While our data suggest that a policy change to use of a large interpreter video network from traditional face-to-face and telephonic interpreting model may have contributed to reducing the time Spanish speakers spent in the ED, we were unable to show a statistically significant impact of this change on care in the Emergency Department. We did find, however, that the risk of patients leaving the ED before receiving appropriate care was significantly reduced among Spanish speakers, compared with English speakers, after the implementation of the video-interpreting network in one of our study hospitals.
Other studies of the impact of the provision of professional interpreter services on ED care have found somewhat more consistently positive results. Hampers and McNulty (2002)
documented a reduction in resource utilization, cost, and rates of admission to the hospital when professional interpreters are employed in a pediatric ED (Hampers and McNulty 2002
). Bernstein et al. (2002)
showed that patients who received interpreter services in a Boston ED were comparable to English-speaking patients in LOS, testing, and procedure completion. In addition, interpreted patients in this same study had higher adherence to primary care and specialty-clinic referrals than either of the other two groups (Bernstein et al. 2002
). We may not have found similar results because of differences in region of the country and study population or because this was a large observational study in which we examined what happened to Spanish speakers at the population level, not just those known to get services via the video-interpreting network.
We also may not have found the results we expected because our outcome measures were not as dependent on communication as we had hypothesized. Given the nature of ED practice (established practice patterns, protocols, malpractice avoidance, admission bed flow issues, etc.), there are many other factors that influence our outcomes that probably had a stronger influence on what happens in the ED than accurate communication between clinician and patient. Clearly, our measures were based on hypotheses about how the video network would enhance physician understanding in their communication with patients. In hindsight, we may have found a more significant impact if we examined measures that were dependent on the patient's understanding, such as comprehension and adherence to communicated follow-up. This idea is supported by one of the previous studies that found interpretation did significantly impact likelihood of following up on recommended outpatient referrals and reduce return to the ED (Bernstein et al. 2002
While possible, it is unlikely that the mode of interpretation delivery, via video-interpreting, contributed to our finding less impact than we expected. Our study appears to be the first to evaluate the impact of a shift to video-interpreting on care outcomes. We identified three other studies of interpreter services delivered via video, all of which compared some measure of patient and provider satisfaction with video-interpreting to in-person and telephonic interpreting (Jones et al. 2003
; Locatis et al. 2010
; Napoles et al. 2010
). In all three, encounters in which interpretation was provided via video were rated as high or in some cases higher than the other methods of providing interpreter services. Clearly, this is an acceptable method of interpretation.
Our study has several limitations. First, this was a natural experiment in which we measured the global impact of a change in policy to use video-network interpreter services on health care. We were not able to measure who actually received interpretation via this network. It may be that we would have been able to demonstrate an impact on care in those individuals for whom we could document use of these services. Alternatively, it may be that the ED clinical staff did not choose to use the services. However, we have data that suggest that there was a rapid increase in use of the interpreter services in the ED after implementation of the video network (Paras and Associates, personal communication). More likely is the fact that there are many other factors that influence our outcomes, as previously mentioned, that may have a stronger influence on what happens in the ED than accurate communication between clinician and patient. In addition, there are limitations to using EHR and billing data. As the language, age, and gender data were recorded by clerks/administrative staff for operational rather than research purposes, the data may be inaccurate or incomplete; however, the data were documented to be of high quality in one of the study hospitals. In addition, using ICD-9 codes to identify patients presenting to the ED with chest and/or abdominal pain may not have captured all patients entering the ED with this complaint or led to the inclusion of patients with other initial complaints. Given that the same EHR and billing data and set of ICD-9 codes were used in each hospital, for both English and Spanish speakers, in the period before and after the implementation of the video-interpreting services, it is unlikely that this biased our comparative analysis.
Finally, just because we did not demonstrate a statistically significant impact of a policy shift to provision of interpreter services via a video network on our six outcomes, it does not mean that they have not had a significant impact on the hospitals, clinicians, and patients that used them. Administrators, clinicians, and patients recognize that interpreter services, and their efficient provision through the video network studied, have inherent value and moral weight. Policy makers do as well, as demonstrated by the development and dissemination of both Federal and state policy directing that interpreter services be provided to LEP patients (Department of Justice 1964
; Clinton 2000
; Onecle.com 2006
). Timely, accurate, and unburdened communication is a universal need in health care, and LEP patients are at the greatest risk of not being able to participate in, consent to, and understand their health care interactions. Provision of interpreter services through a video network may be one method for reducing these barriers and making sure that patients who seek care in a hospital will receive the standard of care—communication in a language they can understand.