Respiratory rate(RR) is the only vital sign that is currently measured by direct observation in many high-income countries in many healthcare settings. However, when relying on direct observation by humans, values may be susceptible to measurement error. In particular, a continuous variable like RR may be inadvertently semicategorised due to rounding and/or multiplying up counts from shorter periods to give estimated counts for a full minute.

This could have an important impact on early warning scores, as these scores rely on the aggregation of accurately measured physiological parameters to trigger specific warning thresholds. Systemic errors in the measurement of RR component could therefore negatively impact patient care if it meant a deteriorating patient did not meet the defined threshold. Seventy-six per cent of UK hospital trusts receiving acute medical admissions use the National Early Warning Score—2 (NEWS-2).1 2 Introduced in 2017, it uses six simple physiological parameters which are respiration rate, oxygen saturation, systolic BP, pulse rate, level of consciousness and temperature. These are aggregated by a simple scoring system that uses predefined thresholds to give a composite …