Abstract
Objective This feasibility study aimed to assess the acceptability of using smartphone notifications to modify the medication beliefs of people with gout. We evaluated the feasibility and acceptability of a smartphone application using the Technology Acceptance Model. We explored adherence rate differences and outcomes between the intervention and control groups.
Methods Fifty-two patients with gout who were prescribed allopurinol were randomly assigned to either active control (n = 24) or intervention group (n = 28). Over 3 months, both groups used the study app on their smartphones. The active control group received notifications about general health advice, whereas the intervention group received adherence-targeted notifications. The feasibility and acceptability of the smartphone app was measured through semistructured interviews. Adherence rate was assessed through serum urate levels and missed doses at 3 timepoints: baseline, 3 months (post intervention), and 6 months (follow-up).
Results The smartphone app demonstrated high feasibility, with strong participant retention and compliance. The participants expressed high levels of satisfaction with the app’s user-friendliness and content, highlighting its acceptability. Both groups showed a significant reduction in missed doses over time (P < 0.05), but no significant differences in serum urate levels were found between the groups. Patients who received adherence-targeted notifications reported finding it more convenient to take allopurinol and expressed higher overall treatment satisfaction throughout the study.
Conclusion Adherence-targeted notifications have the potential to be an effective and scalable approach to supporting medication adherence in patients with gout. Further research is needed with larger samples to refine the components of the intervention and explore its optimal implementation.
Intentional nonadherence, which involves the conscious decision of patients to not follow their prescribed medication regimen, poses a critical challenge in the management of chronic conditions.1 Intentional nonadherence extends across the 3 stages of adherence: initiation and implementation (period of persistence), followed by discontinuation (period of nonpersistence).2,3 Understanding the complex nature of intentional nonadherence is crucial for developing effective interventions. In this regard, Weinman et al introduced the Intentional Non-Adherence Scale (INAS), a valuable tool that provides a comprehensive framework for measuring intentional nonadherence and gaining insights into patients’ decision-making processes.4
Gout management is particularly affected by intentional nonadherence to urate-lowering therapy (ULT), leading to significant limitations in the therapy’s effectiveness.5 Intentional nonadherence not only compromises patient health and quality of life but also underscores the need to address and modify patient beliefs and perceptions regarding medication.6 Emad et al conducted a study using the INAS and identified 4 subscales as key factors contributing to intentional nonadherence in ULT for gout.7 These factors include the following: (1) the desire to avoid reminders of the condition and resist the illness to maintain a sense of normalcy; (2) the inclination to test the limits of treatment effectiveness in terms of using as little medication as possible; (3) concerns regarding medication, such as potential side effects and the risk of dependency; and (4) the perceived sensitivity to medication affecting the body’s natural healing processes. These findings shed light on the multifaceted nature of intentional nonadherence and its effect on gout management,7 particularly in the implementation phase, where individuals’ perceptions of medication necessity and concerns play an important role in influencing adherence.3
In recent years, smartphone-based interventions have emerged as a promising strategy to tackle medication nonadherence, with previous studies demonstrating their potential in promoting behavior change and enhancing treatment outcomes.8-10 However, there is a dearth of research specifically focusing on intentional nonadherence among people with gout and the effectiveness of smartphone interventions in addressing this issue. Although some studies have explored the use of smartphone reminders to enhance adherence, few have investigated the potential of targeted messages that directly address patients’ beliefs and perceptions regarding their medications.11
Therefore, the main objective of this study was to assess the clinical and technical feasibility of using smartphone notifications as an intervention to modify patients’ beliefs and perceptions toward their medications in gout management. More specifically, we examined the feasibility of collecting outcome measures such as adherence rates using serum urate (SU) levels, and questionnaire responses from people with gout. We also evaluated the acceptability of this smartphone-based intervention to the people with gout using the Technology Acceptance Model, a widely adopted theoretical framework that analyzes perceptions of the intervention’s usefulness and ease of use, which influences behavioral intention and technology adoption.12 Further, we sought to explore any potential differences in adherence rates and other outcome measures between the intervention group, which received the targeted smartphone notifications that addressed determinants of intentional nonadherence, and the active control group, which received notifications about general health advice, at 3 different timepoints (baseline, post intervention [3 months], and 6-month follow-up). In addition, we conducted a longitudinal assessment to evaluate how these notifications influenced the outcome measures within each study group over time.
