Background
Stroke is a major public health issue worldwide and especially in industrialised countries [
1]. Incidence rates range from 73 to 223 cases/100 000 citizens/year according to region [
2,
3]. The disease is the second, sometimes even leading, cause of death in many countries [
4,
5]. It is also a major cause of disability and cognitive decline, of worsening of quality of life, and of dependency in one in two patients [
6]. In 2008, in France (population, 64.3 million), 105,000 people were admitted to hospital for stroke and 31,000 for transient ischemic attack (TIA), with 24 to 38 stroke-attributable deaths occurring per 100,000 persons per year [
7].
Stroke Care Pathways (SCPs) have been set up in North America and several European countries [
8‐
13]. Their general framework embraces multidisciplinary care, care coordination and overall cooperation, and includes the following components: first contact with providers, notification of emergency medical services (EMS) and their response to ensure speedy access to stroke units, acute and sub-acute stroke care, access to rehabilitation and chronic care centres, and provision of home care for dependent patients. Patient access to well-organised, multidisciplinary care in stroke units has been shown to improve quality of service and ultimately to reduce stroke morbidity/mortality rates [
14]. Nevertheless, there is little sound evidence for an unambiguous causal relationship between mortality and process measures in stroke care [
15]. The development and maintenance of complex cross-boundary SCPs therefore remains an important challenge for clinicians, healthcare managers, and policymakers [
16,
17]. Fragmented pathways, suboptimal care coordination and poor staff collaboration have hampered the translation of major advances in diagnosis and treatment into clinical practice, resulting in wasted resources and disappointing outcomes [
18‐
20].
The aim of this study was to identify the barriers encountered by healthcare professionals in France when implementing SCPs. Barriers are defined as the causes of discontinuities encountered in daily practice, a continuity being defined as “the delivery of services by different providers in a coherent, logical, and timely fashion” [
21]. The study was part of a national project for the development of quality indicators (QIs) for care pathways for use by French regional healthcare agencies. The development of QIs specifically for hospital use requires a review of the literature and of clinical guidelines. The development of QIs for care pathways requires in addition a qualitative analysis of existing care pathways in order to highlight regional differences and reveal the main factors influencing quality of care [
22].
Discussion
Our study has identified barriers that health professionals consider to be major obstacles to effective implementation of a SCP. These barriers were identified by analysis of 4 SCP types and related to three main quality concerns: lack of resources, lack of training and public information, and lack of coordination between staff or healthcare facilities. Problems relating to best clinical practices, as translated into evidence-based protocols, were not considered.
The barrier most often mentioned was lack of resources, confirming that France needs to invest more in stroke management despite noteworthy results in some parts of the country [
34,
35]. The objectives of the ongoing 2010–2014 national Stroke Plan are not only to increase the number of stroke units and to introduce coordinators to increase SCP efficiency, but also to improve training and public information, and to extend stroke care beyond prevention and primary care by setting up a coordinated set of programmes and actions for social care [
32]. Between 2010 and 2030, stroke prevalence is estimated to increase by 25% and direct medical treatment costs for stroke care to triple [
8]. However, in the current political and economic climate, it is unlikely that more cash will be forthcoming for public health spending. Continuity of care will therefore probably decline as resources dwindle (fewer primary care physicians, more part-time physicians, and greater divisions between hospital consultants and GPs). These shortcomings will need to be offset by better organisation of care.
The interviewees considered that patients and public should have better knowledge of stroke symptoms, and that triage nurses and young residents should be better versed in stroke care in general. Information campaigns on the early recognition of stroke symptoms and on the seriousness of the disease need to be set up nationally, regionally, and for special population subgroups (e.g. relatives of acute stroke patients) [
36], despite the current lack of evidence for the effectiveness of such interventions [
37]. QIs developed for regional health agency use should take account of these knowledge and information needs, and should be developed from a holistic perspective that covers personal risk factors for stroke, warning signs of stroke, EMS activation, need for follow-up after discharge, and medications management, and could also include the consumer perspective [
38] and patient experience [
39].
The third barrier to be identified was poor coordination between staff and between healthcare facilities. The issue of poor coordination has already been raised by physicians [
26,
40,
41] and has been shown to impact on both quality and costs [
42]. Reported examples of poor coordination are lack of treatment follow-up, conflicting information from providers, delayed information transfer following hospital discharge, and unavailability of relevant information during patients’ scheduled visits. Care coordination is increasingly seen as an evidence-based aspect of high-quality health care delivery but needs to be integrated both horizontally (peer-based and cross-sector collaboration) and vertically (patient pathways transcending organisational boundaries) in order to result in healthcare benefits [
43]. There is, however, no consensus about what constitutes care coordination, how to build an evidence base for care coordination [
44], and how to develop coordination measures [
45].
