Clinical predictors of hospital admission in acute lower respiratory tract infection in 2 months to 2-year-old children.

ABSTRACT

BACKGROUND AND OBJECTIVE: Acute lower respiratory tract infections (ALRI) are a common cause of paediatric emergency visits in young children. We studied risk factors for hospitalization and developed a clinical score for predicting hospitalization among 2 months to 2-year-old children with ALRI. METHODS: We conducted this prospective cohort study in the paediatric emergency department of a tertiary-care teaching hospital in India. Consecutive children, aged 2 months to 2 years with ALRI were enrolled from 15 December 2011 to 14 December 2012. A total of 26 a priori identified, putative risk factors were studied among enrolled children. We determined independent predictors of hospital admission (primary outcome) through multi-variable logistic regression analysis and assimilated them into a clinical risk score using regression coefficients. RESULTS: A total of 240 children (130 admissions) with ALRI were enrolled. Eleven clinical risk factors, which displayed association with hospital admission on univariate analysis (P < 0.1), were entered into multi-variable logistic regression analysis. Five factors retained independent association and were incorporated in a predictive score for hospitalization tachypnoea (score of 5), chest retractions (score of 3), temperature > 37.8°C (score of 3), SpO2 < 92% at room air (score of 4), GCS < 15 (score of 6). Area under the receiver operator characteristic curve was 0.80 (95% CI 0.75-0.85, P < 0.001). CONCLUSION: Five clinical risk factors-tachypnoea, chest retractions, fever > 37.8°C, SpO2 < 92% and GCS < 15-independently predicted hospital admission in infants with ALRI. A novel clinical score predicting hospital admission is presented.