The Rural Tax: Out-of-Pocket Costs for Patient Travel in British Columbia

Background: A signicant concern for rural patients is the cost of travel outside of their community for specialist and diagnostic care. Often, these costs are downloaded to patients and their families. Methods: Online retrospective provincial survey seeking to estimate the out-of-pocket (OOP) costs and associated experiences of rural patients traveling to access health care in British Columbia. Respondents were surveyed across ve categories: Distance Traveled and Transportation Costs, Accommodation Costs, Co-Traveler Costs, Lost Wages, and Patient Stress. Results: On average, costs for respondents were $777 and $674 for transport and accommodation, respectively. Patient perspectives obtained from this survey expressed a number of related issues, including the physical and psychosocial impacts of travel as well as delayed or diminished care seeking. Conclusions: These key ndings highlight the existing inequities between rural and urban patient access to health care. This study can directly inform policy related efforts towards mitigating the rural-urban gap in access to health care.

The survey instrument design was guided by previously conducted qualitative focus groups and a quantitative survey conducted by the Centre for Rural Health Research aiming to determine the health care priorities of rural citizen-patients and communities [16]. Participants were recruited through the research centres' outreach strategy, which includes advertisement through rural community-speci c social media pages, local newspapers and chambers of commerce.
Additionally, patient partners assisted in recruiting participants by circulating hardcopies of the survey to members of their home and adjacent communities, who may not have been able to access the survey online. The survey consisted of 68 closed questions (Likert scale/yes-no) and three open-ended questions.
The instrument was pilot-tested in a small study sample (n = 56; ~5% of the total sample) from November 16 -December 4, 2019. Analysis of pilot-phase data revealed the need to capture more broadly on costs incurred when traveling to access health care. Based on this, we expanded the inclusion criteria to assess the costs incurred to access any health care covered by the provincial Medical Services Plan in BC, including surgical or specialist care and on-going treatment. The Medical Services Plan is the provincial health insurance program for BC residents and covers health care bene ts including required medical services, diagnostic services, and some supplementary bene ts [17]. The updated version of the survey was available on December 4th, 2019.
Eligibility criteria included: (i) rural citizens who were living in one of BC's Rural Practice Subsidiary Agreement communities for at least six months prior to participating in the survey, (ii) at least 19 years of age, and (iii) traveled from their community to access care (or escort someone who needed care) within the previous two years. Respondents were asked to report expenses for the most recent health care event that required travel (e.g., a surgical procedure, and/or cancer care). A few respondents included trips for multiple issues in one survey response, as travel was likely required for more than one condition at the same time. A descriptive analysis was performed to calculate frequency, proportions and the average cost per person. Six respondents reported outlying individual travel or accommodation costs and were removed as to provide reasonable estimates for average OOP spending. All costs are reported in Canadian Dollars as of 2020.
This study was co-funded by the Health Economics Simulation Modelling Methods Cluster, BC SUPPORT Unit and the Joint Standing Committee on Rural Issues, through the larger context of the Rural Surgical and Obstetrical Networks program, which works to stabilize and enhance surgical and obstetrical services in rural communities across BC. The study was approved by the Behavioral Research Ethics Board at the University of British Columbia (Certi cate Number H19-00445).

Results
A total of 381 rural citizens participated in this survey across rural BC ( Figure 1); this included 56 respondents from the pilot test sample. A total number of 339 participants started but did not submit the survey; their data was not included in our analysis.

Quantitative ndings
Quantitative ndings from the survey are grouped into seven categories: distance traveled and transportation costs, accommodation costs, costs by Health Authority, co-traveler costs, system-level supports, lost wages, and patient stress.
(1) Distance travelled and transportation costs Figure 2 provides the cost in reference to the distance travelled. The average transport cost per person was $777, and the average distance traveled per person to receive care was 1966 km. Among respondents who reported having pre-operative visit(s) for surgery (n=167), most (87%) traveled outside of their community for their visit(s). This was also the case for respondents reporting post-operative visit(s) (n=95) -88% had to travel for their visit(s).
Airplane tickets were the most expensive type of transportation and cost on average $1581. The most common type of transportation expense was gas, with 86% of respondents reporting this expense. Figure 3 provides more information on transportation costs.
(2) Accommodation costs More than half of survey respondents (58%) reported paying for accommodation. These costs averaged $674 per person and represented the second most expensive type of OOP spending. Half of the respondents reported hotel expenses, which was also the most expensive type of accommodation. Although BC Cancer Agency offers subsidized housing for patients, a high average number of nights resulted in the highest average total costs for this accommodation type ($2205). Table 2 provides more information on accommodation costs.  (3) Co-traveler out-of-pocket costs Family members or friends accompanying patients also incurred signi cant OOP costs. Thirty-one percent of respondents reported separate co-traveler transport and/or accommodation expenses: 13% reported co-traveler transport costs, averaging $1077 per person and 25% reported co-traveler accommodation costs, averaging $862 per person. Refer to Appendix 2 for more information on co-traveler transportation and accommodation expenses.
Most respondents (80%) traveled with someone who was not a health care professional. A spouse was the most common travel companion, followed by a child. While in the community of care, 18% of respondents had someone other than a co-traveler visit them. In total, 85% of respondents had a co-traveler and/or a visitor.

