Systematic review of factors influencing decisions to limit treatment in the emergency department

Background Emergency physicians are frequently faced with making decisions regarding how aggressive to be in caring for critically ill patients. We aimed to identify factors that influence decisions to limit treatment in the Emergency Department (ED) through a systematic search of the available literature.

Design Prospectively registered systematic review of studies employing any methodology to investigate factors influencing decisions to limit treatment in the ED. Medline and EMBASE were searched from their inception until January 2019. Methodological quality was assessed using the Mixed Methods Appraisal Tool, but no studies were excluded based on quality. Findings were summarised by narrative analysis.

Results 10 studies published between 1998 and 2016 were identified for inclusion in this review, including seven cross-sectional studies investigating factors associated with treatment-limiting decisions, two surveys of physicians making treatment-limiting decisions and one qualitative study of physicians making treatment-limiting decisions. There was significant heterogeneity in patient groups, outcome measures, methodology and quality. Only three studies received a methodology-specific rating of ‘high quality’. Important limitations of the literature include the use of small single-centre retrospective cohorts often lacking a comparison group, and survey studies with low response rates employing closed-response questionnaires. Factors influencing treatment-limiting decisions were categorised into ‘patient and disease factors’ (age, chronic disease, functional limitation, patient and family wishes, comorbidity, quality of life, acute presenting disorder type, severity and reversibility), ‘hospital factors’ (colleague opinion, resource availability) and ‘non-patient healthcare factors’ (moral, ethical, social and cost factors).

Conclusions Several factors influence decisions to limit treatment in the ED. Many factors are objective and quantifiable, but some are subjective and open to individual interpretation. This review highlights the complexity of the subject and the need for more robust research in this field.