Review methods

Research questions and aims

The aim of this systematic review was to identify, appraise and synthesise existing research evidence regarding the outcomes and impacts of service reconfigurations that have the effect of increasing the time and/or distance for patients to reach an UEC facility. A list of potentially time-sensitive conditions requiring treatment at a UEC facility was developed in advance (see Inclusion and exclusion criteria). The list prioritised conditions more likely to be affected by service reconfiguration or requiring a decision whether or not to travel further to reach a more specialist facility. However, this list was not intended to be exhaustive.

We also aimed to examine the available evidence regarding associations between distance to an UEC facility and outcomes for patients and services, together with factors that may influence (moderate or mediate) these associations. A moderating factor was defined as one that may directly alter the relationship (e.g. weather/climate). A mediating factor was defined as one that acts indirectly and may help to explain the relationship (e.g. patient age, acting indirectly by influencing the patient’s risk status). Mitigating factors were those that could potentially reduce the influence of other factors on outcomes (e.g. starting treatment in the ambulance rather than at the hospital).

In our synthesis of the identified factors, we also used the terminology of ‘influencing factors’ (which influenced outcomes via travel time) and ‘explanatory factors’ (which influenced outcomes directly). These factors were broadly classified as patient related, context related or service related.

The research questions were as follows:

• What is the evidence regarding effects on patients of service reconfigurations that increase the time/distance to an UEC facility?
• What is the evidence regarding associations between time/distance from an UEC facility and outcomes for patients requiring UEC?
• What is the evidence regarding effects on the health system of service reconfigurations that have the effect of increasing the time/distance to an UEC facility?
• What factors might mediate, moderate or mitigate the effects of increased distance to an UEC facility on patient outcomes and/or the health-care system?

Literature search and screening

A comprehensive literature search was conducted in February 2019. The search was developed on MEDLINE and utilises diverse medical subject heading (MeSH) and free-text terms. The search comprised four broad facets: (1) emergency care, (2) rural or island services, (3) service reconfiguration and (4) potentially relevant emergency conditions. The search was limited to papers from 2000 to February 2019 and in English. The MEDLINE search was translated to the other databases. The following databases were searched:

• MEDLINE via OvidSP (1946–present).
• EMBASE via Ovid (1974–present).
• Cochrane Database of Systematic Reviews via Wiley Online Library (2003–present).
• Cochrane Central Register of Controlled Trials via Wiley Online Library
• CINAHL (Cumulative Index to Nursing and Allied Health Literature) via EBSCOhost (1981–present).
• HMIC (Health Management Information Consortium) via OpenAthens (1983–present).
• Web of Science (Science Citation Index and Social Sciences Citation Index; Clarivate Analytics, Philadelphia, PA, USA).

All of the references were imported into EndNote [Clarivate Analytics (formerly Thomson Reuters), Philadelphia, PA, USA] and then automatic and manual deduplication was conducted.

The MEDLINE search is provided in Appendix 1 with details of how the different facets of the search were combined.

Additionally, a search was conducted for current reviews of telehealth. The search was conducted on the databases listed above and was limited to 2009–19 and research published in English. After deduplication, there were 550 references. The MEDLINE search strategy is provided in Appendix 1.

Citation-tracking of the included reconfiguration studies was performed on Web of Science and Google Scholar (Google Inc., Mountain View, CA, USA) in April 2019. Web of Science identified 52 new references and Google Scholar identified 63.

Given the diffuse nature of the topic and the associated terminology, the reference lists of all included articles were manually screened to identify additional studies.

Search results were stored in a reference manager system (EndNote X8.2) and imported into EPPI-Reviewer software (Evidence for Policy and Practice Information and Co-ordinating Centre, University of London, London, UK) for screening, data extraction and quality assessment. The search results were screening against the inclusion criteria by one reviewer, with a 10% sample screened by a second reviewer. Uncertainties were resolved by discussion among the review team.

Inclusion and exclusion criteria

Population

The population was adults or children with conditions that required emergency treatment including, but not limited to, acute myocardial infarction (AMI), stroke, major trauma, severe exacerbations of asthma, chronic obstructive pulmonary disease or complications during pregnancy and the neonatal period. In practice, included studies included data on any patient wishing to access an UEC facility.

Intervention

Studies looking at changes to the delivery of health-care services (service reconfiguration) that may have an effect on the time or distance for patients to access an UEC facility were included. The review included reconfigurations that have an effect on access to any UEC services including ambulance services, maternity services and hospital EDs. The review also included studies evaluating changes to service delivery that aim to mitigate negative effects of living at a distance from an UEC facility. These included, for example, new forms of services providing UEC at the scene, such as first responders or specialist centre retrieval services.

Given the substantial volume of research on telemedicine/telehealth, particularly for patients living in rural areas, we decided not to conduct a review of this literature. To contextualise the evidence identified, however, we provide a brief narrative summary of key review-level evidence in this field.

Comparison

1. Studies comparing outcomes in groups of people travelling different distances/times to receive UEC, or
2. studies comparing outcomes before and after a service reconfiguration that has an effect on time/distance to UEC. Studies with no comparator were included if they met the other inclusion criteria.

Outcomes

Any outcomes for patients were included, including mortality/morbidity, travel time by ambulance or private care, or other perceived or measured effects, as well as outcomes or impacts on the health-care system such as non-transportation, emergency admissions, increase or decrease in contacts/service use. Transportation by helicopter as an outcome was excluded because of its limited applicability in the UK (not funded by the NHS and, therefore, any findings would not be relevant to commissioners).

Setting

The setting was the UK and other developed countries with relevant health-care systems. Absolute travel distances and density of population (which will affect distribution and density of health-care facilities) were also taken into account in assessing applicability of findings to the UK. In particular, studies of ‘remote’ health care from countries such as Australia were fully considered for relevance.

