How well do out of hours (OOH) clinicians recognise the end of life? A population based study of OOH service contacts in the last 30 days of life.

Talk Code: 
Gail Hayward
Rachel Brettell, Rebecca Fisher, Helen Hunt, Daniel Lasserson
Author institutions: 
University of Oxford; Oxford Health NHS Foundation Trust


Out of hours (OOH) palliative care is challenging, and a priority for further research. Previously we established that OOH primary care services contribute significantly to end of life care, but this analysis was restricted to patients coded as palliative by the service. It is uncertain how many patients at the end of life seek help from the OOH service but are not recognised to be at the end of life, as, currently, no mechanism exists to alert the service to deaths. This study aimed to describe the proportion and nature of patients who died within 30 days of contact with a UK OOH service, specifically to compare those who were not labelled with palliative care codes with those who were.


Mortality data for Oxfordshire for the year 2015 (population 600,000) was linked with Oxfordshire OOH service care records using section 251 approval. Patients dying within 30 days of OOH contact whose care record included ‘palliative’, ‘terminal illness’ or ‘terminal care’ codes (the ‘expected death’ group) were compared to patients whose record did not include these codes (the ‘unexpected death’ group). Using SPSS v22, we compared demographic data, clinical and service use data and causes of death.


Of 5193 patients who died in Oxfordshire over 13 months, 1531 (29%) had contact with the OOH service in the 30 days prior to death. Of these, 994 (64.9%) did not have palliative codes recorded; this ‘unexpected’ group had a higher mean age, (83.5 (sd10.88) vs 80.8 (sd 11.68)) and greater proportion were female (56.7% v 53.8%) compared to the ‘expected’ group. Patients not coded as palliative were over 10 times more likely to be admitted directly to hospital or sent to the emergency department. Causes of death were significantly different between these groups: cancer was the most frequent cause in the ‘expected’ group (42% vs 12%), five times as many deaths were due to infection in the ‘unexpected’ group and twice as many due to cardiac causes.


This is the first study to establish the nature of the population contacting the UK OOH service with end of life needs. Two thirds of patients dying within 30 days of OOH contact were ‘unexpected’ in terms of the clinical coding by treating clinicians; this group differed in cause of death and outcomes of consultations compared with those whose death was expected. The lack of a palliative code for this group could be due failure to recognise patients at the end of life, a reluctance to use the code for patients who do not have a malignancy, or truly unexpected deaths. A mixed methods approach is now needed to explore how clinicians might be supported to both identify and code OOH end of life needs.

Submitted by: 
Gail Hayward
Funding acknowledgement: 
This work was funded by the Oxfordshire Health Services Research Committee