Pneumothorax in Children: A Study of 47 Pediatric Cases at Mali Hospital 2017-2020 ()
1. Introduction
Pneumothorax is defined as a sudden irruption of air into the pleural cavity [1] [2]. A distinction is made between primary spontaneous pneumothorax and pneumothorax secondary to respiratory pathology (cystic fibrosis, asthma, infectious lung disease), trauma (penetrating wound, pulmonary contusion, closed glottis trauma), or iatrogenic origin (biopsy, ventilation, pleural puncture, vascular catheterization, or placement of a central line [3] [4]. Pneumothorax is a rare condition in children.
The incidence of pneumothorax is estimated to be between 3.4 and 18 cases per 100,000 in men and between 1.2 and 6 cases per 100,000 in women, 2 - 4 and a recent age-stratified longitudinal study using data from the National Institutes of Health Research in Taiwan reported an increasing incidence over time. In a large retrospective analysis in the United States (1997 to 2006), spontaneous pneumothorax increased from 2.68 to 3.41 per 100,000 over 10 years [5].
In sub-Saharan Africa, data from the literature on pneumothorax are scarce. In Senegal, Nghahane BH et al. reported in their studies a prevalence of 6.93% of pneumothorax in a population of 1053 hospitalized patients [6].
The diagnosis of pneumothorax is both clinical and radiological. It is a medical and surgical emergency because it can be serious in the case of large pneumothorax or tension pneumothorax, thus threatening the child’s vital prognosis [7] [8].
The pediatric department of the Mali Hospital does not have data on pneumothorax in children, which is why this study was initiated to investigate the clinical, etiological, and therapeutic aspects of pneumothorax in children aged 0 to 15 years.
2. Materials and Methods
2.1. Study Setting
The study took place in the neonatology unit of the pediatrics department of the Mali Hospital.
2.2. Type of Study and Study Period
This was a retrospective study from January 1, 2017, to December 31, 2019, and prospective from December 1 to 31, 2020, descriptive of gas effusions from the pleural cavity in the pediatric ward of Mali Hospital.
2.3. Study Population
Our study concerned all children between one month and 15 years of age hospitalized in the pediatric ward of the Mali Hospital for pneumothorax. This population had been grouped into 4 classes: [<1 year], [1 - 4 years], [5 - 9 years] and [10 - 15 years].
2.4. Inclusion Criteria
The following were included in our study:
· All usable hospitalization medical records of children aged 1 month to 15 years admitted for pneumothorax.
· Children aged one month to 15 years were referred by other facilities for pneumothorax.
2.5. Non-Inclusion Criteria
Not included in the study were:
· Records of premature newborns and children over 15 years of age;
· All inoperable medical records of children admitted for pneumothorax.
· Records of children aged 0 to 15 years with isolated or mixed fluid effusion.
· Records of children aged 0 - 15 years whose parents declined participation in the study.
2.6. Study Parameters
The variables measured were:
· Sociodemographic variables such as age, gender, and residence of the children.
· Clinical and therapeutic variables such as: reasons for hospitalization, clinical signs of respiratory function, causes and elements of medical and surgical treatment, duration of hospitalization, and patient outcome.
· Paraclinical variables such as blood count abnormalities, C-reactive protein values, and radiological characteristics.
2.7. Study Method
2.7.1. Clinical Method
1) Operational definition:
· Complete pneumothorax: complete pleural detachment with retracted lung at the hilum [9].
· Partial pneumothorax: partial pleural detachment visible at the apex and in the axillary regions [9].
· Traumatic pneumothorax: this is the presence of an air effusion in the pleural cavity following trauma to the thorax [2].
· Iatrogenic pneumothorax: it is an air effusion in the pleural cavity following invasive medical procedures [2].
· Compressive pneumothorax: pneumothorax under the tension of great abundance [9].
· Pneumomediastinum: the abnormal presence of air in the mediastinum [10].
· Hyperleukocytosis: white blood cell count greater than 13 G/L [11].
· Leukopenia: white blood cell count below 4.5 G/L [11].
· Hyperplateletosis: platelet count greater than 440 G/L [11].
· Thrombocytopenia: platelet counts less than 160 G/L [11].
· C-reactive protein positive: serum level above 5 mg/L [12].
· C-reactive protein negative: serum level less than or equal to 5 mg/L [12].
