BackgroundInequality in palliative care provision is an ongoing problem for underserved groups, such as those with heart failure (HF) placing a burden on patients, their caregivers and health services due to frequent, often avoidable hospitalisations. Our realist synthesis of integrated palliative care (PC) and HF literature found that integrated HF and PC services work best when service providers are motivated and have the opportunity and capacity to support behaviour change. However, we identified significant knowledge gaps with most studies completed in United States of America (USA) and based primarily on the views of nurses and physicians. We developed strategies to help services provide integrated PC and HF services but identified the need for United Kingdom (UK) primary data to better understand the context-specific implementation of palliative care and HF care.MethodsThis project will use co-design and realist evaluation to generate data from five PC and HF integrated services in the UK, purposively sampled to provide variation in geography and service design (Research Registry 10624). The research comprises three work packages (WPs). WP1 will deliver a realist evaluation of each site including documentary analysis, observations and semi-structured interviews with service providers and users. WP2 uses co-design methods to develop a guide to help others set up and improve integrated PC and HF services. Data from both WPs will be analysed and synthesised using a realist logic of analysis. WP3 will facilitate the development of a community of practice to support those who wish to set up, sustain and embed integrated PC and HF services.DiscussionThis realist evaluation of a complex intervention will improve understanding of how to tailor and implement integrated PC and HF services. The co-designed 'how-to guide' and community of practice will facilitate knowledge translation and ensure that evidence-based guidance is provided to assist in service development.

Background:Frailty affects around 10% of people aged over 65 years, increasing to 65% of those over 90 years. This number is increasing. Older people with frailty are projected to become the largest future users of care services as they near end of life. Living with frailty increases vulnerability to sudden deterioration, fluctuating capacity and mortality risk. This leads to complex needs, requiring integrated care, and an approach orientated towards living with, as well as dying from, advancing frailty. However, accessing care in a timely manner can be difficult. Aims:To develop a sustainable, cross-sectoral partnership to: identify priorities to improve integrated care delivery, and care transitions, for older people with advancing frailty develop organisations in which to conduct research submit study proposal(s) for funding. Objectives:To establish Partnership infrastructure and identify key contacts across palliative and end-of-life care. To understand the strengths, weaknesses, barriers and enablers of research readiness and clinical services for people with advancing frailty. To support provider services to become research ready. To establish Partnership-wide research questions and develop research proposals. Activities:The Partnership brought together experts, by profession or experience (n = 244), across specialist palliative and geriatric care and local government, to improve the delivery of integrated care for older people with advancing frailty as they near end of life. Members included older people with frailty, unpaid carers, health, social and voluntary care professionals and academics, across the East Midlands, South East England and South West London. A survey of key contacts (n = 76) mapped and scoped the Partnership's strengths, weaknesses, barriers and enablers of services for people with advancing frailty, and service providers' research readiness. Forty-six key contacts responded. Most worked in the East Midlands (59%), in health care (70%) and in the community (58%). Survey findings were used to develop a service framework and to create a short list of potential research questions. Questions were refined and prioritised through coproduction with frail older people (n = 21), unpaid carer representatives (n = 7), health, social and voluntary care professionals (n = 11) and care home representatives (n = 3). The question chosen for bid development focused on ensuring what matters most to older people with frailty informs service development. This bid is currently being written. Partnership members were also supported to develop research readiness and enhance meaningful patient and public involvement by the development and curation of multiple resources. Reflections:This work was challenging. The Partnership enabled the collaboration of diverse stakeholders and fostered opportunities to improve end-of-life care for older people with advancing frailty. However, the fluidity of the workforce, lack of finance to buy-out key contacts' time, limited service integration across sectors, lack of common language and concepts across sectors, need to build research understanding and readiness, and minimal evidence of engaging frail older people approaching end of life in determining service provision and research, made achieving the initial goals overly ambitious. Nevertheless, the Partnership developed a service framework for older people living and dying with advancing frailty, and is currently coproducing a clinically applied, translational research proposal. Funding:This article presents independent research funded by the National Institute for Health and Care Research (NIHR) Public Health Research programme as award number NIHR135262.

