Abstract
Objective To examine patterns of prednisone prescribing for rheumatoid arthritis (RA) management in primary care settings; and to explore experiences and perspectives of family physicians (FPs) and people living with RA regarding prednisone use.
Design Mixed-methods study with sequential exploratory design.
Setting Canada, particularly Alberta.
Participants Overall, 546 patients with RA in primary care electronic medical records in southern Alberta after excluding short-term prednisone prescriptions, 33 people living with RA from across Canada, and 16 primary care providers (14 family physicians, 1 clinic manager, 1 pharmacist).
Methods Qualitative interviews and focus groups guided the analysis of primary care electronic medical records linked with provincial pharmaceutical dispensing data. Patterns of prednisone dispensing over 12 years, factors associated with prednisone use before and after RA documentation, and qualitative experiences with prednisone therapy were examined.
Main findings Each year between 2008 and 2019, at least 40% of patients with established RA received at least 1 prednisone prescription. Of patients receiving both prednisone and disease-modifying antirheumatic drugs (DMARDs) (41%, n=210), a subset (n=92) received prednisone before starting DMARDs, with 66% continuing prednisone for more than a year after DMARD initiation. Median time between first prednisone prescription and DMARD initiation was 124 days (interquartile range=13 to 1150 days). Three main qualitative themes were identified: prednisone is used as bridging therapy during the wait time to see a specialist, patients and providers have concerns about long-term use and side effects, and systemic barriers affecting access to guideline-concordant care exist and particularly impact male patients.
Conclusion Current prednisone prescribing patterns in RA management reveal divergence from guidelines recommending short-term use, suggesting systemic barriers to guideline-concordant care. Improving outcomes requires addressing both clinical needs driving prednisone use and systemic barriers perpetuating reliance on this medication.
Rheumatoid arthritis (RA) affects about 1% of Canadians and is characterized by joint inflammation, resulting in pain, stiffness, and progressive joint damage if untreated.1,2 Early medication initiation is crucial to prevent permanent damage.3-5 In Canada, family physicians (FPs) are initial points of contact and care coordinators for people living with RA (PLRA) in need of rheumatology services.6 However, FPs face important challenges in RA management, including timely diagnosis, specialist referral, and initiating appropriate treatment,6-9 often exacerbated by long wait times for rheumatologist consultations.7-12
Glucocorticoids are used by FPs to manage RA symptoms, especially while patients await specialist care.8,13,14 Canadian Rheumatology Association (CRA)15 and European16 guidelines endorse short-term use of glucocorticoids combined with disease-modifying antirheumatic drugs (DMARDs) for active RA, but advise against continuous use past 3 months. The guidelines recommend tapering and discontinuing, when possible, especially after achieving remission or low-disease activity. CRA guidelines suggest short-term glucocorticoids (≤6 weeks for flares) with the lowest effective dose when initiating or changing DMARD therapy.17 They advise against routine long-term (>6 months) low-dose glucocorticoids as DMARD adjuncts, emphasizing persistent disease activity should trigger DMARD adjustment rather than continued glucocorticoid use.
Prednisone, a glucocorticoid with anti-inflammatory and immunosuppressive properties, is widely used for inflammatory, autoimmune, and allergic conditions, including RA.15,18 Our study aimed to provide comprehensive understanding of glucocorticoid use, specifically prednisone, in RA management in Canadian primary care. Primary objectives included the following: estimating the prevalence of prednisone use among PLRA; describing current dispensation trends; exploring patients’ and primary care providers’ experiences with prednisone treatment; and identifying improvements in prednisone prescribing and management for patients in primary care.
METHODS
We used a sequential exploratory mixed-methods design, wherein qualitative research guided subsequent quantitative analysis of linked administrative health and primary care electronic medical record (EMR) data.19 Research ethics approval for the study was obtained from the University of Alberta in Edmonton, the University of Calgary in Alberta, the University of Winnipeg in Manitoba, Laval University in Quebec, and Memorial University of Newfoundland in St John’s. Our interdisciplinary team included FPs, a rheumatologist, physiotherapists, epidemiologists, and a medical anthropologist.
Qualitative interviews
We conducted semistructured interviews with 33 participants with RA recruited through Arthritis Research Canada’s Web-based volunteer noticeboard (n=21) and family medicine clinics in Alberta, Manitoba, Quebec, and Newfoundland and Labrador (n=12). Eligible participants 19 years or older self-identified as diagnosed with RA by a physician in Canada. Recruitment continued until thematic saturation was achieved.20 The sample reflected key characteristics of the Canadian RA population, though western Canada and those with higher education are overrepresented (Table 1).
