Data sources
I: Qualitative phase
Purposive sampling was used to select organizations with a variety of program usage patterns (low, medium and high), and of different sizes (small and large) and types (e.g., hospital, community health centre). All selected sites (n = 9) had at least one provider and one manager participate in the qualitative phase. The refusal rate during recruitment was 10 %. Semi-structured interviews were conducted with 31 program stakeholders who had experience with OPI; they included two members of the TC LHIN leadership, 10 service managers, 17 service providers, and two administrative staff. The four interview guides, one for each type of stakeholder,focused on experiences with the OPI program from introduction through implementation.
II: Descriptive cross-sectional quantitative phase
Surveys for providers and patients were developed based on themes identified in the literature as well as those uncovered in Phase I. All surveys were web-based and self-administered. Both providers and patients were provided with web links to complete the survey, though patients were also given the option of completing a paper survey that they returned to the research team using a postage-paid envelope. Surveys were distributed to all 34 sites that had signed on for and had used the OPI program at the time of the evaluation. The organizations’ managers recruited service providers to complete the provider survey, and service providers recruited patients to complete the patient survey. This convenience sampling approach resulted in a sample of 127 providers (representing 30 of the 34 sites, or 88 %) and 41 patients. The provider survey contained closed- and open-ended questions about changes in communication strategies, frequency of usage, impact on patients, care appropriateness, and program satisfaction. The closed-ended patient survey asked about their satisfaction with the interpretation services, opinions about its impacts, and what they would do if it were not available. The patient survey was translated into the top 10 languages accessed through the OPI program at the time of the evaluation.
Results
Characteristics of respondents
Effects on use of alternative interpretation sources prior to and after OPI
Half (52 %) of the providers reported ‘often’ or ‘always’ relying on assistance of family and friends to interpret prior to the implementation of OPI, and an additional 37 % reported using them at least some of the time. Nearly one-quarter (23 %) used untrained administrative staff for interpretation help ‘often’ or ‘always’, and over one-third (35 %) relied on other providers (refer to Table
1). One-tenth (11 %) requested patients to bring their own interpreters to appointments, and 6 % relied on other patients to help with interpretation ‘often’ or ‘always.’
Table 1
Strategies used “Often or Always” prior to and after OPI implementation
Assistance of patient’s family and/or friends who speak needed language | 52 | No data available |
Face-to-face professional interpreters | 37 | 24 |
Assistance of other providers who speak needed language | 35 | 16 |
Assistance of administrative staff who speak needed language | 23 | 11 |
Asking patients to bring their own interpreters to appointments | 11 | 4 |
Assistance of other patients who speak needed language | 6 | 7 |
Volunteer language interpreters | 4 | 2 |
Referrals to other agencies | 3 | 2 |
Interview respondents described several ethical and practical concerns with relying on family or staff. Issues related to family members, for example, included lack of training (“you don’t know what they’re translating; they’re not bound by any sort of ethical training”) and sensitive situations for which they do not want to involve family or friends. Similarly, health care providers worried about non-clinical staff’s capability (“regular communication can be quite different than communication pertaining to medical issues”) as well as taking them away from their own work (“when you’re pulling other nurses to come and translate for you, when you’re pulling housekeeping… you’re pulling them away from their work.”)
When asked to compare their interpretation practices prior to and after OPI implementation, these providers revealed a marked decrease in the use of ad-hoc, non-professional strategies. Reliance on other providers or administrative staff, for example, decreased by about half, and requesting patients to bring their own interpreters decreased even more substantially. The use of face-to-face professional interpreters also decreased considerably, from 37 to 24 %, primarily due to conveniences associated with OPI such as not needing to book ahead of time.
Effects on health care quality and access
Providers and patients expressed similar perspectives on the overall quality of care with OPI (84 % of providers; and 85 % of patients reported it was ‘improved’ or ‘significantly improved’), and on several aspects of patient engagement, including overall comfort level (both, 72 %), relationships between providers/patients (72 %; 68 %), and disclosure of patients (68 %; 72 %)(refer to Table
2). On the whole, about three-quarters or more of both groups reported overall improvements on all of the measures of patient engagement with their provider and their care. One exception was just half of patients saw improvements in their privacy after OPI was available, compared with two-thirds of providers who did, though this difference was not statistically significant.
Table 2
Effects of OPI on health care quality and patient engagement: perspectives of providers and patients
Overall quality | |
Overall quality of care | 84 | 85 |
Patient engagement with Providers/Care | |
Patient’s comfort level | 72 | 72 |
Relationship between provider and patient | 71 | 68 |
The disclosure of patients | 68 | 72 |
Patient’s privacy | 67 | 51 |
Patient autonomy | 78 | - |
Patient engagement | 78 | - |
Patient access to you organization | 73 | - |
Understanding of information given during appointment | - | 87 |
Likelihood to ask questions during the visit | - | 84 |
Likelihood to recommend the health care organization to other friends and family who speak the same language | - | 82 |
Ability to schedule follow-up or future appointments on time | - | 75 |
Ability to follow health care provider’s instructions | - | 74 |
Questions asked only of patients reveal additional aspects of ‘quality’ and ‘access’; for example, four-fifths or more said they had a greater understanding of information given them during their appointment (87 %) and were more likely to ask questions (84 %); three-quarters reported an increased ability to follow through on providers’ instructions and on scheduling follow-up appointments. A majority (82 %) also said they would recommend the organization to other family members and friends as a result of its OPI capability.
