The role of acculturation in migrants’ use of gynecologic emergency departments

To examine whether acculturation of migrant patients is a predictor of non‐urgent use of gynecologic emergency departments (GEDs).

To reduce health inequalities, migrants should be included in studies routinely. Specific research should examine what improves migrant health and how to implement the findings in the healthcare system. 7,8 Unfortunately, there is an underrepresentation of migrants in health research, 2 and the studies that are available are often hard to compare owing to differing or unclear definitions of migrant status. 3 To account for heterogeneity in the migrant population caused by differing adaptation and integration processes, the concept of "acculturation" has been used in research. The concept was originally introduced by Berry 9 to depict an individual's orientation toward the society of their origin and the society of their settlement.
One of the crucial dimensions in assessing healthcare equity is the utilization of healthcare services. 10 Specifically, the non-urgent utilization of emergency departments (EDs) is a growing problem. 11,12 Overuse and overcrowding of EDs has been identified as a worldwide concern 13 for both patient safety and quality of care, 14 and Germany is no exception. 15,16 A nationwide survey recently reported a substantial increase in hospital visits during non-business hours over the course of the previous 11 years, from a third of survey participants in 2006 to almost a half in 2017. 17 Higher use of EDs among migrants than among non-migrants has been observed in Europe. 3 As compared with native populations, migrants in Europe show a higher rate of ED consultations for gynecologic or obstetric reasons than for other medical specialties, 12 but very little is known about these non-urgent visits to the gynecologic division of an ED or a gynecologic ED (GED).
In addition to the scarcity of data regarding GED use by migrants, the comparability of studies is low owing to ambiguity about what constitutes a non-urgent or urgent case. 2 Because there is no standard definition, researchers must choose proxies and decide whether they want to include the patients' perspective of urgency.
The aim of the present study was to examine non-urgent use of GEDs on the basis of both health system-defined criteria and patientdefined criteria, taking migration status and acculturation into account as possible influencing factors.

| MATERIALS AND METHODS
The present prospective observational cross-sectional study of GED use was conducted among women attending the GED of the Virchow Campus of the Charité University Hospital, Berlin, Germany, between July 17, 2017, and September 30, 2018. The study was part of the project "Emergency Utilization of Migrants and Refugees" (EUMaR), which was approved by the Ethics Committee of Charité University Hospital. All participants provided informed written consent.
Women attending the GED were eligible if they were at least 18 years old and were responsive and oriented. Women were excluded if they had been seen by a GED physician before the interview, were referred from another department within the same hospital, were pregnant with a gestational age over 24 weeks, had acute symptoms that needed immediate treatment, or had been raped.
Standardized interviews were conducted from 8 am to 11 pm on any day of the week. If consent was given, a structured 30-minute questionnaire interview was conducted by trained female interviewers in a secluded area of the GED waiting room to protect privacy.
The validated questionnaire was developed for the EUMaR project (Supplementary File S1) and was available in five languages (German, English, Arabic, Turkish, and Russian) to reduce language-based selection bias. The selection of these languages was based on the language preferences of migrants in a previous study. 18 To reduce selection bias, eligible women were approached in the chronologic order of their registration. The questionnaire contained 50 questions on health, sociodemographic data, and, if applicable, migration-related information.
Acculturation was assessed by using the validated Frankfurt Acculturation Scale (FRAKK), 19 which consists of 20 items on integration into social and societal networks, language use, traditions, media habits, and emotional attachment to the host culture (Supplementary File S2). Each item represented a statement such as "I feel accepted by German society." Participants were asked to rate these statements on a Likert scale from 0 ("very untrue of me") to 6 ("very true of me"). The total score or "Acculturation Index" (AI) was categorized by using cut-off points of the sample defined by the 33.3 and 66.6 percentiles to define "low," "medium," and "high" acculturation, thus providing an ordinal independent variable. High AI indicates a strong orientation toward the host culture and weak orientation toward the culture of origin. 19 The criteria for non-urgent GED use were defined as (1) transport to the hospital other than by ambulance; (2)  health, and a dichotomous variable for employment were added to the model in subsequent steps. In all logistic regression analyses, the study group was women with a migrant background (hereafter referred to as "migrants"); by definition, this group comprises migrants or direct descendants of migrants. 1 The reference group was non-migrants. The level of significance was set at a P value of less than 0.05.

