Search strategy

The systematic search for this review was conducted in August 2019, using the following databases: Cinahl, Cochrane Library, Embase, Medline, PsycINFO, PILOTS, Web of Science (ISI), OPENSIGLE, and WHO: International Clinical Trials Registry Platform /ICTRP. Since relevant material is often published by torture survivor centers or other gray-literature sources, a scan of relevant websites was included in this process. The PICO framework was used to develop the search terms relevant to the PICO question [52]. Medical Subject Headings (MESH) with related text-based search terms were used with a combination of terms and concepts, in accordance with databases’ concordance [53, 54]. Citations from relevant articles and systematic reviews were also screened. A full description of the search strategy is provided in S2 Table.

This study’s search strategies yielded 6,329 citations, which were imported into the Covidence literature-screening software [55], and 1,790 duplicates were removed. Two independent reviewers (ACS and AKB) assessed 4,539 titles and abstracts against the eligible criteria. Any disagreement was resolved by consulting a third reviewer (EKG) until reaching a consensus. A total of 213 full-text articles met our inclusion criteria and were retrieved. Eight studies were included in the final review. Fig 1 presents the search strategy and the article review process.

The eight included studies represented six countries, were published between 1993 and 2018, and comprised data from 290 participants. Their data collection was based on semi-structured, in-depth interviews [56–59], and focus groups [60, 61]. One study used a case study approach [62], and one study used storytelling as a data collection method [63]. All the studies included in this review were conducted in high-income countries (i.e., Canada, England, the United States, Denmark, Switzerland, and Ireland) and presented refugees and asylum seekers’ healthcare experiences.

Data extraction and synthesis

Two researchers were involved in the data extraction process (ACS and AKB). A summary of evidence was processed for each study. Dallam’s Healthcare Retraumatization Model [32] was chosen to create a framework for our data synthesis, and the model’s four sub-processes were used to create a framework of our identified themes.

Data from the included studies were coded toward the thematic framework, and line-by-line coding relating to refugees’ experiences provided the descriptive themes. We interpreted data that could not be included in the framework using interpretative and inductive analyses [64]. Using inductive analysis, we moved from specific data to a general theoretical understanding. An extensive text was condensed into a summary of findings, and in this process, we established links between our findings and our research objectives [64]. Two authors read this text in detail, familiarized themselves with its content, and gained an understanding of the themes covered in the text. The authors identified and defined sub-themes and themes. The sub-themes were derived from multiple readings of the raw data. Using inductive coding, the sub-themes were created from actual phrases and meanings in specific text segments. As a result of this analysis, we developed categories as a framework that summarized the raw data and conveyed key sub-themes. Different and overlapping descriptive sub-themes were noted and summarized.

To improve the overview and to ensure a comparison of the findings with Dallam’s Healthcare Retraumatization Model [32], we grouped the sub-themes into main themes. This deductive approach ensured an ability to benefit from an existing model, and our inductive analysis ensured an ability to develop a model that could explain the re-traumatization process of torture survivors based on the underlying structure of their treatment experiences for somatic conditions. This process is illustrated in S1 Fig. Some of the sub-themes could have been simultaneously included in more than one of the themes; however, to create a clear and highly rigorous system, we placed each sub-theme into only one theme. Although our findings could have been grouped differently, we found the entities in the units we established to be appropriate for adaptation using Dallam’s Healthcare Retraumatization Model. The data extraction and synthesis processes are illustrated in S1 Fig.

Assessment of methodological quality

Appraising different qualitative studies in relation to the specific methodology used in research is recommended because researchers’ chosen approach is linked to the “outcome” of their research [65]. Therefore, a quality assessment of our included studies was performed using the Joanna Briggs Institute (JBI) Standardized Critical Appraisal Checklist for Qualitative Research [66]. This tool evaluates studies’ methodological quality according to a 10-object scale, accounting for coherence between philosophical perspectives, data collection methods, and qualitative analysis—including studies’ representativeness of participants and their voices, ethical criteria, and researchers’ influence on interpreting the research (S3 Table). The purpose of this appraisal is to assess a qualitative study’s methodological quality.