ObjectivesOur work aims to critically review the use of anticipatory medicines in our inner-city hospice community population and whether our current practices are fit for purpose.MethodsRetrospective audit of community palliative care patients at the end-of-life prescribed anticipatory medicines within a 3-month period. Anticipatory charts and case notes reviewed. Intervention included updating local guidelines, local teaching for medical and non-medical prescribers and sharing results nationally. Eighteen months later, reaudit was performed assessing impact.ResultsIn total, 76 patients included. 75/76 (99%) were prescribed an analgesic, antiemetic, antisecretory and anxiolytic. 49/76 (64%) were administered ‘as required’ medications at home. Haloperidol was the favoured antiemetic (88%), costing our hospice ~£2000/month. Case note review highlighted prescribing and administration issues. Reaudit showed a reduction in prescriptions of antisecretory (by 57%) and antiemetic (by 50%), with a wider range of antiemetics (levomepromazine 47%, haloperidol 35%, cyclizine 14%, metoclopramide 3%) indicating individualised prescribing. Those without an antiemetic prescribed did not later require one dispensing.ConclusionOur work challenges the orthodoxy that an analgesic, antiemetic, antisecretory and anxiolytic medication must always be included for effective anticipatory prescribing. Antiemetics may not be universally required and individualised prescribing was cost-effective and safe at a local level. Further work evaluating the impacts of altered practice on patients, caregivers, professionals and in other community settings is required.

Background:Older people with severe frailty are nearing the end of life but their needs are often unknown and unmet. Systematic ways to capture and measure the needs of this group are required. Patient reported Outcome Measures (PROMs) & Patient reported Experience Measures (PREMs) are possible tools to assist this. Aim:To establish whether, and in what ways, the needs of older people living with severe frailty are represented within existing PROMs and PREMs and to examine the extent to which the measures have been validated with this patient group. Design:The scoping review follows the method of Arksey and O ' Malley. Results:Seventeen papers from 9 countries meeting the inclusion criteria and 18 multi-dimensional measures were identified: 17 PROMs, and 1 PROM with PREM elements. Seven out of the 18 measures had evidence of being tested for validity with those with frailty. No measure was developed specifically for a frail population. Using the adapted framework of palliative need, five measures covered all five domains of palliative need (IPOS, ICECAP-SCM, PDI, WHOQOL-BREF, WHOQOL-OLD). The coverage of items within the domains varied between the measures. Conclusion: Existing PROMs and PREMs are not well designed for what we know about the needs of older people with severe frailty. Future research should firstly focus on adapting and validating the existing measures to ensure they are fit for purpose, and secondly on developing a better understanding of how measures are used to deliver/better person-centred care.

Background Post COVID-19 and with anticipated winter pressures, there was a clear need to ensure older people with frailty had timely assessments, with access to high quality end-of-life care and support. Aim To provide a rapid, coordinated response, to improve quality of care and decision making between patients, families and health care professionals across hospital and community settings and reduce hospital admissions from care homes (CH). Methods The CCG funded a dedicated rapid response end-of-life care team: 2 senior nurses (1.6 WTE), medical consultant (0.2 WTE) working alongside existing services. They coordinated inter-agency working through face-to-face and digital communication, prioritising face-to-face review and involvement of relatives. We trialled new approaches to triage requests for support and provided interventions in CHs including quick access to medications, fluids and/or diuretics. Results Over 11 weeks, 138 patients were assessed (mean age 88 ±7.5 years, 68% female). Half had diagnosed dementia, half had clinical frailty score of 8. Two thirds were in nursing homes. 39% were referred by their GP, 24% by the CH and 18% from hospital on discharge. 50% were seen face-to-face on the day of referral; 81% within a week. In 17.4% acute hospital conveyance was avoided by shared decision making between GP, LAS, CH staff and families. Treatment escalation plans were initiated for 56% and updated for a further 32%, including generation of shared electronic care plans. For 50%, single conversations with relatives were needed to support decision making; for a third this was more complex with multiple discussions over time. 36% died, with only 2 deaths in hospital. Whilst COVID affected 18 patients, only 2 of those died. Conclusion Flexible, integrated working enabled patients to receive care and die supported in their usual residence. Care homes need focussed resources and support to hold uncertainty in this disadvantaged group who traditionally don't get access to specialist palliative care.

