Abstract
Objective To examine the effect of biological sex on wait times to first rheumatology appointment in a central triage system before and during the coronavirus disease 2019 (COVID-19) pandemic.
Methods Deidentified data of all referred patients between November 2019 and June 2022 were extracted from the electronic medical record. Variables, including time from referral to first appointment, biological sex, referral period, urgency status, age, and geographic location were collected and analyzed.
Results Twelve thousand eight hundred seventeen referrals were identified. Wait times increased by 24.23 days in the peri-COVID period (P < 0.001). In the pre-COVID period, there was no significant difference in wait times by biological sex or age. Triage urgency was a predictor of wait time, with semiurgent referrals seen 8.94 days (95% CI −15.90 to −1.99) sooner than routine referrals and urgent referrals seen 25.42 days (95% CI −50.36 to −0.47) sooner than routine referrals. In the peri-COVID period, there was a significant difference in wait time by biological sex with women waiting on average 10.03 days (95% CI 6.98-13.09) longer than men (P < 0.001). Older patients had shorter wait times than younger patients, with a difference of −4.64 days for every 10-year increase in age (95% CI −5.49 to −3.78). Triage urgency continued to be a predictor of wait time.
Conclusion Women and younger patients appear to have been affected by wait time increases during the COVID-19 pandemic. This finding should be further investigated to determine its pervasiveness across other specialities and to better understand the underlying cause of this finding.
Arthritis and other rheumatic diseases are common reasons for presentation to primary care. Nearly 60 million Americans, or 1 in 4, have a form of arthritis, with similar data in Canada, where over 6 million have arthritis.1,2 With fewer than 400 full-time clinical equivalent rheumatologists in Canada, or fewer than 1 rheumatologist per 100,000 population,3 timely access to rheumatology care is challenging. Some centers in Canada report wait times upwards of 1 year from time of referral,4,5 further compounded by delays in patient presentation and referral. However, timely referral is paramount for optimal treatment, as evidence demonstrates early initiation of disease-modifying antirheumatic drugs for inflammatory arthritis leads to better disease control and prevents disease progression.6
To that end, central triage systems have been designed to prioritize incoming rheumatology referrals that would most benefit from early assessment. The central triage system at the University of Alberta in Edmonton, Canada, has been previously reported as a robust system to correctly assign referral urgency in a simple, cost-effective manner.7
The coronavirus disease 2019 (COVID-19) pandemic has had a significant effect on the healthcare system. In rheumatology, much of the focus has been on the effects of immunosuppression on COVID-19 outcomes and treatments in established patients with autoimmune inflammatory rheumatic diseases.8,9 What remains less clear is the effect the pandemic has had on initial access to care for rheumatology services and if this effect has affected men and women equally. Here, we conduct a retrospective cohort study examining the effect of biological sex on wait time to first rheumatology appointment in a group practice that employs a central triage system.
METHODS
As previously described,7 the Division of Rheumatology at the University of Alberta established a central triage system in 2009. Twelve rheumatologists participated during the study period. In brief, each week, 1 rheumatologist reviews and triages all incoming patient referrals. Triage is based on the referral letter and information provided, although a rheumatologist may choose to look up investigation results to supplement their triage decision. There is no standardized evaluation or scoring system used. The triage rheumatologist will assign a preassessment diagnosis and urgency status.
Urgency status is assigned based on the preassessment diagnosis and need for early access to care. Urgency is categorized as urgent, semiurgent, or routine, with ideal target wait times of < 7 days, 6 weeks, and no target, respectively. Urgent consults are often those with organ-threatening disease and are facilitated through direct physician-to-physician communication. Semiurgent consults are those with a suspicion for an inflammatory process, such as inflammatory arthritis and inflammatory connective tissue diseases. Routine consults are those for noninflammatory conditions, such as osteoarthritis or soft tissue ailments. Once urgency is assigned, a central triage administrative clerk contacts the patients consecutively for an appointment with the rheumatologist in the group practice who best meets the urgency target window.
