Objectives To synthesise the published literature on the patient experience of the medical primary–secondary care interface and to determine priorities for future work in this field aimed at improving clinical outcomes.
Design Systematic review and metaethnographic synthesis of primary studies that used qualitative methods to explore patients’ perspectives of the medical primary–secondary care interface.
Setting International primary–secondary care interface.
Data sources EMBASE, MEDLINE, CINAHL Plus with Full text, PsycINFO, Psychology and Behavioural Sciences Collection, Health Business Elite, Biomedica Reference Collection: Comprehensive Library, Information Science & Technology Abstracts, eBook Collection, Web of Science Core Collection: Citation Indexes and Social Sciences Citation Index, and grey literature.
Eligibility criteria for selecting studies Studies were eligible for inclusion if they were full research papers employing qualitative methodology to explore patients’ perspectives of the medical primary–secondary care interface.
Review methods The 7-step metaethnographic approach described by Noblit and Hare, which involves cross-interpretation between studies while preserving the context of the primary data.
Results The search identified 690 articles, of which 39 were selected for full-text review. 20 articles were included in the systematic review that encompassed a total of 689 patients from 10 countries. 4 important areas specific to the primary–secondary care interface from the patients’ perspective emerged: barriers to care, communication, coordination, and ‘relationships and personal value’.
Conclusions and implications of key findings Patients should be the focus of any transfer of care between primary and secondary systems. From their perspective, areas for improvement may be classified into four domains that should usefully guide future work aimed at improving quality at this important interface.
Trial registration number PROSPERO CRD42014009486.
- PRIMARY CARE
- QUALITATIVE RESEARCH
This is an Open Access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/
Statistics from Altmetric.com
Strengths and limitations of this study
This review employed rigorous and established methodology specific to qualitative studies throughout.
The quality of the included studies was assessed using a published framework (Critical Appraisal Skills Programme, CASP) and found to be uniformly high.
The independent analysis of full articles by two researchers generated themes, concordant between reviewers, in a transparent and reproducible manner and was considered helpful in maintaining a coding balance.
Given reciprocal translational analysis provides summaries in terms that may have already been used in the original literature, there is acknowledgement that the synthesis will tend towards the privileging of a priori over in vivo codes.
Patient preferences and experiences were synthesised from varying primary–secondary care interfaces in differing healthcare systems serving different populations, but are gleaned only from the published literature. Study findings may therefore not be generalisable to all situations.
In numerous countries, primary care is the usual first point of professional contact for patients with a medical complaint. Many conditions are managed in primary care though some require more specialised medical expertise or treatment, necessitating access across the primary–secondary care interface to specialist attention, usually in hospitals as inpatients or outpatients.1 ,2 For patients who frequently transit the primary–secondary care interface, such as those with chronic conditions, coordination between the different disciplines is essential for the delivery of quality care.3 Since primary and secondary care clinicians often have different perspectives and can act in separate ‘professional tribes’, it is important that any inconsistencies across the primary–secondary care interface do not impact on the effectiveness and safety of patient transitions.4
Although manifestations of the primary–secondary care interface across the world are diverse and the disciplines involved differ, similarities can be identified in most healthcare systems.5–7 In countries where general practice (or family medicine) is well developed, there are many similarities in the functions and characteristics of the primary–secondary care interface-based system, with general practitioners (GPs (or primary care physicians—PCPs)) usually acting as ‘gatekeepers’ to secondary care.2 ,5 In many nations, primary care is largely delivered by non-medically qualified practitioners, with less formal access to secondary care, and in others there is direct access to primary medical care provided by specialists, such as paediatricians, gynaecologists, specialists in internal medicine and cardiologists.2 ,5 ,8–11
In countries with ‘gatekeeping’ primary care systems, there has been increased focus on the interface between primary and secondary care,12–17 highlighting the importance of better relationships between hospital and community, and between specialist and PCP, for the benefit of patient care.2
Qualitative studies describing experience at the interface have highlighted the importance of good access to patient-centred care (ie, that which is respectful of and responsive to individual patient preferences, needs and values),18 ,19 but the amount of patient-focused work is limited.