Background Older people with frailty living in care homes (CHs) are vulnerable to sudden deterioration and death. Unnecessary and unwanted conveyance to hospital could be avoided by robust treatment escalation plans (TEP). However, plans are only useful if accessed at times of crisis when CH staff feel challenged to avoid risk. Aim To introduce a system to improve communication of TEPs between GP, Hospice, Hospital, CH teams avoiding repetition and dependency on electronic systems. Methods Palliative care consultants worked with the Lead GP for CHs in one London borough to 1) integrate electronic record systems; 2) capture key data from conversations; 3) translate information to a patient specific colour coded format; and 4) ensure display and access of TEP by the CH staff in real time. This solution is being trialled over 3 months with the intention that people will receive care in the place they have requested and staff feel more supported. Results TEP generation through EMIS shows a large deficit in the current number of recorded TEPs: 69% of CH residents had TEPs, 32% were patients on the Gold Standards Framework with death anticipated <1yr. Working as part of a pilot project to facilitate excellent end-of-life care in CHs, data showed that the majority of residents/patients and their families are realistic about long-term future when given opportunity to engage in discussions about TEPs. Patients and families were keen to avoid hospital, but importantly wanted reassurance they would receive care that would maintain comfort and dignity. Case studies during this time indicated that TEPs were less likely to be adhered to out-of-hours with poor access to agreed TEPs in CH setting. Conclusion We believe a simple TEP chart generated by the GP, with information shared by the hospice, can be accessed by the CH team to help advocate for CH residents to receive the care and treatment they want.

Context. People with advanced heart failure have supportive and palliative needs requiring systematic assessment. Objectives. We aimed to assess the validity of the Needs Assessment Tool: Progressive Disease - Heart Failure (NAT: PD-HF). Methods. Secondary analysis of routinely collected patient data from a specialist palliative care-heart disease service improvement project. NAT:PD-HF, the Integrated Palliative care Outcome Scale (IPOS), and patient and/or carer-report data were collected. Concurrent validity between NAT:PD-HF items and comparison measures (Kendall's tau; kappa); construct validity via known-group comparisons; predictive utility of NAT:PD-HF for survival (multivariable Cox hazard regression model). Results. Data from 88 patients (50% men; mean age 85; median survival 205 days; 64% left ventricular systolic dysfunction) were analyzed. Prevalence- and bias-adjusted kappa values indicated moderate agreement for physical symptom needs (k: 0.33 for patients, 0.42 for carers). Substantial agreement was observed for patient and/or carer psychological symptoms, and information needs (k = 0.6). NAT:PD-HF distinguished between patients with different survival, comorbidities, functional scores, and palliative Phase of Illness with moderate to high effect sizes. NAT did not predict survival when adjusted for mortality risk score and functional status (2+ needs HR: 1.52, 95% CI: 1.01-1.74). Conclusion. The NAT:PD-HF is a valid tool for clinician assessment of physical, psychosocial, and information patient and/or carer needs. (C) 2022 American Academy of Hospice and Palliative Medicine. Published by Elsevier Inc. All rights reserved.

