Acute surgical admissions at the end of life–an analysis of non-operative hospital deaths

Background: Patients with advanced illnesses are often admitted with acute surgical emergencies. There is currently no evidence characterising such admissions. We aimed to evaluate emergency patients, managed non-operatively, who died during the same admission.

Methods: This single-centre retrospective, observational study collected data points for a 12 month period including age, prior documented do not resuscitate order (DNAR), existing cancer, Charlson Comorbidity Index, frailty, surgical diagnosis, interval from admission to death and care given. Patients who underwent surgical intervention were excluded. Non-parametric tests were used for statistical analysis.

Results: A total of 72 patients were included in this study, of which 68.1% died within 6 days of admission (median 4.0 days). Patients with visceral perforation, obstruction, bowel ischaemia or known malignancy were more likely to die within 6 days than those with pancreatitis, sepsis or new malignancy (median 2 vs 7 days, p < 0.001). Patients with frailty (2 vs 4 days, p = 0.017) and existing DNAR (3 vs 4 days, p = 0.048) died more rapidly than those without. Age and comorbidity index did not impact time to death.

Conclusion: Frailty, surgical diagnosis and existing DNAR were predictors of shorter admission to death interval, while age and comorbidity index were not. This has implications on inpatient palliative care service planning.