2) Conduct of the study
· The retrospective study consisted of reviewing the medical records of children aged 0 to 15 years with pneumothorax in order to include in our study those that were usable. Data were collected using a survey form developed for this purpose. This questionnaire mentions the socio-demographic characteristics of the parents and the child, as well as the clinical and paraclinical characteristics and the outcome of the children.
· The retrospective study concerned children aged 0 to 15 years who came to the outpatient clinic or who were referred by other structures suffering from cough, respiratory discomfort, chest pain, and fever. The parents or guardians of each child were thoroughly interviewed regarding identity, residence, age, previous antibiotic and/or non-steroidal anti-inflammatory treatment, vaccination status, and tuberculosis infection.
Then a general physical examination was performed to assess the severity of the respiratory distress in order to make an urgent therapeutic decision.
2.7.2. Paraclinical Method
1) Radiological examinations
At the end of the clinical examination, a standard frontal chest X-ray was requested to confirm the diagnosis. The X-ray machine used was a WDM brand, the year 2009, with an AGFA DRY STAR 5503 printer and a digitizer for semi-digital image development.
a) Technique of performing the examination:
The examination was performed without preparation.
Infants and children: the examination is performed in dorsal decubitus position, and the plate is placed in the back. We used a low voltage in anteroposterior incidence. The criteria for success were mandatory for it to be interpretable. It should be:
· A frontal radiograph: equal distance between the medial border of the clavicles and the line of the spinous processes.
· A deep inspiratory radiograph: 6 anterior costal arches above the diaphragmatic dome in children and 5 anterior arches for infants.
· A radiograph neither over nor underexposed: visualization of the retro cardiac pulmonary artery tree.
b) Reading: The images were interpreted by a senior radiologist.
In some situations, we used transthoracic ultrasound or thoracic CT scan to eliminate other diagnoses or to look for underlying pathologies.
2) Biological and bacteriological examinations:
Biological and bacteriological examinations were then requested for etiological investigations. These were blood count, and C-reactive protein of blood culture in case of fever (Axillary temperature ≥ 38˚C).
· Blood cultures
Blood cultures are used to search for pathogenic germs after inoculation of blood on specific culture media. Sampling should preferably be done before any antibiotic therapy and at the time of febrile peaks. When several samples are taken, the blood cultures should be taken first to avoid contamination.
Before sampling, place a tourniquet and clean the puncture area with 70˚ alcohol and Betadine or Dakin. Collect sufficient blood from the blood culture bottles, respecting the broth/blood ratio. The blood culture bottles contain an aerobic (BACT/ALERT 40 ml) and anaerobic (BD BACTEC 25 ml) nutrient broth that allows the growth of aerobic and anaerobic bacteria. Label the vials and place them in the oven at 37˚C.
· Blood cell count
The counting of blood elements is done by hematological automatons after sampling whole blood on a tube containing an anticoagulant, preferably EDTA type. The leukocyte count can be done manually by reading blood smears stained with May Gruwal Giemsa.
Before the sampling, place a tourniquet and clean the puncture area with 70˚ alcohol. Perform a venipuncture at the elbow or the back of the hand. Fill the collection tube and immediately mix the blood with the anticoagulant by turning the tube over several times for a few seconds. Depending on age and hemoglobin level, anemia has been classified as mild, moderate, and severe (Table 1).
· The C-reactive protein assay
C-reactive protein is a marker of inflammation. The blood sample is taken in a dry tube without additives or a tube containing sodium heparinate. Prior to collection, place a tourniquet and clean the puncture site with 70˚ alcohol. Perform a venipuncture at the elbow or the back of the hand. Fill the collection tube and immediately mix the blood with the anticoagulant by turning the tube over several times for a few seconds. If the sample is taken in a dry tube, allow the blood to coagulate completely before centrifugation.
2.8. Bioethical Aspects
The authorization of the hospital management and the parents of the children has been obtained for the use of the data. The anonymity and confidentiality of the patients were respected in accordance with the rules of medical ethics and the legislation on biomedical and scientific research. There was no conflict of interest in this study. The integrity of the data was respected.
2.9. Data Capture and Analysis
Data were collected from the hospital medical records of children aged 0 to 15
![]()
Table 1. Classification of anemia in children according to age by the World Health Organization [13].
years. They were entered and recorded with Excel 2016 software and analyzed with SPSS Statistics version 25 software.
