Clinical Features That Predict Treatment Failure And Death In Children Presenting At Kenyatta National Hospital With Severe And Very Severe Pneumonia On Admission

Abstract

Background: Pneumonia kills more children than any other illness in the world and is a significant problem in communities with a high rate of under - five mortality placing a huue burden on families and health systems.1 As of 2005. more than 150 million childhood pneumonia cases were estimated to occur every year in the developing areas of the world and acute respiratory infection is responsible for an estimated 1.9 million childhood deaths each year.2 Methods: Children between the ages 2 months to 59 months coming to the Pediatric Emergency Unit (PEL) at the KNH were screened for signs of pneumonia (cough and difficulty in breathing) and lower chest wall in drawing. The principal investigator, who worked with the other colleagues in the pneumonia study, was stationed at the PEL. Explanation of the purpose of the study and the procedures involved were given to the guardians/parents of the children and a written consent was sought from them. Of the children who presented with signs and symptoms of pneumonia, those with wheeze who responded to bronchodilators were excluded together with those ineligible for treatment. Emergency care such as oxygen and administration of fluids was instituted without delay arising from study procedure. Sociodemographic and clinical information was collected using a pretested questionnaire. At baseline, pulse oximetry and blood cultures were taken. The patients were followed up in the wards at 24 hours. 48 hours and 72hours whereupon the clinical signs and symptoms were checked for improvement or deterioration. Clinical failure was defined as persistence or worsening of signs and symptoms, clinician decision to change antibiotics or death at or before 48 hours. Results: A total of four hundred and eighty seven children aged 2 to 59 months who presented to the pediatric emergency unit with cough and difficulty in breathing were assessed. A total of 385 children were admitted to the study. Of these. 171 (44.4%) had severe pneumonia while those with very severe pneumonia were 214 (55.6%). The proportion of children with treatment failure was 28.1%. Treatment failure rate was higher in those with very severe pneumonia at 39.7% compared to those with severe pneumonia 13.5%. One hundred and seventy one children with severe pneumonia 3(1.8%) died while 27(12.6%) with very severe pneumonia died. The clinical correlates of treatment failure in children with severe pneumonia included history of previous treatment which was associated with a 5 fold higher odds of failing treatment with a P=0.056. Grunting and level of consciousness less than A were associated with a 2 4 and 4.8 fold l i increase in the odds of failing treatment respectively. The ability to drink and presence of wheeze were associated with a better outcome in treatment failure OR 0.5 (0.2-0.5). OR 0.4(0.2-0.8) respectively. Cough, wheeze and ability to drink were associated with better outcomes as regards to death OR 0.1 with a P value of 0.029: 0.4 with a P value of 0.048 and OR 0.1 with P value of 0.01 respectively. Conclusions The rate of treatment failure was 28.1% with 39.7% in the very severe pneumonia and 13.5% in the severe pneumonia group. Mortality was low within the severe pneumonia group therefore the correlates could not be assessed. Association of treatment failure in the severe pneumonia group included a borderline association with history of previous treatment. In general, wheeze and ability to drink were associated with better outcomes in treatment failure and death while grunting and reducing level of consciousness were associated with increased odds of failing treatment or dying in the very severe pneumonia group. Level of consciousness and ability to drink showed co-linearitv. that is one varies with the other, they are not independent predictors of treatment failure and death. Recommendations Children with very severe pneumonia who present with grunting or reduced level of consciousness need to be prioritized because of the increased risk of treatment failure and death. Children with grunting, decreased level of consciousness and inability to drink should be prioritized within the wards for close monitoring and frequent clinical reviews and may require more supportive care in a higher dependency unit.