Sex-related differences in health-care use in adult patients with inflammatory arthritis in Ontario, Canada: a population-based cohort study

Abstract

<h2>Abstract</h2><h3>Background</h3> Timely access and appropriate use of health-care resources is crucial for the management of people with inflammatory arthritis. Sex (biological) and gender (sociocultural) are important health determinants that influence different aspects of care, such as referral to specialists, performance of diagnostic tests, continuity of care, and prescription patterns, which ultimately affect disease outcomes. We aimed to compare health-care use between males and females with inflammatory arthritis. <h3>Methods</h3> This population-based study was conducted using health administrative data of Ontario, Canada. We included Ontario residents aged 20 years or older with valid health insurance and diagnosed with rheumatoid arthritis, ankylosing spondylitis, and psoriatic arthritis using Ontario health administrative data (the Ontario Health Insurance Plan Claims Database). Sex (the primary predictor of this study) was coded as male or female in the database. Health-care use (the study outcome) was assessed at least once per year for 3 years before and after an inflammatory arthritis diagnosis and compared between males and females using regression models, after adjusting for demographics and comorbidities. Prescription dispensation was assessed at least once per year for 3 years after the diagnosis in individuals aged 66 years or older at the time of diagnosis. Results were expressed as female to male adjusted odds ratio (OR) with 95% CIs. <h3>Findings</h3> Between April 1, 2010, and March 31, 2017, 41 277 people with rheumatoid arthritis (28 575 [69%] females), 8150 with ankylosing spondylitis (4160 [51%] females), and 6446 with psoriatic arthritis (3483 [54%] females) were analysed. Males were significantly older (mean 60·4 years [SD 14·2]) than females (mean 57·1 years [15·2]) in the rheumatoid arthritis cohort. Across the cohorts, multimorbidity was more common in females (75–81% <i>vs</i> 58–66% in males) as was depression (8% <i>vs</i> 4%) and osteoporosis (4–5% <i>vs</i> 1%), whereas cardiovascular disease was more common in males (7–15% <i>vs</i> 4–7% in females). Females were more likely to visit family physicians (adjusted OR ranging from 1·10 [95% CI 1·05–1·15]; p<0·001 to 1·15 [1·04–1·28]; p<0·01) and rheumatologists (adjusted OR from 1·32 [1·22–1·43] to 2·28 [1·92–2·70]; p<0·001). Females also had higher odds for receiving imaging before a diagnosis in all 3 cohorts (adjusted OR from 1·15 [1·10–1·20] to 1·18 [1·12–1·23]; p<0·001 for x-rays and adjusted OR from 1·27 [1·09–1·48] to 1·44 [1·15–1·80]; p<0·01 for ultrasound examinations). Similarly, laboratory tests were more likely to be ordered for female patients before diagnosis than for males (adjusted OR from 1·10 [1·05–1·15] to 2·17 [1·89–2·50]; p<0·001) across all cohorts. These sex differences were most notable among people with rheumatoid arthritis and ankylosing spondylitis, and in the prediagnostic period. Males were more likely to visit the emergency department immediately before an inflammatory arthritis diagnosis (adjusted OR from 0·76 [95% CI 0·72–0·81]; p<0·001 to 0·87 [0·80–0·95]; p<0·01). After diagnosis, females were more likely to remain in rheumatology care (adjusted OR from 1·12 [1·01–1·25]; p=0·04 to 1·24 [1·19–1·30]; p<0·001). Prescription dispensation of non-steroidal anti-inflammatory drugs (adjusted OR 1·14 [1·04–1·25]; p<0·01) and opioids (adjusted OR 1·46 [1·27–1·68]; p<0·001) was higher in older females (aged 66 years or older) with rheumatoid arthritis and conventional disease-modifying antirheumatic drug use was higher in older females with ankylosing spondylitis (adjusted OR 1·66 [1·24–2·22]; p<0·001). <h3>Interpretation</h3> Females with inflammatory arthritis are more likely to use health care than males, which might indicate sex-related biological differences in disease course or gender-related sociocultural differences in health-care access, health-care seeking behaviour or patient–provider interactions. <h3>Funding</h3> Enid Walker Graduate Student Award for Research in Women's Health (partial).