Each study in the present review was independently assessed for quality by two reviewer authors (ACS and AKB), and disagreements were resolved by referring to a third review author (EKG). The summary of our findings also includes an overall assessment of the findings’ confidence, according to the GRADE-CERQual (Confidence in the Evidence from Reviews of Qualitative Research) [67, 68] to assess how much confidence should be placed in findings from the data synthesis, based on four components: methodological limitations, the included studies’ relevance, the data’s coherence, and the data’s adequacy. An assessment based on judgments of each of the four CERQual components was summarized at four levels: high, moderate, low, or very low. This assessment is presented in S4 Table.

Synthesis of findings

Table 2 outlines the included studies’ characteristics and the healthcare experiences presented in each study.

Five main themes were elicited from our findings: (1) invisibility, silence, and mistrust; (2) healthcare providers’ attitudes and a lack of perceived quality in healthcare; (3) disempowerment; (4) avoidance; and (5) satisfaction and gratitude. To develop a healthcare re-traumatization model for torture survivors, based on an existing model, we adapted four of the grouped findings to the four interactive sub-processes of Dallam’s Healthcare Retraumatization Model: “hypersensitivity to threats to safety,” “exposure to triggers,” “post-traumatic reactions,” and “avoidant coping” [32]. Table 3 shows the grouped findings and their adaptation to the Dallam’s Healthcare Retraumatization Model concepts. Subsequently, the themes were used to create a re-traumatization model for torture survivors in healthcare via a comparison with the Dallam’s Healthcare Retraumatization Model. The fifth group of findings, satisfaction and gratitude, is not framed in the healthcare re-traumatization model since it represents resources for avoiding re-traumatization. We discuss all of our findings below and use them to create an explanatory model for refugees and asylum seekers’ experiences that can cause re-traumatization in somatic healthcare, which Fig 2 presents.

Negative experiences: Contributing factors in re-traumatization

Positive experiences: Protective factors against re-traumatization

Quality appraisal

Our judgments across the 10 items of the JBI risk-of-bias tool for each individual study are presented in S3 Table. Across all included studies, we found partialities related to internal validity, and the results of the methodological quality assessment are outlined in the table.

Our overall CERQUAL assessment of confidence was graded as moderate because of moderate concerns regarding methodological quality, relevance, and coherence (S4 Table).

Invisibility, silence, and mistrust as hypersensitivity to threats

Participants described feelings of not being heard or seen, causing disappointment and reinforcing mistrust. As Behnia described [35]—and in line with our findings—even many years after trauma, survivors experience the world as a dangerous place where people in positions of authority and caring roles, such as doctors, nurses, and interpreters, cannot be trusted [35, 60, 61, 63]. Consequently, survivors may perceive interactions and situations during treatment as threats to their safety [56, 63]. In a case Jennings [33] described about how victims of sexual abuse become re-traumatized through invisibility in healthcare, we see similarities with the current review’s findings on healthcare providers not considering survivors’ special needs or actual stories as important, which may make survivors feel invisible. Our findings suggest that feelings of fear and insecurity caused by threats during war, flight, and torture cause a hypersensitivity that emerges in interactions with healthcare providers. Factors that reinforce feelings of insecurity include survivors’ experiences of lacking trust and recognition among healthcare providers. We can adapt these findings to Dallam’s Healthcare Retraumatization Model [32], which describes hypersensitivity to threats to safety as the main factor determining survivors’ re-traumatization. Our findings also find support in a case study by Jacobs and Iacobino [36], which described how distrust makes survivors view interactions and healthcare settings as threatening [36]. Our findings indicate that survivors of war and torture expect to be met with understanding and acceptance among healthcare providers and to receive help in turning their vulnerability into a resource for receiving desired healthcare and social care [56, 58].

Healthcare providers’ negative attitudes and a lack of quality healthcare as triggers

Several of our findings described aspects that healthcare survivors found triggering. When they feel that they did not have the same access to care or deserve the same attention as other people [58, 63], survivors may develop feelings of discrimination [69].

According to Dallam [32] and Hokland’s [34] descriptions of how triggers work, these feelings can trigger associations, dynamics, and memories about aggressors during war and torture, which may lead to re-traumatization. When survivors enter healthcare settings, they are automatically exposed to numerous challenges that can work as triggers [32]. This view supports our findings of several challenges that survivors have identified, including healthcare providers’ disinterested attitudes, which serve as triggers. Other studies have connected similar challenges to refugees’ healthcare encounters [70, 71]. According to Richey [72], healthcare providers may hesitate to involve themselves with survivors because of inexperience with this patient group and the complex problems they present. Such sensorial triggers as medical equipment [62, 63], waiting times, short consultations [58–61], language problems [59, 60, 63], and tactile stimuli [56, 58, 63] that were documented in this review have also been described by other authors [30, 73].

