J. Hicham et al. [24] obtained an even lower mean duration of 7 J. This difference in results can be observed justified by many factors. The time taken to take charge is an important element. It is associated with the patient consultation time. These two facts cause a certain delay in the installation of a drain and/or the performance of the surgery and lengthen the patient's stay. Also, the more invasive the pleural evacuation method, the more it would tend to lengthen the stay. Kondov et al. [34] make this observation in a retrospective study: individuals having undergone drainage, decortication or thoracoplasty had respectively 11, 4 J, then 23.3 J and 42.2 J of hospitalization on average. In addition, the existence of comorbidities, the presence of numerous areas of encystment of pleurisy and even the presence of gram-negative bacteria would tend to prolong the stay. Ghosh et al. [42] make the same observation in a prospective and observational study.

Purulent pleurisy is still a rare pathology in our country, in particular in our surgical department at the CHUL. It is an easy to diagnose pathology because it is essentially clinical.

Staphylococcus aureus is the most common germ. The common radiological lesion is that of a "white lung". Thoracic drainage, coupled with antibiotics is the therapeutic means that offers the most satisfactory results. In the event of drainage failure, decortication thoracotomy is still our only alternative, but provides good results.

Moreover, it is clear that diagnostic and therapeutic approaches still need to be developed. Many factors relating to our structures (insufficient technical facilities, high cost of certain services) and to the patient (self-medication, financial difficulties) further limit the management of purulent pleurisy and the expected results.

1. Davies H.E., et al. "Management of pleural infection in adults: British Thoracic Society pleural disease guideline 2010." Thorax, vol. 65, no. 2, 2010, pp. ii41–ii53.

2. Light R.W. Pleural Diseases, 6th ed., Philadelphia: Lippincott Williams & Wilkins, 2013, 504 p.

3. Ferre A., Dres M., Azarian R. "Pleurésies purulentes." EMC - Pneumologie, Elsevier Masson, Paris, 2011, 6-041-A-40.

4. Light R.W. "Parapneumonic effusions and empyema." Proc. Am. Thorac. Soc., vol. 3, 2006, pp. 75–80.

5. Dagnra A.Y. et al. "Nature et sensibilité aux antibiotiques des bactéries isolées des pleurésies purulentes à Lomé." Méd. Mal. Infect., vol. 3, 2003, pp. 327–330.

6. Tayewo F.F. "Profils cliniques et étiologiques des pleurésies hospitalisées au service de pneumologie de l’hôpital Ibn Nafis de 2005 à 2009." Thèse Doctorat Méd., Univ. Cadi Ayyad, Marrakech, 2011, no. 79.

7. Rachidi M. et al. "Le profil bactériologique des pleurésies purulentes au CHU de Marrakech." Rev. Fr. Lab., no. 509, 2019, pp. 77–80.

8. Perfura Yone E.W. et al. "La pleurésie purulente non tuberculeuse de l’adulte à Yaoundé (Cameroun)." Méd. Santé Trop., vol. 2, 2012, pp. 35–39.

9. Achi V.H. et al. "Etiologies des pleurésies purulentes non tuberculeuses chez les patients adultes infectés par le VIH dans un service de pneumologie à Abidjan (Côte d’Ivoire)." Rev. Pneumol. Clin., vol. 69, 2013, pp. 121–125.

10. Ondo N’dong F. et al. "Pyothorax: Aspects cliniques et thérapeutiques à Libreville; à propos de 24 cas." Ann. Afr. Chir. Cardiovasc., vol. 2, no. 2, 2007, pp. 124–128.

11. Koueta F. et al. "Pleurésie chez l’enfant: Aspects épidémiologiques, cliniques, paracliniques, thérapeutiques et évolutifs." Clin. Mother Child Health, vol. 8, 2011, pp. 1–6.

12. Garba M. et al. "Profil épidémiologique et pronostic de la pleurésie purulente chez l’enfant." J. Fran. Viet. Pneu., vol. 19, no. 6, 2015, pp. 49–54.

13. Domoua K. et al. "Aspect étiologique et problèmes thérapeutiques des pleurésies purulentes à Abidjan." Bull. Soc. Pathol. Exot., vol. 88, 1995, pp. 199–202.

14. Laishram N. and Ngangom D. "Empyema thoracis in children: a clinical study." J. Evol. Med. Dent. Sci., vol. 5, no. 1, 2016, pp. 46–49.

15. Goyal V. et al. "Empyema thoracis in children: Still a challenge in developing countries." Afr. J. Paediatr. Surg., vol. 11, 2014, pp. 206–210.

