Abstract

Background: Patients with haematological malignancies (HM) receive intensive chemotherapy regimens including HSCT rendering them immunocompromised and susceptible to complications such as, cardio-pulmonary toxicity, hepatic and renal failure, GVHD, sepsis and haemorrhage. Patients (pts) with HM may require critical care (CC) admission. Many studies have looked at the ICU setting for these pts but few have evaluated the use of non invasive cardio-pulmonary support on HDU prior to intensive care escalation or invasive intervention. The Haemato-Oncology Unit at Sheffield Teaching Hospitals NHS Foundation Trust is a tertiary referral centre, serving a population of 1.5 million. Critical care facilities are divided into HDU and ICU. Levels of care provided on HDU include non-invasive respiratory (non-invasive ventilation or continuous positive airway pressure) and cardiovascular (vasopressor/inotropic) support, but not mechanical ventilation (MV) or renal replacement therapy (RRT), both of which are provided on ICU. We evaluated all admissions to HDU and ICU for pts with HM in order to better understand the prognostic markers for this cohort of pts.

Subjects and Methods: We retrospectively analysed consecutive HM (including HSCT) admissions to HDU or ICU over a 66 month period, between September 2002 and March 2008. We analysed predictors of mortality including age, white cell count (WBC), Acute Physiology, Age, and Chronic Health Evaluation II (APACHE II) score, PaO2:FiO2 ratio as well as requirement for respiratory or cardiovascular support or RRT. Unit survival, hospital survival and overall survival at the time of study (OS) were evaluated.

Results: One hundred and sixty two pts (n=107 HDU, n=55 ICU) were admitted to CC during the study period comprising n=191 admissions (n=128 HDU, n=22 HDU transferred to ICU, n=63 ICU). Of the HDU admissions n=38 pts received HSCT, and of the ICU admissions n=28 pts received HSCT. The median age for the entire cohort was 59 yrs (range 16–83 yrs). The median age for the HSCT and non HSCT pts was 50 yrs (range 17–68 yrs) and 63 yrs (range 16–83 yrs) respectively. Of the n=162 pts, n=99 (61%) were male, n=63 (39%) female. Types of HM varied, AML n=48 (25%), Myeloma n=40 (21%), NHL n=36 (19%), MDS n=16 (8%), ALL n=16 (6%), CLL n=12 (6%), CML n=9 (5%), others n=13 (7%). Overall median WBC was 3.6×109/l (range 0.0–746×109/l), median WBC for HSCT and non HSCT admissions was 3.1×109/l (range 0.1–31.6×109/l) and 4.1×109/l(range 0.0–746×109/l) respectively. The overall mean APACHE II score for all HDU admissions was 16.2 (SD ± 5.9) and 23.7(SD±7.2) for all ICU admissions. The overall median PaO2:FiO2 ratio for all HDU admissions was 15 kPa and 12 kPa for all ICU admissions. In total n=61 (32%) admissions required respiratory support (n=35(18%) MV, n=26 (14%) non invasive respiratory support), n=91 (48%) received cardiovascular support while n=25 (13%) required RRT. For all HDU admissions, unit survival, hospital survival and OS was 81%, 55% and 27% respectively compared to 52%, 44% and 13% for all ICU admissions. For HDU admissions escalated to ICU, unit survival, hospital survival and OS was 32%, 32% and 9% respectively. For non HSCT pts admitted to HDU, unit survival, hospital survival and OS was 79%, 54% and 26% respectively, compared to 84%, 55% and 29% for HSCT pts. For non HSCT pts admitted to ICU, unit survival, hospital survival and OS was 53%, 44% and 14% respectively, compared to 54%, 43% and 14% for HSCT pts. An APACHE II score of >30 was associated with a unit survival of 44%. No pts with an APACHE II score >36 survived. Unit survival for admissions requiring no organ support was 87%, one organ support 79%, two organ support 22% and three organ support 58%. Unit survival for pts requiring MV was 23%.

Conclusions: Our study demonstrates that MV, requirement for two organ support, admission APACHE II score >30 are all poor prognostic factors for pts with HM admitted to CC. Interestingly of the n=12 pts requiring MV, cardiovascular support and RRT n=7 (58%) survived to ICU discharge, demonstrating that selected pts requiring multi-organ support may have a better prognosis. Requirement for single organ support was associated with better outcome. Our study demonstrates that the majority of HM pts requiring CC can be appropriately managed in the HDU setting. We aim to develop a predictive scoring system for HM pts requiring CC using the evaluated prognostic factors.

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author