As other studies have suggested, and in line with our findings, trauma survivors perceive more challenges in interacting with healthcare personnel than other patients [74, 75]. Therefore, such mixed triggers as medical examinations [58, 63], tactile stimuli [56, 58, 63], and a lack of privacy—which combine both sensory and relational triggers [32]—may strongly affect survivors. These factors are usually perceived as part of health examinations or healthcare settings; however, for people who have experienced trauma, these factors can act as triggers, causing distress and concerns about receiving inappropriate care [56–61, 63].

Our findings reveal that survivors experience pain as a common health problem, which health personnel do not handle as survivors expected. According to a recent study about Islamic State (IS)- traumatized female refugees, pain is a major symptom among survivors [40]. Survivors have expressed that healthcare providers do not take their pain seriously, causing dissatisfaction and a sense that survivors are receiving low-quality healthcare. These feelings may also serve as triggers and cause potential re-traumatization.

Triggers may overlap with each other, be related to each other, and have multiple effects. Sensory triggers can cause distress because they resemble stimuli that were present at the time of torture, as Dehghan [76] described in a review on the health impact of sexual torture among Iranian and Afghan refugees. These stimuli can involve any kind of bodily sensations or sights, sounds, and smells in a treatment environment. In line with the Dallam’s Healthcare Retraumatization Model, we also found that triggers can result in physical reactions, such as shaking, nausea, or vomiting [63].

Among other findings in this review, survivors’ sense of not being believed when providing information about their situations to healthcare providers during treatment may remind them of interrogation situations, where victims are tortured because they are suspected, distrusted, falsely accused, or forced to confess [29]. When survivors interact with healthcare providers, these associations may cause re-traumatization.

A lack of cultural understanding—also called “ethnocultural discordance” [29]—was reported by asylum seekers in a Norwegian study [77] as one of the most important factors affecting outcomes in survivors’ healthcare. Our findings echo this view since a lack of cultural understanding was one of the triggers our review identified [56, 63]. According to our findings, both the use and eschewal of interpreters serve as potential triggers in the re-traumatization process [59, 60, 63]. A recent study that included Ethiopian asylum seekers in Norway [77] supports this finding, having reported that survivors felt a sense of autonomy and control when they expressed themselves but not when an interpreter spoke for them. Recounting traumatic war incidents is difficult, but it is even more challenging when such communication must be facilitated by an interpreter [78]. Accordingly, our findings suggest that such relational triggers as the imbalance between healthcare providers or interpreters and patients contribute to most of the triggering situations in healthcare. However, the addition of other triggers from other categories—such as a lack of privacy, medical equipment, sounds and smells in the treatment environment, and tactile stimuli during treatment—can intensify survivors’ distress, causing re-traumatization.

Disempowerment causing post-traumatic reactions

As Dallam [32] described, reactions of discomfort during treatment disempower survivors and reinforce PTSD symptoms associated with re-traumatization. According to our findings in this review, survivors are triggered by experiencing disempowerment and physical reactions similar to what they experienced during their original trauma [62, 63]. Survivors can experience flashbacks triggered by aspects of encounters with healthcare personnel. Such flashbacks cause distress and difficulties in handling life [72]. Our findings also include other physical reactions—such as feeling sick, nausea, shaking, disorientation concerning place and time, and uncomfortable reactions to anesthetics, including feelings of losing control. Survivors also expressed concerns about accepting medications [62]. This finding is consistent with recent research about IS survivors’ concerns with taking any medications because they could remind survivors of sedation and immobilization during captivity, thus evoking feelings of disempowerment [40].

The power imbalance between healthcare providers and patients may also replicate the power imbalance between torture survivors and torturers, reinforcing other triggering situations in healthcare. Consequently, and in line with this review’s findings, survivors may become hypervigilant during treatment [62, 63]. Feelings of such power imbalances have also been described in studies about the re-traumatization of sexually abused women [79, 80], which revealed that such dynamics may cause victims to experience health examinations and touching as sexual assaults, evoking a sense of violation [56, 63], anger, and shame [56]. These reactions may make patients struggle fiercely against feelings of losing control, which can manifest as impatience, violent behavior, or silence [63]. These reactions indicate a high degree of disempowerment, which might lead to ethical concerns about obtaining informed consent from patients who are torture survivors.