METHODS
Participants. This was a randomized controlled feasibility study of 52 patients with gout. The participants were recruited from rheumatology clinics between December 2021 and April 2022. Inclusion criteria were patients who (1) were aged ≥ 18 years; (2) had a rheumatologist-confirmed diagnosis of gout; (3) were prescribed allopurinol; (4) owned a smartphone; and (5) were English-speaking. The New Zealand Health and Disability Ethics Committee approved this study (ref. AH23037) and all patients provided written informed consent.
Sample size. The sample size was in line with recommendations for determining the sample size for a randomized feasibility study.13 For the current trial to have the 80% power to detect a 10% change from baseline with 2-sided α of 0.05 and a small standardized effect size (0.1), a feasibility trial with a sample size of 22 patients per treatment arm (1:1) was recommended. We estimated the dropout rate to be 15% (according to a recent study14), and therefore included 28 participants in each arm. Four of the participants who were allocated to the control group withdrew from the study; hence, we finished analyzing data for 52 participants (24 and 28 in the control group and the intervention group, respectively).
Randomization and allocation. We used a simple randomization technique by creating a computer-generated list to allocate participants into the intervention or control arm. This was performed by an independent research assistant not involved in recruitment, assessment, or delivery of the intervention and who had no prior knowledge of the participants.
Intervention procedure. Participants were randomly allocated to either the intervention group or the active control group. The intervention group received smartphone notifications that targeted intentional nonadherence determinants highlighted in previous studies. For example, Emad et al indicated that nonadherent patients often associate discontinuation of allopurinol with the desire to resume a “normal” life.7 In order to address this issue,7 the notification “by taking allopurinol on a regular basis, you can keep doing what matters to you” was included as part of the smartphone intervention in the current study. This targeted message aimed to encourage adherence by highlighting the connection between regular medication use and the ability to maintain a meaningful and fulfilling lifestyle. The active control group received smartphone notifications about general health advice, which included the importance of regular exercise, the effect of good sleep on physical and mental performance, stress management techniques, and the benefits of staying hydrated. All participants received 130 1-way notifications, which means that in the 3-month study period, participants were not expected or asked to respond to notifications. Instead of a constant frequency, participants received 2 notifications per day for the first 8 weeks, followed by 1 notification per day for the next 2 weeks, 3 notifications per week in week 11, and 1 notification per week in week 12. To receive the notifications, both groups were required to download a smartphone application and received brief training on how to read, save, or delete notifications (the text bank is available from the authors on request).
Outcome measures were assessed at baseline, post intervention (3 months from baseline), and then again at 6 months from the baseline, to investigate the effects of notification intervention over time. To obtain a more in-depth understanding of the potential barriers and facilitators at the service delivery and individual level for the uptake of this program, participants from both groups attended a semistructured interview either in person or online (Figure 1).
Recruitment and participation flowchart. INAS: Intentional Non-Adherence Scale.
Data collection. After obtaining written informed consent, participants actively used the study app to complete demographic and study-related questionnaires. SU results were obtained through a review of patients’ electronic medical records (EMRs).
Outcome measures.
• Feasibility of the smartphone application. To evaluate technical feasibility, the study collected any technical problems associated with the healthcare innovation. Patients reported some of these issues directly, whereas others were identified by the researcher during the project. In addition, questionnaire responses were collected through the smartphone app, and any technical issues encountered during the submission of answers were identified and addressed.
Clinical feasibility was evaluated through issues encountered during the assessment of outcome measures, including adherence rates using SU levels. Adherence rates were determined by evaluating SU levels retrieved from participants’ EMRs, thus obviating the necessity for frequent blood tests. This information allowed us to identify potential challenges or limitations in implementing the innovation in a clinical setting.
• Acceptability of the smartphone app. To investigate patient satisfaction with the app, all participants were asked to rate their overall experience with the app, including willingness to continue the app, ease of use, understandability, usefulness, and tone, language, timing, and frequency of notifications, on a scale of 0 to 10, with 0 being the least favorable and 10 being the most favorable score.
Analysis of study groups.
• Adherence. Rate of adherence to allopurinol was assessed by measuring SU levels and the patient-reported number of missed doses at baseline, post intervention (3 months from baseline), and again at 6 months from baseline, focusing on the implementation phase of adherence.
