^# Marc S Hoffmann is a third year medical student at the University of Pennsylvania School of Medicine.

^* Julie M Linton and Lee Shearer are fourth year medical students at the University of Pennsylvania School of Medicine and will graduate in May, 2007.

^Ω Keira A Cohen is a second year medical student at the University of Pennsylvania School of Medicine.

⁺ Nancy K Etzel is currently pursuing an MSN at the University of Pennsylvania School of Nursing.

Background

In 2000, the Census Bureau estimated that 45 million people in the United States - nearly 20% of the population - spoke a language other than English at home (1). Approximately 43% of those people reported speaking English less than “very well” (1). As the Latino and Asian populations in the United States continue to expand, the number of Americans with Limited English Proficiency (LEP) is expected to increase. Associated with this rise is the need for appropriate health care services that can meet the needs of patients with LEP.

Patients with LEP face both linguistic and cultural barriers to receiving adequate health care. Recent studies show that these barriers translate into health care disparities, including poorer health status (2), decreased access to care (3,4), and decreased quality of care (5). Although language-concordant providers are often ideal, increasing linguistic diversity makes it unrealistic to request that hospitals always provide doctors who speak the same language as the patient. Medical interpretation offers a more practical means to facilitate effective communication between patients and providers. A variety of individuals currently provide these services, including family members and other ad hoc interpreters (such as untrained bilingual hospital staff), professional telephone interpreters, contracted interpreters from outside agencies, and on-site interpreters.

As medical interpreting is a complicated undertaking with important ramifications for patient care, it is critical to assess the abilities and training of interpreters. Several recent studies indicate the importance of using trained interpreters rather than ad hoc interpreters in both the pediatric and adult setting, as they are associated with fewer medical errors of clinical significance (6) and higher patient satisfaction (7,8). One study showed that perceived quality of the interpreter is strongly associated with patients' assessments of quality of care overall (9). Another study specifically demonstrated that when hospital-trained interpreters were used rather than telephone or ad hoc interpreters, parents in a pediatric emergency department were more satisfied with interactions with hospital personnel and quality of care (8).

Title VI of the 1964 Civil Rights Act mandates that hospitals provide medical interpretation, and most states also require that interpreters be trained and demonstrate language proficiency (10). However, while providers are expected to provide medical interpretation services, these services often are not available (11) and generally are not reimbursed (12). The lack of funding not only limits the ability of hospitals to meet the needs of LEP patients, but also downplays the importance of medical interpretation in patient care. Inadequate interpretation services thus represent a significant shortcoming in the delivery of health care to immigrant and minority communities. A recent position paper by the American College of Physicians encourages increased use of interpretation services to reduce racial and ethnic health disparities and specifically suggests that providers seek volunteer interpreters to reduce the costs of these services (13).

Located in the culturally and linguistically diverse setting of urban Philadelphia, the hospitals of the University of Pennsylvania Health System (UPHS) are faced daily with the challenge of serving LEP patients. Census data for 2005 show that 20% of the population of the City of Philadelphia speaks a language other than English at home. Of this population, 47% speak Spanish and 20% speak an Asian or Pacific Islander language. In keeping with national statistics, 40% of Philadelphia's Spanish speakers and 60% of Philadelphians who speak Asian and Pacific Islander languages speak English less than “very well” (14). Last year alone, the Hospital of the University of Pennsylvania (HUP) served patients speaking over 25 different languages.

UPHS currently uses a variety of interpreter services, including telephone interpreters, contracted interpreters, ad hoc interpreters, and trained volunteer interpreters. Trained volunteer interpreters represent an important component of UPHS's language services because they facilitate patient-provider communication, decrease use of untrained and ad hoc interpreters, ease the financial burden of providing face-to-face interpretation, and reduce dependence on bilingual staff who must take time away from their assigned duties to interpret. This intervention thereby promotes the basic rights of LEP patients while increasing hospital efficiency.