Using a metaethnographic approach (a method for synthesising qualitative research studies) described by Noblit and Hare,20 this study aimed to identify what patients perceive as important markers of care quality at the primary–secondary care interface. Such an approach may generate greater understanding than a single empirical study and may be helpful in determining the direction of future work to improve clinical outcomes.21 ,22
The study was registered with the PROSPERO database, registration number CRD42014009486.
The seven-step model of metaethnography described by Noblit and Hare20 was used. The first step involved a clear statement of the specific research question (What do patients perceive as important markers of care quality around the primary–secondary care interface?) and the contribution it will make to the field (to generate new insights, achieve greater understanding of the issues facing patients at the primary–secondary care interface than with a single empirical study, and determining targets for future research with the aim of improving patient outcomes). In step 2, a search strategy was devised to retrieve articles related to this aim. The search was focused to locate primary studies that met the following criteria.
Eligibility criteria for study inclusion
Studies were eligible for inclusion if they met all of the following criteria:
Employed qualitative methodology (focus groups or interviews);
Explored patients’ perspectives;
Targeted the medical primary–secondary care interface (ie, at the interface between PCP and secondary care hospital specialist);
Full research papers (ie, not an editorial, conference poster or abstract).
Study exclusion criteria
Studies were excluded if any of the following were present:
Did not explore patients’ perspectives;
Did not focus on the medical primary–secondary care interface (eg, studies focusing on the interface between social care and tertiary care would not be included);23–27
Not full research papers (ie, an editorial, conference poster or abstract).
Information sources and search strategy
Electronic databases were searched using database-specific terms and validated methods for retrieving qualitative studies (EMBASE (OVID 1974 to 30 July 2014), MEDLINE (OVID MEDLINE 1946 to 30 July 2014 with daily update), CINAHL Plus with Full text (EBSCO host accessed 30 July 2014), PsycINFO, Psychology and Behavioural Sciences Collection, Health Business Elite, Biomedica Reference Collection: Comprehensive Library, Information Science & Technology Abstracts, eBook Collection (EBSCO host, last accessed 30 July 2014), Web of Science Core Collection: Citation Indexes and Social Sciences Citation Index (database inception to 30 July 2014), and grey literature sources ((Open SIGLE (opensigle.inist.fr), last accessed 7 August 2014), (Health Management Information Consortium 1979 to May 2014 incorporating Kings Fund Information and Library service), (National Technical Information Service http://www.ntis.gov/, last accessed 7 August 2014) and (PsycEXTRA http://www.apa.org/psycextra/, last accessed 7 August 2014)) to identify literature using qualitative methods (focus groups or interviews) exploring patients’ perspectives of the primary–secondary care interface.23–27 No language or date of publication limits was applied to the search. Refer online supplementary appendix 1 ‘Search terms’ for specific detail of search used for each database. Authors of included studies were contacted to determine any key papers in the field of interest not identified by our own search strategy.
One author (RS) examined titles and abstracts of all retrieved citations for eligibility according to the above criteria. The full-text articles of any abstracts classified as definitely or potentially suitable for inclusion were retrieved and analysed independently by two authors (RS and JC) against predefined inclusion and exclusion criteria with differences resolved by consensus.28 See online supplementary appendix 2 ‘Excluded studies’ for details of and reasons for study exclusion. Reference lists of all included studies were scrutinised for eligibility according to inclusion/exclusion criteria. Main authors of all included papers were contacted to explore the potential for any studies considered important to them that may have been missed in our search strategy.
No quality filters were applied prior to inclusion of studies in the systematic review. However, the Critical Appraisal Skills Programme (CASP) tool for assessing qualitative research29 was applied postinclusion giving insights into the methods used for data collection and analysis (see online supplementary appendix 3 CASP review).