Background Dementia care is complex, especially at the end-of-life. People with advanced dementia suffer from burdensome symptoms; quality of living and dying is enhanced by palliative care (Eisenmann, Golla, Schmidt H et al., 2020). Emphasising personhood (Kitwood, 1997) should be integral to all interactions, providing a safe and nurturing environment in which the person is encouraged to express themselves. This reflects every individual's intrinsic uniqueness and the interconnectedness of human beings. Research from cultural arts interventions has indicated that professionals from the visual and performance arts are well equipped to see the person behind the condition and focus on possibilities for meaningful relationships in the here and now (Swinnen, 2016). Can this translate into everyday care within hospice care? Aim To promote innovative, creative approaches in dementia care, allowing the development of everyday methods to enhance connectivity and care for people living and dying with dementia. Method A programme of immersive, interactive activities and education raising awareness and stimulating discussion, debate and change has been organised (June 2021). Accessible to all members of the hospice team: clinicians, non-clinicians, volunteers, fundraisers and trustees, encouraging the principle that dementia care is everybody's business. Opportunities to engage in creativity to support connection learning with insightful arts and health practitioners/therapists. These include music events, theatre performances, creative sensory workshops exploring senses, movement, texture and sound, interactive exhibitions and a virtual experiential experience. Results Response/learning from these events will be captured using film, questionnaires and focus groups and used to inform and update the hospice's dementia strategy. Key messages will be presented at conference with development of ongoing engagement and future research plans. Conclusion Sharing experiences and displaying new, creative possibilities will aid understanding of those with barriers of memory and communication, unlocking gateways to connection with people living with dementia is possible.

Background The aim of this study is to describe the characteristics of patients receiving community palliative care and the impact of community palliative care on their symptoms and concerns using routinely-collected patient-level outcomes data. Methods Secondary analysis of routinely-collected clinical and outcomes data (patients‘ demographics, episodes of care, palliative Phase of Illness, and symptoms/concerns using the Integrated Palliative care Outcome Scale; IPOS) from all patients cared for in the community by one hospice, between June - November 2019. We described the characteristic of patients, their episode of care, and their Phase of Illness at first contact with the community service. We compared community patients based on their episode outcome (alive versus died, at end of episode) and the impact of community palliative care over the first Phase of Illness. Caldicott Guardian approval was received for analysis of this anonymized data. Results 881 patients received 916 complete episodes of community-based palliative care between June-November 2019. For 916 episodes of care, the median age of community patients was 79 years with IQR (69 ‘‘ 87); 66% had cancer. Palliative Phase of Illness was stable (17%), unstable (28%), deteriorating (41%) and dying (14%) at the start of episode of care. By the first Phase change; mean ‘pain’ score reduced from 1.3 to 1.2 (30% of 158 patients with moderate/severe/overwhelming pain improved); mean ‘breathlessness’ score reduced from 0.8 to 0.7, (with 39% of 87 patients with moderate/severe/overwhelming breathlessness improved); mean ‘anxiety’ score reduced from 1.2 to 1.0 (with 31% of 84 patients with moderate/severe/overwhelming anxiety improved); and mean ‘feeling depressed’ score reduced from 0.7 to 0.6 (with 42% of 45 patients with moderate/severe/overwhelming depressed mood improved). Conclusions This analysis demonstrates, for the first time in the UK, how community palliative care teams can systematically use routinely-collected outcomes data to demonstrate the positive impact of the care they provide. Funding Supported by Yorkshire Cancer Research (L412)

Background Place of death has been used as a quality indicator. However, evidence on changes in preferences for place of death remains limited. This study describes preferred and actual place of death, plus changes, among patients who died while receiving inpatient and community palliative care. Methods Secondary analysis of routinely-collected clinical and outcomes data between June-November 2019. Inpatient hospice and community settings were included; for each setting, we compared first-documented preferred place of death with the latest-documented preference across the episode of care. We also compared the latest-documented preferred place with actual place of death. Caldicott Guardian approval was received for analysis of this anonymized data. Results 269 patients received 277 complete episodes of care in the inpatient unit, and 510 patients received 527 complete episodes of community-based palliative care. Median age for inpatients was 75 years (IQR 63 '' 83) compared with 81 years (IQR 71 '' 89) for community. 82% inpatients had cancer compared with 63% in the community. 98% inpatients had no documented change in preferred place of death (69% and 27% preferred hospice or home, respectively). 92% community patients had no documented change in preferred place of death (74% preferred home). We compared the latest-documented preferred place with actual place of death. Two thirds of inpatients (66%) preferred and died at the hospice, while 27% preferred home, but died in hospice. In the community, 57% of patients preferred and died at home, while 19% preferred home, but died elsewhere (in hospital or nursing/residential home). Conclusion There is high concordance between preferred and actual place of death among both inpatients and community palliative care patients in this cohort; with the exception of some hospice inpatients who prefer home but die elsewhere. Change in preference for place of death are very infrequently reported. Funding Supported by Yorkshire Cancer Research (L412)