(4) System-level support for out-of-pocket costs
Only 14% of respondents reported having had some of their OOP transport and/or accommodation costs covered by organizations like the BC Travel Assistance Program or the First Nations Health Authority, the body responsible for the administration of health programs and services for B.C. First Nations [19]. Of the 53 people who received transport assistance, 37 did so through the BC Travel Assistance Program (mostly for ferry tickets), and ve through the First Nations Health Authority. Only six respondents (2%) reported having received nancial support for out-of-pocket accommodation spending. Five of these six respondents also reported assistance with transport costs.

(5) Lost wages
For many respondents, time spent away from home meant lost wages. When asked whether they had to take unpaid time off work to receive care, 93 respondents said yes and 56 said no (the remaining did not respond to this question). Those who lost wages missed an average of 17 workdays and an average of $2276 in personal income.

(6) Patient stress
Respondents were asked to rank their overall stress on a scale of 1-10 for their most recent health care visit, where 0 is no stress/anxiety and 10 is the worst imaginable stress/anxiety. The 315 people who responded to this question reported an average stress level of 5.9. On average, those who had higher OOP costs reported higher levels of stress. In addition, higher-income respondents experienced less stress on average. Overall, the stress level was seemingly unaffected by whether or not someone received nancial assistance. See Appendix 3 for reported levels of stress by amount spent and income categories.
The burden of traveling for care was particularly signi cant for maternity patients. Twenty-six percent of respondents reported that they were the caregiver of a child or other dependent. About half (52%) of these respondents with dependent(s) had to arrange for someone to care for their dependent(s) while they traveled to access care. Patient stress is discussed further in qualitative ndings, below.

Qualitative Findings
Participants provided qualitative descriptions of their experience of travelling for care by answering three open-ended survey questions. A research assistant familiar with the context of the study used NVivo qualitative data software to organize themes derived from the data. Patient partners then reviewed the themes to validate them against their own experiences. Participant responses touched on multiple, intertwined themes including: challenges with transportation; the psychosocial impact of travel; the physical impact of travel; and delayed or diminished care-seeking. Each theme is described in detail below.
(1) Challenges with transportation Aside from nancial costs, participants expressed that having to arrange and undertake transport was the most di cult part of leaving their home community to access care. Many participants commented on transportation di culties in relation to their particular geographic environment. For example, many participants discussed the impact of winter road conditions on traveling to receive health care. Several participants recounted being involved in motor vehicle accidents. Other respondents commented that they had to delay care-seeking because they could not drive on dangerous winter roads and could not afford to travel. However, even for those patients who could afford to y rather than drive, some still experienced issues getting back to their community due to winter weather conditions: "Even with being on disability and not having to deal with working around my work schedule, it is di cult to get out of my valley in the wintertime. Flights keep being cancelled and booked solid with no available seats. It's great that the ticket is paid for but when you get stuck in [tertiary centre] for a week 'cause of the ights being cancelled due to weather and no available seats the other expenses can really add up." Other participants in water-bound communities discussed particular geographical challenges for arranging transportation to their health care appointments based on set ferry schedules.
Regardless of geographic location within the province, many respondents reported the expense of 'wear and tear' on their vehicles, such as at tires and the need for early replacement of a car due to unexpected high mileage.
One of the most stark transport challenges for participants in this study, however, was securing transportation back home after an urgent event. (3) Time away from home and the physical impact of travel In addition to the psychosocial stressors of not having social support, many participants expressed a range of other reasons why having to spend time away from home was di cult for them. For example, some participants commented on the challenges of eating out and staying in hotels with speci c dietary or allergy-related concerns. Other participants commented more generally on the impact of having to travel on their physical recovery. One participant noted, "As it [condition that required travel] was due to arthritis the driving was extremely hard on my muscles and joints" while another observed, "With chemo treatments I have no immune system to ght off germs." Several other participants a rmed the di cultly of having to travel directly after a hospital procedure. As one participant noted, "…the most di cult procedure for me was the biopsy and I had to y home with a bleeding and painful wound." Spending time away from home was particularly di cult for families with young children. Challenges included having to miss school to attend their parents' medical appointments and needing speci c types of care from parents that made it di cult to be away from them. For example, one woman described the impact of an unexpected surgery on her husband and young child: "This was an unexpected emergency surgery that happened [the] same day symptoms presented themselves. My husband and son accompanied me to the hospital and when they decided I would require surgery and an overnight stay, my husband needed to head back home with our 10-month-old as he had not prepared for an overnight [stay]. I was also not able to breastfeed due to medications and we had no breast milk on hand. This meant they needed to make the 2-hour trip back the next day to get me and then 2 hours home again. Lots of driving for a small child."

