• haemopoietic stem cell transplant;
  • intensive care unit;
  • prognosis;
  • organ failure;
  • critical illness


Background:  Previous research at our institution (1988–1998) established an intensive care unit (ICU) and hospital mortality between 70% and 80% in haemopoietic stem cell transplant (HSCT) patients requiring ICU admission.

Aims:  This study explored mortality in a more contemporary cohort while comparing outcomes to published literature and our previous experience.

Methods:  Retrospective chart review of HSCT patients admitted to ICU between December 1998 and June 2008.

Results:  Of 146 admissions, 53% were male, with a mean age of 44 years, an Acute Physiologic and Chronic Health Evaluation II score of 28 and Sepsis Organ Failure Assessment score of 11. Fifty-six per cent had graft versus host disease (GVHD), with respiratory failure (67%) being the most common admission diagnosis. All but one received mechanical ventilation. The ICU and hospital mortality were 42% (72% 1988–1998 cohort) and 64% (82% 1998–1998 cohort) respectively. The 6- and 12-month survivals were 29% and 24% respectively for the 1998–2008 cohort. Dying in ICU was independently predicted by fungal infection (P= 0.02) and early onset of organ failure (P < 0.001), while GVHD (P= 0.04) predicted survival. Mortality at 12 months was independently predicted by the acute physiology score (P= 0.002), increasing number of organ failures (P= 0.001), and cytomegalovirus positive serology (P= 0.005), while blood stream infection (P= 0.003), an antibiotic change on admission to the ICU (P= 0.007) and a diagnosis of non-Hodgkin lymphoma (P= 0.02) predicted survival.

Conclusion:  Our study found that acute admission of HSCT patients to the ICU is associated with improved survival compared to our previous experience, with organ failure progression a strong predictor of ICU outcome, and specific disease characteristics contributing to long-term survival.