Knowledge sharing across the boundaries between care processes, services and organisations: the contributions to 'safe' hospital discharge

Stroke discharge

Stroke is the third leading cause of death in the UK and the single largest cause of disability in community settings, where over 50% of strokes result in some form of permanent disability. 32 The annual cost to the NHS of providing stroke care is over £2.8B. 33 The threats to safe discharge for stroke patients are considerable, including falls, the use of take-home medicines, psychological distress and cognitive recovery, access to and use of therapists, and the associated complications of personal care, such as incontinence. 19,21,22 Highlighting the contribution of ‘systems’ thinking, such risks often have their origins in the wider organisation and integration of stroke services. 34

For the last decade, health policies have promoted standards in the organisation and delivery of stroke care, culminating in the National Stroke Strategy. 35,36 These include the introduction of dedicated stroke units, thrombolisation pathways, specialist rehabilitation services and early supported discharge (ESD) services in the community. Each aims to diminish long-term disabilities and promote improved outcomes for patients and families. Evidence suggests, however, that there remain many inequalities in stroke care provision and subsequent outcomes related to age, location, professional skills, organisational strategies and service provision. 37,38 A growing body of research demonstrates the potential benefits of initiatives to support safe and timely discharge, including the use of ESD models. 39 Such techniques bring together service providers and users located in different settings to plan discharge. The National Stroke Strategy,36 for example, sets out the aspirations for ‘life after stroke’ by emphasising the importance of integration between stakeholders to plan and support the transfer of patient care. Similarly, it emphasises the importance of ‘working together’ through stroke networks that bring together stakeholders to review, organise and deliver services. Linked to this, the National Clinical Guideline for Stroke40 suggests stroke rehabilitation should include a MDT of relevant agencies to plan care needs after hospital.

Hip fracture discharge

Each year in the UK around 310,000 patients, the majority of whom are elderly, present to hospital with fractures. Around a quarter of these, up to 80,000, are hip fractures. 41 Projections suggest that, if this continues, numbers will rise from the current figure of approximately 80,000 to 101,000 in 2020. The average age of hip fracture patients is 81 years; 74% are female and they are among the most vulnerable groups to be admitted to hospital. Many have multiple comorbidities, including cognitive and physical impairment, which leads to longer lengths of stay, more complex discharge planning and higher readmission rates. 42 More than 60% have diagnosed dementia. 43 The median length of stay in acute hospital is 26 days, probably an indication of the general frailty of the patients, increased survival rates and increased longevity.

Care of fragility fractures is expensive. Direct medical costs to the UK health-care economy have been estimated at £1.8B in 2000, with most of these costs relating to hip fracture care. 41,42,44 Patient frailty is reflected in the outcome of hip fracture: 10% of people die in hospital within a month, and one-third are dead at 1 year. The fracture is responsible for less than half of deaths,44 but patients and families will often identify the hip fracture as playing the central part in a final illness. Hip fracture can seriously damage quality of life for survivors, of whom only half will return to their previous level of independence. Most can expect long-term discomfort and half will suffer deterioration in mobility, such that they will need an additional walking aid or physical help. Between 10% and 20% of people admitted from home will move to residential or nursing care. Length of hospital stay varies considerably between units, reflecting variability in service structures and provision, such as in early rehabilitation and the availability of downstream beds. After discharge, additional costs for health and social aftercare average £13,000 in the first 2 years.

Like stroke patients, hip fracture patients experience many potential threats to patient safety around hospital discharge, including the risks of falls, social and personal care planning, and management of medicines. Also in common with stroke care, there remains little evidence of the latent sources of risk manifest in these interagency patterns of interaction and knowledge sharing. Many specialities and agencies are involved in hip fracture care, and when discharge is planned and co-ordinated effectively through interagency working, this is shown to reduce length of stay and improve patient recovery. 45–47 Again, this requires effective and timely knowledge sharing between agencies. Alternatively, ‘offsite’ rehabilitation units are commonly aimed at frailer patients who require further rehabilitation; however, the complexities around discharge planning are often more significant given the additional level of care introduced before the patient is ultimately returned home.

Taking the discharge of stroke and hip fracture care as its focus, the study investigates the patterns of knowledge sharing across occupational and organisational boundaries and relates these to the perceived and observed threats to patient safety.

Locating hospital discharge

Hospital discharge describes the point at which inpatient hospital care ends, with ongoing care transferred to other primary, community or domestic environments. Reflecting this, hospital discharge is not an end point, but rather one of multiple transitions within the patient’s care journey. 14,48 The organisation and provision of this transitional care typically involves multiple health and social care actors, who need to co-ordinate their specialist activities so that patients receive integrated and, importantly, safe care. The inherent complexity of co-ordinating a large number of actors, often based in distinct organisations, leads to the view that hospital discharge can be a vulnerable, time-dependent and high-risk episode in the patient pathway.

A prominent example of this complexity is ‘delayed discharge’, where the patient remains in hospital because of the failure to appropriately co-ordinate care between agencies. 27,48 According to Victor et al. ,49 nearly 30% of older people experience some delay in their hospital discharge, which is known to expose patients to additional hospital-related risks, create emotional and physical dependency, incur additional hospital costs and restrict the availability of inpatient beds. In parallel, premature discharge or discharge without appropriate arrangements for onward care can also lead to complications for patient recovery. For example, the 28-day readmission rate for older people has doubled from 103,000 in 2001–2 to 201,000 in 2010–11,50,51 suggesting that more needs to be done to support patient recovery following acute care.

The problems of delayed or poorly planned discharge illustrate the broader challenge of integrating health and social care. 27 Analysing the causes of these delays, Tierney et al. 31 point to a range of common factors, including (a) poor communication between health and social care; (b) lack of assessment and planning for discharge; (c) inadequate notice of discharge; (d) inadequate involvement of patient and family; (e) over-reliance on informal care; and (f) lack of attention to the special needs of vulnerable groups. Reflecting this and other evidence,27 policies have repeatedly sought to improve discharge planning, especially the integration of health and social care agencies. A review of these initiatives is outlined below.

Discharge planning

Improved ‘discharge planning’ has been a consistent recommendation of policy and research. 27,52–54 Over the last two decades, the precise form of discharge planning guidelines has varied to reflect wider health and social care reforms, changing economic imperatives and emerging concerns about care quality. 55–58 Furthermore, they have been developed both locally, by individual care organisations, and nationally, for example by the NHS Institute for Innovation and Improvement, and there is no commonly agreed model. Despite efforts to promote discharge planning, the recent European HANDOVER study found that health-care professions still did not prioritise discharge planning or interagency communication as supporting enhanced discharge. 59 In 2010, the Department of Health published its new discharge workbook, Ready to Go? Planning the Discharge and Transfer of Patients from Hospital and Intermediate Care,30 which outlined 10 ‘steps’ to ensuring a timely, safe and patient-centred transition from hospital, including:

• effective communication with individuals and across settings

• alignment of services to ensure continuity of care

• efficient systems and processes to support discharge and care transfer

• clear clinical discharge management plans

• early identification of discharge or transfer date

• identified named lead co-ordinators

• organisational review and audit

• 7-days-a-week proactive discharge planning.

Effective discharge planning is usually associated with a number of common activities and procedures along the care pathway:14,30

• On admission Prepare detailed and accurate patient record; review assessment information and estimate date of discharge with reference to standard care pathway and complexity of patient circumstances.

• During admission Undertake regular multidisciplinary assessment of patient condition to identify and assess opportunity for discharge; discuss with patient and family ongoing and continuing needs.

• At least 48 hours prior to discharge Inform MDT about estimated date of discharge and review assessment criteria; initiate referrals to community health-care providers and social care agencies; contact agencies responsible for ordering and/or installing patient equipment or home modification; social work/care assessment and referrals; complete referral for social care; finalise care package; order take-home medicines; arrange transport.

• Day of discharge Contact family and carers to confirm follow-up care arrangements; check documentation completion; issue discharge letter to general practitioner (GP); reinforce patient behaviour recommendations and rehabilitation; confirm and finalise transport.

• Follow-up care Initiate social care package and continuing health-care package, where relevant in consultation with GP.

As these policies suggest, a number of specialist roles and activities are promoted as supporting the integration of different agencies. A longstanding objective has been to promote the use of MDTs in discharge planning. 14,53 These are normally organised as formal, usually weekly, meetings between relevant health and social care specialists with the aim of supporting timely communication, inclusive decision-making and continuity of care. Research often describes MDTs as comprising a core team including the named doctor and nurse, occupational therapists (OTs) and physiotherapists (PTs), and representatives from community and social care agencies, as well as family representatives, GPs and other specialist therapists. According to Bull and Roberts,52 MDTs help break down barriers between professional groups and foster a sense of common purpose and trust. Importantly, MDTs provide an opportunity for communication, first between professionals, second with patient and family, and third with community health-care providers. Furthermore, MDTs can help make clear the lines of responsibility for different tasks and create opportunities for individuals to take the lead in co-ordinating the planning process. In practice, however, convening all representatives for individual patients can be challenging in terms of time or resources. 27

A further initiative has been the introduction of discharge co-ordinators. 14,30 These are individuals, usually experienced nurses, who take lead responsibility for both strategic planning and co-ordination of discharge at the interorganisational level. 60 Research suggests that discharge co-ordinators can improve hospital discharge through supporting the integration of different professionals, overseeing and directing planning and addressing emergent problems in a more responsive way. 61 In particular, co-ordinators acquire both deeper understanding of and extended relationships with a wider range of care agencies that help them better navigate and align divergent ways of working that usually delay or undermine discharge. 52,61,62

Integrating care services

In line with the developments in discharge planning, policies have also introduced new or extended statutory powers, financial opportunities and penalties to support more integrated discharge pathways. For example, the Health Act 199963 enabled health and social care agencies to pool resources to codeliver rehabilitation services. Similarly, in 2005, delayed discharge grants were made available to social service authorities across England to develop reablement services. In contrast, the Community Care (Delayed Discharges) Act (CCDDA) 200364 addressed the problems of integration by allowing hospitals to claim financial reimbursements from local authorities where they delayed discharge by not providing timely services. Against this backdrop, a variety of integrated services and new care pathways have emerged to support the transition from hospital to community, but in doing so have extended (and made more complicated) the range of services involved in discharge planning.

One significant development has been the introduction of ESD. ESD is often associated with the care and rehabilitation of mild-to-moderate stroke patients. It enables patients to return home early with a dedicated package of rehabilitation and reablement of a similar intensity to that provided by inpatient care. ESD is shown to reduce the burden on acute providers and support patient recovery. 65 The funding of ESD through joint commissioning between the acute NHS providers, GPs, social services and central government highlights the role of joint working and resource pooling, but there remain variations across the UK, especially in rural areas, where a lack of funding can limit provision. 66,67

Intermediate services provide transitional, ‘step-down’ care between acute hospital and the domestic environment (usually for 30 days). Patients are typically declared as ‘medically fit’ but requiring ongoing care or rehabilitation, for example those at risk of readmission or with complex care needs. Rather than receiving rehabilitation at home or in hospital, intermediate care offers a form of residential, hospital-like care, but with a focus on rehabilitation. Research suggests that intermediate care services have been effective in both reducing financial costs and improving patient outcomes. 68 Owing to their close proximity to patients’ homes and relatives, community (NHS) hospitals or nursing homes are often used for intermediate and post-discharge rehabilitation. Stays in such units can be longer than in other intermediate care services, yet research suggests patient outcomes are generally favourable. 69 The recent Cochrane reviews of long-term rehabilitation in care homes show no evidence of negative health outcomes. 65,70

A similar initiative is the introduction of reablement services. These usually involve a dedicated package of social care to support daily living in the immediate period following discharge (e.g. personal care, cooking and cleaning). They are usually managed and provided by local authority social services, although in some cases they are funded through both health and social care budgets. In 2012, the Department of Health allocated £150M for reablement linked to hospital discharge,30 to be allocated through primary care commissioners working in partnership with social care authorities. Significantly, these services are normally arranged and provided by social services to ease transition from hospital for a period of 4–6 weeks, with the expectation that ongoing social care will be reassessed and provided by other agencies.

A further example of service innovation, with particular reference to end-of-life care, is the introduction of ‘fast-track’ discharges. This normally relates to supporting early discharges from hospital for those patients wishing to spend the last days of life in the community with palliative support. This end-of-life discharge can exemplify effective joint working and rapid prioritisation, whereby the patient can be discharged within 48 hours with all specialist support and medications in place. 71 For example, funding decisions are established post discharge to remove delays; the needs of the patient and family are met by deliberate use of a continuous dialogue with one specialist co-ordinator; and the emphasis is on timely collaborative working to ensure the patient gets home as requested. 72

The threats to ‘safe discharge’

Multiple sources of evidence suggest that care quality can be suboptimal in, or as a consequence of, hospital discharge. 28 In a major telephone survey of 400 patients following discharge, Forster et al. 17 found that nearly 20% reported some form of adverse event, of which 6% were preventable and 6% ameliorable. Research highlights a number of common discharge-related risks associated, for example, with the management of medicines, the provision of appropriate health and social care, incomplete tests and scans, the fitting and use of home adaptation, and the risks of falls, infections or sores. 17–28 The underlying sources of these risks can range from factors related to the patient’s condition or comorbidities, to the assessment of patient need, the availability of specialist resources in the community, and wider organisational and cultural factors. For example, research shows that the patient’s condition, such as hip fracture, and other comorbidities, especially cognitive function and fragility, can represent a cluster of risks, particularly for older patients, that can complicate the discharge process. 73,74 Research also suggests that time of day, week or year can also have an impact on discharge planning and quality. In particular, discharges during the weekend have been shown to increase the likelihood of death compared with those taking place between Tuesday and Friday, accounting for 34% of all post-discharge deaths. 75,76

Although studies highlight the importance of clinical risk in discharge planning, it is not always clear how ‘risk’ is measured. Moreover, the causal analysis of risk is often implicit or an emergent feature of wider trial research. Reviewing the recent literature (Table 1), a number of risks (direct threats to safety) and identified causes (suggested or inferred) are catalogued.

Although the sources of these risks can be complex and variable, research frequently highlights incomplete, inaccurate and inaccessible information as undermining collaborative workings and contributing to unsafe patient discharge. 27–29,86,87 A systematic review conducted by Kripalani et al. 29 found that communication between hospital and family doctor was often partial or missing, relying primarily upon discharge summaries which were often incomplete, lacking in detail and not provided in a timely manner. Similarly, poor communication between the hospital and social care providers is a long-standing risk factor in adverse events. 27,29,88 There remains little extensive research, however, examining the causes of poor communication and adverse events. 29,89 Less is known about how communication breakdowns and patient safety are experienced by patients and carers. 54 A number of studies propose, and in some cases evaluate, interventions to support communication and information transfer at discharge, including structured communication tools, discharge planning guides, discharge checklists, medicine reconciliation guides and patient education strategies. 84,90–94 These suggest that effective discharge planning depends upon effective communication and collaboration between health and social care agencies. 28,86,95 In his analysis of the factors that support or hinder such communication and collaboration, Glasby27 highlights three dimensions:

• occupational factors related to the particular knowledge, cultures and practices of different professionals

• organisational factors related to the working patterns, capabilities and resources of different agencies

• compatibility and co-ordinating factors related to how occupational and organisational factors are aligned, or differences reconciled.

Attention to these and other factors is needed to better understand and enhance communication and collaboration in discharge planning and care transition. Furthermore, greater appreciation is needed of how communication might undermine not only co-ordination but, in turn, safety. In this sense, communication might be seen as a latent (or active) factor that influences the safety of hospital discharge. The next section develops this idea through relevant theory and research on organisational complexity and safety.

The systems approach to patient safety

Current thinking in patient safety is largely informed by theories and research within the fields of ergonomics and human factors. In broad terms, this suggests that performance mistakes are not necessarily brought about by individual negligence, malice or incompetence, but more often by pressures located within the work environment. 96 This line of reasoning makes the distinction between ‘active’ and ‘latent’ errors. The former refers to individual slips, mistakes or omissions that lead to patient harm; the latter to the unsafe conditions that create, enable or exacerbate the potential for active error or patient harm. This can include poorly designed working arrangements, poor defence and early-warning mechanisms or an over-reliance on automation. This approach suggests that risk reduction should attend, not to individual performance alone, but to the upstream factors that make performance error prone, for example by standardising task design, improving team cohesion and communication, alleviating situational ambiguity and recognising the influence of resource management and culture. 96

This ‘systems approach’ to patient safety has been articulated through policies such as To Err is Human1 and An Organisation with a Memory,2 and developed through major programmes of applied health research. 3 For example, it has been used to highlight how a range of ‘task’, ‘team’, ‘situational’ and ‘organisational’ factors contribute to front-line clinical safety. 97 Of specific relevance to this study, this conceptualisation of safety draws attention to the way health care is organised and delivered through a system of interdependent elements interacting to achieve a common goal. 1 Based upon these ideas, various strategies have been promoted to better understand and address the threats to patient safety. These include, for example, the use of incident reporting procedures to enable clinicians to share their experiences of clinical risk and engender system-wide learning of the root causes;98,99 the creation of a safety culture that is mindful of danger, blame free and responsive to organisational learning;100 and a variety of safety-enhancing interventions, such as ICTs or single-use devices which limit unsafe behaviour; checklists, guidelines and the standardisation of practices to reduce variability; and staff training and culture change activities. 3 The ‘human factors’ approach provides a framework (drawing on Vincent101) for conceptualising and investigating the threats to safe discharge (Table 2).

Extending the systems approach

Despite the enormous advances in patient safety research, it has been suggested that the distinctive sociocultural and political dimensions of health care are sometimes overlooked. 6 A related view suggests that the human factors approach tends to, although does not always, focus on what might be considered the ‘clinical micro-system’ or local work environment. Take, for example, the emphasis on team skills, communication aids, decision-support tools and checklists. This approach considers less frequently (and in less detail) the wider cultural, institutional or system factors related to workplace practice. Similarly, research has tended to focus within clinical specialities, departments or units, such as operating theatres, care homes or emergency departments, with less attention to the interconnections between these areas and the wider organisation. 3 Although organisational and societal factors are recognised,96,101 they are not well developed theoretically or empirically as distinct and interconnected levels of analysis.

Attention to the sources of risk and safety between care settings and processes is important for hospital discharge as relates directly to broader organisational and interorganisational factors. It calls for attention to the nature of the relationships and interdependencies between care organisations as patients pass from hospital to community care. These interdependencies constitute ‘system-level’ (latent) sources of safety, and where research does attend to these interdependencies it shows that the organisation of health care is often so complex and non-linear that the idea of creating a reliable, standardised and safe system remains a ‘wicked problem’. 102–106 Extending this line of thinking, Hollnagel107 suggests that although the human factors approach goes beyond a simple (single agent) model of error causation to identify instead environmental causation, it too often neglects the non-linear and more complex dynamic coupling involved in many organisational or system processes.

Developing this perspective, a ‘system’ describes a collection of actors, units or parts that together, and through their various connections, form the basis of a structured and relatively bounded entity. Through combining these constituent elements, systems produce particular effects; in many instances these are positive, but they can also be unpredictable, unanticipated or unsafe. 108 Broadly speaking, ‘complex systems’ have emergent properties that result from the non-linear or complex connections between their heterogeneous units, subsystems, variables or actors. 108,109 Furthermore, complex systems can produce effects that are not always evident from the actions or attributes of the individual parts, owing to the potential for the actions of one component to transform the contexts and actions of others. 108 More than this, complex systems are characterised by the absence of structural design and by patterns of self-organising and adaption. For example, the constituent parts within a system interact according to local rules (e.g. customs or cultures) without any overarching direction or control (e.g. leader). 110

Health-care systems are increasingly identified as complex systems. 102,111–114 The organisation and delivery of health care typically involves a vast number of agencies that span different economic sectors (public, private, voluntary), health domains (primary, secondary, tertiary) and care domains (health, social, personal), with distinct roles and responsibilities (commissioners, providers, regulators) and caring specialities. Moreover, the interactions between these actors are not always well defined, but remain dynamic, non-linear and multifaceted, from micro-level clinical interactions, to the meso-level interactions of service planning or commissioning and the macro-level interactions of policy-making and professional associations. As such, health-care systems can change in ways that are not always easily anticipated by policy-makers or service leaders. Extending this line of thinking, hospital discharge might have been conceived as a complex system involving a network of diverse, often heterogeneous actors, interacting in dynamic and non-linear ways that over time produce unpredictable and unanticipated behaviours or outcomes, both positive and negative.

One seminal analysis of organisational complexity that specifically demonstrates the potential for negative consequences is Charles Perrow’s study of Normal Accidents. 8 This was developed through an investigation of the system failures that contributed to the Three Mile Island nuclear disaster. His analysis suggests that organisational accidents might be interpreted as inevitable for complex organisational systems, not simply because of the high-risk nature of the work or potential for ‘operator error’, but because of the way organisational processes are configured. In particular, his research elaborates how relatively small, isolated or ‘discrete failures’ that occur in one part of a complex system can cascade and escalate into more substantial disasters. In other words, risks do not develop through linear causal chains but through unpredictable interactions. Where organisational activities are ‘tightly coupled’ there are higher levels of interdependence between work processes, meaning that what happens in one area is influenced by, or influences, the work of others. In health care, research has demonstrated how errors and failures from one department have consequences in other tightly coupled activities across the health-care system. 106,115

Although health policies call for greater attention to, and learning about, the systemic threats to patient safety,2 research has largely centred on the latent threats located within discrete and often localised parts of the health-care system. 3 There remains limited understanding or evidence of the threats to, and opportunities for, patient safety located within the wider health-care system, especially in the complex interdependencies and non-linear interactions that exist between care providers located in diverse occupational, organisational and sectoral domains.