Study design. In November 2019, the central triage system moved from eClinician, an Epic-based outpatient electronic medical record (Epic Systems), to ConnectCare, an updated EPIC-based comprehensive universal electronic medical record. ConnectCare allows for specific reporting of patient referral wait times and patient characteristics. We extracted deidentified data of all referred patients between November 2019 and June 2022. Patient-level variables collected included time from referral to first appointment, biological sex, referral period (pre-COVID-19 [before March 2020] or peri-COVID-19 [from March 1, 2020]), urgency status, age, and geographic location, identified using postal code data as urban (within the Greater Edmonton Metropolitan area) or rural (outside the Greater Edmonton Metropolitan area).
Statistical analysis. Baseline characteristics were compared between men and women using 1-way ANOVA for continuous variables, and chi-square tests of independence for categorical variables. Crude analysis using univariable linear regression was performed to examine the differences in wait times by biological sex, age, region, and referral period. Within-group and pairwise comparisons for categorical exposures (triage level) were done using 1-way ANOVA with Bonferroni corrected P values. Multivariable linear regression was performed adjusting for biological sex, age, region, and triage level, stratified by referral period. Effect modification was tested using interaction terms for sex*triage level, age*triage level, region*triage level, and sex*region during the peri-COVID-19 period. Stratified analysis is presented for significant interaction terms. Statistical analysis was completed with Stata/BE 17.0 (StataCorp).
The University of Alberta Health Ethics Board approved this project (Pro00124125).
RESULTS
Twelve thousand eight hundred seventeen referrals were identified with a referral and had their first appointment scheduled between November 2019 and June 2022. One thousand seven hundred thirty-six referrals (13.54%) were received during the pre-COVID-19 period and the remainder were received during the peri-COVID-19 period. Most referrals were women (women: n = 8548 vs men: n = 4269). Women were older than men (mean 51.7 yrs vs 54.0 yrs, P < 0.001) and were more likely to be assigned a lower triage level (P < 0.001; Table 1). There was no difference in geographic location or referral period between men and women. Approximately 43% of patients who were referred lived in the Greater Edmonton Metropolitan area.
Rheumatology central triage referral characteristics.
Wait times increased by 24.23 days on average in the peri-COVID-19 period (pre-COVID-19: 64.65 days vs peri-COVID-19: 88.88 days, P < 0.001). The unadjusted difference in wait times between men and women was 11.73 days (men: 77.72 days vs women: 89.45 days, P < 0.001; Table 2). Other variables associated with increased wait time included younger age (−4.40 days per 10-year increase in age) and lower triage level (routine: 97.74 days, semiurgent: 73.46 days, and urgent: 16.57 days, P < 0.001). Geographic region was not significantly associated with wait time. In the pre-COVID-19 period, once adjusted for age, region, and triage level, there was no significant difference in wait time by biological sex (difference of 2.30 days, 95% CI −5.05 to 9.64) or age (0.02 days per 10-year increase in age, 95% CI −2.06 to 2.01). Triage level was a significant predictor of wait time, with semiurgent referrals seen 8.94 days (95% CI −15.90 to −1.99) sooner than routine referrals and urgent referrals seen 25.42 days (95% CI −50.36 to −0.47) sooner than routine referrals (Table 3A).
Crude predictors of wait time to first appointment.
Difference in wait time to first appointment stratified by referral period (adjusted linear regression model).
In the peri-COVID-19 period, once adjusted for age, region, and triage level, there was a significant difference in wait time by biological sex with women waiting on average 10.03 days (95% CI 6.98-13.09) longer than men (P < 0.001; Table 3A). Older patients had significantly shorter wait times than younger patients, with a difference of 4.64 days for every 10-year increase in age (95% CI −5.49 to −3.78). Triage level continued to be a significant predictor of wait time with semiurgent referrals seen 26.30 days (95% CI −29.22 to −23.37) sooner than routine referrals and urgent referrals seen 86.48 days (95% CI −97.01 to −75.94) sooner than routine referrals. Geographic location had no association with wait time in both periods.
