Emergency department attendance by patients with cancer in the last month of life: a systematic review and meta-analysis

Background: Emergency department visits towards the end of life by people with cancer are increasing over time. This increase has occurred despite evidence of an association with poor patient outcomes, the majority of patients preferring home-based care, and significant overcrowding and capacity concerns for many emergency departments. We aimed to explore factors associated with emergency department attendance by cancer patients in the last month of life.

Methods: We searched Medline, Embase, CINAHL, PsychINFO, and the Cochrane Library from inception to February, 2014, for studies investigating emergency department attendances by adult cancer patients (≥18 years) towards the end of life. No time or language limitations were applied. We performed meta-analysis of factors using a random-effects model, with results expressed as odds ratios (OR) for emergency department attendance. Sensitivity analysis explored heterogeneity.

Findings: 30 studies were identified, reporting three demographic, five clinical, and 13 environmental factors; they included data from five countries and 1 181 842 patients. An increased likelihood of emergency department attendance was found for men versus women (OR 1·24, 95% CI 1·19-1·29), black versus white race (1·45, 1·40-1·50), patients with lung cancer versus other cancers (1·17, 1·10-1·23), and those of lowest versus highest socioeconomic status (1·15, 1·10-1·19). Patients receiving palliative care were less likely than those not receiving palliative care to attend the emergency department in the last month of life (OR 0·43, 95% CI 0·36-0·51).

Interpretation: We have identified demographic (men, black race), clinical (lung cancer), and environmental (low socioeconomic status, no palliative care) factors associated with an increased risk of emergency department attendance. These findings could be used to develop screening interventions and assist policy makers in directing limited resources. Future studies should also investigate previously neglected areas of research, including psychosocial factors, and the emergency care preferences of patients and caregivers.

Funding: LH is a PhD clinical training fellow and funded through project BuildCARE which is supported by Cicely Saunders International and The Atlantic Philanthropies, and led by King's College London, Cicely Saunders Institute of Palliative Care, Policy & Rehabilitation, UK.