Use of a structured palliative care summary in patients with established cancer is associated with reduced hospital admissions by out-of-hours general practitioners in Grampian
- 1Centre for Academic Primary Care, University of Aberdeen, Aberdeen, UK
- 2Department of Grampian Medical Emergency, NHS Grampian, Aberdeen, UK
- Correspondence to Dr Rosalind Adam, Centre for Academic Primary Care, University of Aberdeen, Polwarth Building, Foresterhill, Aberdeen AB25 2ZD, UK;
Objectives Palliative care summaries are used by general practices to provide structured anticipatory care information to those providing care during the out-of-hours period. We hypothesised that the availability of a palliative care summary for individuals with established cancer would influence emergency hospital admission during the out-of-hours period.
Methods Each consultation with Grampian Medical Emergency Department (GMED) is recorded on the ADASTRA software system and the nature of the consultation is Read coded. We retrospectively reviewed consultations between 1 January 2011 and 31 December 2011 which had been coded as ‘neoplasm’ or ‘terminal care’. The availability of a palliative care summary on ADASTRA and admission status were recorded. χ2 Test of association was performed. Binary logistic regression was used for multivariate analysis exploring the effect of a palliative care summary on admission, while adjusting for important confounders.
Results 401 patients with established cancer were identified who had presented to GMED in 2011. 35.7% had a palliative care summary available on ADASTRA. Of the 401 contacts, 100 patients were admitted to hospital. Not having a palliative care summary made admission significantly more likely; χ2=12.480, p=0.001. (OR 2.425, 95% CI 1.412 to 4.165).
Conclusions Availability of a structured palliative care plan can aid decision making in the out-of-hours period and prevent unplanned hospital admissions.
Out of hours general practice care has changed dramatically since the introduction of the new General Medical Services contract in 2004. Responsibility for out-of-hours general practice care no longer rests with the patient's registered medical practice. This responsibility has transferred to the National Health Service (NHS) boards.1 NHS Grampian provides out-of-hours primary care cover via the Grampian Medical Emergency Department (GMED), which employs sessional doctors, salaried doctors and nurse practitioners. GMED assesses around 9000 patients per month and is designed to deal with urgent health needs when general practices are closed.2 Concerns have been raised about lack of continuity of care in the out-of-hours period.3 ,4 Continuity of care and anticipatory care are particularly important in patients with palliative care needs.5 ,6
It is recognised that a substantial number of palliative patients who die in hospital would prefer to die at home.7 The Scottish Government's publication ‘Living and dying well: a national action plan for palliative and end-of-life care in Scotland’ provided guidance and action points for the consistent delivery of high quality palliative care in Scotland.8 It highlighted the importance of pro-active planning for symptom management and end-of-life care. Recommendations were made on improving communication and information sharing between professionals. One aspect of this is the utilisation of existing general practice IT systems.
An emergency care summary (ECS) was introduced in Scotland in 2004. It is an extract from the patient's electronic primary care medical record which contains information on patient demographics, allergies and prescribed medication. The information is updated automatically twice daily from the general practice software system to a central store. It is viewable as ‘read only’ and is available to out-of-hours primary care organisations, NHS 24, accident and emergency departments, and Scottish Ambulance Service. Patient consent for the ECS is implicit, and patients can opt out. The patient must give verbal consent for the information to be accessed for each out-of-hours care episode. The ECS is currently available for over 5.5 million Scottish patients.9
An electronic palliative care summary (ePCS) was introduced in Scotland in 2009 following pilots in Grampian and Lothian health boards.10 It has now been implemented across Scotland and by April 2011 the ePCS was being used in 32% of practices across 11 health boards in Scotland.11 It is available to practices using Vision, Egton Medical Information Systems, and General Practice Administration System for Scotland electronic patient record systems. The ePCS provides a template for data collection under five headings: patient and carer details; patient medical condition; current care arrangements; patient and carer's awareness of condition; and advice for out-of-hours care. The final section includes information on preferred place of care and preferred place of death. The ePCS is completed by the patient's general practitioner or district nurse. In contrast to the ECS, the ePCS contains detailed anticipatory care information, and explicit verbal consent must be given by the patient and indicated in a ‘tick box’ by the completing practitioner before the ePCS can be uploaded. We focused on patients with an established cancer diagnosis and hypothesised that the sharing of anticipatory care information in the form of a structured palliative care summary would facilitate care of the patient in the community and influence decisions on emergency admission to hospital during the out-of-hours period.