• Psychological factors. Patients completed the Treatment Satisfaction Questionnaire for Medication,15 a 14-item scale that uses a rating scale from 1 to 7, and the Intentional Non-Adherence Scale (INAS),4 which consists of 22 items scored on a scale from 0 to 5. These questionnaires were completed at baseline, post intervention (3 months from baseline), and again at 6 months from baseline.
Comparative assessment of study groups. To evaluate the effects of smartphone notifications, we conducted a comparative analysis of outcome measures between the interventional group and the control group at baseline, post intervention (3 months from baseline), and again at 6 months from baseline.
Longitudinal assessment of study groups. To evaluate the effects of the smartphone notifications over time, we conducted a longitudinal analysis of outcome measures for each study group at baseline, post intervention (3 months from baseline), and again at 6 months from baseline. These assessments were designed to identify any potential improvements in medication adherence within the intervention group that might be attributed to the adherence-targeted notifications, as well as within the control group, where health advice could potentially serve as a reminder.
Data analysis. The statistical analysis was conducted using the SPSS version 25.0 (IBM Corp.). Descriptive statistics including medians with ranges and numbers with percentages were used to summarize the clinical characteristics of the study participants. As the outcome measures were not normally distributed, nonparametric tests were employed. Mann-Whitney U tests were conducted to examine differences in adherence and psychological factors between the experimental and control groups. In addition, nonparametric Friedman test was conducted for longitudinal analysis of any differences in outcome measures over time, followed by a posthoc analysis using Wilcoxon signed-rank test, with a Bonferroni correction for multiple comparisons.
In this study, we primarily employed nonparametric statistical methods for data analysis to account for the skewed distribution of several variables. However, for the variable “missed doses within a month,” which exhibited a consistent median value of 0 across all timepoints, we opted to report the mean along with the 95% CI. This decision was made because of the unique distribution of this variable, where most observations had zero values, making it unsuitable for traditional nonparametric tests. The use of the mean and 95% CI allows for better representation of central tendency and provides a more informative summary of this specific variable.
In addition, because of the repeated measures nature of the “missed dose” variable collected at multiple timepoints, we applied a general linear model with a repeated measures design and t tests to assess changes over time and between groups. These statistical approaches were chosen to account for within-subject dependencies and to evaluate differences in missed doses between groups while considering the temporal aspect of the data.
All statistical tests were 2-tailed, and data outliers were excluded from the analysis. A significance level of 0.05 was used to determine statistical significance for all analyses.
RESULTS
Characteristics of the study population. A total of 52 patients with gout, with an average age of 63.6 years were included in the study. Most patients were male (98%), New Zealand European (71%), and married (58%) with university education (71%). On average, the participants had been taking allopurinol for 8 years (Table 1). Among all, 24 of the patients were randomly allocated to the control group to receive some general health advice, and the rest (n = 28) were allocated to the intervention group to receive the adherence-targeted notifications.
Characteristics of the study population.
Feasibility of the smartphone app. Regarding technical feasibility, the study encountered a total of 5 technical issues. Two participants from the control group experienced unstable internet access and were unable to install the app, leading to their withdrawal from the study before it commenced. The research team promptly resolved the issues faced by 2 other participants, enabling them to install the app successfully. Additionally, 1 participant reported difficulties submitting their questionnaire responses through the app, but this was also resolved by the research team.
In terms of clinical feasibility, the present study used participants’ EMRs to collect SU levels, obviating the need for regular blood tests. However, a substantial proportion of baseline data (23%) dated back more than 6 months, and a considerable number of participants had no recorded SU levels at the end of the program (48%) and at the 3-month follow-up (57%).
Acceptability of the smartphone app. Overall, participants showed high engagement and commitment to the research study, with a 0% attrition rate once enrolled. Nearly half of the patients (49%) expressed willingness to continue using the app after completion of the study, and a majority (80%) said they would recommend it to others. Participants rated their experiences with the app on a scale of 0 to 10, where 0 represented the worst and 10 represented the best. The results indicated that the app received high ratings for ease of use, with all participants giving it a score of 10 out of 10. Content understandability received a high rating, with a median score of 10 out of 10, indicating that nearly all participants (98%) found the content of the notifications easy to comprehend. The perceived usefulness of the app was moderately positive, as indicated by a median rating of 6 out of 10 (IQR 5-7). Regarding the language and tone of the notifications, whereas 3 participants (6%) preferred more scientific terms, the majority found the language to be simple and relatable, and all participants found the tone to be appropriate and nonoffensive. In terms of notification preferences, some participants (25%) expressed a desire for flexibility in timing, whereas the majority (86%) preferred less frequent notifications, such as once or twice every 2 weeks.
