Barriers to accessing radiation therapy in Canada: a systematic review

Wait times were the most frequently identified theme in the health system context. Studies investigated the length of wait times[8, 23, 40, 43, 44, 49, 53] and whether the length was actually a barrier to utilization[17, 39, 50]. The most recent wait time data found worsening wait times in Nova Scotia for breast cancer patients in 2004[49], improved wait times in BC between 2004 and 2008[53], and unchanged wait times between 1995 and 2005 at an institution in Toronto[44]. Looking at head and neck cancer patients in the Maritimes, Belyea et al.[40] found that between 74% and 94% of patients were waiting longer for RT between 2007 and 2009 than target standards set by Cancer Care Ontario. Despite the variability in wait times between jurisdictions and over time, subjective interviews and survey data from 3 studies found wait times to be nonetheless be identified as a barrier to RT utilization by either patients or providers[17, 39, 50]. Patients were potentially refusing the modality because of the wait, and providers were often not making referrals due to a perception that treatment would be delayed.

The impact of distance to treatment centres was inconsistent. Three articles support distance as a barrier[20, 33, 37], but only in certain scenarios. Danielson et al.[20] found it to be a barrier only in older and lower income patients and when coming from a larger community. Johnston et al.[37] found that distance was no longer a barrier once the patient had entered the RT system. The influence of distance on utilization rates was confounded by lower referral rates in remote locations, such as the BC interior and rural Ontario[10, 24, 50]. No studies explicitly addressed whether lower referral rates reflected patients’ reluctance to travel for RT or the physicians’ lack of consideration of RT as a treatment option.

Characteristics of RT treatment centres, including their geographical distribution, were important in considering distance as a barrier to RT[21]. Rural counties and those without facilities had lower rates for palliative RT in Ontario[33, 84] and Nova Scotia[30], and for breast and lung cancers in BC[36]. Conversely, this barrier was not seen in prostate cancer in BC[36]. If the initial diagnosis was made in an academic hospital or one with an affiliated cancer centre, or if a patient’s referring physician resided in a county with a cancer centre, the patient had a significantly greater likelihood of receiving RT[19, 27, 33]. Tyldesley et al.[36] argued that rural areas have fewer specialists with adequate knowledge of the indications for RT, and Gray et al.[25] highlighted a lack of patient support modalities. Notably, patient outcomes were not different between the high-volume and low-volume institutions[45].

Themes elucidated in these categories were often interrelated, and tended to cluster around age and socioeconomic status. Referral rates were inversely related to age. While the magnitude of difference varied, older patients were found to be less likely to be referred for RT in a variety of contexts, jurisdictions, and disease sites (Table 3). In most instances, multivariate analyses demonstrated increasing age and comorbidities to be independent predictors of non-referral, and even non-treatment with respect to RT. As noted by Tyldesley et al.[78], “relative decline in the use of radiotherapy for palliative and adjuvant indications is more rapid than the decline in functional status with age in the general population”.

The relationship between income and access to treatment was discussed in eight articles, but the relationship was unclear. Those from lower income areas generally had lower rates of consultation and utilization[20, 30, 33, 37, 62]. However, this relationship was found to be statistically insignificant on univariate analysis in another study[27]. In yet another, it was noted that income extremes had a much greater effect on rural populations in California than on similar populations in Ontario[69]. In some of these studies, there was a confounding variable of distance from treatment centre (often associated with lower income) that was not always controlled for in the analyses. Interestingly, lower income conferred a slight benefit in Quebec, where there was found to be an association with shorter wait times[23].

Other patient-dependent barriers included education level, associated with longer RT wait times[23], and life expectancy, associated with lower rates of consultation and treatment for palliation[30, 37]. Patient refusal was also noted as a barrier[9, 71], though it was ill-defined in the study by Puts et al.[71], as patients were not asked for their reasons for refusing treatment. The authors did note that it was most often older patients living alone, potentially refusing for such reasons as a lack of physician recommendation or not perceiving a need. Chinese women in BC refused adjuvant RT after breast-conserving surgery more than South Asian or Iranian women. This was attributed to the desire expressed by Chinese women “to be rid of their cancer” in a way they did not believe possible with RT[9]. No such differences were found between ethnic groups for surgery, chemotherapy, or hormone therapy utilization. This single study suggests that cultural beliefs, ethnicity, and poor understanding of the nature of RT could be barriers to RT, though these factors were not identified elsewhere in the Canadian literature.

The major themes identified under Provider Factors were lack of referral and poor knowledge or awareness about RT. These themes are integrally related and were often the underlying issue identified in the discussion of other barriers. Higher referral rates were associated with having a university academic appointment, being a specialist in a cancer centre, performing a higher volume of surgeries for cancers potentially requiring adjuvant RT, being more knowledgeable about RT, or having formal training in RT[13, 14, 19, 30, 82]. It was particularly noteworthy to find that physicians lacking certification by the Royal College of Physicians and Surgeons of Canada had higher referral rates for neoadjuvant RT for rectal cancers, being more likely to refer all patients rather than only those with certain clinical indications[19]. In a retrospective chart review in Nova Scotia, Lavergne et al.[30] found the difference between referral and utilization rates varied greatly between certain disease sites. The difference was high in Head and Neck cancers, where there were often concerns about comorbidities, and those referred were often not deemed eligible for treatment. For those who had received previous RT, referrals almost always led to RT utilization. A study in British Columbia noted that “the referral system acts as the initial gateway to the cancer system. The [British Columbia Cancer Agency] can only provide appropriate care to those who are referred”[24].

Low referral rates can be attributed in part to limited RT-related knowledge of the referring physician. Self-reported knowledge was poor in the majority of survey respondents in three separate investigations[13, 34, 82]. Poor self-rated knowledge was also correlated with poor actual knowledge in terms of indications for RT, quality of life considerations, and potential treatment effectiveness. A study in Nova Scotia found that patients referred for palliative RT tended not to be referred a second time, suggesting that providers were under a mistaken belief that RT could only be used once[30]. Another issue for referring physicians was not understanding the RT referral process, or not knowing who to contact to make a referral[13, 34, 85]. The understanding of patient preference relating to RT was also considered. Tucker et al.[14] reported that 51% of Canadian paediatric oncologists believed palliative RT was underutilized for reasons such as family reluctance, distance, doubt about effect on quality of life, and concerns regarding toxicity, but the authors expressed concerns that some of these beliefs might be misinformed. Tyldesley et al.[78] confirmed that “physicians can be inaccurate in the judgment of patient preference” (p479).

Only 16% of family physicians in Eastern Ontario[13] and 62% of Canadian paediatric oncologists[14] reported having had any formal RT training. The scope of this education ranged from a single lecture to a year-long clinical rotation for the family physicians, and lasted a median of 4 weeks in duration for the paediatric oncologists. Samant et al.[13] reported “lack of education [amongst family physicians in Eastern Ontario]…. to be a major barrier to radiotherapy” (p662). Indeed, 94% of those with training made referrals, as opposed to 73% of those without[14]. The majority of providers in various contexts agreed that they might benefit from more information about RT[13, 34], with one study noting that there would be a preference for small group sessions to accomplish this[34]. Samant et al.[13] made the important point that educating family physicians might have a high impact for few resources.