Backgrounds:

The sepsis-3 definitions were developed from databases of adult patients and were neither designed nor validated in children. We sought to validate the performance of age-adapted PMODS, qSOFA, PELOD-2 and SIRS in predicting outcomes for children consecutively admitted to a Paediatric Intensive Care Unit (PICU).

Design:

Prospective, observational study.

Setting:

A single-centre regional PICU in the United Kingdom.

Patients:

656 consecutively admitted children under the age of 16 years were enrolled.

Methods

All children were categorised based on the Sepsis-3 definitions: sepsis, septic shock and no infection. Biochemical and physical parameters were measured within the first 24 hours of PICU admission. The primary outcomes were a composite outcome of 28-day mortality and PICU LOS>3 days, and development of multi-organ dysfunction syndrome (MODS). We derived scores for age-adapted PMODS, qSOFA, PELOD-2 and SIRS in predicting the primary outcomes. The performance of the scores were evaluated using area under the curve (AUC).

Results:

Median age was 1.02 years (IQR 0.29 – 5.02). 367 were post-operative cardiac surgical patients (56%), 105 other surgical (16%), and 184 non-surgical (28%). 351 infectious episodes were described.

In all patients, SIRS was positive in 39.8% of episodes (n=261). 10.5% of children developed DIC (n=69). 123 developed MODS, and D28 mortality was 1.8%. PMODS gave the best discrimination for predicting mortality or LOS, and MODS, in the sepsis and septic shock subgroups. Mortality or LOS>3; sepsis: 0.50 (0.38 – 0.62) and septic shock: 0.67 (0.39 – 0.95) MODS; sepsis: 0.62 (0.49 – 0.75) and septic shock: 0.75 (0.45 – 1.00).

Conclusions/Learning Points:

When using Sepsis-3 criteria, PMODS provides excellent prediction of both in-hospital mortality or PICU length of stay over 3 days, and development of MODS in children with septic shock.

Backgrounds:

The sepsis-3 definitions were developed from databases of adult patients and were neither designed nor validated in children. We sought to validate the performance of age-adapted PMODS, qSOFA, PELOD-2 and SIRS in predicting outcomes for children consecutively admitted to a Paediatric Intensive Care Unit (PICU).

Design:

Prospective, observational study.

Setting:

A single-centre regional PICU in the United Kingdom.

Patients:

656 consecutively admitted children under the age of 16 years were enrolled.

Methods

All children were categorised based on the Sepsis-3 definitions: sepsis, septic shock and no infection. Biochemical and physical parameters were measured within the first 24 hours of PICU admission. The primary outcomes were a composite outcome of 28-day mortality and PICU LOS>3 days, and development of multi-organ dysfunction syndrome (MODS). We derived scores for age-adapted PMODS, qSOFA, PELOD-2 and SIRS in predicting the primary outcomes. The performance of the scores were evaluated using area under the curve (AUC).

Results:

Median age was 1.02 years (IQR 0.29 – 5.02). 367 were post-operative cardiac surgical patients (56%), 105 other surgical (16%), and 184 non-surgical (28%). 351 infectious episodes were described.

In all patients, SIRS was positive in 39.8% of episodes (n=261). 10.5% of children developed DIC (n=69). 123 developed MODS, and D28 mortality was 1.8%. PMODS gave the best discrimination for predicting mortality or LOS, and MODS, in the sepsis and septic shock subgroups. Mortality or LOS>3; sepsis: 0.50 (0.38 – 0.62) and septic shock: 0.67 (0.39 – 0.95) MODS; sepsis: 0.62 (0.49 – 0.75) and septic shock: 0.75 (0.45 – 1.00).

Conclusions/Learning Points:

When using Sepsis-3 criteria, PMODS provides excellent prediction of both in-hospital mortality or PICU length of stay over 3 days, and development of MODS in children with septic shock.