Dana Wiens1, David Robinson1, Irene Smolik1, Cheryl Barnabe2, Hani El-Gabalawy1 and Liam O'Neil1, 1University of Manitoba, Winnipeg, MB, Canada, 2University of Calgary, Calgary, AB, Canada
Background/Purpose: Rheumatoid Arthritis (RA) is a chronic autoimmune disease that requires access to subspecialty care. Although Canada has a universal healthcare system, there are complex and interrelated factors that lead to inequitable healthcare access and delivery. These factors are particularly relevant to Canada's geographically dispersed First Nations People (FN), who bear a disproportionate burden of RA and its complications. In the context of a longitudinal study of RA onset in a FN population, we sought to identify factors that influence access to healthcare in a cohort of FN RA patients and their First-Degree Relatives (FDR).
Methods: A longitudinal cohort of FN RA patients (n = 214) and their FDR (n = 617) was recruited between 2005 and 2017 to participate in a prospective study of RA risk in the FDR (total n = 831). Study participants were recruited in both urban and rural locations in Manitoba, Canada. The study enrollment visit included a healthcare access survey which measured access on a Likert scale (Figure 1). Overall access difficulty was determined using a binary variable, where if any of the 7 questions scored moderate or higher, the individual was deemed to have 'access difficulty', otherwise they were classified as 'adequate access'. Modified health assessment questionnaire (mHAQ) and disease duration were collected for RA patients. Data were analyzed using chi-square test and logistic regression.
Results: Overall, RA and FDR participants living in rural communities reported more difficulties with healthcare access compared to urban dwellers (overall access difficulty, 53.5% vs 33.5%, p< 0.0001), and rural RA patients reported more access difficulty than unaffected rural FDR (Figure 1A, 66.9% vs 50.5%, p=0.002). In contrast, there were no differences reported between RA patients and FDR living in urban locations (Figure 2B). In the entire RA cohort, no differences in access were reported based on disease duration, age, or sex, although those with higher mHAQ scores tended to report worse access. A logistic regression model of the entire RA and FDR cohort suggested that variables which independently associated with healthcare access difficulty were female sex (Figure 2A, OR 1.47, 1.07-2.01), older age (OR 1.51, 1.12-2.04) and living in a rural community (OR 1.99, 1.45-2.71). The model suggested that females living in rural locations, irrespective of an RA diagnosis, were particularly disadvantaged for healthcare access, but also that males with an RA diagnosis experienced substantially more access difficulty compared to FDR males (Figure 2B).
Conclusion: Perceived difficulties in accessing healthcare were reported more frequently in FN RA patients as well as their unaffected at-risk FDR who were living in rural locations compared to those living in urban locations. We also identified sex, age and location of residence-based differences in perceived healthcare access for FN persons irrespective of disease state. In order to achieve equitable healthcare delivery in the context of a universal healthcare system, interventions to address geographic factors, such as transportation and availability of healthcare providers, need to also incorporate complex factors related to sex, gender and age.
Figure 1: Access to care questionnaire.
Figure 2: Poor access to care in First Nation RA patients is restricted to rural communities. (A) Overall access difficulty in RA patients compared to first-degree relatives (FDR) in rural communities. (B) Overall access difficulty in RA patients compared to first-degree relatives (FDR) in the urban setting. Differences calculated by chi-square test.
Figure 3: Poor access to care in First Nations RA patients and first-degree relatives is predominantly mediated by older age, female sex and living in a rural community. (A) Results of a logistic regression analysis that included the variables Age, Sex, RA diagnosis, and timing of the visit displaying OR with 95% confidence interval. Statistically significant OR were observed for age, sex and rural. (B) Effects plots for logistic regression interaction term between Sex and RA diagnosis. Differentially poor access was observed for males with RA, while females with RA and FDR displayed similar access.
D. Wiens: None; D. Robinson: None; I. Smolik: None; C. Barnabe: None; H. El-Gabalawy: None; L. O'Neil: None.
