Racial and ethnic disparities have been shown to exist in quality of care in most medical disciplines, emergency departments and overall health outcomes.1 Patients who speak little or no English receive poorer healthcare and are less likely to receive access to preventive care.2^–5

Adverse events are, however, rarely understood through the lenses of the hospital patient.6 The acute care patient might be culturally and linguistically diverse, and hold a range of ideas about appropriate care and a variety of priorities about what is important in healthcare. In this paper, the term “negative event” is used to indicate an abusive, potentially dangerous or life-threatening healthcare event, as perceived by the patient. The perception might not necessarily match the views of providers, and the event cannot therefore be termed an adverse or unsafe event or error.

While researching the experience of acute patients with limited English, it became evident that some had experienced negative hospital events. This study examined the perception of these negative events, and factors that might prevent or minimise them.

METHODOLOGY

The study uses qualitative techniques based on the work of Strauss and Corbin,7 and Charmaz.8

This study was conducted in 2004 as part of a larger research project examining the experiences of patients from non-English-speaking backgrounds attending the Liverpool Hospital, Sydney, Australia. Patients were initially selected if their medical record indicated they spoke one of nine selected languages other than English. A total of 360 patients were identified as the study population. From this, 278 patients (77%) responded to a multilingual telephone survey conducted by trained bilingual research officers. Trained bilingual research assistants invited respondents from seven of the nine language groups (n = 255), by telephone and letter, to attend a focus group. Focus groups were conducted in seven languages—Arabic, Vietnamese, Italian, Chinese, Spanish, Serbian and Croatian. Paid transport was organised to the Education Centre of the hospital. Culturally appropriate food and drinks were provided at the venue.

Key focus-group questions related to the overall experience of hospitalisation, impact of limited English proficiency, access to interpreters, patient involvement in healthcare and views about healthcare services in their country of origin.

The principal investigator, a white, middle-aged, female, Anglo-Australian with considerable experience in the multicultural health field co-facilitated each group with a bilingual health worker. Each group included a professional healthcare interpreter who interpreted all proceedings, an English-speaking scribe who took notes in English from the interpreted proceedings and a note-taking observer. Proceedings were videotaped, audio-recorded, and scribed. English transcripts were derived directly from the notes and tapes, and then checked for accuracy by the relevant bilingual health worker. Transcripts of all focus groups were read individually by the principal researcher and three research assistants, then reviewed and discussed as a group to develop preliminary themes. Codes were established line by line, and then discussed and amalgamated into core categories (nodes) and related subcategories (child nodes). Disagreements were resolved by discussion and consensus. The process was iterative, with researchers regularly revisiting the raw data and amending the analysis.

Each of the seven transcripts was systematically checked against the codes and refined. The full transcripts and the coded transcripts were then discussed and amended by the relevant bilingual research assistants to ensure capture of cultural issues. This last process provided validation of the initial coding.

All positive and negative comment was identified, and contributing factors were coded. Negative events were separately identified from negative comment, and coded and analysed to assess the nature of the event, contributory factors, and factors which might prevent them.

QSR NVivo Version 2 was used as the analysis tool.

RESULTS

Forty-nine patients with limited English and nine carers of these patients attended focus groups. Sixty-nine per cent had been outpatients attending the Emergency Department. The average length of stay was 2.6 (SD 6) days. The mean age was 74 (6.3) years. Forty-seven per cent were female. Language groups were: 21% Vietnamese, 21% Italian, 17% Arabic, 15% Spanish, 12% Croatian, 8% Chinese and 6% Serbian. Eighty-eight per cent reported that they spoke little or no English. Most had spent a number of years in Australia (range 0–66, average 30.25 (16.8)). In regard to major diagnoses, 31% had circulatory diseases, 15% had respiratory, 10% had musculoskeletal, and 8% had kidney and urinary-tract problems. Fifty-eight per cent had medications prescribed. The 49 patients comprised 19% of the total respondents (n = 255) in the seven selected language groups.

Positive comments about healthcare comprised the most frequent code. Negative events occurred less than 20 times across the seven focus groups. These events were concentrated in the Vietnamese and Arabic groups, appeared in the Spanish group and were not recorded in the Italian, Serbian, Croatian or Chinese groups.

Themes associated with negative events were staff neglect; staff workload and pressure; treatment delays; communication and language issues; information and involvement; medication errors; and diagnostic errors. Statements indicating fear and powerlessness accompanied each of these themes. Some examples follow.