3. Results
3.1. Epidemiological Aspects
3.1.1. Frequency
From January 2017 to December 2020, 5569 children aged 0 to 15 years were hospitalized in the pediatric department of the Mali Hospital, including 47 cases of pneumothorax, or a frequency of 0.84%.
3.1.2. Socio-Demographic Characteristics
Children under 5 years of age were the most affected (80%). The mean age was 2.47 with extremes of 0 and 15 years. The sex ratio was 1.47 in favor of boys. More than half of the children lived in urban areas (61.7%) (Table 2).
3.2. Clinical Aspects
Respiratory discomfort was the most frequent reason for hospitalization (95.7%) followed by cough (78.7%) and fever (61.7%). Clinical examination found pulmonary auscultation abnormalities in all children (100%) followed by hypoxia (91.5%) and signs of struggle (85.1%) (Table 3).
Infectious causes, dominated by bullous dystrophies, are the most frequent (70.2%), followed by traumatic causes (25.5%) and iatrogenic causes (4.2%) (Table 4).
3.3. Radiological Aspects
Involvement of the right hemithorax was more frequent in our children (68.1%). Pneumothorax was compressive in 20 of the 47 children (42.6%) (Table 5).
![]()
Table 2. Distribution of children by sociodemographic characteristics.
![]()
Table 3. Distribution of children according to reasons for hospitalization and signs on clinical examination.
![]()
Table 4. Distribution of children according to causes of pneumothorax.
3.4. Biological Aspects
Anemia was observed in the vast majority of our children (7913%), including 17.6% with a severe form. White blood cell abnormalities were observed in 26 of the 34 children (76.4%). These were hyperleukocytosis in 67.6% of them and leukopenia in 8.8% (Table 6).
C-reactive protein was measured in 21 of the 47 children (44.6%). It came back positive in 76.2% of the children (Table 7).
![]()
Table 5. Distribution of children according to radiological characteristics.
![]()
Table 6. Distribution of the children according to the results of the blood count.
![]()
Table 7. Distribution of children according to C-reactive protein results.
3.5. Therapeutic and Evolutionary Aspects
Surgical drainage was widely used in our children (87.2%). The amoxicillin/clavulanic acid combination was used in almost all of our children (97.9%). We recorded a cure rate of 74.5% and a death rate of 25.5% (Table 8).
![]()
Table 8. Distribution of children according to therapeutic attitudes and intra-hospital evolution.
![]()
Table 9. Distribution of children according to the length of hospitalization.
Hospitalization lasted for one week in more than half of the children (51.1%). The average length of hospitalization was 10.06 days with extremes of 0 and 35 days (Table 9).
4. Discussion
4.1. Epidemiological Aspects
4.1.1. Frequency
During our study period, we recorded a frequency of 0.84%. Our frequency is much lower than that of Randrianambinina et al. [8] who reported a frequency of 62.5% cases of secondary spontaneous pneumothorax at the CHU-JRA in Antananarivo, Madagascar in 2012. It is higher than that of Guyon et al. [1] who counted 11 cases over 10 years in the pediatric emergency room of Montpelier. It is also higher than that of Michel et al. [14] who reported 6 cases over 10 years in children aged 28 days to 12 months, which corresponds to 3.4 cases/10,000 admissions of children of the same age.
4.1.2. Age
In our series, children under 5 years of age were the most affected (80%). The mean age was 2.47 with extremes of 0 and 15 years. Our mean age is lower than that found by Randrianambinina et al. [8] (7 years). Michel et al. [14] in their meta-analysis reported a mean age of 15 years in a series of 58 cases. This could be explained by the frequency of respiratory infections in this age group due to immune immaturity.
4.1.3. Sex
In our study, twenty-eight of the 47 children were boys or 59.6% with a sex ratio of 1.47 in favor of boys. This male predominance has been reported by most authors, notably Randrianambinina et al. [8] (51.04%), Goldman et al. [5] (3.4 and 18 cases per 100,000 for men versus 1.2 and 6 cases per 100,000 for women). However, Guyon et al. [1] found a female predominance in their series with a G/F sex ratio of 0.375. The predominance in our study could be explained by the predominance of males in the 0 - 15 age group in the general population of Bamako.