Applying these ideas to the problem of hospital discharge, it becomes apparent that the complex interdependencies and non-linear couplings between health and social care agencies can be the latent source of poorly timed, inappropriate or unsafe transition from hospital to community care. Policies to support integrated discharge planning have repeatedly looked to introduce improved or more stable couplings between these agencies through, for example, resource pooling and MDTs, and more routine and robust methods of information sharing. 30 There remain, however, enduring compatibility factors27 that inhibit co-ordination, and hence discharge safety, given the persistence of heterogeneous, tightly coupled and interdependent working practices.

Knowledge sharing for safe discharge

Applying the above literatures to hospital discharge, this study investigates how patterns of knowledge sharing among health and social care agencies influence discharge planning and care transition. Developing this view, knowledge sharing is conceptualised as a latent source of safety that, through shaping the patterns of co-ordination, shared decision-making and integrated working, can mitigate system complexity. As such, understanding the barriers to and drivers of knowledge sharing can contribute to the development of new knowledge on the possibilities for improved integration and safety in hospital discharge. These lines of analysis are tentatively outlined in Table 3.

System and organisational selection

The primary unit of analysis was the local care ‘system’, within which patient discharge is planned and organised. This ‘system’ is conceptualised as comprising an acute NHS hospital around which other primary, community, rehabilitation and social care services are arranged. Discharge is seen as the planning and transition of care from the acute NHS hospital to community-based health and social care. In line with this view, the study was undertaken in two geographically distinct English care systems, broadly defined by county boundary. Each has a principal administrative city and conurbation (Farnchester and Glipton), with smaller towns, and villages located in rural areas. For the purpose of maintaining anonymity of participating organisations and individuals, the city, country and organisational names have been changed. Sampling of these geographical areas took into account their relative size, ethnic diversity and urban/rural balance (Table 4). Sampling also considered variations in the configuration of the care systems in these counties. Each was served by a single NHS acute trust; one was a large teaching and research-active health-care provider, operating over three large organisational units (Glipton), and the other was a smaller district general hospital with limited teaching and research activities (Farnchester). Sampling also considered the geographic spread of community-based care services, the proportion of single-handed GPs and structure of social care services. These differences are summarised in Table 4 [data were obtained through National Health Profiles143 and the Office for National Statistics (ONS)142 from 2010 to 2012].

Farnchester NHS Trust comprises two medium-sized district general (acute) hospitals, one located in the administrative city of the county and the other in another market town (Farnchester and Fallow), and a third, smaller community hospital located in a remote small town (Ribble). The hospitals were merged in 2000 under the management of a single NHS trust, with headquarters based at Farnchester Hospital. According to available records, the Trust employs over 7500 staff and treats more than 180,000 emergency patients, 500,000 outpatients and 100,000 inpatients every year. Farnchester Hospital is the main provider of specialist services across the majority of the county, including stroke and hip fracture patients. At the time of data collection, two primary care trusts (PCTs) (east and west of the county) commissioned specialist and acute services. Following the Health and Social Care Act 2012, services are commissioned by four Clinical Commissioning Groups (CCGs) (north, south, east and west), which operated in shadow form during the last year of the study. The Trust works closely with Farnchester Community Healthcare NHS Trust, which was formerly managed by one of the PCTs but became a separate provider of health care. It also works closely with a single unitary local authority for the county, whose social service department assesses and provides social and personal care, or contracts care provisions from private or third sector agencies.

Glipton NHS Trust comprises three distinct hospitals (City, General and District), located in different areas of the same city and also brought under the management of one NHS trust in the early 2000s. The Trust employs over 10,000 staff and provides services to over 1 million residents in the local population. It is an established site for health education, with strong links to the local universities, and the site of an extensive portfolio of clinical research. At the start of the research, the Trust worked closely with two NHS PCTs (city and county), who commissioned the majority of its acute and elective services. Following the Health and Social Care Act 2012, three new CCGs operated in ‘shadow’ form until assuming responsibility for commissioning in April 2013. Two community NHS trusts (city and county) provide continuing health care, rehabilitation and home-based care, including the management of two community hospitals. The social services departments of the city and county local authorities were involved in assessing, providing or commissioning social care through a mixed market of public, private and third sector organisations. This includes a specialist reablement service for immediate postdischarge care. The regional NHS ambulance trust supports the transition of patients both to and from Glipton and Farnchester NHS Trusts, and a private transportation firm is also contracted to provide patient transports from hospital.

Sampling started with the two acute NHS hospitals, and included other primary, community and social care agencies. It was anticipated that the variety and number of agencies involved in discharge planning and care might be considerable, and variable according to patient need. Reviewing the literature, a range of common agencies was identified. Although many of these groups could be identified in advance of the study, such as social services, it was difficult to determine the exact profile for each county. As such, a snowball sampling strategy was used to identify agencies and organisations involved in hospital discharge. This group included:

• GPs and primary care administrative and commissioning units (PCTs, now CCGs)

• community health-care services (some formerly managed by PCTs)

• specialist community in-reach/outreach services (possibly managed by PCTs)

• community pharmacies

• local authority, social services

• social care providers working in private, public or third sector (including reablement)

• intermediate care and rehabilitation services

• residential and nursing homes

• ambulance and transportation services

• voluntary sector support groups.

Through purposively selecting differences in health system configuration and patient group (stroke and hip fracture patients), it was possible to develop comparison along four dimensions: between different hospital types and system properties (e.g. single site vs. multisite), where some degree of control is obtained through looking at the same patient group, and between different patient groups (e.g. stroke and hip fracture patients) within the same hospital (Figure 1).

FIGURE 1.

Selection and basis of analytical comparison.

Patient selection

The research used ‘patient tracking’ to develop a patient-centred understanding of hospital discharge (see Patient tracking). This involves focused observations and interviews with patients and families as they experienced discharge planning and care transitions. To facilitate representation across patient groups, a sampling strategy was devised in consultation with clinical specialists on the project team, patient and public involvement (PPI) representatives and clinical leaders at each site. This recognised key variables and comorbidities known to influence discharge activities, such as cognitive impairment or family situation. Sampling included participants with and without cognitive capacity to reflect the general adult hospital population.

Qualitative interviews

Semistructured qualitative interviews were used to understand how hospital discharge is planned, organised and experienced from the perspectives of different health and social care professionals, patients and family members. Interviews were especially important for identifying the roles and responsibilities of actors within the discharge process, the patterns of knowledge sharing between actors and how actors’ knowledge domains, cultures and organisational context shaped their interactions. The interviews were designed to be conducted in a semistructured, conversational style, giving participants the opportunity to explore emergent issues. All semistructured interviews were guided by a topic guide developed to reflect the study objectives. Draft questions were developed by the project team and representatives from the PPI group and piloted at one research site with three clinicians. These topics (see also Appendix 1) included:

• career biographies and backgrounds

• details of roles and responsibilities, with a specific focus on discharge activities

• an account of the discharge process, including the broad process, planning issues, and working with patients and families

• the role of communication and knowledge sharing in discharge processes

• identification of individuals or groups contacted during discharge activities

• exploratory accounts of knowledge-sharing relationships with identified individuals

• perceptions and experiences of risk and safety

• recommendations and improvements.

A modified guide was developed for patients, family members and carers, based on the advice and feedback from the PPI representatives. This included:

• ways in which discharge plans were discussed and planned with patient and carer

• how the discharge was expected and experienced

• whether or not the plan met the needs of the patient

• ways in which discharge can be improved.

An important feature of the interview questions was that they were structured to generate participant narratives, or stories of discharge processes. These stories were not read necessarily as ‘truths’ but rather as analytical windows into how participants make sense of and give meaning to discharge, thereby highlighting differences in knowledge and culture. These narratives were particular insightful when exploring how participants make sense of the discharge process, and especially the sources of safety and risk. A further feature of the interviews was that they were used to inform snowball sampling, that is, to help identify other potential, unanticipated actors or groups involved in the discharge process and the patterns of knowledge sharing with these actors. For example, all participants were asked to describe the different people, groups or organisations they communicate and share knowledge with in the processes of discharge planning and care transition.

Most participants were invited in writing to participate in the study. This invitation included a participant information sheet and an opportunity to contact the project team for further information. Other participants were recruited during ethnographic observations; for example, where an individual was observed as having an important role they would be asked to participate in the study and provided with a participant information sheet. All participants were asked to confirm that they had understood the participant information sheet and give written consent. The majority of interviews were digitally recorded with the consent of participants and all were transcribed verbatim for the purpose of subsequent data analysis. Table 5 details the interview participants according to location and number, and Table 6 further details the interview participants by their occupational background.

Alongside semistructured interviews, the research involved more informal ethnographic-type interviews with participants, normally as part of observations. These ranged from small interactions or ‘chats’ to clarify an observed occurrence or activity (i.e. asking a participant what they are doing or to explain a technical procedure) to more open, unstructured and lengthy exchanges (i.e. to explain a series of events or a particular situation in greater detail). Many of these interviews occurred when shadowing participants or in other informal settings, such as rest areas or cafés. As such, it was not always feasible to digitally record these interactions and most were recorded as handwritten notes in field journals.

Observations

Ethnographic observations of discharge activities provided the main source of data. Observations focused on how discharge was planned, organised and supported as a series of complex interactions between various health and social care agencies. These involved non-participatory or passive observations144 of everyday practices; that is, the research team did not have a formal, legitimate membership role in discharge activities, but they did build rapport and shared understanding with participants through interacting and engaging in other forms of participatory engagement, such as conversations and non-clinical (technical) interactions. The observations combined different activities to build rapport and understanding:

• Guided tours and structured familiarisation Undertaken at the start of the research over a period of 3–5 days, including ‘walking tours’ with service leaders, introductory meetings with stakeholder groups, informal presentations at weekly meetings and collection of relevant documents, policies and procedures.

• Work process observations More in-depth workplace observations undertaken over a period of 1–6 weeks (depending on the location). These included mapping the temporary and spatial organisation of daily work (schedule of ward rounds, meetings, handovers, discharge times); identifying key events and activities (MDTs, drug rounds); identifying key individuals or groups (discharge co-ordinators, clinical leads); and drawing together these data into a complex descriptive account of the social setting. For the main research settings, i.e. hospital wards, observations were undertaken on a daily basis (3–4 days per week, including evenings and weekends) for 4–6 weeks; in other settings, such as community hospitals or care homes, they were undertaken over 1–2 weeks, and in more dispersed or peripheral settings, such as social service departments, they were undertaken over a shorter time period, i.e. day visits. Some settings were not amenable to observation given the diversity and demands of their workload, such as GP surgeries and community pharmacies, where interviews provided the primary source of data.

• In-depth observations of situational activities, tasks and settings Prolonged follow-up observations of key knowledge-sharing activities, tasks and settings. These observations were usually conducted concurrently with other research activities, including shadowing or interviews. Each setting was observed at least three times and some settings, such as weekly MDTs, were observed up to 10 times. This included observations of:

  • daily/weekly planning and decision-making meetings (MDTs)

  • trauma and assessment ward rounds

  • patient and family consultation meetings

  • community hospital reception

  • nurses’ station ambulance handover

  • social services assessment

  • use of ICT and manual records relating to discharge planning

  • pharmacy ward checking of medications to take out (TTO) on discharge

  • fast-track planning and handovers

  • referrals to multiple agencies by phone.

• Shadowing of individuals To deepen understanding of the roles and contributions of certain individuals or groups, shadowing observations were also undertaken with key individuals (leaders or discharge co-ordinators) or representatives from professional groups (nurses or therapists). These ranged from several hours (ward clerks) to several days (discharge co-ordinators) of observing individuals as they went about their day-to-day work. These individuals included:

  • discharge co-ordinators

  • ward nurses

  • junior doctors

  • consultants

  • hospital social workers

  • ward clerks and administrators

  • OTs

  • PTs.

Using these techniques, it is estimated that the research involved over 180 hours of direct observations and many more opportunistic observations, in a variety of health and social care settings. All observations were recorded, first in handwritten field journals including rich description and separate interpretations. These were later typed up electronically as corresponding text and interpretation, as well as a summary overview of key points.

Patient tracking

A key feature of the research design was the use of ‘patient tracking’ to understand the first-hand experiences of patients and their families/carers as they were discharged from hospital to a community setting. This involves a series of short semistructured interviews (up to five), with patients both before and after discharge, focused observations of their discharge planning and arrangements, and the collection of patient diaries (where possible) (see Appendix 2 for extracts of patient diaries). These activities led to the collection of rich and diverse data that enabled the research team to construct a picture of how patients and family members experience hospital discharge.

In the first instance, patients/families were approached in hospital after admission, usually when an initial date of discharge had been set (within 72–48 hours of planned discharge). Each patient was asked to participate in a series of short follow-up interviews, starting in hospital and then at 1, 2, 4 and 6 weeks post discharge. Although every effort was made to schedule interviews at these times, in some instances they needed to be moved ± 1 week owing to other appointments. Furthermore, not all patients were able to participate in the full series of interviews because of withdrawal (their health had worsened), readmission to hospital or death. Observations were also undertaken with patients, families and carers, on the wards and at home or in a care setting (nursing home) to understand how care was organised and provided. Patients were also invited to keep diaries about their discharge experiences (Table 7). In consultation with PPI representatives, a short, open-text diary was designed and printed and given to each participant with a short description explaining its voluntary use. The diary had a minimal structure and allowed for open text. Specifically, it stated that:

The purpose of this diary is to give the researchers a way of understanding how you are living your life since leaving hospital. The team are interested in what happens each day, especially when you have contact with health and social care workers. Please spend at least 10 minutes each day writing in your diary, but don’t worry if you miss a day.

It also included a number of prompts asking participants to consider which health and social care providers they had interacted with, what they talked about, what kind of decisions were made and who else helped with their care. Participants were asked to keep the diaries for up to 30 days. Diaries were collected following the last interview or earlier if patients needed to withdraw from the study. In most instances, patients withdrew from the study because of continuing illness or moving out of area, in most cases around the time of the second interview.

Focus groups and validation workshops

The research involved a number of focus groups and workshops at each research site, together with a validation workshop at a third hospital site. The focus groups aimed to explore similarities and differences in staff members’ understandings of the discharge process. Each involved structured activities with representatives from professional groups in ward or community settings. Each started by asking participants to individually visualise and draw what they saw as the key stages or activities involved in hospital discharge. These were used to prompt reflection and to gather insight into how discharge was understood by different groups (see Appendix 6). Each group was then tasked with collectively drawing the discharge process on large sheets of A2 paper, using pens and Post-it notes. These usually took the form of flow diagrams, but varied as they were revised and amended. Throughout this process, a member of the research team asked probing questions to clarify the choices made, differences between participants and to surface shared or divergent reasoning. Finally, each focus group involved an open discussion about the sources of risk and breakdowns within the discharge process. Although focus groups were organised with all four sites, the Farnchester hip fracture service was unable to participate in the activity owing to staff shortages and workload pressures. As such, only three of these focus groups were undertaken. Table 8 details the numbers and staff representatives involved in the focus groups.

At each research site a feedback (validation) workshop was organised after data collection to outline preliminary findings, raise any unresolved questions and seek validation on emerging interpretations. These were organised in consultation with service leaders to include staff representatives and clinical leads, usually involving between five and 10 people. They each involved a short visual presentation and provided a written report about the study aims and design, data collection and key findings, including a list of anonymised recommendations from participants for service improvement. These workshops were particularly useful for validating and confirming the early analysis.

An extended ‘validation workshop’ was organised with an additional acute care provider of hip fracture services, not involved in data collection. This aimed to further validate and confirm the research findings through exploring relevance and similarity with professionals and managers from another locality, thereby helping to develop more generalisable lessons. This workshop was organised after all data collection and early analysis was complete. It was attended by the hip services clinical leader and representatives from different professional groups, including nursing, occupational therapy, physiotherapy, and administrative and support staff. The workshop was structured with an overview of the research design and data, including the main points of convergence and divergence across the research sites, and analysis of the common barriers to and facilitators of hospital discharge. Participants were asked to discuss the relevance of the findings to their own service, the growing complexity of patient comorbidities, resource demands in the acute and community settings and the challenges of meeting increased patient expectations. This workshop helped to identify additional lines of analysis to test emerging conclusions.

Overall, the validation workshops identified and help clarify the following issues:

• the range and number of actors involved

• discharge pathways and structure of interactions

• communication and knowledge-sharing problems

• the main perceived threats to patient safety

• contextual factors, i.e. staffing, resource constraints, targets

• wider contextual changes, i.e. CCG priorities

• recommendations for future interventions.

Documents, policies and routine data

Through the study, various documents were collected from each research site. These included formal policies and pathways for discharge planning, patient leaflets and guides, and pictures of posters, fact sheets and other information. These were used to contextualise observational and interview data and provide a point of comparison between expected or planned discharge and observed discharge. The research also aimed to gather routinely collected performance data related to hospital discharge at each site, including data about staffing, bed usage, safety incidents, number of patients, average length of stay and number of readmissions. In three participating sites much of this information was made available by service leaders, and with additional third-party data, specifically health profiles and ONS data. As with the focus groups, however, one site was reported as too busy to help provide these data and, as such, only third-party data were collected. These statistical data were used to help contextualise the organisation and performance of each site, and were not used for formal statistical analysis. The range of data collected included:

• British Heart Foundation Stroke Statistics 2009

• British Heart Foundation Statistics Database 2009–12

• Health Profiles 2012 (Public Health Observatory)

• commissioned discharge support documentation

• annual reviews per authority

• National Hip Fracture Database (NHFD) 201242

• national stroke audit 2011. 32

Admission

At admission, patient care is overwhelmingly concerned with determining an accurate diagnosis, stabilising the patient and commencing initial treatment, with patients usually assessed in the emergency department before being transferred to either the stroke or orthopaedic trauma units. Despite this over-riding focus, the admission stage provided an early opportunity for gathering knowledge related not only to the patient’s immediate care needs, but also to his or her onward care and possible discharge arrangements. In particular, it was an opportunity for clinicians to gather information related to the patient’s wider health and personal circumstances to consider the length of admission, need for rehabilitation and likely discharge arrangements.

At the Farnchester stroke unit, patients were usually admitted by a stroke consultant, junior doctor and ward nurse who would interact with the patient and family to gather and combine knowledge about his or her comorbidities (other health complaints), wider personal circumstances (employment and activities) and domestic arrangements (residence). This line of questioning extended beyond informing initial care planning, to help the medical team determine the level of longer-term rehabilitation and anticipated discharge destination, according to the severity of the stroke and available support structures. A similar pattern was observed at the Glipton stroke unit, but here it was also common for a representative from ESD to join the admitting team (if available), with the aim of identifying more quickly whether or not the patient was suitable for ESD. This was seen as avoiding potential delays in accessing ESD where decisions about patient suitability were primarily made at weekly MDTs rather than at admission.

In contrast, there was little attention to discharge planning during admission at the two orthopaedic units. Rather, patients were usually admitted on to the orthopaedic trauma units by a junior doctor and ward nurse. At this time, clinicians, patients and relatives would interact to determine the cause of the hip fracture, comorbidities and pre-existing health-care interventions. This information would not be explicitly used to inform discharge planning; rather, its primary purpose was to inform surgical decision-making. For example, it was customary for the junior doctor to present each patient at the next orthopaedic trauma meeting, thereby enabling the surgical consultant to determine and schedule surgery. These activities showed almost no attention to issues of discharge, with the primary emphasis on surgical decision-making. At Glipton, it was usual for an orthogeriatrician to meet with the patient prior to surgery. Through a short bedside consultation they would gather information related to the patient’s general health and social circumstances to inform postsurgical care planning, such as palliative care decisions, but there was no clear evidence that this led to early discharge planning.

Ward rounds

In general, the ward round enables the medical team to systematically review the care of each admitted patient through assessing their current health state and recovery; determining ongoing care needs; deciding upon necessary tests or therapies; allocating tasks to the wider care team; and supporting learning. In the course of these routine activities, the ward round represented a significant opportunity for progressing patient care towards discharge, i.e. facilitating the patient’s recovery; allocating tasks directly related to discharge, i.e. ordering TTOs; and, importantly, determining the patient’s readiness or ‘fitness’ for discharge. In other words, the ward round was a major knowledge-sharing and decision-making opportunity in the routine organisation of care.

At the Farnchester stroke unit, ward rounds were led by the lead consultant and involved the participation of junior doctors, the patient’s designated nurse and OT, a ward-based pharmacist and, significantly, patient relatives (if the patient agreed to their involvement). In addition to assessing and reviewing each patient’s ongoing care needs, it was usual for issues of patient discharge and onward care to be explicit (as a standing item) at each ward round. This largely centred on determining the patient’s likely date of discharge and, as this approached, the necessary tasks and activities required to ensure their onward care. This involved a relatively open exchange of knowledge whereby the consultant would lead a process of question and answer with the expectation that others would contribute according to their knowledge and practice domains. For example, the nurse would comment on the patient’s overall well-being and recovery, the OT on his or her rehabilitation, the pharmacists on the patient’s medicines regime, and relatives on their expectations for recovery. Although this pattern of exchange supported a combined and shared understanding of patient recovery, the primary purpose of the ward round still seemed to be to inform medical decision-making. As such, the consultant would use this information to lead decision-making and allocate tasks, i.e. he or she was the primary user of the knowledge.

