To our knowledge, our research regarding the factors influencing the mortality among patients hospitalized in a pediatric intensive care unit in Haïti is one of the first to be published. The average age of patients hospitalized in the intensive care unit of the HBM in our study was 5.73 ± 4.73 years (0.08 to 16 years) and the median age was 5 years. Male patients had an average age of 5.42 ± 4.61 years and female patients 6.24 ± 4.81 years. Which is like the results of the study carried out in Brazil in 2009, in which: the median age of patients was 4.9 ± 2.7 years – 4.6 years in male patients and 5.3 years in female patients (2). Our results were also similar to the study carried out in Europe where the average age was 5 years. Our results were different to the study carried out in Africa by Coetzee, S. where the median age was 7 months. This study used a methodology different from ours (observational study) and analyzed the epidemiological profile of the population admitted to the PICU for a specific factor, namely measles (4). Our results also differ from the retrospective study carried out in Europe by Lanetzki, C. S. et al; where the median age of patients was 2.7 years (5). Our results could be explained by the high mortality rate among children less than 5 years old who are more at risk of dying from preventable and treatable causes mostly due to socioeconomic disparities (3).
Our study revealed a strong predominance of male patients hospitalized in the PICU: 76 patients (63.30%,), with a M/F sex ratio of 1.65. Result similar to the various studies mentioned in our literature review, namely the work of: Araujo et al. in Brazil; Lanetzki, C. et al, in Europe; Kwizera, A. et al, in Africa. (2, 5, 6). Indeed, studies have shown that the mortality rate among boys is higher than girls in most parts of the world. For this age group, this has been explained by sex differences in genetic and biological makeup, with boys being biologically weaker and more susceptible to diseases and premature death (7) .
Regarding the residence area, our study population mainly resided in urban areas with 98 patients (80.33%) and 24 (19.67%) in rural areas. However, in the observational study conducted by O'Callaghan D.J. et al; the infant mortality is slightly higher in rural areas than in urban areas (59% versus 56%) (8).
1/3 of the patients in the study carried out by El Halal M.G. et al., were referred from other hospitals (34.7%). Only 0.7% of admissions came directly from their home (9). The study carried out by Javad Ghaffari et al in 2011 revealed that 49% of patients came directly from their homes and 51% from other hospital structures. This latest study matches ourhas similar results to ours. Indeed, among our study populations 55 patients (45.08%) came directly from their home and 67 patients (54.92%) of admissions were referrals from other health structures (9). This high referral rate could be explained by the fact that HBM is one of the rare hospitals in the metropolitan area, accessible to the population, and equipped with an emergency and intensive care unit.
In our series, the average time to seek care was 1 day. These results are similar to those of the study carried out in Africa by Karin Källander et al., according to which the average time to seek care outside the home was 2 days (4). Results that could be explained by the tendency of our population to resort to traditional medicine before any contact with any medical structure.
Regarding the duration of stay in the hospital, in our study it varied from less than 24 hours to more than 30 days with a maximum duration of 121 days, and an average duration of 12.59 ± 16.15 days. These findings differ from studies carried out in Africa by Karin Källander et al, with an average duration of hospitalization of 3 days (10) and which could be explained by the fact that it is a series of cases carried out on a population of children aged less than 5 years and relates exclusively to admissions due to pneumonia.
Our research also demonstrated that 54.10% of patients admitted to the PICU during this period were intubated and received mechanical ventilation at the time of their admission. These results are almost similar to the study of Navin P. Boeddha et al, who demonstrated that 69% of their patients benefited from invasive ventilation (11). The retrospective study carried out in Iran revealed that 24.60% of deceased patients were intubated during their hospitalization in PICU. (12). Results which differ from our study, where 92.68% of deceased patients were intubated. The differences could be explained by the fact that the study carried out in Iran, although having a similar methodology, had excluded cases of trauma and post-surgical admissions.
The main diagnostic of the patients in our study were classified into different categories: first trauma, found in 43 patients or 35.26% of admission, followed by respiratory diseases, found in 22 patients or 18.04% of admission and third post-surgical admissions, in 17.22% of cases. Our results are similar to the study carried out in 2009 in Brazil by Taisa E. Araujo et al., (2) which demonstrated that respiratory illnesses represented the most common causes of admission to intensive care units (32.7%), followed by post-operative care (30.9%) and trauma cases (12.8%). However, the results of other studies, although all having in common respiratory illnesses as the most common cause of hospitalization; differ from our results. In Uganda, the study carried out by Kwizera, A. et al, demonstrated that the most frequent cause of admission was post-surgical treatments. The percentage was not mentioned in this study (6).
The latest studies carried out around the world have demonstrated that the mortality rate in PICU is now < 3%. Studies carried out in Europe, North America, and Africa, by Agra Tuñas et al., Namachivayam et al., Markita L. and Suttle et al; demonstrated similar results with a respective mortality rate of 2.20%, and 2.38% (11, 13–15). Results completely different from the mortality rate found in our study work i.e., 33.60%. This high mortality rate in the pediatric intensive care unit of HBM could be explained by the generally late delay in seeking care of the haïtian population and by the high rate of trauma due to motor vehicle accidents and falls, admitted during this period; and knowing that HBM is known for the management of emergencies from all causes.