Background Implantable cardioverter defibrillators (ICDs) are increasingly used for prevention of sudden death in people with heart failure (HF). Palliative Care health care professionals (PC-HCP) are increasingly caring for patients approaching end-of-life due to progressive HF or comorbidities. Conversations with patients, families and professionals to facilitate ICD deactivation can be challenging, particularly in community versus hospital settings. Method A survey of PC-HCP was undertaken. Invitations to participate were emailed to hospital and community palliative care teams across the UK. We examined ease of access to appropriate services/personnel from different settings, clarification of whether policies and lines of responsibility were in place, clinical triggers to prompt ICD deactivation and availability of training for staff. Results 97 HCP responded, of whom 46% were in post >10 years. 59% had cared for <5 patients, 16% 5–10 patients and 3 >30 patients with ICDs in the previous 12 mths. Similar numbers worked in NHS or independent sectors (38% each), 24% working across both. 27% were based in primary/community care, 36% secondary care and 37% across both settings. 26% participated in multidisciplinary HF meetings and 72% of these had presented cases for consideration of ICD deactivation. 60% were aware of a site-specific ICD deactivation protocol in their workplace, with less knowledge about linking to a regional or DNACPR protocols or ReSPECT process. Where protocols were in place, 42% contained specific guidance for community-based deactivation. Delays in deactivation were limited by staff resource/availability in both community and hospital settings, particularly out of hours, with only 36% feeling that appropriately trained staff and equipment were available out of hours. Prompt access to a magnet for emergency deactivation differed across settings: hospital (62%), hospice (72%), home (33%) and Care Home (22%). Less than half (43%) agreed that PC-HCP had adequate training to facilitate use of the magnet. Conclusion PC-HCP report significant barriers to ICD deactivation, particularly in the community. Local and regional policy still needs to be standardised across the UK, and education and training provided to enable timely conversations and integrated pathways in place. Further analysis of case vignettes will be presented.

Aim To review our response to support residents and staff in Bromley Care Homes during the COVID-19 pandemic (April 2020–March 2021), reflect on our learning, joint working with colleagues and identify good practice to inform future models of care. Methods We collated clinical activity data and key case reviews of care home patients referred to our service over the year. Contemporaneous notes from formal/informal reflections and debriefs (internal and joint with GPs/CCG/other professionals) were reviewed. Themes from feedback of care home staff and managers (ad hoc and formal focus groups) were included with personal reflections. Results 345 patients were referred from 32 care homes. The majority (45%) in Quarter 1 (first wave), 14% in both Q2/Q3 and 27% in Q4 (second wave). Median age 89(53-110) with 69% >85 years; two-thirds female. 80% had a non-malignant primary diagnosis. Just over half died within the year; median time referral-to-death 17(0-229) days, 81(23%) remained on the caseload April 2021. Key themes in Q1 included: limited effectiveness of virtual assessments, atypical patient presentations, significant impact of social isolation on mental health/function, with families unable to advocate and inconsistent messaging about visiting rights. Care home staff were distressed, burnt out, feeling unsupported. In Q2/Q3 regular GSF meetings with care home-GPs, virtual teaching (webinars/ECHO) and staff 'cascade project' study days helped consolidate learning. The second wave was heralded by an outbreak in extra-care housing; care home-GPs were self-isolating. We led urgent senior clinical review and response. In Q4, daily COVID-19 monitoring meetings were key (representation from CCG, Public Health, Pharmacy, CH-GPs and St Christopher's). Over a month, successive outbreaks were identified and resources coordinated to ensure clear advance care plans, timely review, targeted multiprofessional support to care home staff. Learning from each setting informed changes to approach in the next, including: understanding culture, correcting/enhancing infection control procedures, improved shared decision making and addressing hydration. Conclusion Senior clinical leadership, cross-boundary flexible working and willingness to learn together were vital.

