ABSTRACT: The objective of this study was to determine whether language barriers during the screening interview affected the reporting of asylum seekers' health problems and their referral to further health care. Seven hundred and twenty-three standard screening questionnaires, administered by nurses to asylum-seekers at the time of entry into Geneva/Switzerland between June and December 1998, were reviewed, as well as information pertaining to language use during the interview. Language concordance between nurses and asylum seekers was assessed by considering the presence/absence of an interpreter, the type of interpreter present (trained, untrained), and the nurse's self-assessed proficiency in the language used during the medical interview. Nurses also recorded their own subjective assessment of the overall quality of communication during the interview. More than half of the asylum seekers came from Europe, mainly the Balkan regions, and a third of them from Africa. Most asylum seekers were men (72%). The median age was 26.5 years, and 50% were younger than 25 years. Severe physical and psychological symptoms were reported by 19% and traumatic events prior to migration were reported by 63%. The nurses referred 36% of all refugees to further medical care and 6% to psychological care. Professional interpreters were used in 8% of the interviews and ad hoc interpreters in 16%. Adequate, partial and inadequate language concordance was reported for 54%, 27% and 18% of the consultations respectively. Adequate language concordance was significantly associated with higher reporting of past experience of traumatic events and of severe psychological symptoms, contrasting with much fewer referrals to psychological care when language concordance was inadequate. These results suggest the importance of addressing language barriers in primary care centres in order to adequately detect and refer traumatised refugees. To address this problem, the use of professional interpreters is recommended.
Social Science [?] Medicine 09/2003; 57(3):503-12. · 2.70 Impact Factor