56.10% of deaths occurred in a context of brain death, 41.46% after failure of cardiovascular resuscitation and 2.44%, after a decision of limitation of therapeutic intervention. This differs from the data published by Agra Tuñas M.C. et al, in Spain where 50.7% of deaths occurred after a decision to limit therapeutic intervention; 114 (33.8%) after cardiovascular resuscitation and 52 (15.4%) were due to cases of brain death. And also those published by Markita L. Suttle et al., in the United States of America where 133 patients (70%) died after interruption of life-sustaining treatment, 30 (16%); after a diagnosis of brain death and 26 (14%) following an attempt at cardiopulmonary resuscitation (13, 14).
In our research, the deceased patients were mainly male i.e., 26 patients (63.41%), with a M/F sex ratio of 1.73. Result similar to the study of Rashma R.P. et al, where 52% of the deceased patients were male; and similar to the result of Valavi E. et al, where death was observed in 51.80% of M patients (12, 16). 14 patients (or 34.15%) died in our study, they were less than 1 year old. The results are similar to those of the study carried out by El Halal M.G. et al, where 37.5% of the deceased patients were aged less than 1 year (9). The average age of patients who died in our study was 6.33 ± 4.67 years. This differs from the results of Rashma R.P. et al, where the average age of the deceased children was 3.40 + 4.16 years; note that this study, although having a similar methodology to ours, was carried out on a younger population (1 month to 14 years old). It also differs from the retrospective study carried out in Iran by Valavi E. et al, where the average age of deceased patients was 2.2 years (12, 16).
Our research also demonstrated that the mortality rate decreased with age; 14 deaths i.e., 34.15% aged less than 1 year; 13 i.e., 31.70% aged 1 to 5 years; 9 i.e., 21.95% aged 6 to 12 years old and 5 i.e., 12.20% over 12 years old. (Table 2). This result differs from El. Halal M.G. et al, who found that mortality increased significantly with age. The mortality rate was 0.9%, 8.9%, 12.3%, 10.4% and 17%, respectively, for children ≤ 1 month, < 1 year, 1–5 years old, 5–12 years old and ≥ 12 years old (9).
Among the 41 deaths recorded in our study, 34.14% of patients had a comorbidity at the time of admission. 16 of them i.e., 39.03%, had at least one previous hospitalization. These results differ from those of Agra Tuñas M.C. et al, in Spain where a total of 86 patients (25.5%) had a previous hospitalization, 273 (75%) of them suffered from a chronic pathology and 78 (23%) had a serious disability at the time of admission (13). This difference could be explained by the fact this study was carried out over several years in several hospital centers.
The comorbidities found in the 14 deceased patients were mainly of a neurological aspect with 4 patients, i.e., 28.56% with hydrocephalus, 21.43% i.e., 3 patients with heart diseases, 2 i.e., 14.28%, had hematological disease (sickle cell disease), 1 patient i.e., 7.42% had chromosomal anomaly (Down syndrome). These were also the most frequent comorbidities in the study by Halal M.G. et al: neurological (11.5%), hematological/oncological (11.4%) and genetic (7.3%) (9). Which could probably be explained by the fact that HBM is one of the rare hospitals with a neurosurgery department that serve the adult and pediatric population (9).
The main diagnostics retained in the 41 deceased patients in our work, were dominated by: septic shock in 24.39% of cases knowing that preventable infections still play a major role in the pediatric mortality in Haiti. Meningeal conditions were found in 19.51% of cases, trauma in 17.07% of cases, respiratory illnesses in 14.63% of cases and post-surgical admissions in 14.63% of cases. These diagnostics are different from those published by Agra Tuñas M.C. et al, a retrospective study carried out in Spain for which the most frequent causes of hospitalization were in 32.6% of cases, cardiac and in 22.6% of cases, respiratory illnesses (13). They also differ from the results of El Halal M.G. et al, where the mortality was mainly due to cardiopulmonary arrest in 29% of cases, sepsis 19%, pneumonia 16%, multi-system dysfunction 14%, hepatic diseases in 7% of cases, an inborn error of metabolism 6%, and acute respiratory distress syndrome in 6% of cases (16).
Our research allowed us to identify among the different hypotheses mentioned, two factors associated with greater mortality rate within the PICU of HBM. Indeed, following our various analyses, male gender and intubation were retained as predictive factors of mortality (p = 0.0021, and p = 0.0049) respectively. Intubation was also incriminated in the studies of Navin P. Boeddha et al, and Valavi E. et al. This factor could be explained by the fact that patients generally placed under mechanical ventilation within an ICU are patients in more critical shape than patients who are not (11, 12). Our results are also similar to the studies of Rashma R.P. et al, and Valavi E. et al, (12, 16) where in fact also male sex was identified as a predictive factor of mortality. This could be explained by the fact that our study population is mainly boys with a significant statistical difference as shown in Table 5 (poled 0.042). Results which should arouse scientific curiosity and constitute an avenue for future scientific research.
The predictors of mortality in a PICU is a subject of great importance in Haiti, but unfortunately informations, studies and even researches are very limited. Indeed, very few studies on this subject have been carried out in Haiti (17).
Although there are only two hospitals in the metropolotain area with a pediatric intensive care unit, this study was carried out in only one center, namely HBM, which is known for the management of medical emergencies, intensive care, and surgical intervention.
We made some recommendations to the Ministry of Public Health and Population, on the importance of the implementation of strategies to promote equitable access to health services, promote the continuous training of medical staff, create new PICUs, and create more rehabilitation centers for patients with physical or neurological sequelae.