Program Description

Penn Language Link (PLL) is a group of trained student volunteer interpreters at the University of Pennsylvania who address the needs of LEP patients by providing face-to-face interpretation within UPHS' clinics and hospitals. Volunteers interpret for a variety of clinical scenarios including clinical encounters, patient education classes, genetic or social work counseling, consenting for procedures, assigning DNR status, and helping patients to understand other paperwork. Currently, PLL is authorized only to provide spoken language interpretation and is not authorized to translate written documents. However, volunteers are able to perform “spot translation” where the interpreter reads the form aloud in the patient's primary language, which facilitates accurate completion of paperwork and enables informed consent. Legally, interpreters are covered by the hospital's insurance policy.

PLL is currently run as a clearing-house, matching volunteers with providers in need of an interpreter for a given patient interaction. Most PLL requests are received from ambulatory clinics or specialty services, such as Obstetrics and Gynecology (Table 1). Many OB/GYN requests are emergent and thus have a significantly lower fill rate than adult outpatient or psychiatric appointments, which are typically scheduled in advance. Clinics or inpatient wards make requests via e-mail or in-person to the group of PLL co-coordinators, who e-mail the request to the group of PLL volunteers that speak the requested language. If no PLL volunteer is available to fill a request, the hospital or clinic uses an outside interpretation service to accommodate the language need. PLL co-coordinators also have a pager and will accept emergency requests from patients or physicians on a case-by-case basis, but PLL volunteers do not take call. In addition to providing services at UPHS clinics, PLL recently established collaborations with several community clinics that provide health services to immigrant communities.

Resources

Funding

PLL began operation at the University of Pennsylvania School of Medicine in 1997 and operated at virtually no cost until 2001 when Pennsylvania state law mandated the training and certification of all medical interpreters. Due to a lack of funds to finance trainings, PLL fell dormant for several years. However, in the fall of 2003, three first year medical students revived interest in PLL. With the support of faculty advisors and the medical school administration, these students secured a $10,000 grant from UPHS in 2004, formulated a new protocol for delivering services, and organized a formal training/certification event. PLL applies directly to the CEO of UPHS each year for a grant extension, receiving $5,000 in 2005 and $10,000 in May 2006. An additional $1500 was secured through an Alpha Omega Alpha (AOA) Medical Student Service Project award (http://www.aamc.org/about/awards/cfc.htm) in 2004, and $150 is provided annually from the Medical Student Government to cover meetings and social events.

PLL was founded on the conviction that all people should have the right to communicate effectively with health care providers. However, the. Butre are equally compelling arguments for the cost savings that PLL generates for the hospital system. Cost savings arguments are a key component in grant applications to hospital administrators responsible for the financial health of an institution: well-intentioned programs that also provide cost savings are both initially more viable and ultimately more sustainable. Cost savings estimates are supplemented by arguments referring to the risk of error when using untrained ad hoc interpreters, as well as improved patient satisfaction when using trained interpreters. In the competitive health care market, patient satisfaction has become increasingly important to hospital administration.

Volunteers

In recruiting volunteers, PLL takes advantage of a culturally and linguistically diverse student body at the University of Pennsylvania, with student volunteers from the medical, nursing, and undergraduate schools. Having multiple perspectives enriches the educational value of the program and reflects the importance of medical interpretation to all levels of patient care. Undergraduate students bring considerable energy to the program and often have more flexible schedules that enable them to take on interpretations during normal business hours.

PLL currently has 63 interpreters available, of whom 33 interpret in Spanish, 11 in Mandarin, 6 in French, 4 in Korean, 3 in Cantonese, 2 in Russian, 3 in Vietnamese, and 1 in Italian. PLL initially selected languages based on a combination of hospital statistics for most requested languages and student volunteers' fluency. In the 2002-2003 fiscal year, the most recent year for which these statistics were available to PLL, the top five languages requested were Spanish, French, Mandarin, Vietnamese, and Russian; Chinese, including both Cantonese and Mandarin, was seventh on the list (15). However, since the languages requested at the hospital are not always reflective of the requests that PLL actually receives, subsequent years' data collected by PLL on its actual interpretations are now used to determine the need for specific languages. For instance, the demand for Mandarin interpreters has increased recently, prompting recruitment of more Mandarin interpreters; nine new interpreters just completed training in October 2006. Availability of students with appropriate language skills is also a factor. Fewer students speak fluent French, Cantonese or Vietnamese, making recruitment for these languages more difficult.

