Burden of Respiratory Disease in Pediatric Intensive Care Unit: Experience from a PICU of a Tertiary Care Center in Pakistan.

Introduction: We aimed to determine the burden of respiratory disease by examining clinical profiles and associated predictors of morbidity and mortality of patients admitted to a Pediatric Intensive Care Unit (PICU) in Pakistan, a resource limited country. We also stratified the respiratory diseases as defined by the Pediatric Advanced Life Support (PALS) Classification.

Methods: A retrospective study was conducted on children aged 1 month to 18 years who were diagnosed with respiratory illness at the PICU in a tertiary hospital in Karachi, Pakistan. Demographics, essential clinical details including immunization status, and the outcome in terms of mortality or survival were recorded. Predictors of mortality and morbidity including prolonged intubation and mechanical ventilation in the PICU were analyzed using the chi-square test or Fischer's exact test as appropriate.

Results: 279 (63.8% male; median age 9 months, IQR 4-36 months) patients were evaluated of which 44.2% were malnourished and 23.3% were incompletely immunized. The median length of stay in the PICU was 3 days (IQR 2-5 days). Pneumonia was the principal diagnosis in 170 patients (62%) and accounted for most deaths. 76/279 (27.2%) were ventilated, and 67/279(24.0%) needed inotropic support. A high Pediatric Risk of Mortality (PRISM) III score, pneumothorax, and lower airway disease were significantly associated with ventilation support. The mortality rate of patients was 14.3%. Predictors of mortality were a high PRISM III score (OR 1.179; 95% CI 1.024-1.358, P=0.022) and a positive blood culture (OR 4.305; 95% CI 1.062-17.448, P=0.041).

Conclusion: Pneumonia is a significant contributor of respiratory diseases in the PICU in Pakistan and is the leading cause of morbidity and mortality. A high PRISM III score, pneumothorax, and lower airway disease were predictors for ventilation support. A high PRISM III score and a positive blood culture were predictors of patient mortality in our study.