A similar arrangement was observed at the Glipton stroke unit, with medical-led ward rounds used on a daily basis to inform overall care planning, but also the planning of care towards discharge. However, it was not usual for relatives to be in attendance, and as such there were fewer opportunities to include information about the patient’s personal circumstances. It was more usual for a speech and language therapist (SaLT) and a dedicated discharge liaison nurse to participate in the ward rounds, thereby providing additional insight. The discharge liaison nurse played a particularly important role, being responsible for recording and co-ordinating agreed tasks related to progressing hospital discharge, as well as providing daily feedback on their completion.

At the Glipton orthopaedic unit, the ward round was structured differently, being split between ‘surgical’ and ‘orthogeriatric’ rounds. Following surgery, each patient’s surgeon, junior doctors and designated ward nurse would conduct a relatively standard daily ward round. Significantly, these tended to focus on surgical recovery, for example wound care, infection and mobilisation, with little or no consideration of onward care or discharge. As the patient recovered from the surgical procedure, the junior doctor, under the supervision of the consultant surgeon, would declare the patient as ‘surgically’ fit. This, however, did not necessarily mean the patient was ready for discharge; rather, this would trigger non-surgical clinical and therapeutic interventions with the explicit aim of preparing the patient for care transition. In line with this approach, each patient received a distinct daily ward round from an additional team made up of an orthogeriatrician, PT and OT. This second team would be tasked with supporting physical recovery and rehabilitation towards an estimated date of discharge, often with reference to meeting the unit’s prescribed fractured neck of femur (NoF) pathway. These additional ward visits provided a more discharge-centred set of daily interactions which would be addressed in weekly MDT meetings, where more formal discharge planning decisions would be reached.

In contrast to all other sites, the arrangements for ward rounds at Farnchester orthopaedic unit were even more fragmented, with no clear structure or pattern of daily ward rounds. Instead, patients were visited intermittently by the surgical consultants according to their ‘clinical days’, with no clear fixed schedule. As such, these meetings provided little opportunity for knowledge sharing with or to the wider clinical team. Instead, the surgeon would usually interact with nursing staff during daily board rounds and conduct relatively ad hoc ward rounds.

Ward-based interactions, board rounds and handovers

In the course of their inpatient stay, both stroke and hip fracture patients experienced a wide range of bedside interactions with clinicians, therapists and health-care assistants (HCAs). These were integral to the general provision of inpatient care, including personal care tasks such as feeding, washing or toileting; the completion of routine observations and assessment; dispensing of medicines; and the delivery of specialised rehabilitation or therapeutic interventions. These routine daily interactions were important because they provided a unique opportunity for clinicians to build rapport and trust with patients through the often intimate processes of direct caregiving. In doing so, clinicians and patients often become able to share more tacit or sensitive information about patient well-being, mental health and expectations for the future that might often be overlooked or difficult to articulate in a more formal setting, such as the ward round. This knowledge would then be relayed, in a way that protected patient confidences or trust, within these other, more formal meetings or situations, thereby enriching decision-making. Although this knowledge often seemed trivial or inconsequential, it could often be integral to developing more personalised or robust discharge plans, for example where the therapist could understand more about the patient’s fears of being left alone at home, or HCAs could identify problems with toileting.

Across all research sites, these interactions were common features of day-to-day care. It was also apparent, however, that where clinical work was more time pressured, or where there were staff shortages, these opportunities were dominated by functional task completion, with less potential to adopt a more holistic approach to patient care. Despite being widely common there were some important differences in the local organisation of care that also seemed to frame differences in these patterns of knowledge sharing. A relatively obvious finding related to the general sense of teamwork or integrated working on the ward. At the two stroke units, ward nurses, HCAs and therapists had a relatively strong and close working relationship. For example, they had co-ordinated working patterns, participated in daily team briefings, shared common office areas, and interacted regularly and routinely in the everyday patterns of hospital work, i.e. in reception areas, rest areas and corridors. More than this, their working patterns were overtly patient centred, with their co-ordinated activities primarily focused on advancing patient recovery, and the patient was often at the centre of their collective interactions. These patterns of daily working appeared to further facilitate rapid and timely knowledge sharing across occupational boundaries and led to more joined-up planning activities around hospital discharge.

At the Glipton orthopaedic unit, there were also close working practices between nurses, HCAs and therapists, but this also appeared to be framed by an implicit division of labour in ward-based care. As described above, in the initial days following surgery much of the patient’s care was centred on surgical recovery; as such, ward nurses were more attentive to issues of wound care and less so towards rehabilitation, and there was relatively little involvement of therapists until the patient was declared surgically fit. Therefore, there seemed to a staggered approach to integrated working. That being said, once therapists started work with patients, the pattern of interaction with ward nurses and HCAs did appear to become close and more integrated, like that observed in the stroke units.

The main difference was at the Farnchester orthopaedic unit, where there was little evidence of integrated or team-based working. In general, ward nurses appeared primarily concerned with the patient’s surgical recovery and then with managing the transition of care to other actors. Therapists at this site were not based on or colocated with the ward but rather worked across the entire hospital with an office based away from the orthopaedic wards. There was also little evidence that therapists and nurses tried to co-ordinate their work, except where therapists attended nurse-led handovers and MDT meetings; however, they did so primarily as observers. This led to a situation whereby nurses and therapists would develop close working relationships with their patients, but with few opportunities to share knowledge with each other. There was also found to be little stability or continuity in the provision of rehabilitation, with frequent changes in staffing making it difficult to build interprofessional relationships on the ward.

An important example of the differences in daily ward-based interaction could be seen with the organisation and use of daily handovers and board rounds. These were routine features of day-to-day care planning, and often helped to allocate and plan the types of daily activities described above. In these situations, ward-based clinicians would come together, usually at the start and end of the day, to identify, allocate and review tasks. These relatively formal meetings could have a particularly important role in co-ordinating and prioritising shared activities related to hospital discharge, such as arranging tests, completing assessments, initiating external referrals, ordering medicines or arranging transportation. At the Glipton orthopaedic unit, ward staff followed a traditional morning handover meeting. This was led by the shift ‘nurse in charge’ and was attended by other nurses, HCAs, an OT and a PT representative (who would feedback to their teams) and the discharge liaison nurse. The lead nurse would read from a printed sheet that detailed each patient’s overall care plan and outstanding tasks to be completed during the next shift. It was expected that each patient’s nurse and therapists would provide additional information to offer a more developed account of patient progress, and would also agree to relevant tasks and co-ordinate their activities. As such, these handover meetings became an important situation for sharing knowledge, especially at the beginning of the day, to help identify and manage issues related to care progression and transition.

Similar daily activities were observed at the two stroke units, but here knowledge sharing and decision-making was co-ordinated through the use of large wall-mounted whiteboards, hence ‘board round’. These boards provided an important knowledge-sharing resource for co-ordinating daily clinical activities. They displayed each patient’s basic details (name and date of birth), his or her location on the ward, and also other handwritten annotations or abbreviations, with the expectation that each professional group uses a different-coloured pen. It was also found that magnet tokens and symbols were used to denote certain procedures or tasks, such as ‘close observations’, ‘awaiting test results’, ‘ready for discharge’ or ‘collect TTOs’. These boards contributed to the co-ordination of day-to-day knowledge sharing as they provided an open, visible and real-time picture of ward-based work. In both stroke units, they were used to co-ordinate daily tasks, for example informing therapists that the patient was ready to start rehabilitation or for junior doctors to order prescriptions. They were both managed by the senior nurse in charge and updated by a wide range of other ward-based therapists and support workers. Although ward staff continued to use other records and documents to co-ordinate daily work, the whiteboard remained a centralised point of reference around which staff would regularly congregate to review progress and share knowledge.

At the Farnchester orthopaedic unit, board rounds were also a regular feature of daily care management. Given the absence of more customary ward rounds, the board round was used by the consultant surgeon to review patients with the lead nurse and, where possible, the patient’s designated nurse and junior doctors. As the consultant did not keep a regular schedule of ward visits it was not always feasible for other clinicians to leave other duties to attend the board round. In this interaction, the surgeon would usually request an update from the nursing staff about the patient’s surgical recovery and review patient progress in much the same way as a ward round, but in the absence of the patient. Notably, many nurses showed a reliance on scrap paper to note tasks rather than relying upon the whiteboard. The primary focus of these meetings was to assess postoperating wound care, patient stability and early mobilisation. A significant watershed in these meetings occurred when the surgeon declared that the patient was ready to start mobilisation, at which point the nursing staff would be asked to instruct the therapist team to assess and treat the patient, and the level of surgical input gradually decreased.

Multidisciplinary team meetings

A major knowledge-sharing situation across all research sites was the weekly MDT meeting. Where daily ward activities supported incremental care planning, the weekly MDT meeting was widely seen as the main decision-making forum, especially with regard to hospital discharge. These meetings facilitated a range of clinicians and stakeholders to review the care plan for each patient and, in doing so, to contribute to the general management of workflow and service management. In line with national policies, MDT meetings were found at all sites, but they also varied in terms of the way they were organised and how they contributed to discharge planning.

At the Farnchester orthopaedic unit, MDT meetings were led by the nurse in charge and were attended by two or three senior nurses from different ward bays, and representatives from the hospital’s OT and PT teams. It was customary for the junior doctors to attend the meeting, but in an observer capacity, and it was expected that the orthogeriatrician should attend, although this was intermittent owing to staff shortages. It was not expected that consultant staff should attend the meetings. Within the MDTs each patient’s surgical recovery was rapidly reviewed. The typical format was for the nurse in charge to introduce each patient according to his or her designated bed (not name or date of birth). This in itself caused confusion if the patient had been moved. The nurse in charge would then review the decisions made by the consultant surgeon, usually in relation to surgical recovery and readiness for discharge planning. It was expected that the ward nurses would offer additional input into outstanding tasks, but there was no expectation that therapists should contribute information, and when some did it was observed that the nurse in charge would quickly close off discussion. As such, these weekly meetings appeared to be dominated by managing nursing work in relation to surgical recovery, based upon the senior nurse issuing instruction for onward care planning. For example, it was usual for the lead nurse to inform the therapist team to start rehabilitation or to prepare the patient for care transition, but not to engage in shared decision-making.

In contrast, the other three sites had more developed and inclusive MDT meetings. At the Farnchester stroke unit, the MDT meeting was led by the lead consultant and was attended by other doctors, the nurse in charge and nurse representatives, the lead HCA, the ward manager, OT, PT and SaLT representatives, the ESD service lead, the community stroke nurse and representatives from the local Stroke Association. This provided a broad spectrum of contributions to care planning. Each patient was introduced with a detailed medical summary and summary of the existing care plan. This introduction alone was normally longer in duration than the entire discussion for each patient in the same hospital’s orthopaedic unit. The ward nurse and therapists would be invited to add to the medical account, for example describing the patient’s psychological and emotional state, and his or her general health and level of rehabilitation. In the early stages of the admission, these discussions tended to focus on initial recovery, and included direct reference to the patient’s suitability for ESD. At this point the ESD representative would talk through his or her assessment. Even at this early stage, the MDT meeting would consider patient discharge and set precise goals for therapist activity. As patients neared the point of discharge, the focus of decision-making would turn to the wider personal and domestic circumstances of the patient, drawing upon the knowledge acquired through day-to-day ward-based interactions. This would include attention to the need for additional therapy or rehabilitation in a nursing home, or whether or not the patient could be cared for at home with suitable support. It was usual for both the community stroke nurse and the Stroke Association representative to offer a view on the level of follow-up care. This input seemed important because hospital-based staff had limited appreciation of the available support services in the community. An important contribution was made by the community stroke nurse who provided detailed knowledge on the range of available services, housing issues and issues of accessibility related to the need for home adaptations. A significant feature of these MDT meetings was that all information and decisions were recorded in a single patient record, rather than in separate files, which resulted in a collective record rather than a medical or nurse record.

A similar pattern of interaction was observed at the Glipton stroke unit. However, two interesting features made it different from the Farnchester unit. First, the meeting was not led by the consultant physician; rather, it was expected that a different member of ward staff would lead the discussion of different patients. The consultant had introduced this format in the hope that it would foster greater appreciation among the team of the contribution of different actors and encourage shared ownership of clinical decision-making. Following this format, each patient was introduced by his or her designated spokesperson, who would review the patient’s history and care plan before similarly inviting contributions from other specialists. A further difference at these meetings was the role played by the discharge liaison nurse, who would provide specialist problem-solving expertise for each patient’s discharge plan. For example, he or she would often clarify legal or housing issues. Although the tasks required for care transition would be delegated to specific individuals, such as ward nurses and OTs, the discharge nurse would also act as a key reference source, for example advising on the availability of social care providers in certain areas or whether or not a suitable nursing home or community hospital was available. A similar issue addressed in these meetings was whether or not patients with mild strokes were suitable for ESD, and whether or not those with more severe strokes needed to be transferred to a community hospital or specialist rehabilitation ward in the hospital. However, as an ESD representative was usually present at the point of admission, the involvement of this representative at meetings normally centred on confirming the ESD service’s proposed care plan.

A similar format was found at the Glipton orthopaedic unit’s weekly MDT meeting, where a discharge liaison nurse was also active in discharge planning. However, there were other subtle differences in the composition and format of these meetings. For example, the meeting was usually led by the nurse in charge, who similarly reviewed each case and invited contributions from other actors. Interestingly, this nurse tended to focus her knowledge sharing on matters related to supporting the patient’s ward-based (postsurgical) recovery, whereas the patient’s readiness for discharge and wider health state tended to be discussed by the orthogeriatrician and the discharge nurse. As such, a tight subgroup was evident that seemed to lead decision-making and discharge planning. For example, it was usual for the discharge nurse to provide recommendations of the proposed discharge destination and the activities needed to facilitate transition, whereas the orthogeriatrician was responsible for determining medical suitability for discharge. An important variation at this site was the explicit use of a discharge checklist, which was held by the discharge nurse to list key activities required to progress discharge. This was regularly used within the meetings to progress unfulfilled activities, and outside the meetings to enable the discharge nurse to identify those staff members who might need assistance. There was some concern from participants that the emphasis within these MDT meetings was on determining the earliest possible opportunity for discharge so as to meet performance targets, i.e. percentage of patients meeting the 10-day pathway, and to free up acute care beds.

Internal transfers

Although the primary focus of the study was to investigate the transition of care to a community destination, the frailer patients with complex care needs were often transferred to another hospital unit or ward where their other comorbidities would be attended to before discharge. These transfers of care did not necessarily relate to the recovery from either stroke or hip fracture, but rather to wider health issues that prevented ongoing recovery in the community. This process was more of an intermediate step or diversion of care, but it involved a particular type of interface with internal rather than external agencies. This type of transfer was more common at Farnchester orthopaedic unit and Glipton stroke unit.

At Farnchester, internal transfers usually occurred where patients with complex care needs were seen as ‘problem cases’ and outside the normal remit of care provided on the orthopaedic unit. This transfer process was normally triggered when the patient’s surgeon declared the hip surgery successful but the orthogeriatrician believed the patient remained too frail to be discharged. These decisions were usually communicated via the senior nurse, i.e. in the weekly MDT meeting. The orthogeriatrician would also consult with the patient’s OT and often family members before recommending that the patient be transferred to the care of the elderly ward. In a daily board round, the senior nurse would then delegate responsibility for organising the transfer to the patient’s designated nurse, who would contact a relevant care-of-the-elderly or general medical ward by telephone to request a bed based upon a summary of the patient’s care needs. When a bed space was confirmed, the patient would be transferred by porter to the new ward, where the patient’s nurse would provide a verbal handover or summary of the nursing record, noting any specific health-related issues. The receiving nurse would then reclerk the patient on the new ward and prepare the patient for the next available ward round. It was recorded that the patient’s therapist team would not normally be informed of this transfer, which meant they would often need to search for their patient across different wards. Once transferred, there was also considerable duplication of assessment and care planning, for example repeated tests and consultations with family members. The patient’s discharge planning was therefore transferred to the other ward where an entirely separate planning process was involved.

At the Glipton stroke unit, patients with complex recovery needs would often be transferred to a specialist long-stay rehabilitation ward. This decision was made at the weekly MDT meeting through discussion between the consultant and the patient’s OT, where it was acknowledged that further hospital-based rehabilitation was needed before transition to a community setting. The transfer of care was usually initiated by the patient’s OT, who would telephone the rehabilitation ward to initiate a transfer. At this time, they would give a detailed summary of patient condition and the existing rehabilitation plan. During the transfer of care the OT would also accompany the patient to the new ward and conduct a more thorough handover to the new rehabilitation team. For example, the patient’s existing therapists would introduce the patient directly to the new nursing and therapy staff, assuring them that they would be well looked after. It was also common for the new rehabilitation team to telephone the stroke unit to enquire about additional issues that might be detected later in their work. Despite being based on different wards and having a different emphasis in their care plans, the two stroke care teams work as part of a larger combined service. As such, there was a greater sense of collaboration and integration across the entire service, whereby knowledge and know-how gathered in the acute care unit would be shared and made available to those in the rehabilitation ward.

Family meetings

Across all sites, family members, friends and carers made a significant contribution to diagnosis, clinical decision-making and longer-term care planning. They had a unique insight into the patient’s wider circumstances and often acted as mediators or translators between the patient and the wider clinical team, especially if the patient did not speak English or where there were pre-existing speech and cognitive difficulties. With regards to planning a patient discharge and onward care, family members were often invited by ward nurses and therapists to help assess the patient’s ability to cope at home, providing specific details about his or her domestic, personal and financial circumstances, such as layout of the home, eligibility for benefits and availability of support networks. It was also the case that at the point of discharge, family members were integral to supporting care transition through enabling carers to access the home, accompanying their relative from the hospital and supporting him or her to attend outpatient appointments. Those patients without relatives or carers not only had limited options for ongoing care, but their discharge arrangements were often underspecified, subsequently requiring urgent home visits from social workers or OTs on discharge. Equally, it was often reported that where family members were absent or unco-operative it could be difficult to transfer patient care. Illustrating the role of family members, visiting hours were commonly used by the patient’s nurse or therapist to speak directly to relatives. Here he or she would explain the patient’s care arrangements and often discuss postdischarge care needs. At the same time, he or she would actively gather additional information about the patient’s wider personal circumstances to enrich care planning. A significant pattern of knowledge sharing at all sites took place between the patient, family member and OT. The OTs often focused more explicitly on post-hospital care, discussing the patient’s rehabilitation plan, ongoing care needs and the ways in which relatives could support recovery outside hospital. In particular, OTs would often work with family members to explain the use of certain equipment or therapeutic exercises. The OT would normally ask additional questions about the patient’s home layout, number of steps, toileting arrangements and bedroom options to develop the rehabilitation plan.

An important variation was found at the Farnchester stroke unit, where the lead consultant had introduced a more family-centred approach to care planning. For example, family members were invited to be present during ward rounds and the consultant would also be available during visiting hours to speak with family members. Also at this site, a formal family meeting had been introduced to enable more structured knowledge sharing and shared decision-making with relatives and/or carers. These were usually scheduled when the patient’s recovery was underway with the express purpose of supporting discharge planning and communicating ongoing care needs. These were also opportunities for giving and receiving information among clinical staff and family members. Family meetings were usually initiated by the consultant or ward nurse, who would give a short, non-technical account of the patient’s condition and care before asking if family members or carers had any questions. The meeting was used by the clinical team to further clarify the information they had gathered to date and to understand more about the patient’s domestic circumstances. Prominent topics of discussion related to the patient’s general health condition prior to the stroke; his or her home arrangements and layout, including toilets and stairs; his or her work expectations or general activities; and the availability of relatives or carers to support recovery. This interaction provided additional knowledge about the patient’s wider circumstances which helped to determine the most appropriate care pathway given his or her home circumstances, i.e. nursing or home-based care, and enabled family members to better understand the responsibilities and tasks they might need to address after discharge.

Referrals and assessment for social care

For the majority of stroke and hip fracture patients, the discharge planning process involved referral to the local authority social services department to initiate an assessment for social care following care transition. In line with the CCDDA 2003, hospital staff initiated the referral to social services by completing a Section 2 request (notifying local social services that a patient requires assessment for social care) and a Section 5 request (confirming that the patient is ready for transfer of care). The Section 2 request would trigger social care assessment, usually undertaken by a local authority social worker who would organise and commission a package of social care according to the patient’s assessed needs, home circumstances and financial arrangements. The Section 5 request is significant because it relates to current regulations which stipulate that a social care package should be in place and ready to receive the patient 24 hours from the notification; delays in the provision of this care can result in financial reimbursement to the hospital. Despite awareness of these broad parameters among service leaders and discharge co-ordinators, it was often the case that ward nurses and therapists were not altogether clear about this process, which might explain the frequent complications and errors in completing these referrals.

At the Farnchester orthopaedic unit, social care referrals were usually initiated by the patient’s OT, following the weekly MDT meeting where the nurse in charge would inform the OT that the patient was approaching discharge. The referral would be completed in the therapist department (not on the ward), where the OT would review the patient’s therapist record and request, where necessary, additional information from the patient’s nurse. The Section 2 referral was faxed to the local authority social care contact centre, which acted as a centralised referral point to validate and process social care referrals for the county. Where a patient with more complex health or personal circumstances was being discharged, the OT might contact the hospital’s discharge liaison service, which could provide specific guidance on legal and financial assessment criteria and the types of services available to different patient groups. The referral would then be reviewed by the social care contact centre, at which point referrals could be returned owing to missing or inaccurate information. Where accurate, the referral would be forwarded to the patient’s locality social work team, determined by his or her place of residence.