| RESULTS
Among 1187 eligible women who were invited to participate by the interviewers, 726 (61.2%) were enrolled in the study (Fig. 1). The study sample for analysis (n=723) was composed of 477 migrants and 246 non-migrants (Supplementary Table S1).
Among the migrants, the AI scores were normally distributed.
There were 163 women in the low acculturation category, 156 women in the medium acculturation category, and 158 women in the high acculturation category. Therefore, the study had 80% power at an α level of 0.05 to detect a difference between low acculturated women and non-migrants of 17% for the proportion of women showing non-urgent GED use (under the assumption that 40% of women in the low acculturation group showed non-urgent use).
Relative to non-migrants, the migrant group had a lower proportion of women with a high school or university education, a higher portion of unemployed or retired women, and a higher portion of married women. Unemployment was highest among migrant women of low acculturation (120/163, 73.6%), followed by those of moderate (77/156, 49.4%) and high (70/158, 44.3%) acculturation (Table 1).
More than one-third of the study sample (263/723, 36.4%) fulfilled the study's definition of overall non-urgent GED use, meeting all five non-urgency criteria ( Table 2). In the regression analysis for overall non-urgent GED use, none of the added variables (age, patient satisfaction with their health, employment status) were significant predictors. Thus, in the combined model, acculturation played no significant role in the overall non-urgent use of GED (Table 3).
Non-urgent GED use as defined by health system criteria (i.e., transport to the hospital other than ambulance, no self-reported referral or recommendation for hospital treatment from a physician, and no hospital admission after GED consultation) was high in the study sample, with 421/723 (58.3%) of the participants fulfilling all three criteria mentioned above (Table 2).
In the logistic regression, low acculturation was a significant predictor of system-defined non-urgent GED use as compared with Thus, low-acculturated women were more likely than non-migrants to make consultations at the GED for system-defined non-urgent reasons. In other words, women with low acculturation were more likely to have no ambulance transport, no referral from a physician, and no hospital admission after their GED consultation (Table 4).
Non-urgent GED use as defined by patient criteria (symptom severity or pain level, ≤6 of 10; and patient urgency estimation, ≤6 of 10) was seen in 429/723 (59.3%) of the study sample (Table 2). In regression analysis, low acculturation was a significant inverse predictor of patientdefined non-urgent GED use with a considerable effect size (AOR, 0.58;

| DISCUSSION
The present study examined whether non-urgent GED use is more prevalent among migrants than among non-migrants, and how this is related to acculturation. It differentiated between patient factors and health system factors to account for the multi-faceted nature of patterns of GED use.
The two categories showed opposite directions of effect; therefore, they T A B L E 1 Sociodemographic characteristics of the study women a . Values are given as mean ± SD or number (percentage). b here is missing data for one of the non-migrant participants for nationality and native language.

