Multi-parameter monitoring for paediatric triage
Researchers: Dr Syed Ahmar Shah, Prof. Lionel Tarassenko
Clinical collaborators: Rob Way (Emergency Department), Dr Matthew Thompson, Dr Susannah Fleming and Prof. David Mant (Dept of Primary Health Care)
Funding: National Institute of Health Research (NIHR)
Children are profligate users of both primary and emergency health-care resources, including over 25% of GP consultations. Appropriate and timely triage of these patients is necessary to identify those who require urgent or life-saving interventions. Advances in technology allow measurements of several physiological variables to be made using non-invasive sensors, allowing clinicians to concentrate on obtaining a history of the problem and starting any emergency treatment that may be required. The pulse oximeter is of particular use in primary and emergency care, as it is relatively cheap and easy to use.
We have developed an automated method to estimate the respiratory rate from the photoplethysmogram (PPG – pulse oximeter waveform) using both the amplitude modulation caused by changes in thoracic pressure and the phenomenon of respiratory sinus arrhythmia, the heart rate variability associated with respiration. Signal quality indices have been used to identify high-quality PPG segments free of motion artefact.
The vital sign values derived from pulse oximetry (respiratory rate, heart rate, blood oxygen saturation) and the temperature recorded with an axillary or tympanic probe have been using a probabilistic framework to provide a novelty score for ranking diagnostic categories and assessing the likely severity of infection.
Although heart rate and respiratory rate are routinely measured in children in acute settings, current reference ranges are not evidence-based. Work done using non-parametric kernel regression (published in The Lancet in February 2011) has enabled us to produce evidence-based charts for heart rate and respiratory rate with respect to age.