Pneumonia has been one of the preeminent causes of morbidity and mortality in children under 5 years of age for several decades (1–3). More than 90% of these deaths occur in developing countries involving two-thirds during the infancy period (1, 4). Estimated pneumonia-related deaths dropped down from 1.75 million in 2000 to 0.80 million in 2018. Still, pneumonia is responsible for 16% of a predicted 5.2 million global deaths among children less than 5 years of age (5, 6).

Several studies revealed that death from pneumonia is attributable to hypoxemia (7, 8), potentially due to respiratory failure. Most of the children having hypoxemia develop respiratory failure mainly due to ventilation-perfusion mismatch, derived from uncorrected hypoxemia (9). Indeed, the role of respiratory failure in childhood pneumonia is less focused in developing countries as all the emphases went to hypoxemia for its dreadful impact on childhood pneumonia-related deaths (10). Thus, understanding the prevalence, predicting factors, and outcomes of respiratory failure among children treated in the hospital for pneumonia or severe pneumonia is also critically important. In a study conducted in Taiwan among H1N1 influenza-related hospitalized pneumonia patients in 2009, 22/96 (23%) patients were found to have respiratory failure, where 10 (45%) patients died and all of the non-respiratory failure patients survived (11). Studies that were conducted among infants and children in critical care units of North American hospitals revealed that a good number of these children developed respiratory failure requiring mechanical ventilation. Total intensive care unit (ICU) admission for 6 consecutive months were 6,403, where 1,096 (17.1%) developed respiratory failure requiring mechanical ventilation within 24 h of development of respiratory failure. Among them, 97 had infantile pneumonia experiencing 1.6% mortality (12). A systematic review and meta-analysis reported that the most common causes of respiratory failure demanding mechanical ventilation in children were pneumonia and severe bronchiolitis (13).

Like other hospitals in developing countries, children in the Dhaka Hospital of International Centre for Diarrheal Disease Research, Bangladesh (icddr,b) often presented with pneumonia or severe pneumonia with other comorbidities of diarrhea, dehydration, sepsis, convulsion, and malnutrition (14). However, we did not find any published data on the predictors and outcomes of respiratory failure in children admitted with pneumonia or severe pneumonia, especially in developing countries such as Bangladesh. Thus, we attempted to understand the prevalence of respiratory failure among under 5 hospitalized children with pneumonia or severe pneumonia. Our further aim was to analyze the predicting factors and outcomes in such children.

During this study period, a total of 4,625 children under 5 years of age with pneumonia or severe pneumonia were admitted to both the ARI and ICU. Among them, we found 36.54% (n = 1,690) had pneumonia, 63.46% (n = 2,935) had severe pneumonia, and 4.58% (n = 212) had respiratory failure. However, the prevalence of respiratory failure was 7.22% among children having severe pneumonia. A total of 4412 study children who did not have respiratory failure and, thus, constituted the comparison group (Figure 1).

Compared to study children without respiratory failure, the children with respiratory failure were more likely to have lack of immunization, having severe underweight, severe sepsis, convulsion, dehydration, anemia, and higher white blood cell count at admission (Table 1).

During hospitalization, 156 (73.6%) children died with respiratory failure, in comparison 58 (1.3%) died among those who did not have respiratory failure and the difference was statistically significant (Table 2).

In the logistic regression analysis after adjusting for potential confounders, severe sepsis, convulsion, anemia, and severe underweight were found to be independently associated with respiratory failure (Table 3).

The trend of incidence for respiratory failure gradually reduced over the years from 2014 to 2017 and the reduction was statistically significant in 2017 (p < 0.001) (Figure 2).

To the best of our knowledge, this was the first study that evaluated the prevalence, predicting factors, and outcomes of respiratory failure in children under 5 years of age hospitalized with the WHO classified (both the clinical and radiological) pneumonia or severe pneumonia in developing country setup. This study identified several important observations: (i) the prevalence of respiratory failure was about 5%; (ii) severe sepsis, convulsion, anemia, and severe underweight were found to be the independent risk factors for developing respiratory failure; (iii) as anticipated, mortality rate was significantly higher among those who had respiratory failure compared to those without respiratory failure; and (iv) the incidence of respiratory failure gradually reduced over the years, but the reduction was significant on 2017 than the previous years.

In this study, the prevalence of respiratory failure among the overall study population was found to be 4.58%; however, it was higher (7.22%) in patients with severe pneumonia. Although we did not have precise data on the prevalence of respiratory failure in such children, limited data revealed that the rate may range from 6 to 23% (11, 12).