Acculturation of migrants
canceled each other out when combined into an overall non-urgency index.
As a result, acculturation of migrants had no significant effect as compared with non-migrants on overall non-urgent GED use in the present analysis.
Nevertheless, low acculturation was a significant predictor of system-defined non-urgent GED visits, and yet a significant inverse predictor of patient-defined non-urgent GED visits. In other words, migrants with low acculturation were less likely to be considered an emergency by healthcare professionals (no arrival in an ambulance, no referral from a physician, and no hospital admission), but more likely to have subjective reasons for consultation (high pain or symptom severity, and high urgency estimation). These findings highlight the intricacy of defining non-urgent use of health system services.
Researchers have used varying definitions of urgency and nonurgency, which makes an appraisal of the present findings difficult.
Nonetheless, studies with a similar approach have reported comparable results. 12,15,18,20 For example, our research group previously developed an inappropriateness index for ED or GED utilization that combined both patient and system perspectives, 20 reporting that ethnicity was not a significant predictor for inappropriate use of EDs, analogous to the current findings based on a composite index. Also consistent with the present findings, two studies that accounted for system criteria found higher rates of non-urgent use among migrants, 12,18 and another that considered only patient criteria reported lower rates of non-urgent use by migrants. 15 In a systematic review, a higher rate of gynecologic consultations and more ED use for low-acuity presentations was found among migrants than among non-migrants in Europe, 12 although acculturation was not examined. More frequent inappropriate ED use among migrants was also observed in a German study, in which the index for non-urgency comprised only health system factors. 18  The reference group was non-migrants. b P<0.05.
Model 1: Acculturation and age as predictors for health system-defined non-urgent GED use. Model 2: Acculturation, age and patients' satisfaction with their health as predictors for health system-defined non-urgent GED use. Model 3: Acculturation, age, patients' satisfaction with their health and employment status as predictors for health system-defined non-urgent GED use.
In the present study, the system-defined non-urgent GED use observed among low-acculturated women might be interpreted as inappropriate GED use due to a lack of knowledge about outpatient alternatives for gynecologic care. A systematic review of utilization of healthcare services by Graetz et al. 3 found that, although ED use was higher, outpatient visits for specialized care were less frequent among migrants in Europe. As suggested by Rechel et al., 2 possible access barriers to healthcare for migrants include limited language proficiency, "unfamiliarity with [the] overall health system, health literacy, social exclusion," and "direct and indirect discrimination." Migrants' inability to seek outpatient care in the first place might explain the low referral rates observed. This would mean that women with low acculturation in particular are non-urgent users of the GED. Given the patient perspective, however, this cannot be the only interpretation of the present results. A recent study in Germany used only the patients' view to determine non-urgent use, reporting that migrants were less likely to state a low subjective urgency, 15 which is comparable to the present findings.
The self-report of high urgency and/or pain levels found in the present study might be explained by various socio-cultural concepts and perceptions of pain and disease. For example, David et al. 21 found that, as compared with non-migrants, migrants had a higher pain load in general, as well as more pain regions and higher pain levels, when making use of EDs. 21 In addition, existing access barriers to outpatient gynecologic care may cause a delay in seeking care, leading to exacerbated pain and higher urgency ratings.
There are at least two possible interpretations of the present findings. If trusting only the criteria of healthcare professionals, it might be concluded that low-acculturated migrants are more prone to non-urgent utilization of the GED, while having a distorted perception of their own urgency. If trusting the patients' own perception, however, the findings may be interpreted as appropriate GED utilization for urgent complaints by low-acculturated women who are insufficiently cared for by the health system, possibly due to bias or language barriers. Presumably, the truth lies somewhere in between.
The study has some limitations. The ratio of migrants to non- health policies, migrant patients should not be neglected as part of the mechanism that leads to overcrowding, either as a population that will presumably respond to interventions specific to their acculturation, or as an underserved population requiring adequate care.

AUTHOR CONTRIBUTIONS
SS collected and interpreted the data, wrote and revised the manuscript, and contributed to data analysis and study planning.
OS contributed to study planning and data collection, conducted T A B L E 5 Logistic regression to identify predictors of patientdefined non-urgent GED use. the statistical analysis, and revised the manuscript. OR designed the study and revised the manuscript. JS contributed to study design and planning, and data collection. MD designed, planned, and oversaw the study; supervised data collection and statistical analysis; and interpreted the data. All authors approved the final version of the manuscript.

CONFLICTS OF INTEREST
The authors have no conflicts of interest.

SUPPORTING INFORMATION
Additional supporting information may be found online in the Supporting Information section at the end of the article.
File S2. Acculturation questionnaire. Table S1. Region of origin of the study sample.