INTRODUCTION

Pneumonia is the leading cause of death of infectious origin in children worldwide and accounts for 22% of all deaths in children between 1 and 5 years of age ¹. In Peru, the incidence of pneumonia in children under five years of age is 10.2 cases per 10,000 inhabitants and has a case fatality rate of 0.9 ².

Scales are used to assess the severity of pneumonia in adults, and the use of these scales has been shown to reduce the frequency of hospitalization and administration of broad-spectrum antibiotics ³. However, there are no validated criteria for assessing the severity of community-acquired pneumonia in children. The World Health Organization (WHO) defines severe pneumonia as a case in which the child is unable to drink, has persistent vomiting, convulsions, lethargy, stridor or severe malnutrition; however, these criteria have high sensitivity but low specificity ⁴. A multicenter cohort study developed a predictive model for severe outcomes such as death, need for mechanical ventilation or use of vasoactive agents in children hospitalized with pneumonia; it showed that age extremes, altered vital signs, presence of retractions and radiographic infiltrates were the most important predictive factors ⁵. However, these results cannot be extrapolated to emergency patients and their clinical usefulness is still unknown.

Scales have been designed to predict mortality in children with pneumonia such as the modified PIRO scale (modified Predisposition, Insult, Response and Organ dysfunction), the RISC scale (Respiratory Index of Severity in Children) and the PRESS scale (Pediatric Respiratory Severity Score). The modified PIRO and RISC scales are good predictors of mortality due to pneumonia in children. The PIRO scale has a greater capacity to discriminate mortality and could be used in more complex health centers ⁶. However, these scales have not been evaluated to predict admission to the ICU (intensive care unit) in children with pneumonia.

Early recognition of the factors associated with adverse outcomes and higher prevalence of admission to intensive care in children with pneumonia in emergency areas is important because it allows timely actions to be taken that can reduce mortality in this age group. Therefore, the aim of this study is to determine the prevalence of admission to ICU and its associated factors in children and adolescents hospitalized with pneumonia at the Instituto Nacional de Salud del Niño San Borja (INSNSB) in 2019.

MATERIALS AND METHODS

RESULTS

DISCUSSION

In this study, we reviewed the medical records of 166 hospitalized patients diagnosed with community-acquired pneumonia at INSNSB. The prevalence of admission to the ICU was 14.5%. Younger age, having complicated pneumonia and higher severity assessed with the modified PIRO score were variables associated with admission to intensive care.

The prevalence of admission to ICU in this study (14.5%) was lower than that reported by Fritz et al., who studied children under 18 years of age hospitalized for pneumonia in three hospitals in the United States enrolled in the EPIC study (Etiology of Pneumonia in the Community Study), and found that 21% of those without bacteremia and 43% of those with bacteremia were admitted to intensive care ¹¹. However, the prevalence of admission to intensive care was lower (8%) in a prospective study of children under five years of age admitted for pneumonia in a hospital in Vietnam ¹². Results have been similar to ours in reports from other Latin American countries; a cross-sectional study in children under five years of age aimed at identifying predisposing factors for severe pneumonia in children hospitalized in a Colombian hospital found a prevalence of admission to ICU of 11.3% ¹³. These differences may be due to several factors, the level of complexity of the hospitals included in the study, the age of the patients included and the criteria for admission to the ICU in each one.

This study found that when children with pneumonia are younger, the prevalence of admission to intensive care in is higher. Similarly, other studies have found an association between younger patients age and higher prevalence of adverse outcomes such as mortality. Thus, Djelantik et al. described that patients younger than four months had higher mortality compared to older children (RR=3.5, 95%CI: 3.0-4.2) ¹⁴. These results are explained by anatomical and physiological factors that are more unfavorable at a younger age, such as the absence of interalveolar and alveolo-bronchiolar ventilation channels that favor the appearance of complications such as atelectasis, greater airway resistance, lower pulmonary compliance and greater compliance of the thoracic cage in children under two years of age, and the immaturity of the innate and acquired immune system.

Complicated pneumonia (parapneumonic effusion, empyema) was associated with a higher probability of admission to intensive care. Results similar to those of Golbart et al., who reported that the presence of empyema is associated with admission to the ICU ¹⁵. It has also been reported that the presence of moderate or large pleural effusion is associated with admission to the ICU in children (OR: 3.2, 95%CI: 1.1-8.9) ¹⁶. Space-occupying conditions in the pleura can lead to respiratory failure due to hypoventilation, which may explain the greater need for ventilatory support in the ICU in these patients.

The modified PIRO score discriminates the probability of dying in children hospitalized with pneumonia and may be a reliable tool for selecting patients requiring admission to intensive care ^17,18. This study found that patients with moderate and high modified PIRO scores were more likely to be admitted to the ICU than patients with mild PIRO scores. These results could be an indication of the usefulness of this score in predicting admission to the ICU; however, we did not find information in the medical records for all the items considered in this score, such as the blood pressure value.

This study found no association between multilobar involvement on chest radiography and ICU admission. Unlike Williams DJ et al., who reported that multilobar infiltrates were associated with severe adverse outcomes such as ICU admission, mechanical ventilation and death ⁵. These discordant results could be due to factors that may alter the interpretation of chest radiograph findings such as hydration status, presence of atelectasis, and time of presentation ¹⁹. For our study, we selected patients who were in the emergency area at hospital admission.

No association was found between INL levels and admission to the ICU. This index was assessed with the first hemogram on admission, a time in which the patient probably had not received antibiotics yet. This result is different from those reported by Lee et al., who found that elevated INL was associated with greater mortality and was a predictor of admission to the ICU in adults; this difference could be due to the immune development of children, in whom cellular immunity predominates over innate immunity, which could explain why they have a greater number of lymphocytes than neutrophils compared to adults ²⁰.

Patients with comorbidities were more likely to be admitted to the ICU only in the crude model. This is consistent with the results published by Koh et al. who found that the presence of comorbidities is an early prognostic variable for unfavorable outcomes in patients admitted to the intensive care unit ²¹.

This study has limitations. Since this was a retrospective study, some variables of interest have not been evaluated, such as vaccination status against pneumonia-causing pathogens, nutritional status or the time the patient spent in the emergency room before going to the ICU. In addition, the blood pressure value was not available for all patients and the alteration of this value is used to construct the modified PIRO score. There are no unified criteria for deciding admission to the ICU in children with pneumonia in the INSNSB. In addition, the sample size is small so there are wide confidence intervals and there could be real associations not found in the present study.

In conclusion, the prevalence of admission to intensive care in children with pneumonia was 14.5%. Younger age, the presence of complicated pneumonia and greater severity assessed with the modified PIRO score were associated with admission to the ICU in children with community-acquired pneumonia. It is recommended that modified PIRO be applied to predict admission to the ICU in prospective studies with larger sample sizes to improve the accuracy of the results.