Comparison of smartphone notification effects on nonadherence between study groups. To look at the potential effects of smartphone notifications on nonadherence rates, we compared the SU levels and number of missed doses (within a month) between the control group and the intervention group at the 3 timepoints (baseline, post intervention, and 6 months’ follow-up). There were no significant differences between the study groups in terms of SU levels and the number of missed doses at each timepoint, suggesting that smartphone notifications did not have a significant effect on nonadherence rate (Table 2).
Differences between the intervention group and the control group in terms of SU levels and number of missed doses at baseline, post interventional (3 months), and 6 months’ follow-up.
Comparison of smartphone notification effects on psychological factors. Next, we investigated the potential effects of smartphone notifications on treatment satisfaction and treatment satisfaction subscales, as well as INAS scores and INAS subscales. We compared these psychological factors between the control and intervention groups at the 3 timepoints (baseline, postintervention, and 6 months’ follow-up). There were no significant differences between the study groups regarding these variables at any of the timepoints, indicating that the smartphone notifications did not have a significant effect on the measured psychological factors (Table 3).
Differences between the intervention group (n = 28) and the control group (n = 24) in terms of psychological factors contributing to nonadherence at baseline, post intervention (3 months), and 6 months’ follow-up.
Effects of adherence-targeted notifications in the intervention group. The third aim of the study was to investigate the potential effects of adherence-targeted notifications on nonadherence and other psychological factors over time. A nonparametric Friedman test was employed to examine differences among repeated measures. When applicable, a posthoc analysis was conducted using the Wilcoxon signed-rank test, with a Bonferroni correction applied to account for multiple comparisons. The results indicated that there were no significant effects of the intervention on the level of SU over time (χ22 = 5.29, P = 0.07). This tentatively suggests that the adherence-targeted notifications did not have a significant effect on nonadherence as measured by SU levels over the course of the study.
However, we found a significant effect of the intervention on the number of missed doses within a month (F2,52 = 4.22, P = 0.02, ηp2 = 0.25). Subsequent posthoc tests using the Bonferroni correction indicated that this was the result of participants being more likely to take allopurinol as prescribed after receiving the smartphone notifications compared to the baseline (P = 0.01). There was also a significant difference between post intervention and 6-month follow-up (P = 0.04). No difference was found between baseline and 6-month follow-up (P = 0.06; Figure 2).
Average number of missed dose(s) within a month for study groups at different timepoints. *** P < 0.05.
In terms of other psychological factors, we found no significant effects of the intervention on INAS scores (χ22 = 1.14, P = 0.56), as well as on INAS subscales (Table 4). However, there was a significant effect of the adherence-targeted smartphone notifications on the extent to which participants considered treatment administration more convenient (χ22 = 14.25, P < 0.001, W = 0.25) and were satisfied with taking allopurinol (χ22 = 6.80, P = 0.03, W = 0.12). This was the result of participants being more likely to be satisfied with how allopurinol works 3 months after receiving the intervention compared to baseline (P < 0.001). There was no significant difference between baseline and post intervention (P = 0.07) as well as post intervention and 6-month follow-up (P = 0.42).
Results of the Friedman test for differences in SU level, number of missed doses, and other psychological factors across time among the study groups.
Effects of general health advice notifications in the active control group. To explore the potential effect of general health advice notifications on nonadherence and psychological factors over time, we conducted a nonparametric Friedman test to compare control group data at 3 distinct timepoints: baseline, post intervention, and 6-month follow-up, followed by a posthoc analysis using the Wilcoxon signed-rank test, with a Bonferroni correction for multiple comparisons. In terms of nonadherence rate, there were no significant effects of general health advice on the level of SU over time (χ22 = 3.67, P = 0.15; Table 4). However, there was a significant effect of the general health advice on the number of doses missed within a month (F2,52 = 3.28, P = 0.05, ηp2 = 0.20). Subsequent posthoc tests using the Bonferroni correction indicated that this was the result of participants being more likely to take allopurinol as prescribed after receiving the smartphone notifications compared to the baseline (P = 0.04). There was no significant difference between baseline and 6-month follow-up (P > 0.99) as well as between post intervention and 6-month follow-up (P = 0.28; Figure 2).
