Use of pulse oximetry in the prediction of ICU mortality

  • Research type

    Research Study

  • Full title

    Use of admission pulse oximetry saturations in the Paediatric Index of Mortality Score: does it improve prediction of mortality?

  • IRAS ID

    191836

  • Contact name

    Padmanabhan Ramnarayan

  • Contact email

    p.ramnarayan@gosh.nhs.uk

  • Sponsor organisation

    Great Ormond Street Hospital for Children NHS Foundation Trust

  • Duration of Study in the UK

    0 years, 3 months, 1 days

  • Research summary

    Approximately 18000 children are admitted to UK paediatric intensive care units (PICUs) following critical illness or high risk surgery each year. Based on how sick a child is at the time of PICU admission we can calculate the probability of a child surviving the PICU stay, using a score known as the Paediatric Index of Mortality (PIM). Although the PIM score is not accurate enough to be useful in individual patients, it is commonly used in research/audit to compare how units are performing and as a measure for severity of illness. One of the numbers used in the score calculation is the amount of oxygen in the patient’s blood (PaO2), usually measured from blood taken from the artery. However, only a quarter of children who are admitted to ICU with critical illness have an arterial blood sample taken at admission, because placing an arterial line can be risky (infection, injury). In children with no PaO2 available at admission, PIM assumes a value of 0 (normal). However, this falsely decreases the PIM, and reduces its accuracy.

    On the other hand, all children admitted to ICU have blood oxygen saturations (SpO2) measured using a skin probe. While SpO2 does not give us exactly the same information as the PaO2, it may be a good substitute. Our research question is: in critically ill children without a PaO2, can the SpO2 improve the prediction ability of the PIM score, compared to using a normal value of 0?
    We intend to use existing clinical data to calculate a relationship between PaO2 and SpO2 from children admitted to PICU over a 30 month period. We will then use this to calculate a modified PIM score in children admitted over a 12 month period to test how good the modified scoring system is compared to the existing system.

  • REC name

    Yorkshire & The Humber - Bradford Leeds Research Ethics Committee

  • REC reference

    16/YH/0180

  • Date of REC Opinion

    28 Apr 2016

  • REC opinion

    Favourable Opinion