Characteristics of people living with rheumatoid arthritis: N=33.
We conducted 3 focus groups with 16 primary care providers in Alberta (14 family physicians, 1 clinic manager, 1 pharmacist), recruited through the family medicine departments of the University of Alberta and the University of Calgary.
Interviews of PLRA were conducted between December 2021 and November 2023 via Zoom videoconferencing software, each lasting 1 to 2 hours. The interview guide, developed through iterative consultation with the study’s patient advisory committee and informed by the Candidacy Framework,21 explored access dimensions and treatment experiences. Hour-long provider focus groups were conducted in-person, where barriers to RA diagnosis and care were discussed. NVivo 14 was employed for analysis22 using a hybrid approach of deductive coding guided by the Candidacy Framework and inductive coding for emerging themes.21,23
Linked electronic medical record and administrative data
Using the Canadian Primary Care Sentinel Surveillance Network (CPCSSN) database case definition (sensitivity 82%, specificity 98%), we identified 597 primary care patients with RA in southern Alberta (aged ≥19 years) who had at least 1 encounter with an FP between 2008 and 2020.24 This case definition is based solely on primary care EMR data and does not require rheumatologist consultation, allowing for examination of primary care management patterns, including patients who may never have accessed specialist care. CPCSSN data from FP clinics participating in the Southern Alberta Primary Care Research Network were linked to Alberta Health Services Pharmaceutical Information Network (PIN) data, including Anatomical Therapeutic Chemical codes, Drug Identification Numbers, and number of tablets and number of days for each prescription to access dispensing information for prednisone and prescriptions ordered by ambulatory care physicians (FPs, rheumatologists, other specialists). However, PIN data do not identify the specific prescriber type for individual prescriptions, limiting our ability to differentiate between FP-initiated versus specialist-initiated treatments. Individuals with chronic obstructive pulmonary disease (COPD) or asthma with prednisone dispensations of 14 days or less were excluded, as these were most likely prescribed for non–RA-related conditions. The final cohort comprised 546 patients.
We conducted descriptive and inferential statistical analyses using Stata 14.25 Dominant themes regarding medication use and influencing factors in the qualitative data informed the selection of variables for analysis in our linked data.
RESULTS
The mean (standard deviation [SD]) age of the EMR sample in 2020 was 55 (16) years; females comprised 72% of the sample and 3.6% lived in rural areas. Mean (SD) age for meeting RA case criteria was 61 (16) years, with median follow-up of 7 years (interquartile range [IQR]=3 to 9 years).
Prednisone initiation and continuation
Prednisone was discussed by 20 PLRA and in 2 provider focus groups. Similarly, in our quantitative findings prednisone accounted for 90% of oral glucocorticoid prescriptions and dispensations for patients with RA, making it the dominant treatment approach.
Each year between 2008 and 2019, around 40% or more of patients with RA received at least 1 prednisone prescription (Figure 1).
Annual prevalence and mean (95% confidence interval) of prednisone dispensed to patients with rheumatoid arthritis for each year: Linked dataset from 2008 to 2019.
Prednisone as a bridge to DMARDs
Interviewees reported receiving prednisone to manage RA flares while awaiting rheumatology consultation, which was prescribed either by their FP or during emergency department visits (Table 2, A,B,C). Likewise, in our quantitative sample, 41% of patients (n=210) received both prednisone and a DMARD at some point (Table 3). Within this group, a subset of 92 patients received prednisone before starting a DMARD, with median time between first prednisone prescription and DMARD initiation of 124 days (IQR=13 to 1150 days), suggesting variable and sometimes extended waiting periods (Table 3).
Qualitative themes and illustrative examples of prednisone use in rheumatoid arthritis management
Characteristics of patients dispensed prednisone and disease-modifying antirheumatic drugs: N=546.
Among 210 patients who received both prednisone and DMARDs, 16% (n=34) discontinued prednisone at DMARD initiation, 18% (n=37) continued but stopped within a year, and 66% (n=139) continued prednisone for more than 1 year after DMARD initiation. For those who eventually discontinued prednisone after starting DMARDs (n=183), the median number of prednisone dispensations was 3 (IQR=2 to 11), with median time from DMARD initiation to final prednisone prescription of 1633 days (IQR=430 to 3094 days), suggesting intermittent use over extended periods rather than continuous therapy.