The major observation of this study was the independent association of severe sepsis with respiratory failure to this study children is quite explicable. Severe sepsis is often associated with severe metabolic acidosis requiring respiratory compensation with rapid and deep breathing (26) that might exhaust oxygen stores in the lungs, as well as in the cellular level of others vital organs of the body and often culminating to global hypoxemia and potentially end up with ventilation-perfusion mismatch resulting in respiratory failure (9). Kristina E Rudd et al. had found in his analysis that there were an estimated 11 million total sepsis-related deaths, representing 19.7% of deaths worldwide in 2017. These sepsis-related deaths were attributable to lower respiratory infections, leading to respiratory failure, which was consistent with this study findings (27).

The association of convulsion with respiratory failure might be related to hypoxemia that potentially cause cellular hyperexcitability in the brain leading to convulsion (28). Higher observation of severe sepsis among the study children having respiratory failure might bolster convulsion in some children (15) in this study population who developed respiratory failure. Similar findings were observed in a study, which was conducted in a Japanese clinics and hospitals between 2012 and 2016 among the 16 million Japanese patients, who were influenza positive. Among them, 1% participants required hospitalization, where 3,361 participants developed respiratory failure and 2,603 participants had seizure disorder; the incidence was highest for influenza-positive patients aged 0–5 years and there were statistically significant decreasing trends over the years in the incidence of all causes of hospitalizations (29).

The independent association of anemia with respiratory failure is noble and interesting. Critically ill patients with hypoxemia might fail to produce interaction between oxygenation, hemoglobin level, and cardiac output that may lead to respiratory failure (30). Moreover, the oxygen-carrying capacity of red blood cell in our critically ill children with anemia might be hampered due to reduced level of hemoglobin and simultaneously, tissue hypoxemia might persist, even though pulse oximeter showed normal level (9). Thus, the patients potentially became exhausted and developed respiratory failure. One study suggesting that anemia is associated with adverse outcomes in patients receiving mechanical ventilation, including poorer survival, increased duration of mechanical ventilation, and increased reintubation rates. However, critically ill patients in respiratory failure, there is lack of the necessary replacement capacity to preserve tissue oxygenation; thus, there is development of progressive anemia among ICU patients (31).

We also observed the independent association of severe underweight with respiratory failure in this study children. A number of previous studies revealed that severely underweight children with severe illness, such as severe pneumonia, were more vulnerable to developing interstitial edema in lungs (32–34), which might lead to development of respiratory failure.

One of our important observations was that the incidence of respiratory failure gradually reduced over the years; however, the reduction was significant in 2017 compared to previous years. It is important to note that we implemented our revised guideline for the management of pneumonia or severe pneumonia, severe sepsis, and hospital-acquired infection or pneumonia in 2016 in the Dhaka Hospital of icddr,b and these interventions might had an additional impact for the significant reduction of respiratory failure in 2017 compared to the previous years.

As expected, we also observed significantly higher deaths among children with severe pneumonia having respiratory failure compared to those who did not develop respiratory failure and the observation was found to be consistent with earlier study findings (14).

Retrospective analysis of the data was the main limitation of this study, which might allow our data to have some non-standardized and missing information. Regarding the diagnostic points of respiratory failure, arterial blood gas analysis is the most accurate test to diagnose. The clinicians of resource-poor settings have performed the diagnosis of respiratory failure based on the SpO₂/FiO₂ ratio, where FiO₂ was calculated adding 4% for each liter of oxygen to the FiO₂ in room air (21%). However, we routinely audited electronic database system in the hospital from where the large data set was composed.

Large data set comprising the cohort for children having the WHO defined clinical, as well as radiological pneumonia was the strength of this study.

The results of this study suggested that the prevalence of respiratory failure was 4.6% in under-five children hospitalized for pneumonia or severe pneumonia and 7.2% in severe pneumonia; however, its incidence reduced over the years. Severe sepsis, convulsion, anemia, and severe underweight have been revealed as the independent predictors of respiratory failure in children hospitalized for pneumonia or severe pneumonia. The case-fatality rate was significantly higher among children having respiratory failure than their counterpart. For better diagnosis of respiratory failure, a study with a large sample size may be undertaken that may help for the reduction of respiratory failure-related childhood deaths. Early identification of these predicting factors of respiratory failure may help our clinicians for prompt management in order to reduce potential deaths in such children, especially in resource-constrained settings.

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author/s.

Ethical review and approval was not required for the study on human participants in accordance with the local legislation and institutional requirements. Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.

SS and MC: conceptualization and formal analysis, and supervision. TAh and MC: investigation. TAh, AB, and MC: methodology. SS, MI, AB, and MC: data collection, analysis, and interpretation. SS, TAl, AB, LS, MS, SN, FA, YJ, GM, HS, MI, KS, SI, and TAh: writing—original draft. SS, TAh, and MC: writing—review and editing. All authors contributed to the article and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.