(4) Rural gaps
Many participants expressed that having to travel for certain types of care was expected as a rural resident. However, they also felt that there were some essential services that should be available in their local community but were lacking. Most notably, there was a perception that many rural communities are lacking an adequate number of family doctors, leading to an over-reliance on emergency services. As a result of the closure of the walk-in clinic in their rural community, one participant even commented that they felt they had no other choice but to pay for private care.
A second rural health gap described by some respondents was the lack of alternatives to in-person specialist visits, such as visiting specialists or opportunities for virtual care. This was perceived by some participants to be the result of inadequate systems planning. As one person commented: Some participants commented that after considering the costs and impacts of travel, they delayed or diminished their health care seeking. One participant said, "My child should be assessed for autism but the trip to Prince George is unaffordable." Delayed or diminished care seeking seemed to be more common among individuals who had to rely on others to take them to health care appointments. Others commented from the perspective of a family caregiver, noting the di culty in ensuring access to recommended care: "I cannot take time off work to get my disabled mother to some recommended medical therapies that are not available in or near my home community." Even the knowledge that a local doctor would likely refer the patient to a distant specialist prevented some individuals from seeking care in the rst place: "I have not gone to the Dr. knowing that they would send me to a specialist far away and we couldn't afford the costs at the time." While many participants commented that they had to budget and plan for costs associated with traveling for health care services, some expressed that they would have to cancel or reschedule their appointments at the last minute due to unexpected inability to afford travel. One respondent noted, "Postponed neurological appointments because I could not afford travel. Credit cards and credit line maxed out" while another commented, "[I] have had to cancel out of town medical appointments due to loss of wages and burden of nding child care." Limitations Like all voluntary retrospective survey studies, we anticipate that BC residents who experienced greater di culty in dealing with the nancial and psychosocial burden of traveling for health care were more motivated to respond, thus potentially limiting the transferability of our ndings to all rural BC residents. A cursory comparison of our cohort demographics with that of the BC rural population showed that proportions of respondents from various income brackets and ethnic groups were generally consistent with those of the BC population, however, the age distribution of survey respondents differed from that of rural BC, with the average age for our cohort (53) being much higher than that of Rural BC (42).
Balancing these potential limitations, however, was the relatively high number of responses and geographical spread of respondents. Regardless, as this is the rst rigorous collection and presentation of comprehensive OOP costs for rural residents traveling to access health care, we feel it provides useful information to an under-explored area of health care experiences.