Implementation

The initial PLL training in March 2004 included a group of 20 student volunteer interpreters from the medical, nursing, and undergraduate schools. The training included didactic instruction, role-play, and a formal assessment to earn certification in medical interpretation. The program rapidly gained momentum, and second and third trainings for 20 volunteer interpreters each took place in October 2004 and April 2005. In September-October 2006, the training was expanded to the 40-hour nationally recognized Bridging the Gap curriculum (www.xculture.org). The certification course covers basic interpreting skills, introduction to the healthcare system, cross-cultural skills, and communication skills for advocacy. Participants receive a 450-page manual, a language-specific medical glossary, culture-specific profiles on health care practices, and a guide to common medications. The curriculum is not language-specific, and thus enables training for multiple languages during one certification course. Breakout sessions in language-specific groups allow for interpreter role plays and discussion of cultural aspects relevant to health care.

PLL is committed to providing high quality interpretation services and therefore a high level of language proficiency is required in order to qualify to take the training course. Prior to enrolling, prospective participants undergo a rigorous language proficiency examination administered over the telephone that is provided by Language Line University (http://www.languageline.com). Having a third party assess language proficiency assures competency, avoids possible conflicts of interest, and is required for the Bridging the Gaps curriculum. If students are not selected to take the certification course, PLL provides suggestions for further development of language skills and offers opportunities for other types of involvement including community work and publicity.

Developing a website has facilitated promotion of the program throughout the hospital and the rest of the community. The IT staff was instrumental in website development, not only pointing out critical components that need to be included on the website, but also programming the website and converting the text into a functional webpage. The site is at www.med.upenn.edu/pll and includes background information about the program, a link to request an interpreter, and contact information for the program leaders.

Outcomes

Results

In the past three years, PLL has received 227 requests for interpretation in six different languages (Table 2). Spanish language requests comprise almost 70% of requests, French and Mandarin each contribute about 15%, and the remainder are distributed among Cantonese, Vietnamese, and Korean. This distribution reflects the growing Latino and Asian immigrant populations in South and West Philadelphia and the rapidly expanding French-speaking West African and Haitian populations in West Philadelphia. The larger fraction of Spanish requests can be attributed both to the higher percentage of Spanish-speaking patients in UPHS clinics and recently established partnerships with Spanish-speaking community clinics. PLL also has interpreters in Russian and Italian, but has never received a request in those languages.

Overall, PLL has been able to fill 74% of requests. The ability to fill requests is directly related to the setting of interpretation and the availability of interpreters. Only five of seventeen Mandarin interpretations were filled before the recent recruitment of nine additional volunteers in October 2006. Since their recruitment, all eleven requests were filled. Additionally, only nine of the initial seventeen requests were for inpatient interpretations, while all eleven recent requests were for outpatient services. Of the 24 Spanish interpretations that were not filled, sixteen were patient no-shows. Of the remaining eight, two were emergency requests and four others occurred before November 2004, when PLL had fewer than fifteen Spanish interpreters. French requests are filled at a lower rate because volunteers are less available - of the six French-speaking volunteers, three are in their core clinical clerkship year and two are pursuing second degrees.

