Quality of health care for refugees – a systematic review

Objective The aim of this systematic review was to identify quality indicators (QI) developed for health care for refugees. Methods We conducted a systematic review of international QI databases such as the Agency for Health care Research and Quality in addition to a systematic search in PubMed, Cochrane library and Web of Science, using the terms “refugee” and “quality indicator”, complemented by a search in reference lists and grey literature. All papers which included QIs for refugees, especially for health care were included. In a first step all existing QIs were screened for their relevance to refugees. In a second step, all health care QIs were extracted. In a final step, these health care QIs were classified into process, structure and outcome indicators. Results Of 474 papers, 23 were selected for a full-text review. Of these 23 publications, 6 contained 115 QIs for health and health care for refugees. The main health care topics identified were reproductive health, health care service and health status. Conclusions Most indicators were indicators for outcome and structure quality, the smallest group were process indicators. Within the area of refugee health care, most QIs that have been found were QIs regarding reproductive health. QI databases do not yet include indicators specifically related to refugees.


Background
Health care for refugees and asylum seekers represents a challenge for the health system of the host country for various reasons. Examples are lack of access to health care in the host country and past experience of trauma which may have caused mental health problems. Furthermore, there may be barriers of communication, language and culture [1][2][3]. As a result of persecution, conflict, generalized violence or human rights violations, 65.3 million people were forcibly displaced worldwide in 2015 [4]. Most refugees were from Syria and over 6.3 million people fled from war [4].
Back in 1995, the UN High Commissioner for Refugees (UNHCR) highlighted the 'urgent need to address the areas of safe motherhood, control of HIV/ AIDS/STD, family planning services, and management of sexual and gender based violence within the overall primary health care services' [5]. However, the assurance of a good quality of health care for refugees and asylum seekers will be more and more important for the health system in the host countries.
A core dimension of health system performance is health care quality [6]. Quality of care can be defined as 'whether individuals can access the health structures and processes of care which they need and whether the care received is effective' [7]. Furthermore, quality of care should be divided into three dimensions: structure, process and outcome of care, which could result in measureable quality indicators (QI) [8]. QIs are important for the assessment of health care and are essential measurement tools for documentation and improvement of quality of care [9]. Measurable QIs for health care for refugees have not been identified until now. It can be assumed that regular QIs are just as valid for refugees as they are for all other patients. However, there are specific refugee situations like health care in refugee camps for which there should be quality assurance too. Therefore, the aims of this systematic review were to evaluate and to extract QIs developed for refugees and asylum seekers.

Systematic review
This systematic review was conducted to find existing QIs, concentrating on those relevant for refugee care, as there are some specific requirements in a typical "refugee situation" as in humanitarian crisis situations, refugee camps, reception centres and health care for refugees and asylum seekers in host countries. Different international and national indicator databases were screened in June 2018. These databases were: the Agency for Health care Research and Quality (AHRQ), the UK's Quality and Outcomes Framework (QOF-UK), the Australian Council on Health care Standards (ACHS), the Scottish Clinical Indicators, the Canadian Institute for Health Information (CIHI), the Dutch National Institute for Public Health and the Environment (RIVM), the RAND Health Quality of Care Assessment Tools (QA Tools), and the German Inpatient Quality Indicators (G-IQI). All databases were searched using the keywords "refugee" or "asylum seeker" to identify potential QIs for this target group.
Furthermore, an additional manual search of grey literature with "Google Scholar" was conducted in June 2018. For this search the terms "quality indicator" AND "refugee" OR "asylum seeker" were used. Additionally, we scrutinized reference lists of included studies and relevant reviews identified through the search. Additionally, we conducted a review by searching PubMed, the Cochrane Library and Web of Science, using "quality indicator" and "refugee" as medical subject headings (MeSH)-terms and as text words in June 2018. The search strategy for PubMed was: ("Refugees" [Mesh] OR "refugees" [All Fields]) AND ("Quality Indicators, Health Care" [Mesh] OR "indicators" [All Fields]). The Mesh term refugee included following terms: Refugee, Asylum Seekers; Asylum Seeker; Seekers, Asylum.
The search strategy for web of science was: (refugee* OR asyl* seek*) AND (indicator). Moreover, we cross checked the reference lists of the publications. If publications contained QIs from other indicator sets, we included the original publication of the mentioned indicator set and excluded the secondary source.
This systematic review was independently performed by two reviewers (KH, DW), who conducted the literature search and review following the PRISMA guidelines [10]. These two independent reviewers screened titles and abstracts initially for potential relevance. If the abstract matched the inclusion criteria, the full article was obtained and reviewed. After selection of potentially relevant articles, full reports were obtained and assessed for inclusion and exclusion criteria. Any disagreement on the eligibility of studies was resolved through discussion to reach consensus or, if required, by involving a third experienced review author.
Overall, the search strategy was defined by the principles of a systematic search and implied free-text keywords and Mesh terms by two reviewers who were well experienced in conducting Systematic Reviews. No medical librarian was consulted.