We shall consider coordination within the pre-hospital setting, hospital setting, and post-hospital setting in turn. In the pre-hospital setting where “time is brain”, our interviewees considered that closer collaboration was required between pre-hospital EMS and hospitals for prompt patient transfer to a facility with appropriate resources. This is an evidence-based measure that has been underscored by the Stroke Systems Work Group in California [
9] and by the Madrid acute stroke care programme [
12]. Admission delays have already been considerably reduced in a regional network in France by close collaboration with pre-hospital EMS [
46]. However, an outstanding problem that the French regional health agencies will need to solve is that the fire brigade, when called out by the patient or by the GP, will not transport a patient to a hospital outside its jurisdiction area. This causes misunderstandings and conflicts between the fire brigade and SAMU. The interviewees also complained of lack of communication and cooperation within the hospital setting, with emergency physicians and neurologists often in disagreement over what was appropriate care and with too little liaising among staff (e.g. between emergency physicians, radiologists, laboratory personnel, neurologists). Proposed corrective actions are team-working, implementation of critical pathways in stroke units [
12,
41,
47,
48], recognition of the key role played by nurses [
49], and the setting up of stroke teams in hospitals with no stroke unit or with scant resources [
50]. Coordination with EMS has led to more widespread use of thrombolytic therapy [
24]. The main coordination problems encountered after patient discharge from hospital were poor information transfer from hospital to downstream staff, on the one hand, and delays due to administrative procedures, on the other [
51]. Solutions to such problems might be holistic initiatives such as, for instance, integrated PRM care pathways [
6,
52] and early supported discharge procedures which have been shown to reduce long-term dependency, admissions to institutional care facilities, and length of hospital stay, at least in a selected elderly group of stroke patients with moderate disability [
53]. In summary, for a SCP to function well, health professionals need to collaborate closely. The introduction in France of SCP coordinators involved in setting up and running SCPs may help promote collaboration.
The barriers to effective SCP implementation depended on geographical location (urban/rural (mountains or plains), distance to stroke unit) and patient background (isolation, socio-economic and educational status). Inequalities in access to healthcare by area of residence have already been highlighted [
8,
9,
26]. In semi-rural areas around Paris (e.g. Seine et Marne), the stroke centre may be as far as 80 miles away. Air transportation has thus to be developed to ensure equitable access for all. Improvement initiatives and their costs will vary according to the healthcare system and the facilities available within the system [
42]. Discrepancies might be reduced and higher quality care might be provided by adopting initiatives developed in other large rural areas (e.g. Alberta’s Provincial Stroke Strategy and Telestroke program) [
13].
Our study has limitations: (i) it is not an exhaustive overview of the implementation of SCPs as the interviewees did not mention some key evidence-based measures facilitating implementation, no doubt because they had not experienced these barriers. Such crucial “facilitators” include the importance of “champions” or “fixers” for successful policy implementation [
13,
54], the benefit of establishing clear, relevant, flexible policies to motivate and focus stakeholders on key issues [
9,
10], and the positive role of health care competition and incentives (not restricted to financial ones) to facilitate a wider adoption of policies and to promote improved quality of care [
29,
55]; (ii) our study relied on interview data and documentation and not on real-world observations. Observations during follow-up of patients with stroke are particularly difficult, especially in emergent situations; (iii) data provided by 44 volunteer interviewees from just four regions of France cannot provide a picture of the overall national context; (iv) we did not interview patients or their next of kin about their experiences or for their views on an ideal SCP; (v) our findings apply to the French healthcare system and may not be more widely applicable even though they are often supported by findings obtained elsewhere.
Our findings have practical implications for the development of QIs for SCPs for use by French regional healthcare agencies: (i) Improvement initiatives will have to focus on flaws in organisational aspects, especially on coordination between staff and between facilities, as the major barriers identified concerned organisational issues rather than best clinical practices [
44]. Developing QIs for networks, in which often competing facilities have to cooperate, is more complex than developing QIs for a single facility. QIs for networks relate to the dynamics of a social practice rather than to reified procedures and standards, and have to confer visibility to routine coordinating actions [
56]; (ii) More than one type of SCP will need to be validated as SCPs depend on local geography but the number should nevertheless remain limited [
57]; (iii) QIs for assessing SCP quality in different settings will have to be metrologically validated. Preliminary work has been carried out at the European level [
58,
59] and in France for the referral of patients from acute to rehabilitation care [
60]; (iv) Recent experiences support the assumption that well-organised, multidisciplinary care can improve quality of service and reduce stroke-related mortality and morbidity rates [
40,
61] for acute stroke [
14,
15,
41,
62] as well as for stroke rehabilitation [
18].
Competing interests
The authors declare they have no competing interests.
Authors’ contributions
KG made substantial contributions to design of the study, participated in the acquisition of data and was involved in the drafting of the manuscript. HL and MF participated in the acquisition of data and in discussions on data interpretation. GN was involved in the conception, design, and vetting of the study and in the drafting of the manuscript. FW and EM oversaw the work (FW from the viewpoint of the regional healthcare agencies and EM from that of overall healthcare management) and critically revised the manuscript. All authors read and approved the final manuscript.