Compared to the cost of having a third party medical interpretation service perform the interpretation, PLL provides services at a significant savings. Since 2003, PLL has provided 169 face-to-face interpretations, logging nearly 300 hours at an estimated net cost savings of $36,800 (Table 2). Each professional encounter costs UPHS an estimated $200. This estimate includes the direct costs of the fee for interpretation as well as incidental costs such as parking, travel, and meals that the hospital must provide. This estimate was generated with the office responsible for interpretation at UPHS and is likely to vary from institution to institution. UPHS is charged even if the patient does not show and therefore no-show encounters are included in the cost savings estimate, but are not counted for the number of interpretations. During its first two grant cycles, PLL saved UPHS a net of $24,600, with a direct savings of $9,600 after subtracting out $15,000 received in grant money. Currently halfway through its third grant cycle ($10,000), PLL has already generated $12,200 in net cost savings, giving a direct savings of $2,200 for the cycle.

Discussion

One principal challenge to the viability of PLL has been the ability to provide a consistent flow of interpretation opportunities to volunteers. This is important not only to keep the organization running effectively, but also to maintain volunteer morale. Three principal factors have contributed to this problem: 1) lack of awareness among hospital staff; 2) natural irregularity in the hospital census; and 3) short notice for ER and inpatient interpretations. There is not a consistent understanding on the part of all hospital staff - including physicians and nurses - about how to access and use interpreters, the importance of interpretation in providing safe and effective care, and the improved results obtained with trained as opposed to ad hoc interpreters. To address this issue, PLL currently is developing a workshop curriculum to train health care professionals in how to access medical interpreters and use them most effectively. A public relations campaign within UPHS may help improve awareness of both the importance of medical interpretation and of available services.

Given the short notice and the irregularity of the hospital census, inpatient and ER interpretations are more difficult to provide. Since interpreters are volunteers with other obligations, the program is not well suited to these types of encounters. Efforts have been made to implement an “on-call” system, but such a system is challenging to implement logistically and previous attempts proved unsustainable. It is not practical for volunteers to take call for a 24 or 48 hour period because most have other obligations that would interfere with their ability to respond to all pages. In addition, the volume of interpretations is not large enough to justify shifts taken in the hospital or ED. PLL's experience has been that volunteer morale diminishes if no interpretations are available during a shift, and many are less willing to volunteer for these types of assignments. PLL has responded by focusing efforts on outpatient clinics that make appointments in advance and community clinics that have regularly scheduled blocks of time dedicated to patients speaking a particular language. Early results indicate that fill rates have improved with the increased fraction of outpatient requests. Recruitment for clinical trials is another potentially fruitful arena where volunteer interpretation services may be used, and PLL is engaged in active discussions with UPHS' clinical research groups to establish a partnership.

In an institution with a regular volume of interpretations and a sizable group of volunteers willing to take call, implementing a call schedule may be extremely effective. However, when initiating a program, the top priority ought to be establishing a viable program with a steady flow of work. PLL's experience has been that regularly scheduled outpatient appointments offer the best opportunities for the growth and development of a volunteer organization. Once the program is established, it may be worthwhile to consider carefully the logistic challenges and potential benefits of an on-call system.

There are several limitations to the cost-savings model presented which may underestimate the value of PLL's services. It cannot account for different lengths of interpretation or different methods of interpretation (the value of face-to-face versus telephone interpretation) and, perhaps most importantly, it cannot assess indirect cost savings that may be generated by foregone diagnostic tests and procedures. In spite of these issues, the estimate of $200 per encounter is reasonably accurate and, importantly, was developed in accord with hospital administration that oversees these expenses. Since hospital administration is our main source of financial support, these individuals need to accept the accuracy of the numbers. Including them in the process of cost estimates is therefore critical.