With respect to other psychological factors, we found no significant effects of receiving general health advice on neither treatment satisfaction scores (χ22 = 0.02, P = 0.98) and the relevant subscales, nor INAS scores (χ22 = 1.36, P = 0.50) and the relevant subscales (Table 4).
DISCUSSION
To our knowledge, this is the first study to explore the feasibility and acceptability of using smartphone notifications as an intervention to modify patients’ beliefs and perceptions toward their medications in gout management. Regarding technical feasibility, the study encountered some minor issues related to installation of the app. However, these technical challenges were quickly resolved, highlighting the importance of providing good technical support during the implementation of smartphone-based interventions. Ensuring stable internet connectivity and user-friendly app interfaces are also crucial factors in enhancing the overall feasibility of smartphone-based interventions.
Despite the successful resolution of technical challenges, the reliance on EMRs for collecting SU levels has limitations. A notable proportion of baseline data was outdated, and a significant number of participants lacked recorded SU levels at various timepoints, particularly at the 6-month follow-up. This resulted in missing data and variability in the timing of measurements. To enhance the feasibility of collecting SU levels in future studies, alternative methods such as real-time remote monitoring or the use of wearable devices could be explored. These approaches may offer more accurate and timely data collection, mitigating the risk of missing information and providing a more comprehensive assessment of SU levels throughout the study duration.
The intervention was well received by the participants, with high levels of actual use, positive attitudes toward using the app, and a perception of ease of use. The absence of any dropouts among the participants is noteworthy and suggests that the intervention was not burdensome or inconvenient. Moreover, nearly half of the participants indicated their willingness to continue using the app after the study ended, indicating a high level of user engagement. The majority of the participants said they would recommend the app to other patients on allopurinol therapy, further underscoring the positive attitudes toward the intervention. The feedback regarding language and tone of the notifications was generally positive, with most participants finding the language simple and relatable, and the tone appropriate. However, a small number of participants preferred more scientific language, indicating that the intervention could benefit from customization options to cater for individual preferences.
The present study also allowed us to investigate differences in adherence rates and other outcome measures over time between the intervention group, which received the targeted smartphone notifications addressing determinants of intentional nonadherence, and the active control group, which received smartphone notifications about general health advice. Our findings showed positive effects of adherence-targeted notifications on medication adherence and treatment satisfaction with allopurinol in people with gout. Notably, our results demonstrated a considerable decrease in the number of missed doses among patients who received these tailored notifications.
The effectiveness of adherence-targeted notifications in addressing the underlying reasons for intentional nonadherence is consistent with previous research.7,16,17 Our study expands upon this knowledge by demonstrating a significant positive effect on patients’ beliefs and perceptions toward their medication through tailored notifications. By effectively acknowledging and addressing factors associated with intentional nonadherence, such as resistance to illness, maintenance of normalcy, testing treatment limits, concerns about medication, and sensitivity to medication effects, our intervention seemed to provide a useful perspective for patients on ULT. Reframing medication as a means to restore normal health and life might have motivated patients to adhere to their prescribed medication.
In addition, our study revealed that patients who received adherence-targeted notifications felt that taking allopurinol was more convenient and expressed greater overall treatment satisfaction. These findings align with previous research that has examined the effects of tailored interventions on treatment satisfaction. Spragg et al found that reassuring information highlighting necessity and long-term effectiveness of allopurinol can increase treatment satisfaction and improve adherence in people with gout.3 Therefore, educational strategies that address the key reasons for intentional nonadherence have the potential to improve treatment satisfaction and medication adherence with allopurinol by engaging patients in gout management.11 These strategies may also help patients understand the potential challenges associated with using symptom-based approaches to test the efficacy of ULT, leading to enhanced treatment satisfaction.