Inclusion criteria and screening procedure
Publications were included if the following inclusion criteria were fulfilled: -QIs were reported -QIs were developed for the target group 'refugees and asylum seekers' -Primary source of QI -Published in English, French or German Quantitative and qualitative research was considered. There was no restriction of time or place, all studies from 1980 to 2018 were included in the systematic review. We included both clinical indicators and indicators for practice management. According to the key characteristics of an ideal indicator by Mainz [9] the described indicators had to be specific and measurable with the numerator and denominator principles. In a first step for full-text review, all publications were included that described QIs for refugees and asylum seekers and extracted these indicators. In a second step, all publications were excluded that identified QIs for refugees but did not have a reference to health care. The final step was to classify health care indicators based on the dimensions put forward by Donabedian [8]. These dimensions were structure, process and outcome quality. Two authors (KH and DW) independently read and extracted the data from each study included. In cases of disagreement or discrepancies, we involved a third review author (JS) to reach consensus.
An overview of the literature identification and selection is presented in the PRISMA flow chart in Fig. 1.
We charted the following data from the included studies: bibliographic details such as author/source, year of publication, title, and included indicators, especially health care indicators.

Results
In the QI databases, no indicators could be found that specifically related to refugees or were established for this target group. Using the keywords "refugee" and "asylum seekers" did not provide any results.
The review of PubMed, Web of Science and grey literature, and publications of reference lists and grey literature revealed 644 papers. The removal of duplicates left 474 papers of which 9 publications were excluded because of publication language, and 23 were eligible for a full-text review . Most represented indicators were indicators of integration [11][12][13]17], indicators of education [11][12][13][14]17] and indicators regarding health care [11][12][13][14][15][16]. Other studies contained indicators relevant to single topics: indicators of acculturation [19], indicators of cultural participation [20], indicators of refugee placement [21], indicators with a focus on "youth in refugee camps" [22], indicators for assessing infant and child feeding practices [23], and indicators for malnutrition [24].

Descriptive analysis
We found 115 indicators in 6 publications that were applicable to health care for refugees. These included 33 indicators concerning structural quality, 26 indicators concerning process quality and 51 indicators concerning outcome quality. Four indicators related to both process and structural quality and one indicator related to both outcome and process quality. Please see Table 1 for details.
The indicators covered three thematic domains: "reproductive health", "health care service", and "health status". "Reproductive health" was assessed by 58 indicators for family planning, maternal and newborn health and HIV/AIDS. "Health care service" included 46 indicators describing access to health care and health management. "Health status" included 11 indicators such as birth rates, mortality rates, and diseases. All health care indicators were listed and sorted according to

Reproductive health
The reproductive health indicators found in the literature covered various thematic domains. There were preventive indicators such as "number of condoms distributed per person per month" [14] and "condom use" [11], indicators regarding pregnancy or maternal and child health and some indicators regarding HIV/ AIDS. The topic of reproductive health was the largest group of indicators identified within our review. Additionally, UNHCR focused on this domain in their publications [14,16], and 8 indicators from the Sphere Project [15] were relevant to this topic. Within this topic, there were 11 indicators for structural quality, 14 indicators regarding process quality and 32 outcome indicators. All indicators regarding reproductive health for refugees are listed in Table 2.

Health care services
Within this thematic domain there were 46 indicators that focused on various topics such as access to health care for refugees, training of staff and processes necessary for safe individual and public health. Most indicators in this group were from the Sphere Project (26 indicators), which defines minimum standards for disaster-affected populations [17]. Within this topic there were 22 indicators for structural quality, 12 indicators regarding process quality and 8 outcome indicators.
All indicators regarding health care services for refugees are listed in Table 3.

Health status
Within this group, there were 11 outcome indicators. They covered topics including "self-reported health status" [11] and incidence, mortality and birth rates [12,14,15]. All indicators regarding the health status of refugees are listed in Table 4.
Most of the indicators were outcome indicators (n = 51; 44.35%), structural quality was represented with 33 indicators (28.7%) and the smallest group were indicators of process quality (n = 26; 22.61%). Four indicators (3.48%) addressed both process and structural quality and one indicator addressed both outcome and process quality.