Patient and provider experiences with prednisone
FPs expressed caution about prescribing prednisone, especially for patients with comorbidities. They emphasized the need for comprehensive education on risks and benefits, preferably from prescribing rheumatologists. PLRA shared providers’ concerns about prednisone’s side effects, including hair loss, mood changes, and impacts on fertility. Despite these concerns, some interviewees readily accepted prednisone given symptom severity, recognizing it as an effective, though imperfect, solution to their immediate pain and functional limitations. Some interviewees continued prednisone even against medical advice, driven by symptom severity (Table 2, D), while RA symptom intensity led some to obtain prednisone through alternative means (Table 2, E). Duration and intensity of prednisone use by patients could be substantial (Table 2, F,G).
Systemic issues and access to care
FPs were the first point of contact for most of the PLRA interviewed (29 of 33), playing a critical role in early management and specialist referral, but described being constrained by system-level barriers. Both PLRA and FPs reported variable wait times for rheumatologist appointments, ranging from less than a month to more than 2 years. Extended waiting periods challenged effective symptom management (Table 2, H). FPs described systemic barriers to specialist access, including capacity constraints and inefficient triaging processes (Table 2, I).
Some FPs had access to Specialist Link, a service in Calgary, Alta, for direct consultation with rheumatologists for advice on DMARD prescribing (Table 2, J), but others in the same city thought it was unavailable for RA. Academic FPs noted they probably had more time than fee-for-service colleagues to contact specialists directly. Correspondingly, treatment patterns in EMR data varied by primary care setting, with DMARD recipients (alone or with prednisone) more commonly treated in academic clinics (56% to 63%) than those receiving only prednisone (38%) or no treatment (47%).
A total of 48 RA patients received only prednisone for more than a year after diagnosis, with no DMARDs. Compared to other groups, this group included relatively more males and those with respiratory comorbidities (Table 3).
DISCUSSION
Our study comprehensively examined prednisone use in primary care RA management, revealing complex interplays between clinical practice, patient experience, and systemic challenges. Discrepancies between practice and current CRA guidelines are evident.
Annual prednisone use by 40% or more of established RA patients exceeds expectations for guideline-recommended short-term therapy (≤12 weeks).17 Even higher rates appear internationally: 54.3% in the COMOrbidities in Rheumatoid Arthritis study across 17 countries26 and 65.3% in the Early Undifferentiated Arthritis cohort over 10 years, the latter showing increased cardiovascular and infection risks with cumulative glucocorticoid exposure.27
Multiple studies demonstrated that glucocorticoid use, even at low doses, is associated with increased risk of mortality and cardiovascular events.27-31 Substantial delays between first prednisone prescription and DMARD initiation may reflect challenges in service provision,8,12 aligning with findings that primary care physicians rarely initiate DMARDs for new patients with RA,8 as well as limited rheumatology workforce capacity.32
Our qualitative findings revealed diverse patient experiences with prednisone, from refusal to perceived dependency. Previous research documents physical and emotional struggles due to prednisone, including Cushing syndrome, weight gain,13 and severe complications like rapid cataract development.2 Despite these concerns, many patients resisted discontinuation,2 self-managed dosages,33 or obtained prednisone without prescriptions.34 FPs acknowledged tension between immediate symptom relief and long-term health consequences, which is consistent with the existing literature.35
Our finding that 66% of patients who received both prednisone and DMARDs continued prednisone for more than a year after DMARD initiation is concerning. This contrasts with a study in the United States where only 29.2% of newly diagnosed RA patients continued glucocorticoids 10 to 12 months following a diagnosis.36 CRA guidelines17 recommend short-term use (8 to 12 weeks) with tapering protocols during DMARD initiation and advise against routine long-term use, instead recommending DMARD adjustment for persistent disease activity. This high continuation rate suggests system-level barriers to optimal care. The pattern of few dispensations (median=3) over extended periods suggests intermittent rather than continuous use.
Overrepresentation of male patients in the subset who received only prednisone for extended periods suggests potential disparities in treatment patterns. Males with RA may exhibit greater reluctance to seek comprehensive care due to stoicism and challenges to masculine identity from chronic illness,37 though these patterns intersect with other social determinants in complex ways that warrant further investigation.38
Moreover, overrepresentation of respiratory comorbidities in prednisone-only patients reflects the common interrelationship between RA and pulmonary manifestations.39 Clinical decisions are complicated when respiratory conditions, including interstitial lung disease as a direct consequence of RA, may independently warrant prolonged glucocorticoid therapy.40 Long-term prednisone use might be appropriate for these patients despite guidelines advising against extended use for articular symptoms alone. Standard protocols for COPD and asthma also incorporate prednisone.41 Current RA guidelines inadequately address implementation of these overlapping recommendations, creating challenges for FPs and other primary care providers balancing risks against benefits for both articular and extra-articular manifestations.
