Taking observations

The previous version of this guideline⁸ identified that healthcare professionals should take observations for acutely ill adult patients. Taking full observations may not be appropriate for all patients, such as those receiving palliative care at the end of life. Recognition of deterioration should not be based solely on taking observations; it can also be identified from clinician or carer concern.

• Physiological observations should be recorded at the time of admission or initial assessment.
• A clear written monitoring plan should specify which physiological observations should be taken and how often.
• Observations should be performed by staff trained to undertake these procedures and who understand their clinical significance, including when to seek urgent clinical assistance.
• In certain settings, regular assessment of staff taking observations should be undertaken, to defined competency standards.
• As a minimum, observations should include:
  • pulse rate
  • respiratory rate
  • systolic blood pressure
  • level of consciousness or new confusion
  • oxygen saturation including percentage/flow rate of administered oxygen therapy
  • temperature.

In specific situations additional monitoring may be required to recognise deterioration, for example biochemical analysis (such as blood glucose or lactate), state of hydration, urine output or pain assessment.

Four high-quality systematic reviews were inconclusive in their comparison of electronic and 2++ manual observation taking. ^{9, 10, 11, 12}

Transcribing and charting observations

Recommendation 8: Observations should be transcribed electronically, charted electronically and displayed electronically and be underpinned by effective information technology (IT) systems, protocols and support to ensure ease of use. Appropriate paper-based systems should be readily available as a safeguard in the event of IT failure.

There was limited high-quality evidence for different types of observation recording. None of the studies identified focused on transcribing or charting observations.

Recommendation 9: Within an acute care setting, consider the use of automated prompts based on NEWS2 or other criteria alongside traditional methods of escalating care (such as direct telephone calls or paging systems). Implementation of such systems relies on adequate staffing resource to manage the generated automated alerts.

Some electronic observation systems collate and display observations electronically but still rely on a bedside healthcare worker manually contacting a co-worker (eg junior doctor) if the observations are concerning (such as by pager or telephone). Other electronic observation systems automatically contact a healthcare worker when a set level of deterioration is met (eg automatically messaging a junior doctor when the NEWS2 score is 7). This is termed ‘automated escalation’. The group considered that automated escalation may be more reliable than a manual system, but may also add to the overall burden of task management of front-line staff. Three studies were identified, all of which suggested an improvement in clinical outcomes, including reduced numbers of cardiac arrests, with automated escalation.^{13, 14, 15} However, they are limited in their quality as before-and-after studies, and do not provide sufficient evidence on which to base a recommendation.

Consensus could not be reached on the automated escalation of care in primary and community care settings.