Intensive Care National Audit context page

The Intensive Care National Audit and Research Centre (ICNARC) run a series of audits. The Case Mix Programme (CMP) audits patient outcomes from adult, general critical care units (intensive care and combined intensive care/high dependency units). It collects data on high-risk sepsis admissions, unit-acquired infections, discharges and transfers, re-admissions and mortality.

Audit website: www.icnarc.org/Our-Audit/Audits/Cmp/About

Metrics

Case ascertainment

Is not reported for this audit.

1. Crude non-clinical transfer proportion

The proportion of admissions which resulted in transfer of the patient to a comparable adult level 3 ICU bed in an adult ICU/HDU in another acute hospital.

Rationale: such transfers are assumed to reflect issues such as patient flow and bed capacity rather than clinical issues (i.e. a need for different or more specialised care). Non-clinical transfer adds to patient risk, prolongs ICU stay and may cause distress to patients and carers.
Guideline/mapping: this metric maps exactly to the Faculty of Intensive Care Medicine/Intensive Care Society Joint Standards Committee Guidelines for the Provision of Intensive Care Services 2015, edition 1.1, operational standard 2.4, page 194
Graph/relative performance: crude funnel-plot cross-section; this shows the position of the unit relative to 95% and 99.8% control limits. The grey zone means that the unit’s performance is not statistically different to the national average. Outside the grey zone means that the performance is different to the national average.
Quality improvement prompt question: ‘orange’ (caution) or ‘red’ (alarm) outlier status should prompt questions about why the non-clinical transfer rate is high and what actions the unit has taken to reduce it.

2. Crude, non-delayed, out-of-hours discharge to ward proportion

The proportion of patient discharges from the unit to a ward in the same hospital between 2200 and 0659, amongst those patients who had been ready for discharge by 1800 that day.

Rationale: overnight discharges from critical care are associated with increased mortality. Patients also perceive it as unpleasant to be moved from ICU to a general ward outside of normal working hours.
Guidelines/mapping: this metric maps exactly to the Faculty of Intensive Care Medicine/Intensive Care Society Joint Standards Committee Guidelines for the Provision of Intensive Care Services 2015, edition 1.1, operational standard 2.12, page 196
Graph/relative performance: crude funnel-plot cross-section as for metric 1.
Quality improvement prompt question: ‘orange’ (caution) or ‘red’ (alarm) outlier status should prompt questions about why the unit is discharging so many patients to the wards out-of-hours and the steps it is taking to address this.

3. Crude delayed discharge proportion

The proportion of available critical care bed days utilised by patients who have been ready for discharge from the unit for more than 8 hours.

Rationale: delay in discharging suitably fit patients from critical care represents inefficient resource use and may reflect underlying capacity issues – both of which are pertinent to the delivery of high quality care.
Guidelines/mapping: this metric maps approximately to the Faculty of Intensive Care Medicine/Intensive Care Society Joint Standards Committee Guidelines for the Provision of Intensive Care Services 2015, edition 1.1, operational standard 2.11, page 196. Mapping is approximate because the audit considers 8 hours to be a more achievable target than the recommended 4 hours.
Graph/relative performance: categorical (binary) classification; poorly performing units are defined as the 5% with the highest rates of delayed discharge.
Quality improvement prompt question: placement within the bottom 5% should prompt questions about why the unit is experiencing difficulty in discharging its patients from critical care promptly and the steps it is taking to address this.

4. Risk-adjusted in-hospital mortality rate, all patients

Rationale: death is the most serious objective outcome following ICU admission. This metric uses case-mix adjustment to ensure that units with sicker patients are compared fairly against those with a less complex case mix.
Guidelines/mapping: there is no national guideline related to mortality after ICU admission.
Graph/relative performance: risk-adjusted funnel-plot cross-section; this shows the position of the Trust relative to 95% and 99.8% control limits. The grey zone means that the hospital’s performance is not statistically significantly different to the national average. Outside the grey zone means that the performance is different to the national average.
Quality improvement prompt question: ‘orange’ (caution) or ‘red’ (alarm) outlier status should prompt questions about actions which the Trust has taken to reduce their mortality rates following ICU admission.

5. Risk-adjusted in-hospital mortality rate, low-risk patients (i.e. for patients with predicted risk of death <20%)

Rationale: death in low-risk patients can be considered more anomalous events which warrant closer scrutiny. The metric should be interpreted in conjunction with metric 4.
Guidelines/mapping: there is no national guideline related to mortality after ICU admission.
Graph/relative performance: risk-adjusted funnel-plot cross-section as for metric 4.
Quality improvement prompt question: ‘orange’ (caution) or ‘red’ (alarm) outlier status should prompt questions about actions which the Trust has taken to reduce their mortality rates following ICU admission.