In qualitative interviews, respondents frequently described these issues – access, quality, comfort, engagement, and autonomy - as closely interrelated. One health care provider, for example, called OPI a “wonderful service, which has increased our access and timeliness in reaching out to our clients in a manner which makes them most comfortable and able to participate in the arrangement of their care.” Another noted, “Interpretation is really important for maintaining the health of the individuals in the community. It is very important they understand instructions really well, and that they express what they want to tell physicians really well to be able to get good health care.”
A large majority of the providers considered OPI ‘appropriate’ for encounters involving supportive (90 %), acute (88 %), and chronic (86 %) care. They raised more concerns about its appropriateness for mental health encounters; noting, for example, the special importance of building rapport and trust, and a need to sensitively interact with patients. Nonetheless, most rated it either ‘appropriate’ (73 %) or ‘somewhat appropriate’ (19 %) in mental health encounters.
Asked to speculate about how the loss of OPI might affect themselves and their organization, 81 % providers said they would struggle to engage with patients, nearly three-quarters noted the quality (74 %) and efficiency (71 %) of care would decrease, and nearly two-thirds reported patients’ access to care (64 %) would be compromised (refer to Table
3). Half (49 %) of the patients said they would revert to asking friends or family members to help them, and one-fifth (17 %) said they would try to interpret on their own and another one-fifth (17 %) indicated they did not know what they would do. One-third (32 %) of patients said it would mean seeking another health care provider.
Health care providers described these impacts in strong language (“a lifeline that is cut, because communication is everything for us”) and warned some patients are especially isolated and would lose access entirely. As one provider commented, for example: “a few clients…don’t have anyone who could translate for them and they don’t speak English, and for them I don’t know what we would do. They wouldn’t be able to access care.”
Table 3
Perceived impact of loss of OPI
Increased difficulty for staff to engage patients | 81 |
Decreased quality of care | 74 |
Decreased efficiency of care (time) | 71 |
Decreased use of phone interpretation | 68 |
Patient access to care would be compromised | 64 |
Increased financial cost to offer interpretation | 60 |
Impact on reputation of organization (organization would no longer be seen as accessible to non-English speaking patients) | 44 |
What would you do if the OPI services are not offered anymore by the organization where you received it?
|
% of Patients (n = 41)
|
Ask a friend/family member to help me with interpretation | 49 |
Find a health care provider who speaks my language | 32 |
Stop going to the organization and find another one that offers interpretation | 20 |
Try to understand what providers say without the help of interpreters | 17 |
Don’t know | 17 |
Satisfaction
A large majority of both providers (93 %) and patients (85 %) reported being ‘satisfied’ or ‘very satisfied’ with OPI services overall (refer to Table
4). Their satisfaction with various aspects of the services, including timeliness of access, and quality and professionalism of interpreters, was also very high and closely aligned. A large majority (more than 80 %) of patients expressed satisfaction with issues related to their communication and comfort with the interpreter and the health care provider. Approximately two-thirds of providers expressed satisfaction with program coordination, training, and materials; most of the remainder rated these as ‘neutral’ (neither satisfied nor dissatisfied) and dissatisfaction ratings were rare.
Table 4
Provider and patient satisfaction with OPI
Overall satisfaction | 93 | 85 |
Timely access to interpreters/available in needed languages | 92 | 83 |
Quality of interpretation | 95 | 88 |
Professionalism of interpreters | 92 | 83 |
Confidentiality of interpretation | 91 | - |
Wait times | 89 | - |
Technology and equipment availability | 83 | - |
Training to use the program | 72 | - |
Program coordination/management | 70 | - |
Program reference materials | 63 | - |
TC LHIN leadership | 55 | - |
Relationship with doctor or health care provider | - | 93 |
Confidence in interpretation | - | 90 |
Your understanding of information provided during appointments | - | 90 |
Your ability to communicate with the doctor or health care provider | - | 90 |
Your comfort level during appointments | - | 85 |
Quality of telephone equipment | - | 76 |
Discussion
This mixed methods study assessed the experiences and perceived impacts of making OPI accessible to a variety of hospitals, clinics, and community health care providers in Toronto, Ontario, a diverse urban centre. As noted previously, most of the research on OPI services has heretofore focused on experiences of physicians and medical residents within hospital settings. This obscures the reality that patients are accessing health care in a wide range of settings from providers of various disciplines. For example, while a majority (59 %) of our patient sample used interpreter services in a hospital setting, 93 % had also accessed interpreter services in a community health facility. This high usage suggests great potential to expand the reach of OPI into a broader range of settings where patients access services. One systematic review found cost to be the primary barrier to use of OPI in community healthcare settings; by removing this barrier for community-based health centres and support programs, this initiative has already seen strong uptake of OPI [
20].