Discussion
This systematic review shows evidence concerning QIs for health care of refugees and asylum seekers. Different databases were used and 115 indicators that related to health care of refugees were identified. These different indicators could be sorted into three topics, "reproductive health", "health services" and "health status" including a categorisation into the Donabedian quality dimensions: process, structure or outcome of care. Most indicators were outcome indicators and focused on mortality and morbidity; process indicators represented the smallest group of indicators. Most of the indicators address items concerning reproductive health and maternal and child health. There are many indicators focusing on these topics due to a marked need for addressing highly prevalent conditions that should be addressed in future research. Access to high-quality reproductive health services including appropriate emergency obstetrics can drastically reduce the number of women who die during or after childbirth, ensuring that mothers and their children enjoy a healthy life. UNHCR applied the principle that reproductive health care should be offered to all refugee women [16]. Quality reproductive health services require that organizations, programs and providers use appropriate technology, have trained staff, and ensure accessible services and respectful care.
Although WHO and UNHCR have highlighted the importance of reproductive health in refugee situations, a systematic review about refugee health status shows no study regarding physical health of women during pregnancy and childbirth. The authors identify a priority need for research in this context [34].
The second largest group of indicators observed in this review are indicators concerning the topic of health care services, such as access to health care. Access to health care is a critical determinant of survival in refugee situations such as disasters. The right to health can be assured only if the health care providers responsible for the health system are well trained and comply with professional standards [15]. The organisation of health care for refugees and access to health care present a challenge for the host countries. Refugees often have no regular access to health care and they struggle with restricted access to health care in their host countries [35]. There should be standards for refugee health care access similar to the Sphere-Standards for disaster-affected populations [15]. Moreover, the provision of health care services for refugees and asylum seekers presents also a challenge for the health care providers. A systematic review shows that health care providers struggle not only with the diverse cultural beliefs and language differences but also with limited institutional capacities which additionally restrict the access to health care [36].
The third topic, indicators for health status, only contains outcome indicators such as morbidity and mortality which are connected to the indicators of reproductive health and health services. The quality of life in the country of origin, the migration process and the conditions in the host country could influence the health outcomes of refugees. Refugees may be more vulnerable to certain diseases or mental disorders than people without such experiences. The migration experience itself could create stress which could influence the health outcomes of migrants in different ways depending on the socio-economic and health conditions in the country of origin [11]. As complement to our systematic review a recently published evidence report concludes that there is a lack of common strategies for health care management of refugees and asylum seekers [3]. Owing to different legal frameworks in the host countries, no general conclusion about the accessibility and quality of health care delivery can be adopted [3]. It can be assumed that generic QIs could overcome this barrier and help to optimize and improve the health care of refugees and asylum seekers. A further review shows that different guidelines for migrant health care are available that range from disease specific to generic guidelines for health care delivery which could have an impact on quality of health care [37]. A systematic use of such guidelines, especially of the developed QIs, in the health care process of this population group is essential to ensure a high-quality of health care.
The main strength of our systematic review was the presentation of numerous and diverse areas in which QIs for refugee care were developed. The search strategy was defined by the principles of a systematic search and implied free-text keywords and Mesh terms by two reviewers who were well experienced in conducting Systematic Reviews. No medical librarian was consulted.
However, we only included publications written in English, French or German. Moreover, because of the clear defined search strategy it could be that some institutes on provincial level in different countries like the Institute of Clinical Evaluative Sciences in Ontario, Canada showed no QI for health care of refugees. Therefore, there might be a selection bias in our findings.

Conclusions
It can be concluded that most indicators stress outcome. It can be assumed that an effective process within health care services supports high-quality of health care and should be the focus of further studies. QIs are an important measurement tool for the documentation and improvement of health care. Further research needs to address explicitly measurable QIs to learn more  There are an adequate number of health facilities to meet the essential health needs of all the disaster-affected population: Sphere (2007) [15] Structure There is at least one Community Health Worker (CHW)/1000 population, one supervisor/10 home visitors and one senior supervisor.
Sphere (2007)   Sphere (2007) [15] Process All health facilities and agencies report cases of epidemic-prone diseases within 24 h of onset of illness.

Sphere (2007) [15] Process
The lead agency produces a regular overall health information report, including analysis and interpretation of epidemiological data, as well as a report on the coverage and utilization of the health services.
Sphere (2007)  about health care for refugees and asylum seekers. Moreover, it can be assumed that a smaller number of indicators can be better implemented in health care of refugees. Therefore, the next step would be the reduction and prioritisation of these 115 indicators e.g. based on the RAND/UCLA Method [38].