GMED contacts are recorded using ADASTRAV.3 software package. ADASTRA is independent from Egton Medical Information Systems, Vision, and General Practice Administration System for Scotland and the systems are not interoperable. Authorised out-of-hours practitioners require a username and password to access ADASTRA. The nature of each consultation is Read coded by the treating health professional. A list of all available Read codes in ADASTRA was reviewed and it was decided that ‘neoplasm’ and ‘terminal care’ were the most relevant codes to identify patients with established cancer. We interrogated the ADASTRA database, applying filters to identify retrospectively all consultations related to ‘neoplasm’ and ‘terminal care’ between 1 January 2011 and 31 December 2011. For the purposes of statistical analysis, we used Community Health Index numbers to exclude repeated consultations by the same patient during the study period. Patient identifiable information was removed from the filtered data. Individual consultations were reviewed in detail by two reviewers to ensure that the patient had an established cancer diagnosis. We collected data on baseline patient characteristics including: age, sex, cancer site and deprivation using Scottish Index of Multiple Deprivation 2009 ranks from postcodes.12 We did not have access to individualised patient prognostic information, such as stage of cancer. We used nationally agreed 5-year survival estimates for cancer site13 as a surrogate measure of prognosis, splitting the patients into good (5-year survival greater than 50%) and poor (5-year survival less than 50%) prognostic groups.
Presence or absence of a palliative care summary was noted. During our study period, some practices were having difficulty submitting the ePCS. A proportion of practices faxed printed palliative care summaries to GMED and these were entered manually onto ADASTRA by a member of administrative staff. For faxed summaries, the primary care practitioner providing the information was identifiable on the record and remained responsible for the content of the clinical information provided. Faxed palliative care summaries were included in our analysis because they contained similar information to the ePCS and both constituted structured anticipatory palliative care plans which were available on ADASTRA for use by the out-of-hours practitioner. The patient outcome was classified as ‘admitted’, where the patient was referred acutely to a hospital or hospice, or ‘not admitted’, in cases where the patient remained at home.
Data were entered into SPSS V.19 statistical package. The χ2 test was used to conduct univariate analysis to test the null hypothesis that there is no significant difference between admission status and absence of a palliative care summary. Univariate analysis was also used to explore the effect of potential confounders (gender, age, rurality, deprivation, reason for call, cancer site and prognosis) on admission. Binary logistic regression was used for multivariate analysis exploring the effect of palliative care summary on admission, while adjusting for important confounders.
Researchers worked with anonymised data. Individual patients were not identifiable and, as such, patient consent was not sought for inclusion in this retrospective audit. The North of Scotland Research Ethics Committee was consulted prior to the undertaking of this work. The work was clinical audit and did not require formal review.
Our database search identified 401 patients with established cancer who contacted GMED between 1 January 2011 and 31 December 2011. A palliative care summary was available for 143 (36%) patients and unavailable for 258 patients. Following assessment by GMED, 100 patients were admitted to hospital. Of those admitted to hospital, 21 had palliative care summary completed. χ2 Test of association for admission status and palliative care summary presence χ2=12.480, p=0.001. Binary logistic regression was adjusted for potential confounders. These are presented in table 1 alongside baseline patient characteristics. The absence of a palliative care summary significantly increased the likelihood of admission p<0.001 OR 2.425 (1.412–4.165).
Our data suggest that in patients with cancer, presence of a palliative care summary reduces the likelihood of hospital admission during the out-of-hours period. To the best of our knowledge, this retrospective study is the first to look at the association between the presence of a palliative care summary and hospital admission. It is intuitively understandable that provision of a structured anticipatory care plan should aid difficult decision making during the out-of-hours period. A reduction in hospital admissions will have financial and resource implications in addition to the potential benefits to the individual patient.
One limitation of this study is that data were unavailable for individualised prognostic information and stage of cancer or presence of metastatic disease. It could be argued that the stage of disease could have a bearing on completion of a summary and might also influence decisions relating to hospital admission.
A surprising finding was that 64% of patients in our study with established cancer did not have a palliative care summary completed, despite Grampian being an area which piloted the ePCS scheme. A palliative care summary relies upon information being collected and submitted by the patient's registered practice and there is variation in completion rates. Timing of summary completion can also vary. The fact that some practices in the area were having problems with the electronic submission process may have affected completion rates during the study period, although procedures were in place so that palliative care summaries could be faxed. Completion rates might be increased by inclusion of the ePCS in National palliative care guidelines. At a local level, GMED staff are providing informal feedback to practices on individual patients who have made contact with the service and who might benefit from ePCS completion. We would advocate that primary care practitioners should have a low threshold for completing an ePCS in patients with a cancer diagnosis, especially if the patient is likely to experience symptoms relating to their cancer or its treatment during the out-of-hours period. While our study focused on patients with cancer, the ePCS can be used for any patient who might have palliative care needs. Pro-active planning for out-of-hours contacts is one way of promoting continuity of care and high quality palliative care and should be encouraged.
The authors would like to acknowledge the secretarial staff at the Grampian Medical Emergency Department who aided with the data collection process.
Contributors This project was designed by PM and AAA. AAA collected data and performed statistical analysis. RA drafted the submission and reviewed the data critically. DT revised the article critically. All authors commented substantially on subsequent drafts.
Competing interests None.
Ethics approval The study was discussed with the regional ethics committee and is a clinical audit, not requiring formal review.
Provenance and peer review Not commissioned; externally peer reviewed.
- Received 20 September 2012.
- Revision received 14 November 2012.
- Accepted 20 November 2012.
- Published Online First 3 January 2013
- Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go to http://group.bmj.com/group/rights-licensing/permissions