Discussion
At the time of pan-Canadian regionalization, Church and Baker suggested that Canada's geography makes it di cult to achieve the economies of scale that make regionalization a functional model for more densely populated jurisdictions [20]. Speci cally, they note, …all in all, regional populations in Canada might be too small to achieve any real economies of scale or to more generally affect coordination of health services" 2 A Ministry of Health appointed Advisory Committee on Ontario's Rural Health Hubs Framework found that insu cient public transportation was one characteristic of rural communities that limits access to care [21]. We have seen diminished support for rural transport through for-pro t services across Canada (for example, through the withdrawal of Greyhound bus services) which exacerbated the growing issue of transport to specialist care in regional communities and, predictably, has disproportionately disadvantaged vulnerable populations with reduced social and nancial capital. In this way, transport (as a proxy to access to care) has become a social determinant of health.
Patient-travel from rural and remote communities to larger centres is a key assumption of regionalized health care systems where patients bene t from regional specialist care. In many jurisdictions, this improves access to such care for most of the population, as they no longer need to travel to larger urban centres, thereby affecting the 'closer to home' advantage. For smaller rural and remote communities, however, regionalization can diminish local access to all, but primary care as regional procedural care must include the caseload of rural residents in order to maintain a viable case volume for specialist call groups.
From a systems perspective, this may be an appropriate cost-bene t calculation with increased rural patient travel being a necessary by-product. There are however, productive ways of ameliorating the effect on rural residents through a reconceptualization of both patient travel and system supports for those instances when travel is necessary.
A consequence of the regionalization of health services that many jurisdictions in BC have undergone in the past two decades has been the attrition of specialists in low-volume communities in exchange for regional concentrations. Although access to care is still prioritized, the default mechanism of achieving this (in non-urgent situations) has been through patient-initiated travel. This is not the case, however, in many instances where regional specialists provide clinical care in smaller communities through regularly scheduled outreach visits. This is usually contingent on having enough accrued patient volume to justify the travel. When this is not possible, optimizing the potential of virtual care, either with or without the involvement of local care providers, also acts to lessen the burden of travel for rural patients. This may involve virtual visits between a specialist and rural patient supported by a local care provider or direct specialist-to-patient care. We have recently seen the capacity for health system adaptation for increased virtual care in the face of COVID-19; instilling the infrastructural resources and work ow patterns into rural and referral communities to support the expectation of virtual care where possible will create a legacy for this adaptation. When framed within these opportunities to reduce non-urgent patient travel, we can recognize the value of a paradigm shift where instead of being the rst recourse to access to care, patient travel becomes the last resort.
Within this framework, the health system could, for example, immediately reduce the need for travel for pre-operative care. In this study, 87% of respondents who reported pre-operative visits reported having to travel for such care (and 88% post operatively) despite respondents' perception of the lack of urgency for in-person visits. Preoperative care could reasonably be offered through virtual care in many circumstances or, in some instances, involving local care providers in a tripartite model with the patient and specialist if hands-on care is required. There are other system-level solutions to minimize the 'rural tax' on patients in accessing health care, such as supported accommodation in regional referral centres. In this study, respondents paid an average of $674 in accommodation costs. There are already models in place providing subsidized accommodation to defray such costs in cases like cancer care or pediatric emergencies that we could learn from.
Further, the importance of social support in optimizing health outcomes must not be underestimated, nor the costs associated with this support. In this study, 85% of respondents were accompanied by and/or had visitors in the referral centre. This additional expense is often covered by the patient themselves, particularly in situations where accompanied travel post-visit is essential. In considering a holistic view of wellness, the mitigating in uence of social support would be wise to consider not just to reduce system costs, but as a compassionate counterpoint to frequent criticism of 'depersonalized health care' [22,23]. In instances where patients struggle to afford travel costs for themselves, let alone escorts, the question of whose responsibility it is to enable access to necessary health care remains.
To our knowledge, this is the rst primary research study to systematically document the nancial consequences of traveling for care for rural residents and, as such, provides important information for health care planners. A broader societal perspective of costs, including costs that are downloaded to individuals and families, is essential to include in health care planning and decision-making, especially given that the impact of OOP cost expenses are most strongly felt by those who lack nancial and social resources. A broad view of cost accounting also includes considering less tangible costs, such as increased stress and anxiety that occur alongside the stress of the medical event. This may be due to not only nancial worries, but also as a result of losing support networks of family and friends when having to travel. If the time out of the community is extended, then there is also disruption to usual routines, which is particularly di cult for caregivers including those with young children. Although we acknowledge that "not everything that can be counted counts and not everything that counts can be counted," through the rich descriptive comments provided by survey respondents, we can start to better understand the consequences of traveling for care.

Conclusion
The results of this survey provide a starting place for discussions on the role of public support for rural residents who need to travel for health care. These discussions must involve key stakeholders from rural communities but also regional representatives and government ministries entrusted with ensuring appropriate access to care, transportation and social development. Bringing the right group together in jurisdictions challenged with patient travel for nonacute health care will provide a starting place for developing a system response to ensure all residents have access to the health care they require, without nancial barriers.