Looking at the findings, receiving general health advice can have a positive effect on medication adherence among people with gout by reducing the number of missed doses. This is consistent with previous research indicating that medication reminders can help prevent missed doses.14,18,19 Although our study did not involve sending specific medication reminders, patients were informed of general health advice through both email and the app. It is possible that simply receiving a notification can serve as a reminder and help decrease the frequency of skipped doses. These results can convey the importance of unintentional nonadherence, which is the failure to take medication due to factors such as forgetfulness, as a common reason for nonadherence.20 It is worth noting that a considerable proportion of the control group had SU levels at baseline that were well controlled and below the therapeutic target for ULT. This baseline characteristic may have influenced the findings related to medication adherence in the control group, as individuals with well-controlled gout may have different perceptions and behaviors toward their medication compared to those with uncontrolled gout. This discrepancy in baseline characteristics is an important consideration when interpreting the results and may indicate the need for further investigation into the specific factors influencing adherence in such populations.
The study had several limitations, including the small sample size and the short duration of the study, which may not have been sufficient to capture the full effects of the intervention. Further, given that the study sample exclusively comprised patients who had been taking allopurinol for an average of 8 years, it is possible that the findings may not fully reflect the experiences of individuals who have been newly prescribed allopurinol or who differ from the study population in other ways. In addition, many of the sample who enrolled in the study did not report any missed doses, indicating that there was no room for improvement in adherence for those individuals.
Moreover, the observed high baseline adherence in our study population may not be representative of the broader population of patients with gout, potentially limiting the generalizability of our findings. This could be attributed to unrepresentative sample characteristics, as our participants had a history of active participation in previous gout-related projects, resulting in heightened awareness of medication adherence. Although we required a rheumatologist-confirmed diagnosis of gout as the primary eligibility criterion, we did not specifically consider the severity or activity level of gout in our participants at the outset of the study as an inclusion criterion.
Although we collected adherence data through patient-reported missed doses, it is important to acknowledge that this method can be prone to recall bias and may not provide a comprehensive view of medication adherence. Future work in this area would greatly benefit from more objective methods of measuring adherence, such as pill count and electronic pill bottle monitoring.
Further, our study underscored the need for more reliable methods of measuring SU levels, as the collection of data from patient EMRs proved to be inconsistent. SU levels were not systematically assessed at predefined intervals in this study, which constrained our ability to comprehensively evaluate the intervention’s effect on SU levels and treatment goals. Although the initial findings appear encouraging, it is important to acknowledge the significant amount of missing data in the SU analysis. This data gap may have limited the statistical power of our study to detect meaningful differences, necessitating a cautious interpretation of the results.
Another limitation to consider is the relatively high level of education observed among our study participants. Hence, the generalizability of our findings to individuals with lower educational attainment may be limited. Thus, it may be necessary to replicate the study with a more diverse and nonadherent patient sample to assess the generalizability of the results. In addition, smartphone interventions rely on access to technology and mobile data, which could pose challenges for individuals with limited financial resources or those residing in rural areas with restricted connectivity. These disparities raise concerns about the equitable distribution and accessibility of such interventions. Future research and implementation strategies should take these equity issues into account, ensuring that the benefits of smartphone-based interventions are accessible to a broader population. Alternative approaches or support systems may be necessary to assist individuals facing technology-related barriers.
In conclusion, the findings of this study suggest that the smartphone-based intervention holds great potential as a patient-centered approach to enhance adherence to allopurinol among people with gout. Despite its limitations, this study provides preliminary evidence for the feasibility and potential effectiveness of such interventions. The high levels of patient acceptability and the minor technical issues encountered indicate that the intervention is feasible and easily implementable in clinical settings. Moreover, the incorporation of educational strategies and tailored notifications addressing key motivators for nonadherence showed positive outcomes. These interventions may be particularly valuable for patients with limited access to traditional healthcare services, given their cost effectiveness and accessibility.
ACKNOWLEDGMENT
We thank Hamish Franklin and Tayla Eveleigh from Artilect Limited for their support in developing the smartphone application used in this study. We also express our gratitude to the participants who contributed their time and insights.
Footnotes
This research was funded by The University of Auckland. The sources of funding for this study played no role in the study’s design, conduct, or reporting.
ND reports grants and personal fees from AstraZeneca, grants from Amgen, and personal fees from Dyve BioSciences, Hengrui, Selecta, Arthrosi, Horizon, AbbVie, Janssen, PK Med, and JW Pharmaceuticals, outside the submitted work. The remaining authors declare no conflicts of interest relevant to this article.
- Accepted for publication October 30, 2023.
- Copyright © 2024 by the Journal of Rheumatology