Background: Heart failure is a complex clinical syndrome affecting an increasing number of the ageing population. Patients and carers require increasing input from specialist palliative care services to both manage symptoms and access support in the last year of life. An integrated clinical service between the local cardiology team at Princess Royal University Hospital and the palliative care team at St. Christopher’s Hospice was piloted for patients with end-stage heart failure in Bromley in Kent, UK. This study explored views of patients and carers who participated in the integrated pilot service.Methods: A qualitative study was conducted in which a convenience sample of patients and carers were invited to participate in focus groups: two bereaved carer groups (n=2, n=2); one patient group (n=4), held between 14th December 2018 and 18th January 2019. Participants were asked to describe their experiences of care received facilitated by a topic guide. Interviews were recorded, transcribed and coded using thematic analysis to identify common themes.Results: Four patients (2:2 M:F) aged between 70 to 87 years and four female carers whom had cared for patients aged between 70 to 96 years who were since deceased, participated in this study. Overall, the service was positively received, and responses were mapped into four key areas; being diagnosed and living with heart failure, referral to palliative care, key helpful components of the care received and finally, unhelpful components of the new service in terms of care. Common themes emerged including understanding of heart failure and its trajectory, communication around palliative care, having a ‘broker’ for the system, recognition of carer’s needs, service responsiveness, and feeling ‘in control’.Conclusions: This qualitative study highlighted important considerations when developing an integrated heart failure and palliative care service. Education about heart failure for patients and carers, but also the integrated multidisciplinary team is crucial to improving detection of deterioration and facilitating communication around Advance Care Planning. The value of the ‘expert-carer’ should also be promoted and supported in chronic conditions. We recommend a focus on development of integrated services that enable joined-up care or single point of contact for patients and carers.

Aim To identify patients suitable for subcutaneous fluid administration (SCF) in the home/care home setting and standardise clinical approach to assessment, decision making and evaluation of response. Methods A standardised decision support tool was developed which supported bedside assessments. Patient/carer information leaflets were given and documentation of discussions/outcomes recorded in relevant clinical records (EMIS or SystmOne). We evaluated clinical response over three days: symptoms, side-effects/complications, observations or bloods and collated case vignettes. Results Between April 2020-Feb 2021, 28 patients were assessed, median age 87.5(43-97) years, 20(71%) female. Place of care: 13(46%) own home, the remaining 15 (54%) resided in care homes/extra care housing. Diagnoses: 19 dementia; 6 cancer, 2 heart failure, 1 Motor Neurone Disease. Frailty scores (Rockwood): median 7(6-9). 17(61%) received SCF. Equipment was sourced equally from hospice and local pharmacy. Four had clear benefit, symptom relief, admission avoidance, alive >1mth post-intervention. Six cases demonstrated psychological benefit for patient/family with no harm/side-effects (all died 4-17 days post-intervention). Two had possible benefit but diuretics were withheld/reduced concurrently (1 alive >1mth post intervention, 1 died 3wks). Five had no clear benefit, but no harm. 6/11 cases where SCF were not given, died within a week (2-8 days). Five with dementia in care home/extra-care housing were managed with mouthcare/encouraging oral fluids (3 alive >1month post-assessment). Fluids were stopped due to side effects in three patients: none experienced ongoing harm. Volumes of fluid given varied. Feedback from clinicians and families suggested the tool facilitated discussion/understanding of rationale for giving/withholding. Conclusion SCF was of clinical benefit for some patients, with minimal harm. The tool increased clinicians' confidence to assess benefits/harms of SCF and address family concerns in a structured way. Further robust data collection is needed to monitor use/outcomes of SCF in a community setting and to better identify patients who may benefit.