Step 3 of the metaethnographic synthesis involved reading the studies. Two authors (RS and JC) read and re-read the included studies, and independently listed the main themes from each article including both first-order (views of the participants) and second-order interpretations (views of authors). Where patients were interviewed with another healthcare professional, the analysis was restricted to the views of the patient where possible. Data were abstracted into standard fields, such as study aims, design, methods, setting and participants (see online supplementary appendix 4 ‘Data extraction template’),30 and entered into QSR International's NVivo V.10 software to assist qualitative analysis and synthesis.31
In step 4, two of the authors (RS and JC) determined how the studies were related to each other by comparing individual study findings, and derived key concepts that reflected the main findings of all included studies. Subsequently (step 5; studies were translated into each other) each study was re-examined and assessed for its relevance to these key concepts. In the same way that primary study moves from descriptive to explanatory analysis, these translations were then synthesised (step 6) to develop third-order interpretations (higher levels of abstraction) to represent the overarching perspective of patients at the primary–secondary care interface. Members of the research team (RS/RB/PW) developed this third-order interpretation or ‘line of argument’ synthesis by listing the translated themes and subthemes (derived from first-order and second-order constructs reported in the primary studies), then review and discussion.
The final step involved expressing the results of the synthesis using tables, figures and text in accordance with the ‘Enhancing transparency in reporting the synthesis of qualitative research’ (ENTREQ) statement (see online supplementary appendix 5 ‘ENTREQ statement’).32
The electronic database search returned 690 citations, leaving 654 after removal of duplicates (see figure 1).
A further 618 articles were excluded after scrutiny of the title or abstract for using non-qualitative methodology (n=226), involving participants other than patients (n=195) or because they did not concern the experience of patients at the medical primary–secondary care interface (n=197).
The full texts of all remaining 36 articles were retrieved and evaluated independently by two authors (RS/JC) against predefined criteria, leading to the inclusion of 20 papers (table 1). See online supplementary appendix 2 ‘Excluded Studies’ for details of, and reasons for, study exclusion.
Included studies originated from 10 countries and comprised a total of 689 patients (range 7–53 per study). Two studies used focus groups alone, 10 utilised solely individual patient interviews and 8 used both methods. The overall quality of the 20 included studies was high, with all articles meeting the majority of CASP criteria. One common weakness was around whether ethical issues had been taken into consideration (not clear in 5 of the 20 studies).34 ,37–39 ,47 The other common limitation was a lack of evidence demonstrating that the relationship between researcher and participants had been adequately considered; particularly whether the researcher(s) had critically examined their own role and potential for bias in formulation of topic guides, data collection, recruitment, study location, and whether they considered the implications of any changes in the research design (not reported in 17 of the 20 studies).35 ,37 ,38 ,40–42 ,44–50
Translation of included studies
Four key concepts that reflected the principal findings of all included studies were determined: barriers to care, communication, coordination, and relationships and personal value (table 2). Within each key concept, subthemes arose that are highlighted in bold.
Barriers to care
There was a strong patient perception that lack of PCP knowledge could be an obstacle to the delivery of good clinical care; “If it's of a serious nature, I certainly wouldn't take the GP's word for it.”33 ,46–48 Conversely for some patients, this apparent knowledge deficiency was not a practical problem, as their PCP referred them on.33
Some patients described the PCP acting as a barrier (the one who controls your entry into the system) they had to be overcome; “I was lucky, I didn't have to go to my GP because I collapsed in church.”34 ,35 ,46
Further system problems were identified, such as perceived delays to specialist clinic appointments, postponements and cancellations. When at specialist clinics, patients described long waiting times, frequent consultation interruptions and a lack of time with clinicians.38–40 ,46 Patients spoke of being disempowered by system complexity,48 bureaucracy resulting in disjointed care (“There is a distance between the hospital and primary care. Each part protects his own territory”4) and personal stress.33
Some patients reported that scheduled specialist follow-up was often unnecessary when they were well and described frustration at having to organise their lives around superfluous appointments.43
There were conflicting patient views of the effectiveness of communication across the interface. Some informants were satisfied, while others cited that inadequate communication between specialist and PCP could lead to provision of contradictory information by healthcare professionals, causing patient confusion.33 ,35 ,37 ,38 ,45 ,47 Use of informal routes of communication was mentioned as effective means of overcoming fragmentation in care.33