A similar process of social care referral was observed at the Glipton stroke unit, where the patient’s OT also took responsibility for making referrals to a similarly centralised social care contact centre. At this site, referrals would usually be made on the ward, because stroke therapists were colocated on the stroke unit. This had the advantage of providing easier access to nursing and medical records and information detailed on the whiteboard, and also for making direct enquiries with nursing and medical staff. In comparison with Farnchester, this seemed to enable more detailed and accurate referrals, where multiple information sources were combined and included into a single referral. It was also found that the unit’s discharge liaison nurse would routinely provide additional support and guidance, especially for more novice staff, and take overall responsibility for more complex patient discharges. A further difference was that the referral form would be telephoned through to the Glipton contact centre, often with a fax sent as a basis of confirmation. This direct contact made it more possible to address any potential inaccuracies and missing information, thereby reducing delays, but also depended upon referrals being made during office hours.

At the Glipton orthopaedic unit and the Farnchester stroke unit, social care (and health-care) referrals were completed by the patient’s designated ward nurse, during normal working hours. The process broadly followed the above interaction, with referrals made in both cases by fax to the social services contact centre. Although the patient’s nurses often had a good understanding of the patient’s care plan, they did not always have details of his or her current rehabilitation plan and general home circumstances. As such, ward nurses at both sites usually worked closely with the patient’s OT to draw together information to complete the referral. This normally occurred after the daily handover meeting at Glipton or the board round at Farnchester, when clinicians were usually available on the ward. A significant problem faced by nurses at both sites related to the difficulties of completing social care (and health-care) referrals alongside their other clinical duties. It was observed that nurses would set aside time each day to complete patient referrals and other administrative tasks, but they were often interrupted or required to give direct patient care. As such, many referrals were completed in a fragmented and disjointed way which seemed to reduce the coherence and timeliness of some referrals. At the Glipton orthopaedic unit, the dedicated discharge liaison nurse would often step in to assist nurses at this time by checking information and completing documentation. It was also usual for the discharge nurse to assume general responsibility for discharges during busy periods or where there were staff shortages, and always for more complex discharges that required more engaged knowledge sharing and negotiation with both social and health-care agencies. Significantly, where both health and social care referrals were made by the same person, i.e. the patient’s nurse, there was less duplication of work and less need to pass on or share patient records. A further observation from the Farnchester orthopaedic unit was that Section 2 and Section 5 referrals were made simultaneously. It was described how delays in the time between requesting social care assessment and initiating social care transfer meant that the unit often over-ran its 10-day care pathway, and by submitting both referrals together it placed pressures on the locality social work teams to respond quickly.

For all sites, social care assessments were completed by a locality social worker selected according to the patient’s place of residence. Previously, social work teams were based in each hospital, not in the community, which participants described as making it easier to complete assessments (i.e. no need to travel to the hospital), and as fostering closer working between hospital and social work teams. With recent local authority changes, social workers were no longer located in hospital but instead in locality teams, which many participants believed had degraded the relationship between health and social care. At both stroke units and the Farnchester orthopaedic unit, social care assessments were completed through direct interaction between the social worker, the patient and family members at the patient’s bedside. The social worker would usually spend time speaking with the patient to complete the necessary assessment documentation, and also with the patient’s designated nurse. At both stroke units the patient’s OT would also join these discussions to share additional insights or to prompt the patient, as they tended to have a more detailed and intimate knowledge of the patient’s personal or home circumstances. This was possible because they were colocated on the stroke unit. At Farnchester orthopaedic unit this was more difficult as therapists visited hospital wards on scheduled rotation. As such, there were often some discrepancies in the breadth of information used to inform the social care assessment.

In contrast, social care assessments at the Glipton orthopaedic unit were completed by telephone. Here it was usual for the social worker to first speak to the patient’s designated nurse or nurse in charge to check the basic patient details and then to speak to the patient. Although this made for a relatively speedy referral, care packages were often basic in design and requiring modification once the patient was transferred to the community setting. Developing this observation, the focus of the social care referral seemed to be on completing referrals quickly, rather than thoroughly, in order to avoid financial penalties. At all sites, the social work team would complete their preliminary assessment and invariably organised a package of 6-week reablement to address the immediate personal and social care needs following discharge, or a short-term placement in a residential home for those with more complex needs. In most cases, it was expected that a further social work assessment would be completed after discharge to determine the patient’s longer-term care needs. As such, the social care referral and assessment process seemed predominately concerned with facilitating a short-term or transitional package with the expectation that further and more detailed assessments would be completed later.

Referrals and assessment for health care

For ongoing health and community-based care a similar process of referral was undertaken, but this could vary according to the configuration of local health-care services, such as the availability of intermediate care beds or community hospitals. For both hospitals, intermediate or step-down services were funded by the local PCT (now CCG) and provided by a local NHS community trust or private sector provider. This often involved a fixed package of 30-day care for specialist rehabilitation provided in a community hospital or a nursing home with dedicated rehabilitation beds. For patients with more complex needs and comorbidities, it was also possible to make a referral for continuing health care at home or end-of-life care.

Across all sites referrals for ongoing health care were completed by the patient’s designated nurses. At the Farnchester site, this involved an initial fax request to the community rehabilitation team, colocated on the hospital site. This colocation made assessment at the patient’s bedside relatively easy, with a community nurse or OT attending to the patient within 24 hours of referral. Being colocated also fostered a close working relationship between acute and community staff and, despite some tension points around workload and timing, the interaction between ward and community staff was based on shared knowledge and understanding of their respective working practices. At the Farnchester stroke unit, fax referrals were usually made mid-morning after the morning board and ward rounds. The rehabilitation team would receive the fax and contact the ward in person or by telephone, with a view to determining the most appropriate rehabilitation care package, typically within a privately run nursing home. The community rehabilitation team would then assume responsibility for liaising with the local nursing homes to arrange an appropriate bed, thereby reducing the demand on ward staff. In contrast, health-care (and some social service) referrals from the Farnchester orthopaedic unit were made during the night shift, on the assumption that there would be less demand for direct patient care and more opportunity for completing routine administration. Although this appeared to be an effective use of nursing time, it was often difficult for nurses to obtain all the necessary information to complete some referrals, because they had no access to the therapist records and were not able to speak with family members. As such, many referrals remained incomplete and were passed on to the day shift. It was also found that many faxed referrals were lost or incomplete because of problems using the fax machine. This was also exacerbated by the lack of administrative support at night (e.g. no ward clerk) and also the inability to contact the community rehabilitation team to check that they had received the referral. As such, many referrals were delayed while the day shift and community rehabilitation service interacted by telephone to remedy problems experienced at night.

At the Glipton site, referrals for continued health care were usually made to one of the local community hospitals or on occasion to nursing/residential homes with rehabilitation beds (usually because of patient residence). Where it was felt that no rehabilitation was possible, patients would be referred directly to nursing/residential homes for longer-term places. Community hospitals provided an extension to both stroke and hip fracture care, having dedicated rehabilitation units for both patient groups. Referrals to the community hospitals were usually made by the patient’s nurse (or discharge liaison nurse), who would book a bed by telephone, providing a brief overview of the patient’s condition, recovery and ongoing care needs. A ward nurse would receive and log the request, which would be reviewed by the nurse in charge or ward manager to assess bed availability and level of care. A similar process would be undertaken with nursing/residential homes, but here it was usual for the discharge liaison nurses to negotiate with care home managers about bed availability and care plans. This would often involve lengthy discussions about clinical risk and patient need before agreeing on a placement; for example, it was usual for nursing homes to turn away patients if they were too sick.

Requests for home adaptations and equipment

An additional external referral related to the ordering of home adaptations or equipment for postdischarge recovery and rehabilitation. This included, for example, commodes, hand/grab rails, stair lifts or ramps, bedding and mattresses, wheelchairs or frames, and tables, chairs and other assistive devices. Across all sites, the general expectation was for the patient’s OT to order this equipment, usually when the discharge date had been confirmed and the level of need assessed. The ordering process varied, however, between sites. At Glipton an electronic (e-mail-based) ordering system was used, whereby the OT would complete an online order form. At Farnchester, the OT would contact the equipment supplier relevant to the patient’s residence via telephone and place an order. Although there were no significant differences between these approaches, telephone-based orders had the benefit of enabling staff to provide more detailed responses to queries, but these also seemed to take longer and relied upon the availability of staff in the supplier agencies.

A number of additional issues seemed to have an impact upon the ordering of equipment. The first related to equipment availability, where it seemed that certain more complex or specialist items, such as a type of brace or mattress, would not be in stock. This would create delays in patient transfer or lead to a situation where the patient returned home without equipment. The second related to whether or not suppliers and fitters had access to the patient’s house to fit certain grab rails or ramps. This therefore involved an additional line of communication with family and relatives to arrange a home visit. The third related to the level of funding for home adaptation and whether certain changes should be funded by health or social care. Again, a prominent example was grab rails and ramps; these were often seen as social care issues, but the delays in assessing patients for financial eligibility often meant that health-care workers ordered the equipment without confirming the details. Fourth, at both sites there had been a recent change of supplier, with new IT systems that were uniformly described as undergoing teething problems. Staff at both sites were learning to deal with new processes for ordering and frequently identified that equipment was currently delayed. Finally, there was fragmentation in the process of ordering equipment, with some responsibility for authorising equipment orders residing with the OT and some with the district nurse (particularly for pressure mattresses). This took significant co-ordination and was identified as a point of daily difficulty in preparing patients’ home environment for discharge.

Day of discharge

In general, the day of discharge involves a frantic series of linked activities that escalate in pace and intensity as care is transitioned. Across all sites, this usually involved the following activities:

• At the start of the day, the patient’s nurse would confirm whether or not the patient was still deemed medically fit and ready for discharge, usually by enquiring with the consultant or junior doctor as a part of the morning ward round. The ward round would also be used to confirm the patient’s take-home medicines, which would then be ordered by the junior doctor for collection or delivery.

• The patient would then be reviewed as a part of the morning board round (handover) where further tasks would be described and allocated to the patient’s nurse, HCAs and the ward clerk.

• Following this, the ward nurse would contact the locality social work or health-care provider to confirm that the care package was in place and to give a final verbal handover of the patient’s current health state, medicines and ongoing care needs.

• The ward clerk would order transportation from the regional ambulance trust, or possibly a taxi, requesting an appropriate time slot. When confirmed, the ward clerk would contact the patient’s family to notify them of the transfer time. The ward clerk or nurse would also follow up on the patient’s medicines and have them ready along with any other personal items.

• Before he or she was collected by the transport service, the HCA and nursing team would prepare the patient by removing any catheters or monitoring equipment, completing the patient’s observations and health checks and, where used, completing a discharge checklist with the patient.

• If the patient was mobile, he or she would be moved to the waiting area, but more often patients would remain near or on their beds. Both hospitals had discharge lounges for patients to wait for transport, but these were only used for patients not requiring regular supervision or care. At around this time the nurse in charge would contact the bed bureau or manager to notify them that a bed would be available later in the day.

• The patient’s nurse would complete and sign off the nursing care documentation and request that discharge letters be issued by the ward clerk.

• Upon arrival of transportation, the patient would be taken to his or her community destination.

Within this common pattern of activity, a number of important variations were detected. First, there were variations in the extent of consultant or medical involvement. At both orthopaedic sites there was relatively limited interaction with the designated consultant because they did not operate a usual ward round (Farnchester) or because the orthogeriatricians operated a variable pattern of ward visits. As such, there was greater reliance on junior doctors to confirm patient readiness for discharge. In instances where a junior doctor was uncertain about the patient’s condition, i.e. where a new blood test showed some anomaly or the patient was reporting new symptoms, it was customary for discharge arrangements to be delayed until the consultant could be contacted. The alternative was that the patient was discharged with potential health issues remaining and requiring further medical attention in the community.

Second, there were frequent problems with the timely ordering, dispensing and completion of TTOs. Across all sites these were completed by junior doctors with varying levels of medical supervision. At the Glipton stroke unit, for example, the stroke consultant closely supervised and monitored these processes, leading to more accurate and timely procedures. In other sites, we found nurses and ward clerks were often required to ‘chase up’ the junior doctors to complete the prescriptions. At Farnchester this was made relatively easier as the hospital utilised an electronic prescribing system that communicated prescriptions to the pharmacy, which then dispensed medicines via an internal delivery service. At the stroke unit, this was further supported by two additional variations: an expectation that all medicines would be ordered the day before anticipated discharge so that medicines were ready in the morning, and the availability of a pharmacist on the stroke unit, meaning that TTOs could be checked and amended before leaving hospital. That being said, it was not uncommon for patients to leave hospital without the complete range of medicines or with prescriptions to be completed in a community pharmacy.

Third, it was also found that in the processes of progressing patient discharge a number of tests and procedures were reallocated to be undertaken on an outpatient basis. For example, the patient would be asked to visit his or her GP in the subsequent 48 hours to request a particular blood test, and would be given a letter of instruction for the GP. Alternatively, he or she would be requested to return to the hospital over the next week for further tests. Although this was not found to cause any significant health-related problems, it did appear to transfer the burden of care to both the community sector and the patients themselves. This was especially problematic for patients living in rural areas and some distance from the hospital, who would need to organise taxis or family assistance to return to the hospital for follow-up tests.

Fourth, an important activity was the organisation and scheduling of ambulance services. A single ambulance provider operated across both areas and utilised a standardised booking system for patient transfers. Across all sites it was usual for the ward clerk to complete this booking by providing each patient’s details and destination, general care needs, specific requirements with regards to mobility issues (for example hip precaution) and the desired time for transfer. The booking clerk would enter this information into a computerised system to identify a suitable time slot for the patient reflecting his or her condition and locality. In the vast majority of patient transfers this system tended to work, but with regular delays in collection and drop-off due to time slippages. As such, patients tended to be collected on time earlier in the day and with delay later in the day. This created particular complications for community hospitals, especially for Glipton orthopaedic patients; here it was reported that patients might not arrive until the evening, at which point there were limited staff to fully admit the patient. As such, the patient would be stabilised on the ward but no detailed care could be initiated.

Fifth, there were variations in the involvement of the patient’s family at the point of discharge. At both stroke sites, it was an expectation of the medical team that relatives be closely involved in helping to prepare and ready the patient on the day of discharge, at which time the opportunity would be taken to provide further information or education to relatives about ongoing care needs. At the orthopaedic sites, however, there was relatively limited family involvement, other than a telephone call to provide notification of the time of transfer. A major concern among medical staff related to the difficulties of contacting family members, especially where they were contacted at the last minute or not given sufficient warning to prepare their relative’s home, and therefore excluded from supporting the transition process.

Finally, there were differences in how discharge documentation and letters were distributed at the point of handover. At all sites, the ward clerk would usually collate the necessary information to complete a generic discharge summary to be sent to the patient’s GP. At Glipton this issued electronically, together with a summary of the changes in the patient’s medicines and other diagnostic test results. In Farnchester, this remained a paper-based summary sheet sent via post. As such, GPs in the Glipton area had a more comprehensive and immediate overview of their patient’s care and transition. In addition, the consultants at each site also reported writing an additional letter to each GP (dictated to and typed by their secretary), which provided a separate medical overview of the patient’s care. For transfers to community hospitals or nursing care homes, the main handover was by telephone between the ward nurse and the service provider, but it was also expected that a similar summary record be issued with the patient and taken with the transport service. This record would be completed by the patient’s nurse providing further details about ongoing care needs and medicines regime. Those in the community hospitals often reported that the quality of these summaries was poor and that they provided limited information about ongoing care needs, requiring additional telephone contact with the acute hospital. As noted previously, however, where this involved an evening transfer this contact could not be made.

Returning home with early supported discharge

Early supported discharge provides a comprehensive package of home-based rehabilitation and support for patients with mild to moderate strokes. ESD is provided by a dedicated team of therapists, nurses, HCAs and social care workers who together provide a range of services to support patient recovery outside hospital and to prepare patients for a return to the activities of daily living, including employment. As the title suggests, it is initiated relatively early within the patient’s care pathway because of the reduced need for acute medical care. As described above, the decision to refer a patient for ESD is typically made within the weekly MDT meeting and involves interaction between the medical team and an ESD representative.

When a patient was transferred home with the ESD team it was expected that the local ESD representative would have devised an appropriate care plan that provided a degree of continuity with hospital-based rehabilitation. Although this package of care can be relatively generic, ESD leads attempted to tailor services to individual needs, for example if a patient required speech and language support or had an employment-related issue. Across both stroke research sites, the broad process of initiating ESD care involved a particular pattern of knowledge sharing between the patient and family, ESD therapists, community nurses and HCAs, and other support groups, especially the Stroke Association. It was typically the case that few community-based professionals had had contact with the patient or relatives prior to discharge and a significant proportion of early interactions involved building trust and developing a shared understanding through exchanging both health-related and more personal information. This would then help nuance and refine the more generic care plan. It was also found that there was limited knowledge sharing between the stroke unit and the ESD team in terms of formal records. Instead, there was greater reliance upon the community stroke nurse (Farnchester) and the discharge liaison nurse (Glipton) in facilitating more personal and direct knowledge sharing between the acute and community sectors based upon their well-developed working relationships with both groups. These actors not only supported the continuity of care or assisted with problem-solving but helped to establish a broader shared understanding, or at least a set of operating expectations, between hospital and community, based upon providing feedback and sharing lessons.

It was also found that both ESD services utilised a relatively standardised and centralised patient record that captured the information sources, decisions and activities of multiple professional groups in a single knowledge resource. This was particularly important for supporting the continuity of care where the roster of therapists and community staff changed. In general terms, there were few reported concerns with either ESD service, and patients found them both to provide a supportive and patient-centred package of care that supported early recovery from stroke. A key consideration was the ability of ESD representatives to accurately determine patient suitability for ESD and follow a relatively rigid assessment procedure in order to reduce potential risks, whereby the patient required more extensive hospital care. Again, this decision-making process seemed to benefit from the strong connections between ESD and stroke units earlier in the care pathway.

Returning home with social care reablement

For both stroke and hip fracture patients, the provision of local authority reablement care was a common occurrence. Reablement usually involves a 6-week package of general personal and social care that supports adjustment to home life and activities of daily living, with the expectation that more therapeutic or rehabilitation activities are provided on an outpatient basis. This included, for example, help with dressing and toileting, domestic chores, shopping, the use of braces or other devices and some rehabilitation. A feature of all reablement services was that they were designed to provide an initial period of intensive and close personal support with the expectation that this would reduce in line with patient recovery and with the patient taking more responsibility for their own care. Two factors regularly seemed to influence this model. The first was where patient recovery and independence did not increase in line with the reduced need for reablement care, because of more complex social and personal care needs. The second was where family members were not available or forthcoming in providing additional care both during and after this initial 6-week period. These circumstances would require additional social work assessment of patient need and financial eligibility for ongoing social care.

For both Farnchester and Glipton localities, the provision of reablement was well received and seen as well organised by patients and families. The provision of reablement relied upon an initial period of intense knowledge sharing and decision-making between the patient, family members and reablement team (usually the team leader) immediately after the patient arrived home. It was common for the team leader to review with the patient the package of care that had been scheduled and authorised by his or her social worker. This schedule of care usually took the form of a three- to four-side document providing basic personal information, an account of the patient’s recent inpatient treatment, the outcome of the social care assessment and details about his or her ongoing care need. It was rare for the reablement service to have access to additional information. Through the interaction with the patient it was often found that the proposed care package was poorly specified or inaccurate. For example, for a number of patients the proposed care plan provided limited information about their complex personal care needs, such as the use of a hip brace or feeding aids, which required specialist skills among the reablement team. Given deficiencies in the proposed care plan, it was usual for the reablement team to initiate a more detailed assessment through close interaction with the patient and his or her family members, and also through assessing the home environment and other domestic arrangements. This pattern of knowledge sharing was therefore important for ensuring patients had a more bespoke package of care that fitted their recovery needs, and which was then documented more thoroughly and reviewed daily by each reablement team.

Given the requirement to undertake a thorough review of patient needs when commencing reablement, an issue that occasionally appeared to complicate care transition was discharge from hospital on a Friday or over the weekend. As was common across all research sites, there was a general desire to discharge patients towards the end of the working week, especially after the mid-week MDT meeting. This could sometimes place pressure on social care agencies as they were expected to process and manage new patients before the weekend. The consequence of this during busy times was that team leaders would struggle to complete a thorough patient assessment and the initial package of care provided, over the weekend, would be generic and not tailored to individual care needs. This led to the possibility that staffing levels or expertise would not always be to the level required by patients with complex needs.

A more common issue experienced by patients returning home, and often managed by the reablement service, related to problems associated with the delivery, fitting and use of home adaptions and other equipment. As outlined above, OTs would often order specialist equipment to facilitate home rehabilitation but there were frequent problems with availability of these in advance of the patient’s transition. Some prominent examples included missing commodes, beds, grab rails and hoists. It often fell upon the reablement team, family or social worker to chase up these items and arrange time for them to be fitted, while patient recovery would often be limited in their absence.

Transfer to a community hospital

At Glipton hospital it was usual for more frail patients and those needing longer-term rehabilitation to be transferred to the community hospital for intermediate care. Upon receiving the patient at the community hospital, ward staff in the respective stroke and orthopaedic units would seek to clerk the patient, admit them to the ward and devise an appropriate care plan. However, at the point of transition there had usually been only limited knowledge sharing between acute and community hospitals. As described above, this took the form of an initial telephone enquiry about patient need and bed availability, and a follow-up call to conduct a telephone handover of the patient happened at a later stage. The patient was also transferred to the community hospital with a short discharge summary of their care and treatment regime.

