Is Spanish Language a Barrier to Domestic Violence Assessment?

Introduction

At least 1 in 4 women in the United States reports having suffered or having been threatened with physical domestic violence during her lifetime. ¹ Such abuse, as incurred by a spouse, former spouse, fiancee, boyfriend/girlfriend, or a cohabitant, is thought to be markedly underreported. Nonetheless, domestic violence remains the leading cause of injury to premenopausal women and results in the death of over 1000 women per year. ² Furthermore, domestic violence appears to have long-term consequences: victims manifest higher rates of sexually transmitted diseases (STDs), sexual dysfunction, chronic pain symptoms, heart disease, cervical cancer, fatigue, and arthritis over time compared to the general population. ^{1,3 –8}

Despite the magnitude and widespread ramifications of domestic violence, few healthcare providers ask about it. Surveying 582 physicians, Rodriguez et al. ⁹ reported most physicians admit to not asking. Although 79% of physicians do ask when a woman has sustained an injury, this percentage plummets to 6% among internists and 17% among gynecologists during routine health maintenance examinations. Why do healthcare providers shun this topic? Physicians cite numerous barriers, including patients' fear of retaliation by the partner, patients' unwillingness to disclose, patients' fear of police involvement, lack of physician training and time constraints, sparse resources to deal with reported domestic violence, and lack of privacy during the evaluation. ⁹ Notably, 56% of physicians also describe feeling ill at ease in broaching this subject with a patient from a different culture and 39% cite an inability to speak the patient's language as factors that preclude a discussion of domestic violence.

This latter observation is an important one. Although numerous studies suggest that domestic violence occurs regardless of the victim's ethnicity and that ethnicity does not protect any group, language appears to have a strong influence on patients' disclosure of sensitive information to healthcare providers. Julliard et al. ¹⁰ conducted a qualitative study among 28 Latina women, 23 of whom described how language barriers had a negative effect on disclosure in general. Their study described how healthcare providers sometimes appeared frustrated with a patient's accent. The involvement of translators led to concerns about confidentiality, particularly if a family member was helping with the translation. Referring to the latter, one patient stated, “That situation is really embarrassing.”

Despite such challenges, assessing domestic violence remains important, as indicated by the endorsement of such practices by several major healthcare organizations, including the American Medical Association, the American Academy of Family Physicians, and the American College of Obstetricians and Gynecologists. ^{11 –13} Thus, the current study was undertaken to test the hypothesis that language poses a barrier to healthcare providers' asking about domestic violence and that Spanish-speaking patients report lower rates of assessment for domestic violence. This study sought also to assess other factors that might be associated with domestic violence assessment.

Materials and Methods

Results

Discussion

This multiclinic study explored the hypothesis that Spanish-speaking women are less likely to have ever been queried about domestic violence. We initially observed the exact opposite in our bivariate analysis: Spanish-speaking women were more likely to report having been assessed for domestic violence compared to English language-dominant women, a finding that was reduced to nonsignificance in the multivariate model. In contrast, we did find that age (based on quartiles), being married, and having at least one pregnancy were directly associated with a greater likelihood of patient-reported lifetime assessment, whereas several other patient-related factors, such as race/ethnicity, mean age, and education level, were not. Taken together, our findings suggest that language is not a barrier but that healthcare providers' assessment for domestic violence remains infrequent.

Of importance, our findings also suggest that certain healthcare providers may be more committed than others to asking patients about domestic violence. For example, in our study, a higher lifetime assessment rate appears to have occurred among patients seen at clinic 3. Although women are not required to always attend the same clinic, many do. It is important to note that all clinics follow the same guidelines, and all are known for providing patient-centered care. ¹⁸ Similarly, providers may be more conscientious in assessing women at postpartum visits or assessing women who have a child, although this interpretation must be viewed with caution, as our assessment question pertained to ever and did not include assessment practices at the time of the current visit and not necessarily assessing at a women's health visit. Nonetheless, the foregoing observations suggest that medical specialty (for example, obstetrics) or certain practice patterns among specific groups of healthcare providers may translate into more pervasive assessment practices. In turn, these observations may also suggest that domestic violence assessment might be a mode of action that can be successfully inculcated into other healthcare providers' practice patterns over time; if some healthcare providers appear to assess the majority of their patients, perhaps others can as well.

How do our patient-reported assessment rates compare with those previously reported by others? Rates of screening for domestic violence remain highly variable, depending on the healthcare provider's medical specialty, the circumstances surrounding the healthcare encounter, the patient's age and socioeconomic status, and the patient's previous history of domestic violence. ^15,19,20 In general, however, patient-reported lifetime rates vary from 2% to 50%, with the latter having been derived from an addiction recovery unit and with most rates being on the lower end of this range. Thus, the rates observed in our study appear to be higher than what one might find in many other clinical settings.

Our study has limitations. First, this study focused heavily on patients' recall of domestic violence assessment as opposed to written documentation or direct observation that a healthcare provider had in fact assessed for domestic violence. Although subtle, this distinction does have implications for healthcare providers' adherence to guidelines. A healthcare provider may in fact meticulously ask patients about domestic violence, but patients might not remember these queries. Indeed, by relying on patient recall, our study might be underestimating assessment rates. Second, our analyses used patients' choice of the Spanish or English questionnaire as an indicator of their dominant/preferred language when, in fact, further questioning revealed that a notable percentage of patients were proficient in both. Our conclusions might have been different—and potentially more indicative of a language effect in multivariate analyses—had we surveyed a group of strictly unilingual patients. Third, this study was undertaken within a group of clinics in southeastern Texas. The regional confinement of our study population raises some concerns about our ability to generalize findings widely. Fourth and finally, the terminology used in our screening questionnaire might potentially lead to patients' underestimating whether they had been ever assessed for domestic violence. The latter term might suggest only physical violence to some patients, not taking into account psychological aggression or even sexual violence from the patient's partner. Future studies on such patient-reported assessment rates might choose to use other, more encompassing terminology and behavior-specific measures.

Finally, one might, in fact, question if routinely asking women about domestic violence is of value. Although not all studies have been consistent, ²¹ at least two recent randomized controlled trials—one that examined assessment followed by discussions and referrals at the discretion of the treating physician and another that identified abused patients and then intervened in an attempt to improve victims' depression—have not demonstrated that assessment or a subsequent intervention improves outcomes among abused patients. ^22,23 Such data have led some to conclude that if assessment does not yield tangible benefits, it should not be pursued. ²⁴ Admittedly, assessing domestic violence is a controversial topic, but we contend that it remains important. The two trials alluded to ^22,23 are noteworthy in that they demonstrate that assessment, at the very least, is safe and yields no harm, such as partner retaliation. In view of the humanistic aspects of the healthcare provider-patient encounter, the importance of learning the multidimensional aspects of the patient's life and her healthcare concerns, and the higher rates of disease among domestic abuse victims, we believe that it behooves healthcare providers to ask about domestic violence, to continue to do so in a language that patients truly understand, and to work toward improving domestic violence assessment rates. Perhaps Bradley et al. ²⁵ put it aptly when they described the role of domestic violence assessment “as a way of uncovering and reframing a hidden stigma.”