Patients believe good communication with their doctor requires time; established trust, understandable information and good clinician interpersonal skills.18 ,35 ,37 ,39 ,40 ,45 ,46 ,48 Conversely, patient stress, inadequate information or a sense of not being listened to increased the perception of poor communication.37 ,38 ,40 ,44 ,47
Certain patients responded to poor communication by becoming more proactive (eg, involving family members to act on their behalf, preparing lists of questions or quarrelling with clinical staff), while others became more anxious and uncertain.18 ,40 However, patients did not lay all responsibility for poor communication on healthcare professionals.38 Some perceived that their own lack of personal knowledge, lower comparative social status (eg, when attending a hospital specialist) or physical condition at time of consultation inhibited their ability to communicate with clinicians.18 ,37 ,38 ,47
Problems with the transfer of information at the interface were considered a significant cause of three major concerns:
Delays in care delivery: “considerable numbers of patients were dissatisfied with the length of time it took for the GP to be given word from the specialist”;38
Frustration: “Separate clinics don't talk to each other or ring each other. I find the whole thing incredible the length of time it takes; it's just been horrendous, waiting weeks to see a consultant to be told ‘I don't know why you've been referred to me’…it can make you feel very insignificant”;18
Inaccuracy, for example, incomplete hospital discharge information. In addition, lost referrals, professional schedule inflexibility, poorly communicated processes and resourcing issues highlighted complex difficulties.18 ,37 ,40 ,45 ,47 ,48
Some patients preferred their PCP to coordinate care since they could see the ‘whole picture’33 ,35 ,37 ,39 ,45 and others described increased confidence about discharge from hospital when they believed their PCP managed the ‘gatekeeping’ role well and would be willing to refer or re-refer when necessary; “I’ve every faith in him…Any problems I’ve got I just pick up the phone. He’s a very good doctor. He gets down to it you know and if he’s not sure he says ‘Right, hospital!’”38
Where formal care coordination roles were not clear, patients and family members often assumed this role in either a proactive or passive fashion.33 Proactivity was associated with provision of medication lists, a felt need to be assertive, initiation of PCP follow-up postdischarge and intentional modification of personal behaviour and clinical information disclosure to facilitate progression through the system; “I have to ‘play act’ when I see a GP. So I have to pretend that I am really ill and about to die before anything actually happens…I don't have a lot of faith in them.”37 ,40 ,44 ,46
Coordination of patient care could be influenced by the role of their family or carer; “I wasn't sure I was going to agree to the hospital tests. It was the family that changed my mind,”35 ,48 and the level of information provided; “lack of information perpetuated patients’ feelings of an imbalance in status and power, and reduced their sense of being involved in their own care.”18 ,37 ,45 ,48
Specialist nurses and those with specific care-coordinator roles were valued by patients in terms of access, liaison with specialist and PCP, thoroughness, and acting as a point of continuity within the hospital system.33 ,36 ,37 ,43 ,46 ,48
Relationships and personal value
Patients appreciated sympathetic ongoing relationships with their PCP; “He tries to help me, he is a really understanding doctor. He understands how I feel. I can really talk to him. He knows how I feel. I tell him where I am having the pain. I relate to him.”38 ,45 ,46 Good relationships with PCPs based on trust and understanding increased patient confidence to progress through the system18 ,45 and was thought central to help seeking. In particular, continuity of PCP was considered essential to build relationships in which patients felt safe to disclose concerns.48 Conversely, patients were less confident when the relationship with their PCP was poor or they were unable to see their usual PCP.18 Informal patient relationships with clinicians were sometimes used to facilitate progress through the system; “A patient with cancer must have connections, otherwise he gets lost in fairyland: go there, come back, wait, and so on.”33
Patients’ personal perceptions of hospital care were related to the quality and consistency of their relationships with secondary healthcare professionals.39 ,43 Patients can sense they are not valued with predictable consequences; “I think you feel a bit like an accessory, you've got this great big medical system and you're not really part of it, the system rolls on whether you're there or not…as a patient I thought the system was there because of you, not you there because of the system. It's this great big wheel of medicine going round and round and you're an insignificant speck.”18 Attitudes of staff in practices and outpatient clinics can make patients feel ‘in the way’ and powerless to challenge failures in the system.18
Patients also noticed tensions in the relationships between primary and secondary care, some expressing the view that PCPs and hospital doctors were not working together; “Is there maybe problems that the doctors are frightened to refer people to the hospitals-you know, they are the small fry and the hospital the big fry?”4 ,34 ,44
Third-order interpretations and ‘the line of argument’
Barriers to care
Many patients perceived low levels of PCP knowledge as being a barrier. Access to the PCP was highlighted as a difficulty. Once in the system, clinician and staff attitudes were described as obstacles. Patients identified organisational and system problems as hindering progress.