For those working in the community, this care transition was widely seen as problematic. In particular, staff often reported concerns about the timing of discharge, especially where patients arrived in the community later in the afternoon or in the early evening when staff were generally wrapping up the end of the working day or dealing with shift handovers. As such, there were often limited time or staff resources to admit and clerk the patient onto the ward, which in broad terms functioned at a different pace compared with acute hospitals. A significant concern was the transfer of patients later in the evening, when the available ward staff could often lack the necessary skills or experience to admit and continue the programme of care for patients with complex needs.

A more common issue related to the lack of opportunity for thorough handover or knowledge sharing between acute and community staff. As described above, a telephone conversation and summary record represented the main sources of knowledge sharing, and ward staff often had too little information from which to continue supporting patient recovery. For example, the information provided often detailed the patient’s medications, broad treatment plan and current levels of rehabilitation, but gave little detail about how they had responded to care, their wishes or preferences, or their wider personal circumstances. There was a general sense of community staff ‘flying blind’ as they worked to provide continuity of care. In many instances, this necessitated a series of repeated interactions to determine the most appropriate plan of care. In such circumstances, the lead nurses would contact the acute ward to request further information, but this was regularly complicated by not knowing the name of the patient’s nurse or because the relevant clinician was not on duty. As such, there was a general lack of familiarity between the community and acute hospitals with little shared understanding of their respective needs. For those in the community hospital it often seemed that the community was regarded as a dumping ground to free up acute beds. The discharge liaison nurses at both the Glipton stroke and orthopaedic units recognised that there were often demands on the community hospitals and often seemed to act as mediators and problem-solvers. In particular, they were found to be on first name terms with the ward managers and senior nurses in the community and would often be contacted by telephone to discuss particular patients. However, not all nurses within the community had this type of relationship and the discharge liaison nurses were not always available to fill the knowledge gaps, especially in the later afternoon and evening.

Transfer to a residential and nursing home (including intermediate care bed)

The transfer of care to a residential and nursing home was evident across all research sites, but was particularly common in the Farnchester area given its lack of community hospitals and the enhanced availability of city-located nursing homes. There were distinctive challenges in securing nursing home placements in rural or registered dementia homes because of a lack of capacity to meet the demand in Farnchester. These provided multiple levels of support for patients leaving hospital, from step-down intermediate health-care support, where a patient would receive a 30-day package of specialist rehabilitation, to more residential care to support personal and social needs, where the expectation might be that the patient might not be able to care for him or herself at home. For many elderly patients, admission to hospital was often from one of these nursing or residential homes where they were already receiving some level of nursing care or assisted living due to a pre-existing condition. For others, their discharge to a nursing home was an alternative to their own home or other domestic setting.

As with other forms of discharge, the transfer of care to a care home was usually arranged by the patient’s designated nurse. As described above, this would usually be initiated after the anticipated date of discharge was confirmed and following instruction from a senior nurse during the daily board round or handover. The patient’s nurse would usually incorporate this within his or her daily tasks and as a part of completing other referral requests. In the stroke and orthopaedic units at Glipton, the discharge liaison nurse provided a key resource in this process through providing information about the available nursing and residential homes, including the types of patients they cared for, the levels of care they could provide and their location within the wider county. At the Farnchester orthopaedic unit this process was typically assisted or taken up by the dedicated community health (rehabilitation) team, who had a more detailed knowledge of the local nursing homes and the services they provided for rehabilitation. At the Farnchester stroke unit a similar support service was provided by the community stroke nurse, who would advise the patient’s designated nurse. However, it was also common for social workers to organise residential care placements where less demanding nursing care was needed but the patient was still not able to return home.

In most cases, the exchange of information largely centred on this initial telephone interaction between the patient’s nurse and the residential/nursing home manager. This conversation involved a broad overview of the patient’s details, primary health condition (stroke or hip fracture), wider health state and comorbidities, and the continuing needs for his or her recovery and rehabilitation. The nursing home manager would usually question and probe this review to determine whether or not the patient was suitable for their specific nursing home; for example, whether or not he or she needed intensive personal care as well as rehabilitation, or specific forms of speech and language rehabilitation. This interaction seemed to focus on whether or not the nursing home had both availability and the necessary skills to care for the patient. This assessment of need and service provision appeared complicated and protracted. Although both relied upon formal hospital records and patient eligibility criteria to assist their decision-making, it also seemed that decision-making relied upon the judgement and experience of the nurse and the nursing home manager.

At the point of transfer, there was also evidence of a lack of knowledge sharing between hospital and care home, especially about the patient’s care plan and personal circumstances. Some nursing homes had tried to remedy this by initiating their own transfer-of-care checklist, which required information regarded as necessary to continue care. For those nursing homes providing intermediate care and rehabilitation, it was also common for community therapists to telephone the acute hospital to obtain further information about the patient’s rehabilitation plan, expectations and activities of daily living. Although they would normally complete their own assessment and care plan, it was seen as important to check the accuracy of their records and to ensure continuity of care. In general, it seemed that the acute rehabilitation teams at both hospitals tended to have relatively positive relationships with the community therapist teams based upon these almost daily telephone interactions. For the hospital therapists this was seen as promoting patient recovery based upon the continuation of their care plan. In the Farnchester area, stroke patients also benefited from the work of the community stroke nurse who would attempt to visit patients in the nursing home following discharge to support their adjustment and recovery. As the community nurse regularly visited the acute ward and worked closely with patients and families prior to discharge, this provided an important source of continuity in the discharge process and often enabled discussion of important information about the patient’s stroke, care plan and recovery. Notably, the Glipton ESD service was resourced to continue care in both residential and nursing homes as part of the rehabilitation plan. This liaison provided opportunities to build strong relationships with certain private providers and share expert knowledge of equipment use with staff.

Primary care support

For all patients leaving hospital, the local medical practice and GP was an important provider of ongoing care and recovery. Across all sites, GPs were notified about their patients’ hospital treatment and discharge back to the community, usually with details about their ongoing care and changes in medicines. Although it was difficult to recruit GPs to take part in the study (see Chapter 3), those who did participate talked about the variable quality of discharge letters and summaries. For example, whereas some of these provided limited descriptive information about the surgical procedure and care plan, others offered more detail about the patient’s general recovery and ongoing needs. This often depended on the individual consultant who completed the letter and suggested a lack of uniformity in completing discharge letters. It was also reported that there were often delays in receiving discharge letters, and often the first a GP would hear about a patient’s admission and discharge was when he or she was contacted by the patient, family member or nursing home manager following care transition. This suggests significant delays and time lags in the knowledge-sharing process between acute and primary care.

General practitioners, and community nurses working in GP practices, were described by most patients as providing an important source of ongoing security and support following discharge, especially in relation to the more medical aspects of recovery and medicines management which other carers might not be able to address. GPs were generally described as helping to explain and manage the ongoing plan of care, especially changes in medicines. For many patients, the GP was often ‘called out’ to address significant changes in the patient’s health, especially where there appeared to be infection or general decline in well-being. However, some patients reported difficulties in accessing their GPs, especially those who lived in rural areas, and greater reliance upon the community nurse for support. An additional source of support for many patients, especially stroke patients, was provided by community pharmacists, who were also described as explaining changes in medicines and reviewing medicines management as the patient progressed.

Community nurses provided an important level of ongoing support for patients, especially when transferred home. In general, they not only attended to ongoing care needs, such as wound care, but were found to act as ‘problem-solvers’, helping to track down missing equipment, ordering additional supplies and providing essential instruction and guidance to family members on how best to support their relative’s recovery. In many ways, community nurses picked up the responsibility of care where ward nurses left off. However, the availability of community nursing support was limited across all sites and visits to the patient were usually limited to two or three per week. Moreover, it was not always clear to the community nurses what levels of support were needed, as there was no standard handover between hospital and community nurses. As such, the community nurses would often need to contact the acute ward to make additional enquiries, which was generally facilitated where there was a designated discharge liaison nurse who would act as the primary point of contact. Further, it seemed that ward-based discharge liaison nurses, where in place, had a relatively positive relationship with these community nurses and the exchange of knowledge between them was generally supportive and enabled continuity of care.

Falls

Falls, including slips, trips and collapses, were described by almost all participant groups as a common safety event in relation to discharge. Falls can occur at almost any time within a patient’s care journey and have a range of causes. It was usual for those in hospital and community settings to describe different situations where falls occurred, but in general these related to falls from the patient’s bed, when using a chair, the toilet or other bathroom facilities, when ascending or descending stairs or steps, and when being mobilised during therapy. It is important to note, however, that many of these accounts described falls as part of general care management, and not with specific reference to discharge planning and care transition. With specific reference to the falls and falls risk factors associated with discharge planning and care transition, a number of specific issues were identified. These were broadly related to processes of ‘transportation’ and ‘resettlement and adjustment’, with the second group also associated with the use of equipment and devices (see Equipment).

In terms of ‘transportation’, two issues were raised by patients and/or family members. First, it was described how falls could happen when being moved or handled between hospital, vehicle and community location. For example, the process of helping patients into the ambulance or back into their own home was often described by patients as challenging because they were still adjusting to their mobility constraints and because carers were seen as unsure how best to move the patient or were unfamiliar with the home layout. This suggested underlying concerns about the knowledge and skills of some support groups, i.e. porters and ambulance crews, to appropriately and safely move patients. Second, it was also described how, on occasion, ambulances used for transportation might not be suitable for certain types of patients in terms of seating arrangements, availability of chairlifts or having the necessary devices to support the care of frail patients. A number of patients on hip precaution following hip replacement described how the wrong type of ambulance was provided as there were no high-seats or supports to enable safe transfer between hospital and home. A number of ward clerks also described how inappropriate ambulances were provided when transferring very frail, bed-bound patients to community hospital, because they could not accommodate a bed. In such cases, it was usual for participants to describe unnecessary delay, rather than an actual fall, but where alternative transportation could not be arranged then the risk of fall remained. In general, patients and their relatives reflected negatively on the transportation process and often referred to falls and other risks associated with both the physical movement of patients and the suitability of the vehicle. When describing these risks it was common for hospital-based participants (doctors, nurses, HCAs, ward clerks) to highlight deficiencies in the regional ambulance service, especially where bookings for a particular time, location and ambulance type were not fulfilled. For some participants, this reflected a communication problem between the wards and the ambulance service, especially where booking systems and the ambulance control did not accurately record or meet provided instructions. Representatives from the ambulance service acknowledged these problems but also described themselves as overburdened by the demands of emergency care and often struggling to manage service provision. It is also important to recognise the potential for missing information at the time of booking, for example if the ward clerk had not been informed about specific needs and could not therefore relay this to the ambulance service.

More prominent were falls associated with ‘resettlement and adjustment’ within a domestic setting, usually home. Three safety events were described within patient and family narratives. The first group comprised trips, slips and falls within the home that seemed, in general, to be associated with everyday activities, such as using furniture, tripping over rugs and internal steps, ascending or descending stairs, using bathroom facilities or getting in or out of bed or a chair. For many participants, these were associated with patient confidence following discharge and the need to readjust to the home setting after prolonged hospitalisation. It was also described how certain aspects of the home layout could be taken for granted rather than moved to reduce the risk of fall, such as the position of rugs and tables. A widely understood risk related to the use of bathroom facilities, which were especially problematic for hip fracture patients. For example, patients often described minor slips or falls when sitting or rising from a toilet seat or when entering the bath. A range of possible risk factors were described as increasing the likelihood of falls around the time of discharge, including cognitive impairment, medicines use or adverse effect, infection [especially urinary tract infection (UTI)], patients’ understanding of their ability to mobilise, lack of support from carers and the use of equipment or devices (see Equipment). It seemed that, for many, these risks were accepted as normal for patients needing to readjust to their home or care home and, as such, extra vigilance was needed in the early days following discharge. However, some also raised the possibility that these risks were heightened by a lack of thoroughness in, or poor communication of, clinical assessments, i.e. where cognitive impairment was not adequately determined, or information on the use of medicines or equipment was incomplete (see Equipment).

The second group of resettlement-related falls risks were associated with the provision and use of home adaptations and equipment to support recovery and mobilisation around the house (see also Equipment). Both hip fracture and stroke patients were commonly provided with a range of grab rails, seat and bed risers, steps, frames, walking sticks and wheelchairs to support their recovery outside hospital. Two specific issues were associated with this equipment. First, patients and carers (including reablement teams) described problems with the availability and appropriateness of equipment and devices. A common example described essential equipment, such as grab rails, walking frames, commodes, steps/ramps, beds and mattresses, being delivered or fitted after the patient had arrived home. In such situations, patients were at risk of fall or accident because everyday activities, such as washing, toileting or walking, were not appropriately supported. As explained below, these delays were often linked to problems in the ordering of equipment, especially in contacting suppliers, or disagreements between nurses and OTs about whose responsibility it was to order commodes, or between health and social care about who should fund certain adaptions, especially steps and ramps. Second, patients, relatives and carers also experienced complications in the use of certain devices, such as hoists or pressure mattresses. For example, one patient described a fall when using the toilet because he did not realise that his weight needed to be evenly distributed on the support frame. These risks are elaborated below, but again participants described shortcomings in supporting patients and carers in the use of equipment.

The third group of falls risks relate, in very general terms, to the quality and appropriateness of the domestic settings to which the patient is transitioned. A number of city-centre patients were discharged to Victorian-era homes with several steps between pavement and door, narrow and steep staircases and relatively small bathroom facilities. In contrast, a number of other patients were discharged to more rural settings, to small cottages or farms. These property types were commonly described by community nurses or other social carer workers as poorly suited to patient recovery, some with outside toilets, making it difficult to accommodate necessary home adaptations and requiring patients to take unnecessary risks. The failure of social work and community health-care teams to recognise and consider these issues in discharge planning was seen by community nurses as putting patients at risk. In many instances, it was suggested that patients needed further rehabilitation at an intermediate facility before returning to these home settings.

In general, falls and falls risks featured more often in the narratives of patients and families, but they were also included in those of community-based carers (reablement teams, community nurses, OTs) following hospital discharge. It was more usual for community nurses to discuss the link between falls and cognition, medicines use or infection, reflecting their enhanced clinical understanding of the possible causes. It was more usual for community stroke nurses (Farnchester) and the discharge liaison nurses (Glipton) to mention property type in relation to discharge planning. There were no variations between sites in terms of perceptions of risk associated with falls as it seemed a general issue related to postdischarge recovery. Drawing on participant narratives and, where possible, cross-referencing these with observational data, falls and falls risks associated with discharge planning and care transition are summarised in Figure 8.

FIGURE 8.

Illustration of falls-related risks.

The above paragraphs describe how participants understood the risk of falls. Relating this to the patterns of knowledge sharing explicated in the previous chapter, we can begin to make a number of suggestions about how knowledge sharing contributed to the source of risk. Although falls can be brought about by a range of local and environmental factors, a number of knowledge-sharing dynamics condition or mitigate this potential. In relation to hospital transportation, the booking process involved an important opportunity for knowledge sharing between ward clerk, ambulance control and ambulance crews. The study found that this could be complicated, i.e. missed or incorrect, when the ward clerk was not on duty and other ward staff would be required to arrange transportation, for example nurses or HCAs. Although booking transport was relatively straightforward, the ward clerks appeared to have a developed understanding of the types of transportation available and an interpersonal relationship with staff at the ambulance service. As such, they were often more accurate in booking ambulances or appreciating the daily demands and schedules of the ambulance service. The absence of the ward clerk and reliance upon other staff could, on occasion, lead to the dispatch of inappropriate transportation or lead to delays. As such, the ward clerk provided an important knowledge-sharing resource on the day of discharge.

In relation to the patient’s general well-being and mobility, and also his or her use of equipment, a key knowledge-sharing dynamic was found in the interaction between the patient, his or her relatives and the ward-based therapist team. Specifically, OTs and PTs had an essential role in assessing the patient’s overall mobility, ability to undertake activities of daily living and cognitive function. They also provided individualised information and training to patients and relatives on the use of specific equipment and devices to be used after discharge. These types of activities relied upon relatively detailed, close and personalised knowledge-sharing relationships in day-to-day ward-based interaction across all sites, for example explaining to patients how best to use walking aids or climb stairs. At some sites, however, this pattern of interaction and knowledge sharing could become more fragmented, less patient-centred and not integrated into other ward-based activities. At Farnchester, for instance, therapist teams worked across the entire hospital and had sessional visits to the orthopaedic unit (note that the stroke unit had a dedicated therapist team). It was observed how OTs and PTs often had limited time with individual patients, could not always co-ordinate their sessions with family visits and were often marginalised from ward-based decision-making, such as MDT meetings. As such, the opportunities for both giving and receiving knowledge to and from patients and ward staff were limited. A similar format of therapist input was observed at the Glipton orthopaedic unit, but here there was closer involvement of therapists in ward-based activities. The two stroke units, however, featured dedicated therapist teams which facilitated closer, regular and more timely interaction with patients and families, and thereby more continuous exchange of knowledge. A related feature at Farnchester stroke unit was the explicit emphasis on family-centred care, including the use of family meetings, which enabled closer working with family members so that they were better aware of their relative’s ongoing needs and use of equipment. A further issue related to the provision of equipment related to ordering and supply, which is discussed below.

A further knowledge-sharing issue was the involvement in discharge planning of specialist staff who could provide knowledge about specific devices, medicines or housing. In particular, a key contribution was made by discharge liaison nurses at Glipton, and also the community stroke nurse at Farnchester, who had detailed, even ‘encyclopaedic’, knowledge of local housing quality, including steps from pavement to door, layout and bathroom arrangements. Their involvement in discharge planning could often avoid the need for home visits, but more commonly enhanced the decision-making processes of therapists, for example when ordering equipment (see Equipment) or supporting patient use of devices. A similar contribution was also found in the role of both hospital- and community-based pharmacists, who could inform patients and family members about the potential medicine-related complications that might reduce stability (see Medicines). Where these actors were not present, or were marginal, it seemed that discharge planning was often more generic and less specific to the needs of individual patients.

Infection

As with falls, infections were identified as a patient safety concern for both stroke and hip fracture patients, especially where both patients and staff were attentive to the problems of hospital-acquired infections, such as methicillin-resistant Staphylococcus aureus (MRSA). Notwithstanding these broader infection risks, two types of infection risk were discussed in relation to hospital discharge, usually by ward and community nurses. The first was associated with ‘extended admission’, where patients would be exposed unnecessarily to hospital-acquired infections because of delayed discharge. For example, ward nurses talked of the risks of contracting pneumonia and Clostridium difficile, especially for frail older patients. As outlined below, clinicians describe a ‘window of opportunity’ for care transition, which once missed could lead to patients being ‘trapped’ within a cycle of infection and recovery, which could have little direct connection to their stroke or hip fracture but reflect other comorbidities and the general risks of admission. Similarly, the risk of pressure sores was discussed for extensive inpatient admissions, where a lack of mobilisation and extended confinement to bed could lead to sores and chest problems. This was usually related to problems with ongoing patient monitoring after delayed discharge, when the patient was seen as ready for discharge and hence levels of direct active medical care were reduced, but monitoring and mobilisation were still required. As such, infection risks were often related to delays in discharge and the changing patterns of care giving (see Timing and scheduling).

The second group of infection risks were associated with ‘follow-up care’, when infection developed or was identified immediately after discharge. A prominent concern, for example, was UTI, which was seen as relatively common among both patient groups. Community nurses, in particular, described the need to monitor patients for signs of increased temperature and confusion in case of UTI, and also the need to involve the patient’s GP to prescribe antibiotics. An additional, but less common, infection risk related to sepsis, again following surgery, which might not develop until the patient was discharged into a community setting. This was raised by community nurses and ward nurses in community hospitals; for example, one Glipton hip fracture patient was found to have developed sepsis due to a surgical clip being left undetected at the time of care transition. A more common problem was with wound care and whether or not the reablement teams had the appropriate nursing skills to thoroughly clean and bandage surgical wounds or to detect potential problems. A more general concern for community nurses was the patient’s adherence to prescribed medicines after leaving hospital. It was described by one pharmacist, for example, that patients often attend the community pharmacy confused about their changed medicines regime (see Medicines). Such confusion and lack of adherence to prescribed antibiotics was seen as contributing to the potential risk for infection.

It was more usual for nurses within community hospitals to highlight the potential for postoperative infection after hip fracture patients were transferred from the Glipton hospital. This seemed to suggest a belief that patients could be transferred too quickly or without a more thorough assessment of need. It was also more common for community nurses in the rural Farnchester area to describe the problems of regularly monitoring patients dispersed across a large geographical area, where it could also be difficult to access GP services. Drawing on participant narratives the infection risks associated with discharge planning and care transition are summarised in Figure 9.

FIGURE 9.

Illustration of infection-related risks.

The above paragraphs describe how participants understood the risk of infection. Relating this to the findings described in the previous chapter, we can begin to make a number of suggestions about how knowledge sharing could frame or mitigate infection risk, often in relation to other risk issues discussed below, such as reducing delays, improving adherence to medicines or identifying at-risk patients. For patients on the ward awaiting discharge, and with delayed discharge, knowledge sharing between ward nurses and other therapists was observed as important for identifying and managing bed sores or pneumonia. Where ward-based activities were more integrated (i.e. in both stroke services) there were more opportunities to share knowledge between clinical groups about the need for therapeutic intervention, such as turning, to limit infection risk. For patients transferred to community hospital and nursing homes, the sharing of knowledge between ward nurses and carers was also important for maintaining monitoring and therapeutic interventions. As shown in Glipton, the sharing of knowledge between these sectors could often be limited or fragmented, leading to subsequent complications in ongoing care following transfer. In care homes and other residential homes, support staff routinely assessed and monitored infection risks and, as such, problems were usually detected early.

