545

Abstract

Introduction: There is increasing concern that cancer patients receive inappropriate active treatment in the last few weeks of life. One in 7 patients admitted to Intensive Care Units (ICU) has a cancer diagnosis but the evidence on their expected outcomes after admission has not been synthesised. Methods: Systematic literature review and meta-analysis on outcomes of solid cancer adult patients admitted to ICU from 2000 onwards using PRISMA reporting guidelines and STROBE assessment criteria. EMBASE and MEDLINE electronic databases searched with independent review by two researchers. Results: Of 33 studies identified, 17 reported outcomes in mixed solid cancers from Europe, South America and North America. Mean ages range 47-70 years; 10/17 general ICUs. ICU mortality range 14.3-57.2%, pooled estimate 33.5% (95% CI 22.9% to 44.1%, n=4066). Hospital mortality range 19.0-76.8%, pooled estimate 53.1% (95% CI 39.4% to 66.9%, n=1499). 11 studies described survival in lung cancer patients after ICU admission. ICU mortality range 20-49%, pooled 38.8% (95% CI 31.7% to 45.9%, n=725); hospital mortality range 40-61%, pooled 52.8% (95% CI 46.8% to 58.9%, n=637). Heterogeneity between studies was high and none reported sex-specific or cancer stage-specific outcomes. Acute physiology scores were correlated with poorer survival. Small numbers of additional studies identified described outcomes after ICU admission in patients with breast, colorectal, head and neck, gynaecological, oesophageal and pancreatic cancers. Conclusions: Published research on outcomes of cancer patients after ICU admission is heterogeneous and lacks sufficient clinical information to guide decision-making. As the number of patients potentially requiring ICU support will continue to rise, primary research is needed to describe outcomes in cancer patients with sufficient casemix and treatment details to be of prognostic value to clinicians.