Clinicians on both sides of the interface require good interpersonal skills and a patient-centred approach in order to communicate effectively. The effect of poor communication with patients (patient uncertainty, anxiety, unwillingness to communicate and the potential for quarrelling with clinicians) should not be underestimated.
Some patients feel able to take on an active coordinating role in order to progress within the primary–secondary care interface. Patients value those healthcare workers with formal care coordinator roles (eg, specialist nurses).
Relationships and personal value
Good relationships with clinicians and staff, described as sympathetic, understanding and trusting promote patients’ sense of being valued and influenced disclosure of concerns, help seeking, compliance and confidence about referral and progress in the system. Patients sometimes used informal relationships with clinicians in order to advance through the system.
This systematic review, to our knowledge the first attempt to synthesise the qualitative literature exploring the patient experience at the medical primary–secondary care interface, has led to a broader description and fuller understanding of the range of challenges that exist at this critical point in care delivery.
Patients encounter multifaceted dynamics at the interface including barriers to care, communication, coordination of care issues and the impact of relationships and personal value.
Comparison with other research
Barriers to care
Patients in several studies perceived low levels of PCP knowledge as a barrier to care, one specific example being the patient experiencing a delayed diagnosis of lung cancer which they specifically related to a lack of PCP knowledge.48 They contrasted the knowledge of the PCP and specialist, seeing the consultant as the ‘expert able to delve deeper’, in contrast to the PCP; “You need specialist input for some things, I don’t think that GPs have enough knowledge.”47 This subtle negative view of PCP knowledge base perhaps belies a lack of understanding of differing roles of PCPs and specialists in a ‘gate-keeping’ interface context. There may be potential for further clarification of this phenomenon, perhaps including a triangulated approach involving patient and peer assessment of PCP knowledge. Clinicians have a professional responsibility to maintain up to date knowledge and skills throughout their working career, and to regularly take part in educational activities that maintain and further develop competence and performance.51
Systems need to be improved, so patients (and their information) can travel seamlessly across the interface between primary and secondary care. Medical notes should be complete, accurate and accessible to all relevant care providers.52 However, ‘information shared between GPs and hospitals when a patient moves between services is often patchy, incomplete and not shared quickly enough’.53 PCPs frequently receive hospital discharge summaries with deficient or inaccurate medicines information and delay in transmission of such information from secondary to primary may adversely affect patients.53–56 For example, in one study, PCPs felt it may have been important to carry out a follow-up home visit had they known that the patient had been discharged from hospital.56 The quality of patient information provided by PCPs to hospitals can also be substandard with regard to previous drug reactions, comorbidities and allergies.53 There certainly remains room for improvement54 ,57–69 and a need to develop and evaluate interventions to develop the content and speed of information sharing between primary and secondary care.70
The effect of poor communication skills on patients on both sides of the interface should not be underestimated: poor communication is an increasingly frequent reason for complaint71 when patients feel ‘disempowered and disengaged’ and patronised by clinicians.72 Inadequate time spent by clinicians with patients is associated with increased likelihood of malpractice litigation, largely due to poor communication.73 ,74 Cultural factors and educational background may influence clinician communication abilities.75 The question of how factors such as age and communication skills training impact on the doctor–patient relationship is interesting, and may merit further research.
Patients vary in their ability to take on an active coordinating role of their progress within the primary–secondary care interface and value professionals adopting formal roles in this sphere. Though there has been some confusion about terminology,76 ,77 the clinical nurse specialist (CNS) is generally in a unique position with access to both PCP and specialist, as well as being available to the patient.78 ,79 This role is probably most developed and researched in patients with cancer where the CNS can improve quality of care, positively impact management, and improve understanding of treatment options and prognosis.80 Such roles may offer good value for money, reducing emergency admissions, length of hospital stay, follow-up appointments and providing support to enable end of life care in a place of the patient's choice.81 ,82 However, further research may be needed before firm recommendations can be made on the widespread value of CNSs in other chronic diseases.83
Relationships and personal value
Good relationships with clinicians were important to patients, influencing disclosure of concerns, help seeking, compliance, and overall confidence about referral and progress in the system.