Introduction Co-ordinate my care(CMC) is the pan-london electronic patient record system for palliative care patients. It facilitates sharing urgent care records across settings, linking services which patients may access(GP, ambulance, district nurses, palliative care). By creating records and working with local services to update records, we aim to improve 'joined up working' for our patients. Aims/objectives To audit use of CMC for patients under the hospice including numbers of patient consented/records created, congruence of information recorded with hospice electronic records and access of records by emergency services. Methods We conducted snapshot audits of records in december 2018, subsequent QI initiatives including staff training and integration of CMC reviews into MDM meetings, with re-audit December 2019. Each audit examined 200 patient records, 100 indicated as 'CMC record created' on our hospice system and 100 indicated to have no record. Patients were randomly selected; 40 from each service:Bromley community, bromley care co-ordination, croydon community, Lambeth/southwark/lewisham community and sydenham outpatients. Information was cross-checked against the live CMC record. Standards included: 90% of patients should have a CMC record, 95% should have documentation of consideration of CMC, 95% of CMC records should have resuscitation and ceilings of care decisions recorded with congruence between systems. We also recorded whether CMC records were accessed by OOH services. Results Comparing data from our initial audit to the re-audit; 61% vs 73% of patients under our care were consented to have records created. Of these >95% had a live CMC record. For those not consented by us 220–40% had a live record across each of the 5 services. Between audits, numbers of patients with CMC records increased from 48–60% to 58–68%. Discussion of CMC with patients increased from 48–53% to 70–93%. Recording of DNAR status averaged >95%. Congruence of DNAR decisions was 82% with ceilings of care 61% congruent. Inclusion of hospice contact details varied across teams (30 to 74%). Records were accessed by out of hours services for 13 of these patients. Conclusion Use of CMC is increasing with access to records by OOH services. Further QI initiatives to drive numbers and quality of records will be discussed with impact of OOH records accessed on patient outcome.

Background/Aims Anticipatory prescribing (AP) of injectable medications at the end-of-life for community patients is good practice to achieve timely symptom control (National Institute for Health and Care Excellence. Care of dying adults in the last days of life, 2015). However robust evidence to support current practice is lacking. We evaluated baseline AP practices at a South London hospice against current local guidelines. Methods This retrospective audit looked at patient case notes and anticipatory charts (June-August 2020) using a standardised proforma including: drug choice, dose, indication and subsequent patient clinical trajectory. Anticipatory drug use, drug wastage and themes from case vignettes are presented. Results 76 patients were reviewed, median age 80 (41-107), 66% male. 52% had a cancer diagnosis. All patients were prescribed and dispensed four medications for: pain, agitation, secretions and nausea/vomiting. There was close adherence to local guidelines (choice of drug, dose). Most commonly used drugs were: morphine 61%, midazolam 99%, glycopyrronium 97% and haloperidol 88%. 94% of patients died within three months (median 9 days). Eleven patients required admission to hospital or hospice. 64% had stats given at end-of-life, 53% for pain, 41% for agitation, 24% for secretions; only 16% needed an anti-emetic. We saw wide variation in prescribing practices for seizures at end-of-life (11 patients). Multiple health care professionals (Clinical Nurse Specialists, District Nurses and the London Ambulance Service) administered stats in and out of hours. Further training is required to ensure appropriate dose escalation and titration. Dispensing all four medications costs approximately £50 per patient; haloperidol alone accounts for 3/5ths of that cost and is not often used. Conclusions Few patients required all four medications, where anti-emetics are needed the choice of drug warrants further review with likely cost savings. At least 50% of our medications could be dispensed by GPs rather than hospice pharmacy. Further education/training will improve individualised AP and tailored administration of medications at end-of-life, including improved confidence around seizure management. We are currently updating our guidelines to reflect this.