Another key limitation to the program is our inability thus far to assess whether medical outcomes have been improved through PLL's interventions. A principal motivation for the program was a moral commitment to improve clinical outcomes of LEP patients. However, outcomes data for medical interpretation are difficult to assess. Hospitals do not routinely record whether an interpreter was used or who performed the interpretation. Randomizing patients not to receive an interpreter would be unethical and, given that patients come with a variety of diagnoses, appropriate control groups to measure medical outcomes are difficult to identify. Studies of patient satisfaction with interpretation services are more amenable to study and have fewer confounding variables; the methods and results of current published studies are summarized in a 2005 systematic review by Flores (16). Rigorous studies of patient satisfaction have used comprehensive survey instruments and compare results among a variety of interpretation strategies (ad hoc, telephone, face-to-face) or compare interpretation to bilingual providers. A study of this magnitude exceeds PLL's current financial and administrative means. However, the existing voluntary survey already used by UPHS to measure patient satisfaction could be given to LEP patients receiving a PLL interpreter. The results of this cohort could then be compared to the hospital average for non-LEP patients. While results would not be conclusive and a selection bias may develop due to the voluntary nature of the survey, such a study would help establish whether patients are equally satisfied with an encounter using an interpreter compared to a language-concordant encounter.

Another potential criticism is that volunteer interpreters provide an inferior service compared to paid professional interpreters. Professional interpreters may have more experience interpreting than PLL volunteers, and this may result in better outcomes. However, PLL volunteers undergo the same training program as professional interpreters, and students bring a background of general medical knowledge and clinical experience that is often superior to that of professional interpreters and which can be helpful in delicate clinical scenarios. Furthermore, as mentioned above, the American College of Physicians specifically suggests that providers seek volunteer interpreters to reduce the costs of these services (13). To the best of our knowledge, there has not been any literature to date examining the efficacy of student volunteer versus professional interpreters.

PLL also contributes to the quality of medical, nursing, and undergraduate education. With the growing understanding of the importance of cultural competency, medical educators have responded largely through didactic learning and case discussion. Although this approach promotes awareness, developing truly effective cross-cultural skills requires experiential learning as well. By providing a venue for students to volunteer as interpreters, PLL offers a hands-on approach to cross-cultural medical education while also contributing a valuable service to the hospital system. Facilitating patient-provider communication demonstrates the critical role of interpretation services in effective care for patients of limited English proficiency (LEP). Helping LEP patients negotiate the healthcare system fosters a greater appreciation of how both access and provision of care are obstructed by language and cultural barriers, which bear particularly heavily on issues of patient safety, quality of care, and patient satisfaction. Additionally, students have the opportunity to advocate for patients, gain clinical exposure, and learn to work with, and as, interpreters. Finally, through involvement with the entire health care delivery team, PLL volunteers are in a key position to lead their peers in their exploration of cultural competency. Such benefits have also been reported by a service-learning program in medical interpretation at Brown Medical School (17).

Conclusions and Recommendations

PLL is a unique student-run service organization that provides much needed medical interpretation services to LEP patients within the UPHS system. The program offers a cost-effective alternative to current systems of medical interpretation while also enabling students to obtain clinical and cross-cultural skills. Key current efforts include educating hospital staff in how to access and optimally use interpreters, and continuing to develop relationships with outpatient and community clinics.

Recommendations for Starting a Student Volunteer Interpretation Program:

Develop several key advocates within the medical school faculty and administration and establish a good working relationship with the hospital administration. Administrative assistance is invaluable in developing an effective and organized program.

Contributors:

MH, JL, and LS drafted the manuscript. All authors revised the manuscript for key intellectual content, read, and approved the final manuscript. MH is the guarantor and corresponding author for this article. Address Correspondence to:

Marc S Hoffmann

4530 Pine St, Apt 1

Philadelphia, PA 19143

Phone: (314) 409-5047

Email: msh3@mail.med.upenn.edu

Acknowledgements:

The authors would like to thank our advisors, Dr. Steve Larson and Ms. Hilda Luiggi, for their impassioned support of this organization. We would also like to thank the University of Pennsylvania Health System for continued financial support of our service organization. Finally, we would like to thank the administrators of the University of Pennsylvania School of Medicine for continuing to support our organization's efforts.

References

Care: A Systematic Review. Medical Care Research and Review, Vol. 62 No. 3, (June 2005) 255-299

Table 1. Requests by Setting, 2003 - Present.

* Includes surgical and medical floors

^**Includes inpatient and outpatient pediatrics

Table 2. Interpretation Requests by Language, 2003 - Present.