Some patients reported tensions between primary and secondary care, expressing the view that PCPs and hospital doctors were not working together.4 ,34 ,44 The influences on this relationship are complex and include political restructuring, patient demands and advances in medical knowledge.84 Professional ‘tribalism’, acknowledged on both sides, has the potential to undermine the effectiveness and safety of patient transitions,4 though good evidence supports that professionalism and desire for best clinical care can overcome these obstacles84 and lead to positive patient outcomes across the interface.84–86
Strengths and limitations
This review employed rigorous and established methodology specific to qualitative studies throughout. The search strategy (see online supplementary appendix 1) validated combinations of qualitative search terms to optimise the list of citations returned,27–30 was extensive to ensure inclusion of relevant articles in the sociology or psychology literature,31 and broad in order to retrieve all articles with important information on the primary–secondary care interface, even if the stated focus was not from a patient perspective. Further, there was no language restriction, and translations of potentially relevant titles and articles were obtained. Given reciprocal translational analysis provides summaries in terms that may have already been used in the original literature, there is acknowledgement that the synthesis will tend towards the privileging of a priori over in vivo codes.20 Further, the quality of the included studies was assessed using a published framework (CASP) and found to be uniformly high.
The independent analysis of full articles by two researchers generated themes, concordant between reviewers, in a transparent and reproducible manner and was considered helpful in maintaining a coding balance.
Patient preferences and experiences were synthesised from varying primary–secondary care interfaces in differing healthcare systems serving different populations,4 ,37 but are gleaned only from the published literature. Study findings may therefore not be generalisable to all situations.4 ,37 ,44 However, the countries represented in the synthesis do reflect a broad range of organisational and funding systems that, while not proportionally representative of the global picture, nonetheless provide findings relevant to individual national healthcare approaches.
Several studies highlighted limitations in the research participant population (either in numbers or representativeness),33 which may also restrict transfer of findings to all patient groups and settings.40 ,41 ,44 ,50 Further, individual studies highlighted potential problems with the distinct participant characteristics of focus groups (eg, that views of quieter participants may have been lost), and the setting for group discussions.34 ,87 ,88 Other potential sources of bias were highlighted in individual studies including interviewer bias,49 recall bias41 and selection bias.40
Though the results represent the real-life pragmatic challenges faced by patients at the primary–secondary care interface, the clinician perspective also requires consideration.
The key areas for patients in the primary–secondary care interface may be classified into four domains: barriers to care, communication, coordination, and relationships and personal value. These findings highlight the separate but inter-relating areas of patient experience that require intervention with the aim of improving patient care. There will not be a ‘one size fits all’ remedy but the domains that have emerged from this review give useful targets to guide the development of interventions that will assist and improve the provision of care to patients across the primary–secondary care interface. Further research may focus on the clinician experience of the interface, and from their perspective how this impacts on patient care.
Review history and Supplementary material
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Twitter Follow Rod Sampson at @SampsonRod
Contributors RS designed the study, undertook the systematic review, carried out the data extraction, analysis and interpretation, and wrote the manuscript. RB provided guidance on qualitative research methods and reviewed the manuscript. JC participated in the analysis and interpretation of results and reviewed the manuscript. PW participated in the design of the review, analysis and interpretation of results, and critical review of the manuscript. RP provided assistance in developing an appropriate search strategy and reviewed the manuscript. All authors had full access to all of the data (including tables) in the study and can take responsibility for the integrity of the data and the accuracy of the data analysis. RS is the guarantor.
Funding NHS Highland Research, Development & Innovation Department, Aberdeen University, and Stirling University provided funds to meet publication costs.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
Data sharing statement Additional data on the assessment of quality of included papers are available from firstname.lastname@example.org.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.