Avoidance

Our findings show that survivors experience several barriers to disclosing their stories [59, 60]. In line with this finding, Eisenman [12] revealed in a study that torture survivors may not volunteer to recount their torture histories due to feelings of guilt, shame, or generalized mistrust. In addition, healthcare providers often fail to identify histories of torture—even when patients believe torture has affected their health [12]. Moreno and Grodin [4] concluded that asking about torture should be part of interviews when caring for survivors. Our findings in this review indicate that approaches to caring for torture survivors should be evidence-based and linked to interview guidelines.

To overcome such barriers, survivors employ strategies of proactive coping [62, 63] to alleviate the effects of stress reactions during triggering. However, some survivors in the studies we examined [56, 63] ultimately avoided healthcare (i.e., avoidant coping) [32], avoided appointments with doctors [59], and stopped telling their stories to healthcare providers [59, 60]. Dallam [32] described avoidant coping’s ability to impede the development of trusting relationships with healthcare providers, interfering with survivors’ ability to seek essential care in order to maintain or improve their health. Torture survivors can regard healthcare providers as frightening because they are afraid healthcare providers might have connections to governments [36]. Concerningly, the coping strategies identified in this review—such as choosing not to seek treatment for serious illnesses, refusing treatments, failing to engage in treatment regimens, and abandoning treatment despite medical advice—can deteriorate survivors’ health and even lead to death.

Satisfaction and gratitude as preventive factors against re-traumatization

Our findings suggest that being seen and listened to, being met with respect and openness, and receiving advice and support are what survivors ask for and expect from healthcare providers [58, 63]. These feelings of satisfaction and gratitude may prevent the re-traumatization process and remedy feelings of unsafety and invisibility. These findings expand on Dallam’s Healthcare Retraumatization Model [32] by including feelings of satisfaction and gratitude as factors that can prevent re-traumatization.

Limitations and strengths

Although this review involved a limitation in the low number of available studies to assess, the included studies that we synthesized represented the experiences of 290 participants with different backgrounds and countries. We found similarities in participants’ concerns and experiences related to healthcare somatic services, which increase the review findings’ credibility. To ensure rigor, a prespecified plan was followed, and the data were rich and reasonably coherent across studies. One strength of the method we used is the possibility to condense extensive and varied raw text-data into a brief, summary format in order to establish clear links between the research objectives and the summary findings derived from the text data. This advantage made the data more available and transparent and, therefore, easier to understand for readers.

Our findings may have been shaped by our assumptions and experiences as researchers in deciding what aspects of the data were more important and less important. Another issue we must address is that inductive analysis can be colored by researchers’ pre-understanding of human beings as physical, psychological, and social beings. Other potential limitations of our systematic review should be considered. Even in interview-based surveys, an accurate assessment of experiences surrounding re-traumatization during healthcare is difficult to ensure. The eight studies that contributed to the present review were conducted over several decades, within different research topics and with different samples—including refugees with diverse experiences from different continents and in different host countries. Because our sample studies were few in number, transferring our findings to all survivors of war and torture is not possible.

Practical implications

Before appointments with survivors, healthcare providers should familiarize themselves with patients’ histories. Such preparation not only helps healthcare providers understand patients’ problems—especially trauma histories—but also shows a genuine interest in patients. Some routine procedures can be extremely stressful for torture survivors, evoking memories of torture. Therefore, preventing or minimizing stressors as much as possible is important. For example, survivors who lack trust may require longer consultations in the process of creating trust. A survivor who was ill-treated by someone in their own cultural group may be more comfortable with a healthcare provider from another background. In some cases, certain procedures—such as small surgical procedures and examinations—should be performed under anesthesia; relaxation, meditation, and other coping strategies can also be helpful in this context. These techniques allow torture survivors to calm themselves when they begin feeling anxious. Importantly, many torture survivors can undergo examinations and procedures without being seriously affected if they are told what the examination or procedure will entail and if they experience empathy. Although some survivors believe that exploring traumatic histories may be re-traumatizing, other survivors expressed that they want their histories to be acknowledged. One of the most important ways for healthcare providers to improve torture survivors’ treatment outcomes is to listen to what they need in order to feel safe. This approach requires that healthcare providers take the time to initiate conversations about war experiences, ask direct questions, and utilize trained interpreters. Acknowledgment from healthcare providers can empower survivors, preventing re-traumatization.