In this administrative health data population-based study, contemporaneous incidence rates and point prevalence of adults living with DM were determined, as well as medication use in Alberta, Canada. The standardized overall average annual incidence was estimated at 2.8–3.0 cases per 100,000 adults between 2013 and 2019, and standardized prevalence was 28.6 cases per 100,000 adults on March 31, 2019, which is greater than most previous reports5,6,7,8,9,10,12,13. The characteristics of adults living with DM were consistent with epidemiologic findings4,5,14; DM was diagnosed in the middle aged, and there were twice as many adult females with DM compared with males. During the first year after the DM incident date, 27% of individuals did not receive any DM-related medications; among those who had medication use, glucocorticoids (prednisone and topical corticosteroids) were found to be the most commonly used drug class, and other common systemic therapies included methotrexate, hydroxychloroquine, mycophenolate mofetil, and immunoglobulin G which is in alignment with empirically-based recommendations and treatment guidelines for DM from other countries15,16,17,18. Findings from this study provide a contemporary description of the epidemiology and medication use of adults living with DM in Alberta, Canada.
The reported incidence and prevalence of DM varies widely due to the limited number of studies conducted in this area, the different methodological approaches, and the rare nature of the disease. Reported incidence ranges from 0.1 to 10 cases per 100,000 person-year and prevalence ranges from 1.97 to 21.5 per 100,000 population5,6,7,8,9,10,11,12,13. Large population-based cohort studies have reported a DM incidence of 1.4 (2003–2008) to 1.7 (2004–2008) cases per 100,000 person-years in the USA12,13, and 6 to 10 cases per 100,000 person-years was estimated in Norway between 2003 and 201211. To our knowledge, this study is the first to report the incidence of DM in Canada, and results indicate that incidence among the adult population in Alberta may be higher than the USA, but lower than Norway.
Applying a similar administrative-based case finding algorithm as Bernatsky et al.7 who identified the combined prevalence of DM/PM in Quebec, Canada at 21.5 per 100,000 individuals in 2003, we found a somewhat higher prevalence of DM in Alberta, Canada; Smoyer-Tomic et al.13 also applied a similar algorithm to that used by Bernatsky et al.7 and estimated the prevalence of DM to be lower in the USA at 9.2 cases per 100,000 in 2008. Variations in the case definition (including/not including codes for PM or the specialist type included the algorithm [rheumatologists only, or also including dermatologists and internists]), the timeframe over which the case definition was applied (Bernatsky et al. used a 14 year period from 1989 to 2003, we used a 17 year period from 2002 to 2019, and Smoyer-Tomic et al. used a 4 year period from 2004 to 2008), and the administrative databases used (Bernatsky et al. and our study utilized databases from publicly funded health care systems in Canada that cover all residents of a province, and Smoyer-Tomic et al. utilized a commercial database for those covered under private insurance plans in the USA, along with Medicaid and Medicare data from federal and state programs that cover those ≥ 65 years, some < 65 with certain conditions, and some individuals with limited income and resources) may have contributed to these varying results7,13. Geographical variation of the disease may also exist. Dobloug et al.11 found that although the overall average combined prevalence of DM/PM was 8.7 cases per 100,000 individuals in 2012 in southeast Norway, the counties included in the study varied from 6.4 to 16.4 per 100,000. Discrepancies in access to specialist care may also play a role. Bernatsky et al.7 found that a greater proportion of individuals living in rural areas met the case definition for DM/PM from hospitalization data compared with urban areas, and Dobloug et al.11 proposed that the varying density of rheumatologists across southeast Norway may have been a contributing factor to the wide variation in prevalence observed across the counties7,11. Collectively, it appears that a number of factors may be contributing to the wide variation in reported prevalence of DM.
In the current study, 27% of adults living with DM did not receive any disease-related medication during the first year after the incident date. This is of clinical importance as delayed initiation of appropriate treatment following onset of DM symptoms has often been reported to be an indicator of poor outcome19, although not in all cases20. Considering that we did not have clinical evaluative information in this study to confirm diagnosis date, it is possible that these individuals had not yet received a confirmed diagnosis of DM and/or were receiving treatment for a different condition. Diagnostic delay of idiopathic inflammatory myopathies has been reported, with a median delay of 15.5 months for diagnosis of DM21,22; a number of factors have been shown to contribute to this delay including complexity of clinical features, misdiagnoses, and incorrect treatment22. As with all retrospective administrative data based research, it is also possible that the individuals who did not receive a DM-related medication in this study were misidentified having DM and included in the cohort. With that said, an established case definition was used to identify adults living with DM (adapted from7,23) that reported a high sensitivity (88.4%, 95% CI: 75.5, 94.9) and specificity (96.4%, 95% CI: 94.9, 97.5).
While determining optimal pharmacotherapy for DM is challenging given the small number and varying outcome measures of randomized controlled trials, guidelines recommend glucocorticoids as first-line therapy15,16,17,18. In this study, glucocorticoids were the most commonly used drug class, and similar findings have been reported for those living with DM/PM24. Considering that large numbers of individuals may develop complications associated with long-term use of glucocorticoids, additional immunomodulatory agents are often utilized1,25,26. Studies indicate that concomitant treatment with steroid-sparing drugs reduces the glucocorticoid dose required for remission induction, lessens the relapse risk during glucocorticoid tapering, and decreases the adverse effects of glucocorticoids3,27. Concomitant drug use was common in this study, and four of the six most used concurrent drug class combinations included glucocorticoids with various immunomodulatory agents.
An important strength of this study is the large population-based design. However, this study is also subject to several limitations. While an established case definition was used to identify adults living with DM (adapted from7,23), retrospective administrative claims-based studies use data as opposed to medical records, and therefore there is a potential for misclassification of the study cohorts or outcomes. The different clinical and pathological phenotypes DM cannot be comprehensively identified using administrative data, and were therefore not reported. The Pharmaceutical Information Network (PIN) and Laboratory Information System (LIS) databases only provide information on prescription medication dispensations, and therefore may not represent actual medication uptake by individuals. Additionally, it is not known whether DM-related medications were being taken specifically for DM or for other conditions. Use of over-the-counter supplements and/or medications, prescription medications provided in a hospital or secondary care setting, and other non-pharmacotherapy management were not captured within the administrative data, and therefore not reported.
Findings from this study provide a contemporary description of the epidemiology and medication treatment patterns of adults living with DM in Alberta, Canada. Considering that Alberta may have one of the highest rates of DM among adults, further research on the burden of disease is warranted for planning within the health care system.