For patients transferred home, access to primary health-care services was particularly important, for example GPs, community nurses and pharmacists. In general, where patients had easier access to their GP, community nurse and community pharmacy, often due to colocation of services and proximity to home, it was easier to share knowledge about potential infection concerns and seek advice or treatment. Where patients lived in rural areas, for example around Farnchester, it seemed that patients had more fragmented and partial relationships with their community health-care team. That being said, across all sites community nurses fulfilled a proactive role in monitoring infection risks and other care issues, such as nutrition, continence, wound care or medicines adherence. However, where community nurses were stretched in workload, especially in rural Farnchester, there was less opportunity for patients or family members to share their concerns in more immediate, direct and personal ways. Interestingly, it was found that GPs had a more reactive involvement in monitoring infection risk, in that their intervention was usually prompted by a community nurse or family member once infection concerns were identified. It was seen as rare for GPs to take a more proactive role in postdischarge care, often because they received little detailed or timely information about patient need from the acute hospital.

Medicines

Medicines were one of the most discussed risk issues across all participant groups. Although these were highlighted across the entire care pathway, specific problems with prescription, dispensing and adherence were also described in relation to discharge planning and care transition. These broadly fell into two categories: first, those related to organising and providing take-home or TTO medicines before discharge, and second, those related to medicines use and safety following discharge.

The perceived risks associated with TTOs resemble more common safety issues associated with prescribing and dispensing medicines, but highlight particular issues associated with hospital discharge. Prescribing errors are well recognised within the wider patient safety literature and occur when, for a variety of reasons, the prescriber (usually a medical doctor) requests, for example, an inappropriate medicine, the wrong dosage or the wrong combination of medicines. At discharge, such prescribing issues were often discussed in relation to the work of junior doctors, who tended to be responsible for prescribing TTOs. A number of accounts were given where the wrong medicines were prescribed; where medicines were prescribed that were unsafe for certain patients, i.e. those with pace makers or allergic reactions; where insufficient medicines were ordered, i.e. the patient would run out after discharge; or where medicines were not ordered in sufficient time to be ready for discharge. A very common situation, for example, was for relatively standard or regular medicines, such as furosemide, not to be prescribed because it was assumed that an existing prescription existed and the records were not checked.

A number of common accounts were given for why TTOs might be incorrect or unsafe. First, junior doctors could find it difficult to access all the relevant information about often unfamiliar patients, especially their previous history of medicines, because patient records were poorly documented. This made it difficult to find relevant information or meant that the information was simply not available. Second, it was common for other staff groups, including ward nurses, clerks and hospital pharmacists, to highlight poor handwriting or ineffective channels of communication as a problem with prescribing. Thirdly, junior doctors may not have the necessary prescribing knowledge to accurately prescribe oxygen and pre-emptive palliative care drugs, leading to delays in managing fast-track patients. At Farnchester hospital electronic prescribing was seen to greatly improve the quality of prescribing and communication between departments. Finally, there were the broader error-creating pressures associated with the wider organisation of work, especially excessive workload, limited work planning and the general lack of support and guidance. For example, junior doctors often seemed to be thrust into new prescribing roles with relatively little guidance, and expected to rush TTOs ‘urgently’ to facilitate a last-minute discharge, with little time for reviewing records in detail or seeking support. At the Glipton stroke unit all TTOs completed by junior doctors were checked by the stroke consultant prior to being issued as a basis for quality control and also to support learning.

Risks were also described in relation to the dispensing of medicines within hospital, especially the checking, transporting and storage of medicines before being issued to the patient. A prominent narrative at the Farnchester orthopaedic unit was for patients to be given incomplete medicines or even the medicines of a different patient. It was explained that TTOs were ‘dropped off’ at the ward in large batches, and stored in a general side room before being sorted according to patient name. This made it possible for multiple packages and dosset boxes to become mixed up or separated. At the point of discharge it was observed that ward nurses would enter this side room to obtain patient medicines, but because of poor storage, limited handover within the pharmacy department and time pressures, medicines would often be missed or mixed up. A further problem occurred when large dosset boxes were requested late on Thursday in advance of a Friday or Saturday morning discharge. Although the prescribing information was usually accurate, there was insufficient time to complete these orders, suggesting that ward staff were not knowledgeable about the work schedules and demands in the pharmacy department; in turn, this would delay discharge until the Monday or require the patient to leave hospital with incomplete medicines. More generally, it was described how ‘last-minute’ prescribing would often mean that discharge could be delayed or, again, patients might have incomplete medicines. At Farnchester hospital the use of electronic prescribing and 24-hour dispensing seemed to enable more balanced workload planning and reduce the pressure on dispensing at the point of discharge.

Medicines risks were also discussed in relation to postdischarge recovery at home or in a nursing home. First, participants described possible adverse physical (side) effects associated with inappropriate medicines use, i.e. where contraindications existed or where medicines were missing. Some examples included TTOs containing antibiotics despite penicillin allergy, or aspirin despite the patient also taking warfarin. If these errors in prescribing and dispensing were not identified in the hospital prior to discharge, the potential side effects would invariably be detected in the community by nurses, pharmacists, GPs and care home staff. This was often complicated by polypharmacy where patients were prescribed multiple medicines to deal with complex health needs; this made it difficult for community staff to understand health-related problems associated with medicines, for example confusion or decline in health or well-being.

Second, community nurses and GPs also described problems with patient adherence to prescribed medicines. Common examples were failure to complete the prescribed course of antibiotics, reverting to medicines regimes used prior to admission or not taking certain medicines in the expected way. For example, community nurses reported concerns about their patients not taking calcium medicines as required because of the complex instructions; they were not sitting for sufficient time to allow absorption on empty stomachs. Problems with adherence were linked to a range of factors, including lack of instruction or guidance by hospital staff, difficulty in remembering what medicines to take and when, and physical complications in opening medicines (dosset box) or being able to swallow (after stroke). Importantly, adherence appeared to be based upon key knowledge-sharing interactions between patients (and relatives) and ward nurses at the point of discharge, and with community nurses following discharge. Where these interactions were rushed, there was often an absence of knowledge sharing and therefore adherence could be incomplete.

Third, risks were associated with the reconciliation of new and old medicines in the community and the extent to which medicines use was monitored and supported. Patients’ and family members’ narratives revealed a burden of responsibility to follow a prescribed medicines regime, but often with limited understanding of the medicines prescribed. A significant factor, therefore, was whether or not patients, family members and other carers felt supported by community and primary care professionals. Community nurses were often described as providing essential support in explaining ongoing patient care needs and in monitoring the provision, use and physical effects of medicines. Similarly, GPs were often described by patients as providing essential support in reviewing new medicines and altering existing medicines regimes after leaving hospital. Further still, community pharmacists were seen as a readily available source of information to describe new medicines and support adherence. However, participant narratives also highlight common problems with accessing these specialists, especially in rural areas where access to a GP or pharmacist was difficult. As described above, it was also suggested that GPs often took a reactive, rather than proactive approach to managing patient medicines, i.e. attending to the patient after problems were detected. Again, GPs related this to a lack of communication at the point of discharge, especially in terms of receiving timely or complete summaries following discharge, and relying upon the medicines held by patients to determine the new medicine regime (which could be inaccurate if not all medicines were dispensed).

As with other aspects of patient care, medicines were seen as a significant source of risk and potential unsafe care, especially in relation to the processes of prescribing, dispensing, use and adherence. At the time of discharge, however, these processes were seen as potentially exacerbated by a general lack of continuity of care, where medicines are prescribed and dispensed in one setting but taken and monitored in another. The handover of medicines management between secondary and primary care professionals appears integral to patient safety, but across all sites there was relatively limited direct communication between these groups; instead, the patient and his or her medicines became the primary basis for knowledge sharing. These processes are also often complicated by the very nature of discharge itself, which, as shown in the previous chapter, can be a complex process involving lots of last-minute activities undertaken by a multitude of actors, and not always well co-ordinated. Medicines-related risks were common to all sites, but problems of follow-up care and adherence were particularly marked in more rural areas, especially Farnchester; at the same time, however, this hospital also benefited from more accurate dispensing of TTOs because of its electronic prescribing system. Drawing on participant narratives these medicines-related risks are summarised in Figure 10.

FIGURE 10.

Illustration of medicines-related risks.

As indicated above, effective knowledge sharing before and after discharge is essential for medicines safety. For example, accurate and complete prescribing of TTOs is based upon the ability of junior doctors to effectively access, understand and use recorded knowledge about patients’ pre-existing medicines, allergies, contraindications and ongoing care needs (which is itself dependent upon the accuracy of record keeping throughout patient admission). At hospitals with multiple and fragmented care records, i.e. the orthopaedic units, the initial prescribing process seemed more complicated than that observed where a single patient record was used, i.e. the stroke units. The prescribing process then relies upon knowledge sharing between prescriber and hospital pharmacy, where timely and accurate exchange is essential. At Farnchester, the use of an electronic prescribing system appeared to facilitate this process over and above the use of more traditional, paper-based prescriptions at Glipton. These ICTs also enabled prescriptions to be made 24 hours a day, and so potential delays in the process could be avoided. In contrast, at the Glipton stroke unit it was found that a senior doctor would usually check TTO prescriptions for their accuracy and handwriting, thereby providing an additional level of quality control. The study found, therefore, that the availability and distribution of knowledge materials (records) and also knowledge systems (ICTs) could enhance the quality of TTOs.

Knowledge sharing was also important at the point of dispensing medicines, especially when deliveries were made to the ward or medicines were issued to patients. At all sites, there were no rigorous methods or systems for checking medicines at the time of delivery, often because of time and work pressures. It was customary for medicines to be checked at the time of giving them to the patient, which could create further delays if items were missing or incorrect. More significant, however, was the level of knowledge sharing with patients and relatives prior to discharge. At all sites, for example, it was usual for the patient’s designated nurses to explain medicines use. The study found that this interaction could easily become fragmented or piecemeal during busy times or when discharge had been delayed. This would lead to patients being discharged with relatively poor understanding of their medicines and would lead to potential problems with adherence. Following discharge, it was also found that access to a GP or, more often, community pharmacy was important for providing patients with additional information about their medicines and medicines use. As described above, those living in more rural and remote areas usually had limited access to these services and sources of knowledge.

Clinical procedures

In the days and hours leading up to hospital discharge it was usual for patients to undergo a range of assessments, procedures and tests to confirm readiness for care transition and inform onward care planning. Common examples included blood tests to clarify the absence of infection, electrocardiograms to confirm the patient’s cardiovascular health, and scans to provide a final image of the patient’s hip to facilitate subsequent comparison and analysis of follow-up scans after discharge. Although concerns about these clinical procedures and tests featured less than those regarding other safety issues, such as medicines, they were identified by clinicians and patients as relatively common in occurrence if not significant in impact. In broad terms, the failure to complete these procedures was seen as allowing possible underlying health concerns to go undetected, leading to complications following discharge. A major concern, for example, was stroke patients discharged without a computerised tomography (CT) scan or, more commonly, other comorbidities not thoroughly investigated prior to discharge. For example, several ward nurses commented that certain blood tests might not be completed on the day of discharge because the patient was already confirmed as leaving and it was assumed the test would be completed in the community. Similarly, one SaLT described how patient swallowing tests were not always reassessed prior to discharge, and so the proposed care plan might not be accurate. More commonly, it was described that patients were regularly discharged in need of almost immediate appointments for outpatient scans or tests, as these could not be completed on an inpatient basis within the time frame available prior to completing discharge (for example, because of delays in booking scans on the day of discharge).

Similar issues were identified by patients and relatives following discharge, normally in relation to follow-up outpatient appointments. A number of participants described not receiving appointment letters, which necessitated that they contact the hospital directly to arrange an appointment. A particular problem concerned patients whose first language was not English, who could not easily understand appointment letters without translation. More specific problems were found among patients accessing community orthopaedic physiotherapy, where it seemed that bookings often went astray, especially in the Farnchester area. It was also found that patients would sometimes present at outpatient clinic to find that not all tests could be completed on the same day, especially where they involved specific diagnostic imaging, thereby requiring a second hospital visit. This was especially draining for frail patients who relied upon the hospital ambulance service, where additional risks were incurred (see Falls), and for those who lived in more remote rural areas. In contrast, clinicians also talked about the problems of patients not attending follow-up clinics and putting their own health at risk.

Failure to complete clinical procedures, tests or scans, or to attend follow-up appointments, did not necessarily constitute a direct threat to patient safety, but created the possibility that health concerns or problems might go undetected or without sufficient medical intervention. There were no clear indications that missed tests had led to patient harm, although one patient did restroke after discharge and had missed a scan before discharge. Furthermore, outpatient appointments were often an important opportunity for patients to discuss their progress and to seek additional guidance on recovery. As such, failure to maintain a follow-up relationship could undermine adherence to clinical instruction and recovery. Drawing together the patient narratives, the risks associated with clinical procedures are summarised in Figure 11.

FIGURE 11.

Illustration of risks associated with clinical procedures.

The planning and completion of clinical procedures, tests and scans either before or after discharge was framed by a number of knowledge-sharing issues. First, the decision to initiate or request a specific test or assessment was often taken as part of daily ward-based interaction, such as the ward round, or alternatively in weekly MDTs. Variations in how these decisions were communicated and recorded could determine how quickly and thoroughly requests were actioned. During the initial period of inpatient care, it seemed that these tests were escalated in priority, but as the patient neared discharge, especially after a date was set and readiness confirmed, the priority given to completing tests seemed to decline. It seemed that staff assumed these patients were healthy and ready for discharge and, as such, clinical attention could be given to more frail patients. This would mean that some more routine tests or scans would be overlooked in daily planning activities, such as ward and board rounds. For example, they might be recorded in the patient notes, but the allocation of responsibility might be left unsaid or vague. In other words, it seemed that the cultural value of diagnosis was absent, as these tests usually dealt with confirming readiness for discharge, and therefore there was less emphasis on or importance attached to these requests.

A range of more obvious knowledge-sharing issues was identified in relation to daily ward-based planning activities, especially handovers and board rounds. As described in the previous chapter, these helped to co-ordinate routine activities, such as ordering and completing tests. In sites where these interactions were more inclusive of different staff groups and based upon shared decision-making, i.e. in both stroke units, it seemed that staff worked together to identify and complete tasks, thereby reducing the potential for procedures to be overlooked. In other sites, clinical interactions seemed more fleeting or fragmented. Moreover, there were seemingly important differences in how requests for clinical procedures were recorded during these interactions. At the Glipton orthopaedic unit, for example, staff often relied upon scraps of paper or memory to record activities, which seemed to explain why certain tasks were overlooked until the last minute. In contrast, the use of whiteboards, for example, helped staff record and monitor task completion and could be reviewed regularly and easily throughout the day. The availability and use of ICTs also shaped the booking and provision of tests and scans. At all sites, ICTs were increasingly used to request appointment slots and to receive test results. In general, this appeared to facilitate the process, but it was also observed that interpersonal skills and the ability to negotiate were also important determinants of the successful booking of requests. In particular, ward clerks and some experienced ward nurses at Glipton would often contact the pathology or radiology department, for instance, to request a different slot or to rush requests. These interactions were facilitated by personal connections and familiarity which enable the ward staff to circumvent the procedures prescribed by ICTs, and expedite test results to enable discharge. Furthermore, the Farnchester stroke unit had recently procured several imaging and specialist computer terminals to reduce dependence on shared hospital resources. Finally, arrangements for outpatient tests and procedures often relied upon the quality of record keeping at the time of discharge and the detail of the discharge summary. At Glipton, it was customary for all ward clerks to independently log requests for outpatient appointments (in addition to the discharge summary) and to action these on the day of discharge. At Farnchester, however, the consultant’s discharge letter and nurse’s discharge summary provided the primary triggers for onward requests for appointments and often became delayed as they were processed alongside other competing tasks by the medical secretary or the patient’s nurse.

Equipment

Equipment and other devices are integral to the continuing recovery and rehabilitation of hip fracture and stroke patients before and after discharge. For the vast majority of patients they provided invaluable support, yet, linked to the discussion of falls, equipment availability and use were widely described as a source of risk following discharge. In terms of the availability of equipment, a common occurrence was for patients to arrive home without essential items being delivered in advance of care transitions, especially commodes, wheelchairs, walking frames and foot stools. In some cases, items were delivered up to 48 hours after discharge, meaning that patients lacked appropriate support, and either resulting in confinement to bed or increasing the possibility of a fall. A significant concern for patients was toileting and the desire to avoid using pads or bedpans, or soiling the bed. In other instances, equipment was delivered in a staggered or fragmented way. A common example was for a bed frame and mattress to be delivered several days apart. In one instance, a patient bed was delayed by over 9 days, requiring the patient to sleep in an armchair; he passed away in his armchair before the bed could be delivered. A linked concern was the delivery of incorrect equipment, for example a walking frame supplied when the patient required a wheelchair. A common problem was for grab rails and other handles to be fitted after the patient had returned home.

A second set of issues related to equipment use by both patients and their carers. Clearly, inappropriate use of equipment runs the risk not only of falls, but of additional health hazards such as trapped fingers. Common, but seemingly trivial, examples included not locking brakes on wheelchairs, not using walking sticks appropriately or not being able to use adjustable beds. Although hospital OTs described their efforts to prepare patients for recovery outside hospital, including advising them on how to use basic equipment such as walking sticks or wheelchairs, it was felt that some patients did not thoroughly understand or value the importance of lessons provided. Cognitive impairment was often highlighted as making it difficult to explain equipment use. A further concern was patient reliance on relatives or carers who were also not adequately trained or experienced in the use of certain devices. For example, relatives or carers were often required to assist with certain feeding tubes and bathing aides, but typically had little or no training. Moreover, some reablement teams described how they had not been trained in performing certain moving and handling tasks or in the use of certain equipment, such as rotundas. Inappropriate use of such devices introduced additional risk to patient recovery.

Equipment risks were often pre-empted by OTs, ward nurses and other therapists in both the stroke and hip fracture units, but it tended to be community nurses, patients, relatives and reablement staff who described incidents of actual or potential harm. The problems with equipment supply and fitting were generally more common in rural areas where there was usually a longer delay compared with city or urban locations. The use of electronic systems seemed to be associated with reduced delay in supply and fitting, possibly because of the benefits of electronic stock control and workload management. Drawing on participant narratives, equipment-related risks associated with discharge planning and care transition are summarised in Figure 12.

FIGURE 12.

Illustration of equipment-related risks.

As above, drawing upon the issues described in the previous chapter, knowledge sharing played a significant role in the provision and use of equipment. As described, the ordering, delivery, fitting and even maintenance of equipment was dependent on the patterns and quality of knowledge sharing between hospital staff (OTs and ward nurses), equipment suppliers (private firms), fitters and delivery drivers (often another private firm) and relatives and carers (to provide access to the home). Although the interaction between these groups was often effective, there were common breakdowns across all sites. In particular, the interaction between the hospital staff and the suppliers could be complicated by whether or not the ordering and booking system could accommodate bespoke needs or requests. At Glipton it was usual for staff to use computer-based ordering systems which made ordering possible across a wider time frame and supported more rapid ordering, so equipment could easily be ordered at the last minute despite increased costs. At Farnchester, urgent orders were not supported by the hospital, despite costing less than the resulting delay in discharge. However, the computer system could make it difficult to address complex or unique needs or specialist equipment. At Farnchester, the use of a telephone service made it easier to address and explain specialist requests, but the process was often protracted and dependent upon the working hours of the equipment suppliers. A further knowledge-sharing issue related to the level of experience and knowledge of the person ordering the equipment. It was found that discharge liaison nurses and most OTs were highly skilled at navigating the ordering process, but some relatively new nurses often struggled to understand the process. This was especially the case for more complicated and unusual devices, where a discharge liaison nurse would usually know a specialist supplier who served the region.

In terms of patients’ and carers’ use of equipment, the ward-based and therapeutic relationship between the OT and the patient was integral. In all sites, OTs worked closely with patients to support their onward recovery, but it was not always possible to involve relatives or carers in this process. At the two stroke units this was more common because the therapist teams were dedicated to the units and could be available to meet with family members. As such, they could more easily share knowledge on the use of equipment to support use in the community setting. In the orthopaedic sites, however, this was less common and there were few formal opportunities to explain to family members how they might support their relatives following discharge. It was also found that where therapist teams rotated between different wards and were not dedicated to an area, there was generally less time to build familiarity and work closely with patients. As such, there was less instruction on how to use equipment. In general, there was no direct communication between the hospital OTs and the social care reablement teams, which meant that any guidance or instruction on the use of specialist equipment was missed.

Timing and scheduling

Turning to risks associated with the broader organisation and planning of care, participants often talked of problems broadly related to whether or not discharge occurred at ’the right time’ or during what some called the ’window of opportunity’. For example, participant narratives highlighted both ‘premature’ and ‘delayed’ discharge as potential threats to patient well-being, where a patient can either be discharged while requiring acute care or remain in hospital despite the possibility that his or her recovery might be better supported in the community. Reflecting on these narratives, it often seemed that delayed discharge, in particular, was invoked as a risk perhaps because it was a wider organisational priority that had penetrated the wider culture of health and social care, especially given legislation and financial inducements to reduce delays. As such, it was not always clear whether descriptions of delayed discharge were directly or necessarily related to patient safety, but the coming together of these two organisational priorities seemed significant for participants.