The interaction terms sex*triage level and age*triage level were significant in the adjusted linear regression model (P = 0.002 and P = 0.001, respectively), but the interaction terms, sex*region and region*triage level, were not significant (P > 0.05). As there was effect modification seen with triage level for the association between sex and age and wait times during the peri-COVID-19 period, stratified analyses are presented for these predictors in Table 3B. Females waited on average 14.15 days (95% CI 9.19-19.10) longer than males for routine referrals, and 6.32 days (95% CI 2.90-9.75) longer for semiurgent referrals. Patients waited on average 5.90 and 3.48 days fewer for every 10-year increase in age for routine and semiurgent referrals, respectively. There were no significant differences in wait times by sex or age for urgent referrals (P > 0.05).
Difference in wait time to first appointment stratified by urgency status (adjusted linear regression model).
DISCUSSION
This study demonstrates that wait times for new rheumatology referrals have increased significantly for women compared with those for men during the COVID-19 pandemic. Although increased wait times as a result of the COVID-19 pandemic have previously been reported,10-12 the inequality of this adverse effect on women more than men has not.
In the central triage system, patients are offered appointments consecutively based on their urgency status alone, regardless of their personal characteristics. Therefore, from a central triage perspective, age and biological sex should not affect time from referral to first appointment. Further, this difference in wait time by biological sex was not seen before the COVID-19 pandemic. Although this study does not examine the cause of this discrepancy, multiple studies suggest the COVID-19 pandemic has affected women and men differently. Specifically, women appear to have been more affected by loss of employment or fewer hours of employment compared to men, and women parents appear to have taken on more of the burden of childcare during the pandemic, as lockdowns led to children completing school online from home, requiring home supervision.13,14 These differences may affect the ability for women to accept more timely appointments.
Similarly, once adjusted for covariates, older age was not associated with shorter wait time before the COVID-19 pandemic. However, during the COVID-19 pandemic, patients waited on average 4.6 days fewer per 10-year increase in age. Although there is no explanatory literature around this finding, younger patients with children may have also been disproportionately affected by childcare issues during the pandemic, preventing them from accepting a timely appointment.
It was reassuring to see geographic location did not affect wait times, particularly as previous studies have demonstrated higher rates of inflammatory arthritis in rural areas of Alberta compared to metropolitan centers.15 However, it is possible a more granular examination of data beyond the simple urban and rural divide used here may have demonstrated different results.
One limitation of this study is that the pre-COVID-19 time window is relatively short, owing to the change in the electronic medical record at the center in November 2019. Although this may affect our power to detect differences between groups in the pre-COVID-19 period, there were no other changes in our triage process. In this study, we dichotomized the period of referral into pre-COVID-19 and peri-COVID-19. However, we recognize the peri-COVID-19 period is more complex, with multiple waves of increased COVID-19 infections and varying degrees of public health restrictions requiring a detailed review of the infection rates and health restrictions over the triage period, which was outside the scope of this study. Thus, the effects shown for the peri-COVID period may not be consistent through this period and may benefit from further study in the future.
Women and younger patients appear to have been especially affected by wait time increases during the COVID-19 pandemic. This finding should be further investigated not only in other rheumatology centers but in the healthcare system in general, as this inequality may be pervasive across specialities. Most importantly, studies need to be done to understand root causes of this disparity in biological sex and age so that interventions can be implemented to repair it. Further, as the COVID-19 pandemic evolves, it will be important to continue to monitor inequalities in access to care.
Footnotes
The authors declare no conflicts of interest relevant to this article.
- Accepted for publication March 2, 2023.
- Copyright © 2023 by the Journal of Rheumatology
This is an Open Access article, which permits use, distribution, and reproduction, without modification, provided the original article is correctly cited and is not used for commercial purposes.