16. Guyon G. et al. "Les pleurésies purulentes de l’enfant: expérience montpelliéraine." Arch. Pédiatr., vol. 12, 2005, pp. S54–S57.

17. Dass R. et al. "Empyema thoracis: analysis of 150 cases." Indian J. Pediatr., vol. 78, no. 11, 2011, pp. 1371–1377.

18. Massika L.L. et al. "Suppurations pleurales chez l’enfant: aspects épidémiologiques et étiologiques." Méd. Afr. Noire, vol. 37, no. 2, 1990, pp. 24–28.

19. Dusemund F. et al. "Characteristics of medically and surgically treated empyema patients: a retrospective study." Respiration, vol. 86, 2013, pp. 288–294.

20. Shomaker K.L. et al. "Impact of an evidence-based algorithm on pediatric empyema." Pediatr. Pulmonol., vol. 46, 2011, pp. 722–728.

21. Koshy E. et al. "Impact of PCV7 on childhood hospital admissions in England." Thorax, vol. 65, 2010, pp. 770–774.

22. Desmuraux A. et al. "Complications suppuratives des pneumonies de l’enfant." Arch. Pédiatr., vol. 14, 2007, pp. 1298–1303.

23. Hassan I. and Mabogunje O. "Paediatric empyema thoracis in Zaria, Nigeria." Ann. Trop. Paediatr., vol. 12, 1992, pp. 265–271.

24. Hicham H. et al. "Surgical management of pyothorax: 172 cases." Open J. Thorac. Surg., vol. 8, 2018, pp. 50–56.

25. Subay K. et al. "Traitement chirurgical des pleurésies purulentes: 416 cas." Méd. Afr. Noire, vol. 38, no. 7, 1991, pp. 499–502.

26. Bremont F. et al. "Traitement de l’empyème pleural chez l’enfant." Arch. Pédiatr., vol. 3, 1996, pp. 335–341.

27. Almaramhy H.H., Allama M. "Indicators for surgical intervention in thoracic empyema in children." Saudi Med. J., vol. 36, 2015, pp. 1061–1066.

28. Lin T.Y. et al. "Etiology of empyema thoracis in Taiwanese children." Pediatr. Infect. Dis. J., vol. 32, no. 4, 2013, pp. 419–421.

29. Hernandez-Bou S. et al. "Pediatric parapneumonic pleural effusion." Pediatr. Pulmonol., vol. 44, 2009, pp. 1192–1200.

30. Schultz K.D. et al. "Changing face of pleural empyemas in children." Pediatrics, vol. 113, no. 6, 2004, pp. 1753–1740.

31. Malhotra P. et al. "Empyema thoracis: tuberculous vs non-tuberculous comparison." Respir. Med., vol. 101, 2007, pp. 423–430.

32. Carter E. et al. "Management of children with empyema: drainage not always necessary." Pediatr. Pulmonol., vol. 45, 2010, pp. 475–480.

33. Alao M.J. et al. "Pleurésie chez l’enfant à Cotonou: étude clinique." Mali Méd., vol. 25, no. 4, 2010, pp. 47–51.

34. Kondov G. et al. "Surgical treatment of pleural empyema-our results." Sec. Med. Sci., vol. 38, 2017, pp. 99–104.

35. Sakran W. et al. "Thoracic empyema in children: clinical and microbiological analysis." J. Infect. Chemother., vol. 20, 2014, pp. 262–265.

36. Kumar A. et al. "Empyema thoracis: A short term outcome study." Indian Pediatr., vol. 50, 2013, pp. 879–882.

37. Mandal K.C. et al. "Empyema thoracis in children: five-year experience." J. Indian Assoc. Pediatr. Surg., vol. 24, 2019, pp. 197–202.

38. Kundu S. et al. "Adult thoracic empyema: Analysis of 75 cases." Lung India, vol. 27, no. 4, 2010, pp. 196–201.

39. Semenkovich T.R. et al. "Current state of empyema management." Ann. Thorac. Surg., vol. 105, 2018, pp. 1589–1596.

40. Meier A.H. et al. "Rational treatment of empyema in children." Arch. Surg., vol. 135, 2000, pp. 907–912.

41. Maffey A. et al. "Outcome in children with pleural empyema." Pediatr. Pulmonol., 2019, pp. 1–6. https://doi.org/10. 1002/ppul.24255.

42. Ghosh S. et al. "Factors affecting drainage duration in non-tuberculous empyema." Med. J. DY Patil Vidyapeeth, vol. 11, 2018, pp. 318–323.