Mortality of in-patient medical admissions to a DGH Critical Care Unit (2019)

Type of publication:
Conference abstract

*Sunil Tailor, *Ade Adiamou, *Omu Davies, *Roger Slater

State of the Art 2019. Intensive Care Society Conference. Birmingham December 9-12.

Increasing clinical demand and reorganisation of hospital services may result in “stable” medical in-patients being “cohorted” into hospital locations without ready access to critical care services on site. The NEWS tool is recognised to provide warning of deterioration and a trigger for escalation of care [1]. Current EWS tools perform well for predicting death and cardiac arrest within 48 hours, although the impact on in-hospital health outcomes and utilization of resources remains uncertain [2]. In a retrospective case review, we examined the outcome of medical patients who had been hospital in-patients for more than 48 hours, whose condition deteriorated, requiring admission to our critical care unit (Princess Royal Hospital). The aim was to identify “red flag” observations for their deterioration and to measure mortality (overall: hospital + 30-day) in this patient group. During 2016-17 we identified 51 patients, of whom 39 patients were analysed because of a complete data set. We classified 48% (19 patients )as hot: defined as having a NEWS of 5 or more on one or more occasions in the 48-hours prior to ITU admission.
52% (21 patients) were classified as warm; defined as having a NEWS no greater than 4 on one or more occasions in the 48hr prior to ITU admission.
Red flag parameters (individual score of 3) were: Respiratory rate <8 or >25; SpO2 < 91%; temperature < 35C; SBP < 90 or > 220 mmHg; HR < 40 or > 131; AVPU: VPU.

  • The most common predictive red flag indicator for ITU admission was a raised respiratory rate.
  • 11 of the 39 patients did not score any red flag indicators. The mortality of this group was 33%.
  • The mortality rate of the 39 patients was 66%. The mortality for hot and warm classifications was 71% and 46% respectively.
  • The majority (>80%) of patients were aged 51-80 years .
  • The mean duration of ITU stay was 7.4 days (warm) vs 9.5 days (hot).
  • Over the same time period, ICNARC data demonstrated that 80% of all admissions to the ITU were non-surgical; overall ITU and hospital mortality was 15.7% and 22.5% respectively.


  • Medical inpatients with a persistently high NEWS of 5 or more during 48 hrs prior to ITU admission had a very high 30-day mortality despite ITU care.
  • 11 patients had no red flag indicators in the 48 hours prior to ITU admission despite their subsequent deterioration.
  • There are significant resource implications in managing such patients.
  • Care must be taken in defining stability in respect of medical in-patients in order to avoid later deterioration.

Limitations: Not all patients could be analysed because of incomplete data.


  1. Thompson R. National Early Warning Score (NEWS). Report of a Working Party, Royal College of Physicians,UK, July 2012.
  2. Beth Smith M, Chiovaro J, O'Neil M, et al. Early Warning System Scores for Clinical Deterioration in Hospitalized Patients: A Systemic Review. Ann Am Thorac Soc 2014; 11:1454-1465.

Link to full-text [no password required]