Where patients were discharged prematurely, participants described how patients could return home still requiring more extensive rehabilitation, without which recovery would be compromised and patients could, theoretically, be placed at increased risk of fall or relapse. Community nurses and some reablement teams commented that some severely sick and frail patients were discharged and the level of care needed was often beyond their skills or abilities. Another common concern was situations where patients seemed to have recovered sufficiently from their primary stroke or hip replacement, but other comorbidities were not sufficiently addressed prior to discharge. More broadly, premature discharge framed or conditioned many of the other, more direct risks discussed above. In other words, where discharge was brought forward or seen as too early, it often reflected a situation in which staff felt unable to complete the range of necessary tasks to assure a safe discharge, such as ordering equipment, TTOs or clinical tests. As such, premature discharge was a latent risk conditioning factor. For participants, premature discharge was often related to wider organisational priorities and demands, especially bed shortages. As observed at Glipton, service managers and bed bureau representatives would often attend daily board rounds to review bed availability and encourage the transition of care. In these interactions, more senior clinicians would often need to share complex patient-related and clinical information to explain why certain patients should not be discharged early. Similarly, weekly MDTs were often used to balance different clinical opinions to ensure that relevant comorbidities and health issues were addressed when setting the date of discharge, but where these meetings were more limited or exclusive, i.e. within the Farnchester orthopaedic service, there were fewer opportunities to draw together these different views.

As described above, delayed discharge featured prominently in the narratives of virtually all participants. Delays were described as creating the opportunity for patients to be exposed to additional risks such as infection, or to become anxious, depressed or even institutionalised. Again, it was not always clear how delayed discharge had a direct impact upon patient safety, but it did represent a contextual factor that complicated the co-ordination of care. More experienced participants, especially consultants, sisters and social work managers, described how patients could easily become depressed through prolonged hospital stay, leading to general lowering of independence, increased incontinence, agitation and immobility. A number of stroke consultants talked of the risks associated with prolonged stay which had relatively little direct association with the primary medical condition, such as hospital-acquired infections, delirium, incontinence and weight loss. More broadly, it was described how progressive rehabilitation and recovery, in particular with respect to activities of daily living, would be stalled owing to delayed discharge, potentially leading to lower-than-anticipated longer-term outcomes of recovery and rehabilitation.

The sources of delay were wide-ranging and highlight the complexity of the discharge process. At the point of making an external referral, for instance, OTs discussed a tendency for social services contact centres to question relatively trivial documentary inconsistencies, thereby holding up the discharge process. Similarly, social workers would comment that referrals were often inadequate because of missing information or difficulty in accessing ward-based staff to clarify information. At the point of care transition, further issues were identified as creating delay. It was common, for example, for health-care teams to blame social services for holding up discharge by not putting in place necessary provision or for querying referral notices; in turn, social services described limited capacity for providing social care or reablement because of increased demand (and diminished funding). It was suggested that patient discharge could be delayed because essential home adaptations and equipment had not been delivered. In these instances, the community nurse or social worker would request that transition be paused while the supplier was contacted to expedite the delivery and fitting of devices. Similarly, the transition of care to community hospital could be blocked because of limited capacity, which could cause acute services to miss pathway targets, especially the orthopaedic unit. At the same time, community hospitals and nursing homes often described the larger acute hospital as pushing patients out to the community without completing necessary checks or sharing information.

The risks associated with timing were therefore more indirect and less obvious than those associated with other identified safety issues, such as medication or equipment. That being said, it seemed that timing issues framed these other risks. Where patient discharge was delayed owing to problems with ordering equipment or planning social care, for instance, this created two broader changes in the provision of care. First, it framed a wider shift in clinical emphasis whereby acute patient care was regarded as complete, with little need for further care. In these cases, the patient was often seen as ‘bed blocking’ for more urgent cases and, in parallel, clinical attention or care seemed to diminish because ward staff seemed to feel that their responsibilities had been transferred. As such, changes in patient condition, such as new illness or relapse, were not always detected as readily as during the early stages of admission. Second, when these delayed patients did reach the point of discharge there seemed to be greater urgency to transition their care, and less attention given to the thoroughness or completeness of this discharge. Again, it seemed that staff had figuratively almost discharged the patient and therefore gave less attention to this process. This was more common for the two orthopaedic units and less so for the stroke services, possibly because of the stroke patient’s ongoing need for direct care.

Broader still, the risks associated with the timing of discharge also appeared to reflect differences in the discharge means for different groups. National policies suggest that discharge should not be seen as an end point, but rather a transition of care. For many participants, however, discharge meant very different things, which framed underlying tensions in the planning and scheduling of care. Specifically, for many hospital staff hospital discharge signalled the end of their direct involvement in patient care, and the passing on of responsibility to other, more relevant, agencies. As such, hospital discharge seemed to be located, in terms of both work planning and cultural norms, towards the ‘end’ of both the working day and the working week. It seemed, for example, that work planning was geared towards Friday discharge to make space for admissions over the weekend and new cases to be discussed at the ‘start’ of the following week. This could be seen with the weekly MDT which, for all research sites, was in the middle of the week, with the intention of reviewing new admissions and planning the discharge of existing patients. In contrast, for those in the community setting, hospital discharge signalled the start of their clinical or caring activities with the patient and was generally preferred to start at the beginning of the working day or week.

The mismatch in preferences and perceived meanings of discharge, as either the end or beginning of care, seemed to frame many of the issues outlined above. For example, discharge delays often came about because social care services were not ready to accept new patients on Friday afternoons or over weekends. More importantly, this created the possibility of relatively minimal or poorly specified care immediately following discharge where many of the required assessments or start-up meetings for organising ongoing care were delayed, usually over a weekend. Many patients and family members described feeling coerced and abandoned immediately following discharge because the community nurse, social care team or GP was not ready to support the patient until the start of the following week. Drawing on participant narratives, the risks associated with timing of discharge are illustrated in Figure 13.

FIGURE 13.

Illustration of timing-related (latent) risks.

As outlined above, knowledge sharing was integral to determining and supporting timely care transition. Three specific knowledge-sharing dynamics stand out, which broadly correspond to the three phases of patient care outlined in the previous chapter.

In the acute care phase, the planning and scheduling of patient discharge was shaped by a range of cultural and organisational factors, such as at which stage a date of discharge should be set, but also by the range of actors involved in determining these decisions. At both stroke units, decision-making was broadly inclusive, with initial decisions made at the point of admission and regularly reviewed in daily and weekly meetings. At the orthopaedic units (especially Farnchester) discharge decisions were largely made by the surgical team in consultation with nursing staff, with limited involvement of others. Such differences appeared important when determining the patient’s readiness for discharge and the level of care needed following discharge. Where divergent views and perspectives were regularly brought together, decision-making seemed more thorough and robust; this helped to guard against premature discharge, in particular, but also allowed co-ordination with external services, whereby OTs or liaison nurses could relay difficulties in ordering or arranging ongoing care. Where decision-making was more fragmented and carried out by a limited range of clinicians, issues or problems associated with specific tasks or activities were not always adequately addressed. An important issue was whether or not these sites had the opportunities, and also the collaborative patterns of working, to address discrepant views between nurses, therapists and doctors.

A further set of problems related to the patterns of knowledge sharing between health and social care agencies. As described in the previous chapter, these processes were often protracted and complicated and revealed underlying cultural and organisational differences between these sectors. A challenge for social workers was meeting the growing demand for rapid referrals and the limited opportunities and time available for making robust assessments and care planning. A significant issue described at all sites was the shift from hospital- to locality-based social work teams. Although this made it easier for social workers to visit patients following discharge at home or in residential care, it made it difficult to assess need when in hospital or to meet with family members. As such, the opportunities for knowledge sharing with social workers seemed to have reduced. Furthermore, hospital staff described social workers as having limited regular and sustained contact with ward-based staff and rarely attending weekly MDTs. As such, the sharing of knowledge between health and social care was difficult and often fractious, leading to misunderstandings and accusations of delay. At Glipton, hospital staff described a strategy of issuing both referral notices simultaneously, therefore requiring social services to respond more quickly in order to avoid penalties. However, this also seemed to have encouraged further queries at the point of referral as a strategy to delay the process. Where services utilised a discharge liaison nurse, it seemed that many of these issues or tensions were resolved as these nurses acted as a regular contact point for social work teams to clarify issues or to arrange patient visits.

A third knowledge-sharing issue related to the involvement of family members and other community-based carers in the decision-making process. Where services had a more explicit and formal approach to involving patients and their relatives in decision-making, it was more possible to identify and address wider complicating factors associated with patients’ domestic or personal circumstances. Some significant issues related to family disputes, access to housing, the levels of support available and the financial arrangements for onward care. Involvement of relatives early and regularly within the planning process not only enabled more thorough knowledge sharing, but also increased trust and family involvement in care. As described above, helping relatives to understand the use of equipment or adherence to medicines was important. Equally, building trust with family members and gaining their involvement in care planning ensured that often crucial basic issues were addressed by the day of discharge, such as having the home prepared, providing access to equipment suppliers, enabling fitters to access the house to fit steps or grab rails, and being available at the time of discharge to support transition. One problem much discussed among nursing staff was that of family members going on vacation during their relative’s hospitalisation, which could delay discharge.

Communication

Notwithstanding the more direct or manifest safety events, such as medication errors, or the more ubiquitous concern with delays, communication problems and breakdowns featured widely across the narratives of participants. In part, this is arguably a reflection of the type of questions asked during interview and the over-riding focus of the study. That being said, three specific communication issues were widely described by participants as framing discharge safety, while the broader challenges of knowledge sharing are discussed below.

The first, and very broad, communication issue related to the quality of interpersonal relations and teamwork both within and across different care settings. Most participants talked of the importance of an open and inclusive approach to everyday clinical activities and discharge planning. It was described how the quality (safety, timeliness and completeness) of clinical assessments, care planning and task allocation depended upon the quality of face-to-face and interpersonal interactions. Where services had more regular and inclusive opportunities for interaction and communication, for example ward rounds, board rounds or weekly MDT meetings, participants usually described more possibilities for communication, which also framed a clearer understanding of what others contributed to the care process, a greater sense of teamwork and co-ordinated working practices. For example, daily board rounds were widely seen as providing a structured opportunity for people to come together and participate in shared decision-making by carving out time not typically found amidst daily tasks and activities. This opportunity was absent at the Farnchester orthopaedic unit, where clinical groups remained more isolated and independent, and where communication and teamwork were seen as problematic by all participant groups. Similar issues were also noted in relation to the interaction between health and social care teams, where the opportunities for communication and collaboration appeared to have been lessened by recent changes in social work organisation. Additionally, the reliance upon telephone consultations and regular changes in social work staff meant that it was difficult to build lasting and mutually beneficial relationships between these different communities. Looking across the research sites, there were marked differences in local cultures and the importance or value given to teamwork. In the stroke units, there was a clearer sense that teamwork and open communication were essential for patient care and discharge planning, whereas in the orthopaedic units there was a clearer distinction, and lack of communication, between surgical and non-surgical groups. Across all sites, however, staffing pressures, shift patterns and also the wider workload pressures were frequently described as undermining the potential for teamwork and robust communication.

A second and more specific communication issue was linked to the quality of handovers between clinical shifts, between organisational units and between care agencies. Participants described how handovers should provide the key information necessary to maintain and advance patient care, but too often involved relaying basic information which lacked necessary accuracy or clinical detail. Prominent examples of handover failure included the transfer of daily tasks or observation duties between ward shifts at the Farnchester orthopaedic unit; communication during the transfer of care of patients to community hospitals at Glipton; and, more broadly, the handover of care to reablement services. In each of these common examples, participants – usually those receiving the responsibilities for care – described how it was usual for communication to be problematic, with many gaps in their understanding of patient need, including highly important information related to observations or infection risks, as well as more general plans for onward care and rehabilitation. The use of standardised handover checks at all sites appeared to have improved daily communication patterns, and the regular use of board rounds, especially at the two stroke units, was described as improving general communication, task allocation and teamwork. Discharge checklists were used at the Glipton research site, but these seemed to be reserved for more complex patients, when managed by the discharge co-ordinator. There were no checking or documented procedures for supporting communication commonly used during care transition.

The third broad communication issue relates to the quality of documented knowledge, which could influence a variety of discharge planning activities, from prescribing to ordering equipment. A range of documents were regularly used during discharge planning and care transition, yet they were described as being variable in quality, with subsequent implications for the safety of discharge. Discharge letters often featured in the narratives of consultants and GPs as a formal basis for communication between the two. Although these were described as offering a clear, usually medically focused summary of the patient’s care in hospital, anticipated care needs, medication changes and future appointments, they were widely described as a problematic and limited form of communication. For GPs, they were seen as patchy or variable in quality, often providing limited detail, and were usually received long after the patient had left hospital. Some GPs seemed to regard them as a record of past activity, whereas others did seem to use them as a basis for ongoing care planning. Hospital consultants also described a sense of dissatisfaction with discharge letters which were, for some, an outdated and outmoded basis for interactions where instead shared or integrated ICTs should be used. As such, discharge letters were commonly seen as offering limited benefit to the advance of patient care.

Hospital (patient) records varied considerably between sites, which in turn seemed to shape a number of safety-related issues, such as the ease of sharing knowledge between clinical groups, making referrals or ordering equipment. At most sites, patient details and clinical activities were detailed in more than one document or record, usually a ‘medical record’, a ‘nursing record’ and a ‘therapist record’. This arrangement reflects the distinct knowledge requirements and recording procedures of different clinical groups. However, it meant that the knowledge contained within these different sources was typically dispersed and not easily reconciled as a basis for communication or shared decision-making. In MDT meetings, for example, it was usual for different clinical participants to read from their distinct records. A further issue was whether these records were colocated or stored in different hospital locations. At Glipton, therapists were permanently based on the stroke unit and stored their records in a shared office, making it easier for other clinicians to access this information. At the Farnchester orthopaedic unit, however, therapist records were stored in a hospital therapists’ department, making it difficult for ward staff to access relevant information when contacting, for example, external agencies. The main difference was at the Farnchester stroke unit, where a single patient record had been introduced that compiled the different records of each specialist. Although this could make it difficult for staff to simultaneously access the often bulky document, it ensured that all relevant information was recorded in one place.

Information communication technologies are widely described as enabling safer working through more detailed and accurate management of knowledge. A range of ICTs were described across all research sites, including general summary records, prescribing systems, equipment ordering and discharge summaries. Although these were usually described in positive ways, there were reported problems associated with accessing computers in order to complete prescriptions or update records, and in terms of whether or not different ICT systems were integrated. For example, GPs talked about the lack of integration with hospital systems, while one ward clerk described how she needed to use at least five separate computer packages to process one hospital discharge.

A further documentation issue related to the process of making referrals, both within the hospital for tests or scans and, more commonly, to external agencies such as social services. As described above, these issues represented common concerns related to delayed discharge or missing assessments. The problems described often related to underlying differences in understanding about care, or indeed the patient. Those in the health-care setting, for example, tended to talk about the need for detailing information relevant to the patient’s ongoing care needs, whereas those in the social care setting talked more about information required to understand his or her home, personal and financial circumstances. Referral letters generally offered little opportunity to provide sufficient detail of either type of knowledge, and therefore became a focal point for blame, requiring more personal or direct interaction between health and social care staff. For virtually all participants, referral processes were described as unsatisfactory and not enabling a robust or well-developed assessment of patient care needs. Drawing on participant narratives, the communication risks associated with hospital discharge are illustrated in Figure 14.

FIGURE 14.

Illustration of communication risks associated with hospital discharge.

As suggested above, communication features as a risk concern more broadly across the care pathway and often as a less explicit risk issue or one that is taken for granted. As such, participants often talked of communication breakdowns more in relation to other risks, such as in ordering equipment or making referrals. Moreover, the underlying dynamics or patterns of knowledge sharing were also revealed as framing or conditioning many of these risks (or sources of safety).

Roles

First, the study shows how certain key actors or observed roles made significant contributions to discharge planning and care transition, not necessarily in terms of their input into individual tasks, but rather through helping to co-ordinate the entire process, mitigate system complexity and facilitate knowledge sharing. In particular, discharge liaison nurses, community stroke nurses and ESD teams worked across occupational and organisational boundaries, often in the form of in-reach or outreach services. 53,61 Their contribution to discharge safety can be summarised along four lines. First and foremost, they supported the sharing of knowledge within and across different boundaries, helping actors in one setting to recognise and use the knowledge of another. This could be seen, for example, in the way community stroke nurses reached into the hospital setting to provide knowledge about community services, or discharge liaison nurses reached out to community health-care teams to co-ordinate onward care. Second, these actors often became, in themselves, repositories of knowledge, whereby they accumulated and recombined different bodies of knowledge found dispersed across the wider system. Operating as repositories of knowledge, they were often sought out by other actors for the purpose of rapid problem-solving and for guiding decision-making. For example, discharge liaison nurses were often sought out by others to navigate complex discharge processes. Third, through their developed knowledge of the wider care system and its stakeholders, these key actors often assumed a co-ordinating role by supporting the work of others. For example, discharge liaison nurses often supervised other ward nurses and therapists when organising discharge, or administered key paperwork. As such, they became informal or distributed leaders of discharge who would direct, supervise and monitor the work of others. Finally, these actors became the champions or advocates of the patient. They worked to develop a combined or holistic view of each patient, through pooling the knowledge and experience of different actors and working closely with patients and families. They also provided a sense of continuity for patients, working with them early in admission, through discharge planning and often providing follow-up care in the community (this was especially true of the community stroke nurse). This proximity to the patient provided a further level of insight or know-how, but meant that patients also had a source of knowledge or guidance when confronted with difficult decisions.

As well as the more obvious discharge liaison roles, the study found that other ‘hidden’ or backstage actors made similar contributions in discharge planning and care transition. This was exemplified by the taken-for-granted work of ward clerks, porters, voluntary workers, domestic staff and ambulance drivers. These actors also supported routine information exchange between groups, for example retrieving records, passing on memos or chasing up referrals. More than this, they acted as a repository of knowledge and contextual know-how of the wider care system. They also performed hidden ‘repair’ work, by double-checking referrals, chasing up missing documentation, reminding clinicians to complete tasks and dealing with the consequences of delayed or inaccurate communications. These hidden actors represented the ‘small cogs’ that ensured the overall operational performance of the system.

Elaborating the significance of these roles, knowledge brokers and boundary spanners are those actors who ‘get the right information, to the right people, at the right time’. 117,149 Knowledge brokers are able to work within and across engrained occupational and organisational boundaries, either through their roles within more than one group or by having extended social relations with a wider range of groups. 149 Through these relationships, knowledge brokers can identify, translate and support the use of knowledge between groups, thereby supporting learning and integration. As such, they are effective translators of knowledge as they are able to recognise, articulate and reframe the knowledge and working practices of one group, so it can be understood and acted upon by others. In the context of this study, these key actors possessed a form of ‘architectural’ knowledge of hospital discharge, i.e. they recognised how the wider system and its inter-related parts fit together, unlike other stakeholders who tended to have a more ‘component’ view of their individual contributions within the system. This architectural knowledge was integral to how they co-ordinated different activities, mediated gaps between staff groups and developed more collaborative practices. This enabled these actors to co-ordinate the complex discharge process, ensuring that decisions were reached, recorded and acted upon. As such, they reinforced the standards of effective discharge through promoting good practices and, in particular, placing the longer-term needs of the patient at the centre of their activities.

During the course of the research, the study found evidence of these roles being scaled back or discontinued as a result of funding constraints, leading to more fragmented knowledge sharing, reduced co-ordination and growing uncertainty in discharge planning. Although the reduced dependence on single actors can promote more shared and collective responsibility for hospital discharge, it seemed to reduce the overall co-ordination and coherence of discharge planning, limited active problem-solving and inhibited knowledge sharing across boundaries. As such, the study cautions against the loss of knowledge brokers and supports the creation of further opportunities for actors to bridge the gaps between acute and community care.

The study also found that a number of seemingly important groups were often marginal or had limited involvement in collective discharge planning and care transition. These groups included social workers, some community health-care services and GPs. Social workers, for example, appeared to have a highly limited, narrow or technical role in assessing ongoing care needs, checking the patient’s eligibility and arranging a package of care. There was little evidence, however, of social workers actively working with others in the hospital or community setting to produce more comprehensive or joined-up decision-making. Similarly, GPs were largely decoupled from the planning of patient discharge, despite having primary responsibility for patient well-being and continuity of care in the community. Few agreed to participate in the study, but those who did suggested that the existing patterns of communication with the local hospital provided only a partial and descriptive understanding of ongoing patient care needs. Moreover, they suggested that such information was often communicated after the patient had returned home, thereby placing the GP in a reactive position. As such, community nurses often acted as the primary contact point and conduit of information between the patient and the GP following discharge.

Relational opportunities

Second, the study found that discharge planning and care transition occurred through a complex series of linked activities and situations. These temporarily stable and spatially bounded situations or networks provided the primary opportunities for knowledge sharing, and together provided the framework or constellation of activities through which hospital discharge was progressed. Despite many of these activities and situations being informed by discharge planning policies,14,30 the study found significant variations in their form and function as well as the connections between them. In particular, it finds a number of interesting differences in the use of both formal and informal opportunities for knowledge sharing and, more broadly, the importance of physical or functional proximity in enabling knowledge sharing.