Staff neglect

I was then very frail, I had heart problems, I had lung problems and I was supposed to have the breathing tube connected to me all the time and I told the nurse when she transferred me to a new room, “Do not pull it out!” She said, “It’s only a short distance, don’t worry.” She turned off the machine during the transfer despite my request . . . I had big problems breathing . . . I was so short of breath, I was gasping for air, literally, and I wet myself. I got so distressed I even wetted myself and then the nurses knew it. . . . even after they had connected the machines, I was still struggling with my breathing and they saw it, it was so obvious, and still they left me there. They left me there struggling. (ID 5V Vietnamese woman)

Staff workload and pressure

After the draining that gave me some relief which didn’t last for long. . . . After every 15–20 minutes if no one came and helped me with that, I would not be able to breathe and I could even pass away. Again I want to stress that I am not complaining about this. I just am speaking out. This could be due to a lack of staff, shortage of staff. (ID 8V Vietnamese man)

Treatment delays

. . . this kidney problem I went to the Emergency at around 5, half past 5 pm, and it was not till midnight that I was seen by someone, seven hours! . . . Now when I hear the word “Emergency,” I get panicked! I get scared! (ID 4V Vietnamese Woman)

Communication and language

When I developed pneumonia and they had to operate on me and then I lost my consciousness, they did not call the interpreter for me. (ID 5V Vietnamese woman)

Information and involvement

When the doctor comes I don’t know what (has) happened to me, what I’ve got . . . it scares me everyday. . . . (ID 8 S Spanish woman)

Medications

I had a clot in (my) lungs and the nurse wanted to give me a tablet just for the heart. I said . . . “No, No, No! Don’t give me this thing!” (ID 11A Arabic woman)

Diagnosis

The third point I want to make, even now I still don’t know whether I have got cancer or something else. (ID 8V Vietnamese man)

“The Happy Migrant Effect”—de-emphasising negative events

During my stay and my treatment in hospital, I was very compliant, but one thing I found strange, they gave me an injection at 9 pm, and after I was frothing and I lost my consciousness . . . Within half an hour, all the doctors were swarming over me, bringing machines . . . Two days later, one nurse came and gave me exactly the same injection. I told her this injection had caused me problems two nights before. She said this was the doctor’s instruction and she had to carry it out. (ID 7V Vietnamese man. He said he had then lost consciousness and needed to be resuscitated, yet was happy with the care outcomes)

Except for some language problems and some problems with medications, tablets and medications, apart from those problems everything has been good. (ID 5V Vietnamese woman)

We can’t complain much about things here because when we compare it to our homeland, everything is good, and we can get everything. (ID 3A Arabic man)

DISCUSSION

“The Happy Migrant Effect” is a useful emergent concept, which explains, in some patients, an acceptance of a negative event in healthcare delivery. It might reflect the extreme powerlessness related to being seriously ill and not being able to communicate, a very positive comparison of healthcare in the new country compared with that remembered, experienced, or expected in their country of origin, patriotism for the new country, cultural norms that proscribe acceptance, politeness or social desirability, and sometimes a degree of self-denigration for not having learnt English. For some patients with limited English, there was also a fear of reprisals if they spoke out. The effect might also be related to being elderly or of poor socio-economic status.

Social desirability may be more pronounced in cross-cultural contexts, especially with people in lower socio-economic brackets, older people and those not well integrated into the dominant culture.9 Perceptions and meanings may be further complicated by filters associated with powerlessness, negative experience of oppressive regimes and cultural traditions with varied proscriptions related to social desirability, self-disclosure and politeness.10^–12 Certainly, some patients believed that an unsafe event had occurred but did not consider that this necessarily marred their hospital experience. Some patients did not expect to have a say in their care and in healthcare decisions. Others believed they had received poor, neglectful or even physically abusive treatment but were either unable or unwilling to complain, or constructed the event as minor.

While “The Happy Migrant Effect” was pervasive in some of the focus groups, it was clearly not a universal construct, with some patients expressing profound dissatisfaction with their hospital experience. A contributory significant factor to “The Happy Migrant Effect” might be a lack of awareness of healthcare complaint mechanisms and avenues for resolving problems. The reduced likelihood of patients with limited English complaining has been supported in recent studies about complaints.1314 Under-representation in complaints does not necessarily signify satisfaction with services; rather, it may indicate poor access, cynicism about complaining or poor knowledge about complaints procedures.15 When patients with limited English are actively asked to comment on services, they may then complain at a higher rate.16

Interestingly, patients did not generalise a negative event as meaning that the hospital was of poor quality. This observation is consistent with the findings of Burroughs et al,17 who studied patient concerns about medical errors. Those who had a particular concern did not then have a “global” concern about the possibility of multiple errors. Most patients with limited English expressed enormous appreciation for the Australian healthcare system, often accompanied patriotically with bold declarations of love of Australia, its perceived values and ideals, and confirmatory comments about their decision to migrate. Many perceived the countries they had immigrated from as having poorer-quality healthcare.

One implication of “The Happy Migrant Effect” is that the health system might lack important data. If patients with little or no English are accepting of what others perceive to be serious health-service flaws, then there is every possibility that some negative events may not be formally recognised. In this case, processes and procedures that may prevent such events are less likely to be undertaken. This may help explain why ethnic disparities in care and in hospital outcomes are elusive and difficult to address. In all patients in this study where problems with healthcare were reported, the events were considered by the researchers to be largely preventable by appropriate language facilitation, patient and family involvement, and provider respect and compassion.

Healthcare services have much to learn from listening to the hospital stories of patients with limited English. Sage18 wisely urges physicians to “take a step away from the community of experts approach . . . and move a step closer to the patients.”