As we consider implementation of OPI into more community-based settings, we also need to heed the voices of a more diverse group of healthcare providers; our sample, for example, included nurses, social workers, case managers, and care coordinators. With rare exception in this study, provider and patient experiences with, and opinions about, OPI strongly converged. While the literature consistently shows patient satisfaction related to OPI, provider satisfaction tends to be mixed [
9,
25]; the level of convergence we saw may be attributable in part to the diversity of providers included in our study and/or the absence of preferable alternatives in community-based settings. In general, this study suggests more research is needed to understand the interpretation needs and experiences of patients and providers in more diverse healthcare settings.
Our mixed methods approach and diverse participant sample also revealed a broad range of positive impacts that OPI has on health care service delivery; again, prior research has tended to focus on a relatively narrow set of outcomes (e.g., ease/cost of use, communication quality, satisfaction). This study found providers and patients reported improved patient-provider relationships, comfort and privacy, but also an increased capacity and likelihood to schedule follow-up appointments, follow health care providers' instructions, disclose information, ask questions, and recommend the health care organization to family and friends. Patients and providers largely agreed on this broad range of positive impacts associated with patient autonomy and health care accessibility. While this provides a general endorsement for OPI, it also suggests that a better understanding of the range of benefits it can provide would help to inform and customize the role that OPI plays for patients and providers in a variety of settings. It is also notable that providers’ interview responses revealed strong interconnectedness between these impacts, exposing a far more nuanced and multifaceted nature of general concepts like health care access and quality. Measures of quality, for example, should arguably encompass issues such as a patient’s comfort disclosing to a provider, understanding of information the provider offers, and a likelihood to ask questions during the visit.
A large majority (86–90 %) of these providers considered OPI appropriate for encounters involving supportive, acute and chronic care. This is a more expansive endorsement than found in the extant literature, which generally concludes that OPI is most acceptable for simple, brief appointments not requiring visual communication, such as administrative, ancillary, or follow-up meetings [
20,
29]. And, while the providers in this study raised more concerns about OPI’s appropriateness for mental health encounters, most still deemed it appropriate (73 %) or somewhat appropriate (19 %). Asked to explain ratings other than ‘appropriate,’ most of the concerns related to a personal preference for face-to-face interactions, a need for better understanding of mental health symptoms, and the cumbersome nature of the technology. Much of the provider (again, primarily physicians) resistance discussed in the literature references a range of implementation problems, such as rooms not wired for telephone use, lack of interpreter training, and long wait times; however, misconceptions about its use, aspersions about its value based on past experiences, and/or contentment with the status quo are nearly as common [14, 21, 24). Indeed, one author concluded that a key factor which impedes research and “stymie(s) professional debate” on the effectiveness and impact of telephone interpretation is “the persistence of (usually unsubstantiated) myths and stereotypes of TI” [
35]. The relatively greater openness of providers toward OPI in this study, all of whom had used it in practice, suggests it may have broader applicability than previously assumed, particularly when preferable alternatives are unavailable. It is also noteworthy that all but one of the providers reported, during qualitative interviews, that access to OPI had either no impact on their workload or had decreased it overall. In general, though, it is critical that reasons underlying resistance be exposed and better understood if we are to effectively address them.
Frequency of reliance on family and friends – though we know it is fraught with issues – was used pre-OPI at least some of the time by 90 % of these providers – 52 % used them often or all of the time; reliance on other ad hoc solutions such as other providers or administrative staff were also common. Asked what they would do should OPI no longer be available, 81 % of providers said they would struggle to engage with patients, and nearly as many noted overall quality and efficiency of care would decrease and/or that patients access to care would be compromised. Half of patients said they would revert to asking family/friends, and a strong minority reported their access to care would be negatively affected. This, and providers’ comments that some patients would lose access entirely, reinforces the importance of finding a coordinated solution to interpretation needs within all of these health care settings. Failing to offer access to quality language interpretation services is clearly incompatible with a commitment to health equity.
This sample had broad representation from the organizations, providers, and patients accessing OPI as part of this initiative, and the two-phase exploratory design enabled development of well-tailored survey instruments. This evaluation also had several limitations, including a lack of qualitative input from patients due to logistical barriers. The use of convenience sampling means it is possible that survey respondents differed from non-respondents on important characteristics, and the completion of surveys primarily occurred in non-acute health situations, which may also have affected results. Our strategy of asking the intervention group to rate their experiences before and after the OPI program implementation is also prone to recall bias. On the whole, however, these methodological decisions enabled us to obtain information from a larger variety of respondent types and healthcare settings than we otherwise could have; future research should expand on these promising findings.