Study design

Scoping work undertaken for this review found two types of relevant studies.

1. Studies of changes to travel distances/times/outcomes following changes to the health-care system (designated ‘reconfiguration studies’). These studies are generally of observational or experimental design including before-and-after/longitudinal, cohort, case–control, or randomised designs.
2. Studies reporting the relationship between distance and outcome for particular groups of patients in a particular health system/setting without an actual change to service delivery (‘association studies’).

The initial scoping also identified a third group of studies that used population-level data to examine associations between population mortality/morbidity and distance to the nearest hospital. The review is investigating immediate access to UEC, thus studies that only provided data for whole populations rather than particular groups of patients were excluded.

Any identified mixed-methods or qualitative studies that reported perceived effects on patients or services of reconfigurations that increased time/distance to UEC were included.

Other inclusion criteria

• Literature published since 2000.
• Literature published in English.
• Grey literature in the form of service evaluations or reports from the UK.

Other exclusion criteria

• Studies that describe reconfigurations or initiatives without providing any quantitative or qualitative data.
• Conceptual papers and projections of possible future developments.
• Studies conducted in low- or middle-income country health systems.
• Theses, conference abstracts, articles in professional magazines, books and book chapters.

Data extraction and quality/strength of evidence assessment

We extracted and tabulated key data from the included studies, including study design, population/setting, results and key limitations. The full data extraction template is provided in Appendix 2. Data extraction was performed by one reviewer, with a 10% sample checked for accuracy and consistency.

Quality (risk-of-bias) assessment was undertaken using The Joanna Briggs Institute Checklist for Quasi-Experimental Studies. This nine-question checklist was chosen because of its relative simplicity and because the questions are applicable to a wide range of non-randomised study designs. Quality assessment was performed by one reviewer with a 10% sample checked for accuracy and consistency. Details of the quality assessment tool can be found at http://joannabriggs.org/research/critical-appraisal-tools.html (accessed 1 November 2019).

Evidence synthesis

We performed a narrative synthesis structured around the prespecified research questions and outcomes. The included primary studies were classified into three groups (i.e. reconfiguration, association and mitigation studies) corresponding to the prespecified research questions. We first described the characteristics of the groups as a whole. We then summarised the results in terms of the types of patients included (e.g. general UEC population, AMI, trauma, etc.), again considering reconfiguration, association and mitigation studies separately. Further analyses examined the results in terms of setting to assess the relevance of the evidence as a whole to the UK health-care system and to rural as against urban and suburban settings. Given the current background of changes to services in the UK NHS, we also summarised information from the studies about how health systems implemented service reconfigurations and subsequent adaptations to increased travel distances/times affecting emergency medical services (EMS) vehicles as well as patients’ own transport.

Mediating, moderating and mitigating factors identified by study authors were extracted from included studies and used to populate a logic model incorporating effects on patient outcomes via time to access UEC (‘influencing factors’) and directly (‘explanatory factors’).

Summary tables were generated from extracted data using the reporting function of the EPPI-Reviewer program. The overall strength of evidence was assessed using a previously described method.3 Evidence was rated as ‘stronger’, ‘weaker’, ‘inconsistent’ or ‘very limited’ based on study numbers and design. Specifically, ‘stronger evidence’ represented generally consistent findings in multiple studies with a comparator group design or comparative diagnostic accuracy studies; ‘weaker evidence’ represented generally consistent findings in one study with a comparator group design and several non-comparator studies or multiple non-comparator studies; ‘very limited evidence’ represented an outcome reported by a single study; and finally, ‘inconsistent evidence’ represented an outcome where < 75% of studies agreed on the direction of effect. All studies included in the review were included in the analysis of overall strength of evidence.

Public and patient involvement

We had input from our Evidence Synthesis Centre Public Advisory Group during all stages of the review. In the early stages, a meeting was convened to outline the topic of the review and to seek input regarding the questions and focus from a patient and public perspective. There was considerable interest in this topic among group members, and discussion regarding the high profile nature of this area and importance to members of the public. In particular, the experiences of local reconfigurations were described and discussed, and challenges for individual patients when reconfigurations occurred, particularly where travelling was not straightforward, and there were limitations in available transport options. This input provided the team with valuable insights into the concerns and worries of local communities, and highlighted the need to consider potential impacts on differing local communities. Our group members emphasised that distance and travelling time were very different concepts, with traffic conditions and road networks of key importance in time to access UEC, which focused the team on carefully examining the nature of the measurements used within the identified literature. The group also highlighted the potentially positive impacts on the health-care system of reducing ease of access to care, in terms of reducing unnecessary use of accident and EDs, which provided the team with an important area to consider when examining the literature.

Towards the final stages of the review, a further meeting was convened to examine the emerging results, and to consider the findings in regard to key messages for members of the public, and implications for the health-care system. At this meeting the key results were presented by a member of the team. There was considerable discussion regarding the applicability of the findings to the UK system, with members of the group wishing for greater information regarding the country of origin of the evidence underpinning the findings. There was consensus that it was important that in our reporting, we highlighted where evidence came from countries with very different health-care systems, or very different geography to the UK as this could be crucial to interpreting the findings. This input was therefore very helpful to the team in consideration of interpretation of the findings. Members also emphasised that it should be made clear where there was insufficient evidence available and uncertainty, as this was important for members of the public to be aware of. This input informed our reporting, and will be considered during future dissemination activities. At this meeting members also drafted the Plain English summary to be included in the final report.

Registration and outputs

The protocol was registered as PROSPERO CRD42019123061 and is also available via the HSDR programme website (www.journalslibrary.nihr.ac.uk/programmes/hsdr/164717/; accessed 16 May 2019).