Patients in academic clinics showed higher DMARD use and lower rates of prednisone monotherapy, suggesting academic FPs may have better specialist access and communication pathways,42 more resources and time for complex care coordination,43 or different patient populations with varying health care access patterns.44 Many FPs expressed hesitation regarding long-term prednisone or DMARD prescription due to misdiagnosis concerns, adverse effects, and potential symptom masking that might complicate rheumatologist assessment, which is consistent with research indicating that long-term corticosteroid prescribing remains substantial despite risks of masking poor disease control and significant adverse effects.45
Implications for practice and policy
Our findings that some FPs were unaware of available consultation services (Specialist Link) and the 124-day median delay between prednisone initiation and DMARD therapy highlight the need to strengthen existing communication tools and develop systematic collaboration approaches accounting for practice constraints, which would be consistent with CRA guidelines.42,46
Given patient reports of obtaining prednisone through alternative means and continuing use against medical advice, patient education regarding prednisone is essential to align with patient-centred care principles2 and CRA and European guidelines on shared decision-making.15-17
Substantial treatment delays (median=124 days) underscore the need to reduce barriers to access rheumatologist care in a timely manner to minimize prolonged prednisone bridging therapy, supporting CRA guidelines for short-term, low-dose use during DMARD initiation or adjustment.17
The 66% continuation rate beyond guideline-recommended timeframes suggests that enhancing dissemination of CRA glucocorticoid guidance in primary care is necessary. Despite existing protocols,17 FPs encounter implementation barriers. Targeted knowledge translation strategies and practical tools, integrated within clinical pathways, could standardize prescribing while accommodating primary care realities.
Limitations
A key limitation of this study was attributing prednisone use specifically to RA management. Despite excluding short-term prescriptions (≤14 days), misattribution may persist among patients with multiple indications. Our data do not capture clinical decision-making nuances, medication dosages, adherence patterns, or differentiate between continuous versus intermittent use. It also lacks prescriber identification, constraining cross-setting prescribing pattern analysis. Elevated prednisone dispensation rates during 2008 to 2010 may be a sampling or data quality artifact. Patients receiving neither DMARD nor prednisone (25%) may represent those with well-controlled disease in remission, patients receiving treatments not captured in our dispensing database (such as hospital-administered biologics), those with contraindications to these medications, or potentially undertreated patients. Recruiting providers primarily from academic settings may limit transferability of qualitative findings, though many participating FPs practice in both academic and community settings with similar approaches to patient care. Ultimately, we cannot determine whether prolonged prednisone use reflected patient preference, clinical necessity for extra-articular manifestations, or systemic constraints.
Conclusion
Disparities and guideline implementation challenges persist in prednisone dispensation for RA management. Despite its value as bridging therapy, our data reveal prolonged use beyond initial symptom management, particularly among male patients. FPs face difficulties balancing immediate symptom relief with early DMARD initiation amid extended specialist wait times. Future research should evaluate interventions reducing these disparities, including supportive care models for primary providers managing RA during specialist referral periods. Research should also aim to develop guidelines recognizing the complexity of managing patients with conflicting indications for glucocorticoid therapy and examine communication patterns between primary care providers and specialists in these cases. Improving outcomes necessitates addressing both immediate clinical needs driving prednisone use and systemic barriers perpetuating reliance on this medication. This requires expanded rheumatology resources alongside innovative care delivery approaches to serve diverse populations and reduce access inequities.
Footnotes
Acknowledgment
The authors thank those with rheumatoid arthritis who shared their experiences. The authors acknowledge the use of data from Alberta Health Services, the Canadian Primary Care Sentinel Surveillance Network, and the Southern Alberta Primary Care Research Network. The results and conclusions are those of the authors and no official endorsement of the data providers is intended. This work was supported by the Arthritis Society of Canada (Strategic Operating Grant 20-0000000018) and a Mitacs Accelerate Postdoctoral Fellowship (Dr Anh N.Q. Pham).
Contributors
All authors contributed to conceptualizing and designing the study; to collecting, analyzing, and interpreting the data; and to preparing the manuscript for submission.
Competing interests
None declared
This article has been peer reviewed.
- Copyright © 2026 the College of Family Physicians of Canada
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.