Long held assumptions of poor prognoses for patients with haematological disease have meant that intensivists are often reluctant to admit them to the intensive care unit (ICU) During the recent Covid pandemic this was heightened further given the increased pressure on the UK's ICU beds. This led us to evaluate our utilisation of our ICU over the last 6 years in order to rationalise our own requests and justify continued ICU support during this pandemic. If ICU access was temporarily withdrawn as it was to a variable extent peak pandemic, how would it have affected outcome of acute leukaemias or aggressive malignancies.


Our Institution which is a regional cancer centre admits haematology patients from a population of 750,000. We analysed 65 admissions over last 6 years. The clinical notes were reviewed and we obtained the following data: haematological and other diagnoses, chemotherapy, transplant status, apache score, lactate level, age, laboratory values, level of organ support, ICU and hospital discharge status, overall survival or time to follow up if alive, plus ability to proceed with treatment post ICU.

Predictors for overall survival were assessed using Xlstat and Stata software.


Of 65 ICU admissions from 54 patients, median age was 62 (range 19-86), Apache score median of 8 patients had more than 1 ICU admission (2-4). Reason for admission was predominantly infection/sepsis post chemotherapy with requirement for circulatory or respiratory support 54 episodes out of 65. Other reason for admission included 3 episodes of plasma exchange in unstable patients, airway support for 2 patients with intracranial haemorrhage, support during major haemorrhage and 2 episodes with pulmonary embolism, 2 episodes sedation for seizure control, ventilatory support for differentiation syndrome in APML.

Organ support

Six (9%) of admissions to ICU didn't recieve any specific ICU level treatments including 1 case that was unsalvageable on arrival, 3 admissions just attended to facilitate urgent plasma exchange. Patients receiving single organ, double organ and triple organ support were 28 (43%), 22 (34%), 6 (9%) respectively. Admissions receiving non-invasive ventilation 29 (47%), invasive ventilation 24 (37%), vasopressors 33 (51%) and filtration 8 (12%).


  • Acute Myeloid leukaemia 25 (38%) episodes including 3 APML patients,

  • Acute Lymphoblastic leukaemia 3 episodes

  • Lymphoma 19 episodes (29%)

  • Myeloma 8 episodes (12%)

  • Others 5 (8%)

  • Post allogenic stem cell transplant 1 patient, 2 episodes


The overall survival of all the admissions a mean survival of 445 days is illustrated. 28 (43%) admissions survived to hospital discharge and 16 (25%) survived for a year or more. We compared the outcome of patients admitted to ICU 2013-2015 versus admissions post 2016 in Figure 1 which illustrates a 9% improvement in 1 year survival.


The improvement in survival over the last 6 years is clearly illustrated and consistent with other reports (Outcomes and prognostic factors in patients with haematological malignancy admitted to a specialist intensive care unit, British journal of Anaesthesia 108 (3): 452-9 (2012)). The recent longer survival reflects advances in treating haematological malignancies as well as the targeting sepsis campaign.

Out of 119 patients with acute leukaemia we treated with induction chemotherapy, 20 required admission to ICU, 8 of these acute leukaemia patients survived for over a year. Admission to ICU is therefore responsible for 7% of our acute leukaemia patients long term survival.

Following ICU admission 17 patients (31%) of them were able to receive further chemotherapy including 6 stem cell procedures.

Age, 2 or more organ support and Apache 2 score were significant by univariate Chi-squared test in being risk factors for mortality pre-hospital discharge. Multivariate analysis showed 2 or more organ support had an odds ratio of 6 for death within a year as did intubation independently with odds ratio of 3.8 both with p values under 0.05. Neutropenia or thrombocytopenia didn't significantly impact on mortality. No scoring system or prognostic factors reliably separates out those for whom admission to ICU is futile.

Our findings challenge the reluctance to admit haematology patients during and post the covid pandemic in the uk.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.

Sign in via your Institution