The study shows how formal situations, such as weekly MDTs and ward rounds, greatly facilitated knowledge sharing. In general, these represented a designated opportunity or ‘hub’ for sharing knowledge, addressing relatively common or regular tasks (i.e. agreeing the date of discharge or initiating referrals),47 formulating decisions and allocating resources. In short, these formal meetings were often the key decision-making points around which other activities or decisions were organised. That being said, the potential for these formal situations to contribute to discharge planning and care transition depended upon the range of actors involved, how they interacted with one another and the assumed purpose and ethos of the meeting. Specifically, knowledge sharing and shared decision-making was more apparent where these meetings were organised as more open and discursive opportunities to explore different perspectives, and where the over-riding purpose was to inform more comprehensive and patient-centred care planning. This was particularly evident in the stroke service. In contrast, others involved only a limited range of actors, were centred on information-giving and were primarily used to co-ordinate uniprofessional activities. This was exemplified by the Farnchester orthopaedic MDT meetings. Such differences explain the relative contributions of MDTs, for instance, to overall discharge planning, and where these situations offer limited opportunities for knowledge sharing, they also fail to contribute to the overall progress of hospital discharge. In such circumstances, stakeholders need to find alternative means of sharing knowledge. As such, the study suggests that MDTs should remain an important feature of discharge planning but based upon some form of inclusive representativeness of the wider workforce and on the assumption of collective decision-making.

Notwithstanding the contribution of weekly MDT meetings to discharge planning, a universal finding related to their position or timing within the weekly flow of care. In all sites, they were organised towards the middle of the week, for example Wednesday afternoon, on the assumption that this enabled more comprehensive assessment of newly admitted patients and the planning of care for those about to be discharged. However, this could make it difficult to escalate, arrange and complete discharge planning for patients by the end of the week, while creating time pressure on both ward staff and social work teams to complete tasks. Furthermore, it led to the increased likelihood that patients would be discharged on Friday with limited weekend care or would be kept in hospital until the following week. This not only highlighted a mismatch in the timing of work scheduling between health and social care but, more broadly, a mismatch between the standard working week (Monday to Friday) and the complexity of patient care, which not only spanned more than 10 days, but remained relatively constant over weekends and bank holidays. Aligning clinical decision-making to the patient pathway, and reducing reliance upon one midweek MDT meeting might alleviate these pressures, especially where more informal and daily knowledge-sharing opportunities are available.

Alongside these more formal situations, the study therefore highlights the contribution of more informal and daily knowledge-sharing opportunities. These included everyday encounters during ward-based care, handover, at reception areas, in corridors and during rest breaks. In these less structured situations, clinicians, patients and families often appeared to develop a sense of rapport and understanding that enabled the free flow of experiential know-how, related, for example, to psychological well-being, personal care or domestic circumstances. For example, it was often during the intimate acts of toileting or personal hygiene that patients would share apprehension about the ability of their relatives to provide this level of care. In other words, through these situations clinicians were able to develop a richer understanding or awareness of patients and their relatives concerning their ongoing care needs and, in particular, the use of certain devices and equipment to support their recovery outside hospital. These informal situations became a basis for experiential learning or education which then supported self-management and adherence. 150 These situations also enabled the flow of know-how between clinicians, whether working together in direct patient care or interacting during board rounds or other handovers. Through these everyday interactions, clinicians came to appreciate the expertise, practices and contributions of others and, consequently, more collaborative practices could emerge. The contributions of these informal knowledge-sharing opportunities to everyday knowledge sharing and integrated practice were reinforced in settings where they were absent or less common, i.e. where board rounds were infrequent or clinicians worked in a more sequential and siloed way, because these sites exhibited less shared understanding of discharge processes and more fragmented working.

The study suggests, therefore, that both formal and informal relational opportunities contribute to hospital discharge, but more significantly, it suggests that the interplay or connection between these formal and informal situations is integral to overall discharge planning and care transition. Specifically, the research shows how more routine and informal day-to-day interactions between clinicians, patients and families enabled the sharing of more tacit and experiential know-how about the patient’s personal circumstances. This experiential know-how was typically used to inform immediate decision-making and clinical care, but it was also relayed or shared within other, more formal situations, such as the ward round or the weekly MDT meeting, to enrich other forms of explicit knowledge and to enhance overall decision-making. In other words, the recombination of explicit information and experiential know-how was made possible through actors’ participation in multiple formal and informal situations. This suggests that where actors are confined to only the informal (or formal) situation, and not permitted to participate in both, there is less scope to draw together and use both the implicit and explicit forms of knowledge.

Developing this line of analysis, the study suggests that knowledge sharing relies upon open and reciprocal interaction where stakeholders are colocated or have ‘functional proximity’ in formal or, more often, informal situations. This everyday proximity enables actors to share their unique know-how, develop mutual understanding and common working practice, build trust and rapport and, in general, work more collaboratively. 117,126 This could be seen, for example, where therapists were permanently based on the acute wards, had routine interactions with other clinical groups, and more easily shared knowledge and collaborated with coworkers. Problems associated with a lack of colocation could be seen where hospital-based social workers had been transferred to community localities, which not only restricted opportunities for more routine interactions between health and social care teams, but fractured the sense of trust and understanding between these groups. Similarly, where therapists were based within a central hospital department, rather than on the given ward or unit, there was less evidence of knowledge sharing and collaboration with ward nurses. Although periods of service reconfiguration inevitably require adjustments, the loss of this functional proximity in discharge planning seemed to exacerbate the tensions between health and social care. As such, the study encourages the promotion of more routine and regular contact or functional interaction between different agencies around common problems, including the introduction of both formal and informal settings.

Culture and ethos

The study shows how each of the occupational groups and organisational units involved in discharge planning and care transition appeared to have, to varying degrees, a distinct culture or ethos of discharge. That is, they had distinct ways of making sense of discharge, prioritising or valuing discharge, allocating or taking up responsibilities for discharge and involving the patient in discharge planning. Differences and similarities in these cultures inherently framed the possibility for knowledge sharing and integrated working; for example, where actors held different priorities for discharge they were less likely to place the same value on key activities in the discharge planning process. This culture, or the ethos of discharge, is important because it frames the possibility for integration and knowledge sharing. Elaborating this idea, four specific features are emphasised as framing knowledge sharing.

The first relates to the way different groups made sense of or gave meaning to hospital discharge. A key area of difference was whether discharge was understood as an ‘end point’, a ‘beginning point’ or an ‘ongoing’ feature of patient care. Many hospital staff, for example, saw discharge as an end point that was positioned at the ‘end of the day’ or the ‘end of the week’. In contrast, community-based groups saw it as the beginning of patient care, which was reflected in the assumption that it should start at the ‘beginning of the day’ or the ‘beginning of the week’. This simple mismatch in understanding explained many complications in the integration and co-ordination of acute and community care, including when knowledge should be shared and when discharge should be scheduled. These mismatches often seemed to frame many of the breakdowns in communication and delays in care transition. Where different groups saw discharge in similar ways, often as a single stage in an ongoing care process, there was greater emphasis on working together to share knowledge and hand over responsibility.

The second cultural feature relates to the relative priority or value given to hospital discharge within the work and organisation of different care providers. Some groups, especially surgeons, appeared to prioritise the initial period of acute patient care, including diagnosis and surgery, but placed less emphasis on planning for onward care. For others, hospital discharge was a high priority that triggered or determined the pattern of their work, and was central to their overall work and identity, i.e. discharge co-ordinators and social workers. These differences did not simply reflect when or how discharge planning should commence, but rather the underlying significance or value of discharge to both the patient and the wider care system. When clinicians valued discharge highly as a key stage in the overall care pathway, they were more likely to value knowledge sharing that supported discharge planning and care transition. Further, a mismatch in the priority given to discharge helps explain some of the tensions observed between different groups, especially in regards to their lack of interaction or involvement in discharge planning.

Following on from the above, the third feature reflects whether discharge was regarded as ‘somebody’s business’ or ‘everybody’s business’. In some sites, it was apparent that discharge planning was a shared responsibility of the entire health and social care team, with the expectation that all groups participate in knowledge sharing and shared decision-making. In other settings, discharge responsibilities were delegated to individuals or subgroups, with the expectation that this freed up the work of the wider clinical team. Where discharge responsibilities were devolved or delegated it seemed to signal to others that there was little or limited need to participate in decision-making, and as such it reflected a less integrated or collaborative workplace. This could mean that knowledge gathered through informal situations was not valued as important or worthy of sharing in other, more formal situations. Although the delegation of discharge responsibilities to a single group (discharge co-ordinators) can help with workload management, a more inclusive approach seemed to support more collaborative working and mutual learning.

The fourth cultural feature relates more broadly to the relative value given to patient (and family) involvement in discharge planning. Arguably, patients are at the centre of their own care journeys and represent the only point of continuity between different care settings. In other words, they can be both the focus and medium of knowledge sharing. Where stakeholders actively involved patients and relatives in care planning, knowledge sharing allowed not only for a more bespoke and patient-centred care transition, but also for learning among clinicians where different members of the care team co-ordinated their distinct activities rather than addressing sectional or uniprofessional concerns. Although all sites claimed to recognise and value the contribution of the patient and family to discharge planning, only one site appeared to proactively involve these stakeholders in formal situations, including the use of designated family meetings and the routine use of daily clinical encounters to engage patients in the planning of their own care. The importance given to the patient at this site reflected a broader culture of inclusive and integrated working that supported knowledge sharing and discharge planning.

Together, these four cultural features highlight how a ‘culture of collaboration’ frames discharge planning and care transition. An ethos of discharge based on collaboration was more evident where actors shared broadly common understandings or meanings of discharge and similarly valued the importance of discharge – not just within their own work, but also for the wider service and individual patients – and where there was an assumption that all stakeholders, including patients, should participate in supporting discharge planning and care transition. Looking across the research sites, it is doubtful that such a culture of collaboration was evident across the wider care system, yet there was evidence of this culture in different organisational pockets. Furthermore, it seemed that where services utilised designated discharge co-ordinators, had enhanced levels of functional proximity and actively involved the patient in discharge planning, they also exhibited a strong sense of collaboration and integration based upon a broadly shared ethos of discharge.

Organisation factors

Finally, the study highlights the wider significance of organisational and interorganisational factors as influencing both the patterns of knowledge sharing and hospital discharge. Although each study site had a particular organisational and system-wide configuration, including systems of commissioning and governance, the study suggests that four common factors appeared to have an impact upon knowledge sharing and relate more broadly to the extent of integration and collaboration in discharge planning and care transition.

First, the discharge policies, procedures and pathways can have an important structuring influence on the patterns of knowledge sharing, especially through stipulating when decisions should be reached and communicated, and in some cases defining the anticipated pathway of patient care. 14,24,30,47,53,30,62,94 In general, the recommended use of MDT meetings and national guidelines for initiating social care referrals were strongly followed in all sites. 47,30 At the same time, the findings suggest that slavish adherence to such policies can potentially compromise the ability of stakeholders to work flexibly or to adapt to complex cases, especially patients with comorbidities or complicated personal circumstances. That being said, two particular policies or procedural features appeared to be particularly beneficial in supporting knowledge sharing. The first was the use of a discharge checklist or summary. 94 Although all sites had developed discharge summaries, these were rarely used, and often seemed to be completed post hoc as a bureaucratic exercise. 94 At the Glipton orthopaedic unit, however, such checklists had been modified by the discharge co-ordinator and were regularly used prospectively to collate key information, record outstanding tasks and co-ordinate the discharge process. It appeared that national frameworks and tools require some degree of translation and modification to fit with local practices and, more importantly, need to be valued by a local champion in order for them to be taken up.

The second procedural feature related to the setting and use of an expected date of discharge (EDD) to direct and co-ordinate discharge planning activities. 30 The study found that where an EDD was set, usually within 48 hours of admission, clinical work appeared to be more explicitly directed towards discharge planning, in the sense that it brought together and galvanised disparate groups towards a shared or common target. When adopted and made visible, for example on whiteboards, a preliminary EDD acted as a shared goal and point of reference against which different groups could focus their work. Sites that set an EDD tended to use it as a flexible rather than hard target, given that patient needs often changed, but it still helped to focus attention. For example, in daily interactions or weekly MDTs, the EDD became an important marker and point of reflexive assessment, i.e. ‘are we still on target?’. EDD setting can be seen as triggering consequential collective activities and knowledge sharing when embedded as part of the cultural norms of practice. At the same time, pressures to meet more formal organisational or managerial targets, such as the 10-day NoF pathway, were frequently considered as compromising professional practice. Although national policies and procedures can structure and guide discharge planning, the most significant contributions to knowledge sharing appeared to be the use of MDTs, the potential use of discharge planning tools and setting an EDD.

Second, and linked to the above, the study found that the availability, design and use of various technological and material resources strongly influenced the potential for knowledge sharing. This included more traditional paper-based records, as well as a large variety of information technologies, such as telephone, fax machine, computer software and hospital bleeper, which were all used on a daily basis for communicating between groups. In general, well-co-ordinated, easily accessible and legible patient records were integral to knowledge sharing. Where such records were developed along uniprofessional lines or in duplicate, or were stored within different organisational units, the sharing of knowledge became difficult. Conversely, where records were colocated and treated as an open resource, i.e. as a single care record, knowledge sharing was more apparent and discharge planning more co-ordinated. Furthermore, the excessive demands of completing paperwork were frequently described as delaying discharge activities, creating confusion and leading to communication breakdowns. As such, fewer and more straightforward methods of documentation might further simplify discharge planning. A significant resource was the use of whiteboards to help share knowledge and co-ordinate daily activities. These provided more real-time access to information and helped to structure daily planning of work. More than this, they acted as a type of departure board whereby each patient’s progress, date of discharge and outstanding tasks could be easily monitored and updated. Although these lacked the detail of patient records or ICTs, they often provided the primary ‘go-to’ source of information about a patient before seeking additional sources of knowledge either by referring to the records or consulting with other members of staff. In short, records and materials need to provide timely, rapid and robust access to knowledge within the parameters of routine day-to-day practice.

With particular regard to the use of ICTs, the study found no consistent approach. Where systems were used, i.e. for prescribing or ordering equipment, they generally supported less personalised but more timely knowledge sharing. For example, it was possible to request scans, medicines or equipment ‘out of hours’. However, a problem across all research sites was the fragmentation and duplication of computer programmes. For example, ward-based staff often needed to switch between different systems to prescribe drugs, order equipment, complete letters, gain access to hospital information or bring up referral forms for community health services. This took time and also required staff to retain different sets of login information. In some locations there was limited access to computer terminals, and queues and interruptions were a common occurrence. Furthermore, focusing on the processes of information technology could be seen as taking attention away from the content of communication, increasing the communication workload and the possibility of confusion or inaccuracy. This could be seen to contribute to errors in record keeping that were subsequently passed around the system.

Third, the study found that patterns of leadership influenced the general pattern of discharge planning and care transition. At all sites, there was an over-riding sense of medical leadership where doctors led MDTs, ward rounds and other key decision-making situations. In general, this type of medical leadership was important when drawing together and completing different sources of knowledge around more complex or ambiguous issues. 151 In many instances, the medical doctors provided a level of authority for decision-making, as well as for structuring the overall process. However, where doctors were less active in daily activities and offered less support for collaborative working, there also appeared to be less overall structure and coherence to discharge planning, for example within the Farnchester hip fracture service. Medical leaders could also have a strong influence on the ethos and culture of patient care in general, and discharge in particular. As noted above, services were characterised by different assumptions and customs around the importance of collaborative working, which often stemmed from the approach fostered by the senior medical leaders; this was particularly evident at the stroke units. A further example of medical leadership was seen where doctors interacted, and sometimes came into conflict, with service managers or administrators in regards to escalating ‘early’ patient discharges in order to meet service-level targets, for example to reduce delayed discharges or increase bed availability. Here, medical leadership was found to articulate the continuing needs of acute care, thereby countering the demands of bed managers or ward managers.

In less formal ways, examples of distributed leadership were also observed in day-to-day problem-solving around discharge. 152 For example, discharge co-ordinators and more experienced nurses provided support and guidance to more junior colleagues when they struggled to understand discharge procedures. In other settings, more ‘reluctant leaders’ (ward clerks, social workers, ESD leads) were found to ‘step up’ and ‘fire-fight’ in situations that seemed to be spiralling out of control, especially when a series of activities had not been completed on the day of discharge, for example organising TTOs, ordering transportation and contacting the family. The patterns of knowledge sharing supported by these informal leaders not only supported the discharge of individual patients but, more broadly, promoted collaborative learning and shared understanding among the wider care team. In the community settings, however, leadership seemed particularly absent. Although different social work teams, social care agencies and health-care teams had designated ‘team leaders’, these roles seemed either administrative in focus (social workers) or focused on localised clinical care. As such, there was little over-riding sense of leadership for ongoing care planning, except by some community stroke nurses, and knowledge sharing tended to remain isolated within individual care teams with no over-riding sense of collaborative or joined-up care.

Finally, the study highlights the influence of wider organisational priorities and resource constraints in shaping (both positively and negatively) discharge planning and care transition. Across all sites, there was pressure to improve performance in line with measures enshrined in national audit documents, chiefly the National Sentinel Stroke Audit (NSSA) and the NHFD. These identified factors such as length of stay, timeliness of treatment, involvement of relevant professionals and other best practice guidelines by which the organisation’s performance on particular conditions would be judged, with future financial implications for the organisation. These measures appeared to shape the broader organisation and management of ward-based priorities and planning; for example, skill mix, care pathways and availability of resources were partially linked to the priorities set out in the national measures. Such targets framed discharge planning and could, in many instances, help focus activities towards a common goal in much the same way as an EDD. However, this seemed to depend upon the type of medical leadership, for example where leaders could translate organisational targets into more clinical-facing priorities, and where they would mediate tensions between organisational and professional goals.

There was also important variation in priorities between health and social care organisations. Whereas hospital priorities were focused on progressing individual patients along a particular pathway, social services viewed discharges on a more collective level, with generic population-based guidelines relating to response time, fulfilment of care preferences and funding levels. As a result of this, social workers placed a greater value on assessment of ongoing need rather than patient pathway targets. This meant that there was a lack of shared understanding around the urgency or requirements of particular cases. In general, there was little evidence that health and social care agencies really appreciated the local pressure and organisational constraints that each faced, and which made integrated working complicated.

There were also pronounced financial and staffing constraints across all health and social care settings as a consequence of current spending restraints. This framed changes in the provision of social care services and the funding of specialist roles, especially discharge co-ordinators. More broadly, it was observed that staff levels placed considerable demands on routine working practices and limits on the opportunities for different staff groups to work openly and in collaboration around discharge activities. This is an inherent feature of contemporary health-care organisation, and calls for creative ways of using scarce resources to better co-ordinate the networks of activity necessary for effective care delivery.

Summary of key learning points

Knowledge sharing is likely to be enhanced, system complexity mitigated and discharge safety improved where:

• discharge co-ordinators, and other boundary spanners, are appointed to co-ordinate discharge planning and facilitate the exchange and use of knowledge across organisational and occupational boundaries between otherwise disconnected groups

• different health and social groups, and other stakeholders, are colocated and/or afforded functional proximity in day-to-day work, and local (informal) learning is actively utilised in more formal decision-making settings

• different stakeholders share the view that hospital discharge is an ongoing but crucial stage in a complex care pathway and it is everybody’s business, including the patient’s, to support and contribute to discharge planning and care transition

• locally adapted discharge procedures and checklists are used in conjunction with simplified, streamlined and shared record keeping and ICT systems

• formal and distributed leadership supports integrated working, and goals and targets are aligned to changeable patient- and service-level needs.

Overview . . .

Can you help me understand the discharge process? Maybe talk me through what you see as the steps for patients on your ward/you see in the community?

When do you start thinking about discharge plans?

Which main factors do you consider?

What happens when delays occur?

Are there any particular days which are more difficult? Why?

Are there any particular times of day which are more tricky? Why?

Are there any particular ways in which you work around these anticipated difficulties?

What happens when patients are all ready to go but have to wait for transport?

Do you get to hear if a patient is readmitted soon after discharge? If so how does that make you feel?

Planning issues . . .

And in your day-to-day work, what input or responsibilities do you have in terms of patient discharge?

Can you identify a point during the discharge process at which you have completed your task?

How do team meetings, staff briefings or other group activities help with the discharge planning? What do they help to do?

It seems like there are lots of forms and paperwork in the discharge process? What do they aim to do? How do they support you? Are these easy to complete or use?

Do you follow the guidelines to the letter?

Are there any parts of your job which don’t fit into any of the discharge guidelines?

Do you have your own way of doing things?

What kinds of technologies or computer systems are used during the discharge process? What do they aim to do? Are they easy to use? Does everybody use them in the same way?

Patients and families . . .

Do families understand their role once a patient is fit to go home?

Do patients understand that they are fit to go home?

What about immobile end-of-life patients? Are they accompanied by a nurse?

What happens when a patient self-discharges?

The wider context . . .

In addition to yourself, who do you see as the main people involved in facilitating or managing the discharge process? How do they go about their work and what do they contribute to the process?

What are the key organisational factors that shape the discharge process, such as management approaches or resources? Can you give me an example?

As far as you know, what interventions or changes have been used to improve the way people work with each other across the hospital?

As far as you know, what interventions or changes have been used to improve the way people work with each across between the hospital and the community?