4.2. Clinical Aspects
Respiratory discomfort was the most frequent symptom and reason for hospitalization in our children (95.7%) followed by cough (78.7%) and fever (61.7%). Except for the fever, our data are close to those classically described in the literature by Le Clainche [15] and reported by most authors in their series Randrianambinina [8] and Guyon et al. [1]. This could be explained by the fact that infectious causes were predominant in our study (70.2%) followed by traumatic (25.5%) and iatrogenic causes (4.2%). Our data are close to those of most African authors. Randrianambinina et al. [8] showed in their study conducted at the CHU-JRA in Antananarivo in Madagascar that infectious bronchopneumopathies were the main causes of secondary pneumothorax (62.5%). Diallo et al. [6] also found in their series carried out on young adults at the Ignace University Hospital in Conakry, a high frequency of pneumothorax secondary to pulmonary tuberculosis (72%) and bacterial infections (22%).
4.3. Radiological Aspects
The standard chest X-ray was the diagnostic tool for pneumothorax in all our children as reported by most authors (Le Clainche, Labbé, Guyon et al.) [1] [2] [15]. The majority of our children (68.1%) had right hemithorax involvement, contrary to what was reported by Randrianambinina et al. [8] who observed in their series left hemithorax involvement in more than half of the children (68.75%). The pneumothorax was compressive in 20 of the 47 children, or 42.6%.
4.4. Biological Aspects
In our sample, 34 of the 47 children had a blood count or 72.3%. The blood count revealed microcytic anemia in the vast majority of our children (7913%) of which 17.6% were severe. White blood cell abnormalities were found in 26 of the 34 children (76.4%). These were hyperleukocytosis in 67.6% and leukopenia in 8.8% of children. C-reactive protein was measured in 21 of the 47 children (44.6%). It was positive in 76.2% of the children. The predominance of microcytic anemia, hyperleukocytosis, and elevated C-reactive protein in our series can be explained by the high frequency of infectious causes, particularly infectious bronchopneumopathy.
In our study, surgical drainage was widely practiced in our children (87.2%), and a probabilistic antibiotic therapy based on amoxicillin/clavulanic acid in almost all our children (97.9%). This therapeutic protocol allowed us to have an average hospital stay of 10.06 days with extremes of 0 and 35 days.
Our rate of use of surgical drainage is close to that of Randrianambinina et al. [8] who performed it in 75% of the children in their study. However, Diallo et al. [6] largely used exsufflation in children (60%) in their study. The choice of surgical drainage in our children could be explained by the predominance of compressive forms in our sample as recommended by some authors [16] [17].
We recorded a cure rate of 74.5% and a death rate of 25.5%. Our data are close to those of Diallo et al. [6] who recorded a cure rate of 80% and a death rate of 20% in their study. Our high mortality rate could be explained by the fact that 83% of the children in our series are less than 5 years old, 59.6% of whom are less than 1 year old. It could also be explained by the frequency of severe forms: compressive form (42.6%) and bilateral form (2.1%).
Currently, the technique of choice is surgical video-thoracoscopy which is less painful and as effective as the traditional thoracotomy [14]. It allows exploration of the entire pleural cavity and the pulmonary surface through an easy and non-destructive approach. The bullectomy can be performed with automatic stapling-section forceps, then the pleura is mechanically ablated under visual control. Two drains are put in place at the end of the operation and maintained for 3 to 5 days. The hospital stay is less than a week. This technique allows a definitive cure in 98% - 99% of cases [14].
5. Limitations
During the course of our study, we identified a few limitations, including:
· The size of our sample.
· The retrospective nature of part of our sample.
· The non-feasibility of bacteriological examinations.
· The low socioeconomic level of our children’s families.
In spite of the limitations of this study, the results obtained allowed us to highlight the problems of pneumothorax in children in our department.
6. Conclusion
Pneumothorax is an uncommon disease in the pediatric department of the Mali Hospital. It is a medical and surgical emergency. Children under 5 years of age are the most affected. Its diagnosis must be evoked in the presence of dyspnea and confirmed by a frontal chest X-ray. The main etiologies are infectious bronchopneumopathies. Surgical drainage associated with broad-spectrum antibiotic therapy provides a good cure rate. Its mortality rate is still very high despite the progress in its management in our hospital. The reduction of this mortality rate requires the prevention and treatment of respiratory infections in children, the implementation of good bacteriological diagnostic tools, and the use of other surgical techniques such as surgical video-thoracoscopy.