Association between patient adherence and treat-to-target in gout: A cross-sectional study

The implementation of a treat-to-target (T2T) approach has been widely recommended for achieving optimal outcomes in gout treatment, as substantiated by a wealth of compelling evidence. However, a paucity of knowledge exists regarding the barriers hindering effective T2T management in China. This study seeks to investigate the factors contributing to treatment failure within the context of the T2T strategy. A cross-sectional, multi-center investigation was conducted, involving the completion of electronic questionnaires by outpatients undergoing urate-lowering treatment for a duration exceeding 6 months. These questionnaires encompassed demographic information, disease-related conditions, comorbid conditions, and management. The study analyzed factors associated with serum uric acid levels exceeding 360 µmol/L, poor disease control, and poor medication adherence. A total of 425 valid questionnaires were collected, representing 90.8% of the patients. The T2T implementation rate was 26.82% (n = 114). Factors linked to serum uric acid levels surpassing 360 µmol/L included moderate medication adherence (odds ratio (OR) = 2.35; 95% confidence interval (CI) 1.17–4.77; P = .016), poor medication adherence (OR = 4.63; 95% CI 2.28–9.51; P < .001), and management by general practitioners (OR = 0.60; 95% CI 0.37–0.97; P = .036). The rate of well-controlled patients was 14.35% (n = 61). Predictors of not well controlled encompassed the presence of tophi (OR = 2.48; 95% CI 1.17–5.61; P = .023), general medication adherence (OR = 2.78; 95% CI 1.28–6.05; P = .009), poor medication adherence (OR = 6.23; 95% CI 2.68–14.77; P < .001), and poor patient’s perception of gout (OR = 4.07; 95% CI 1.41–13.91; P = .015). A poor medication adherence rate of 55.29% (n = 235) was observed, with lower rates of poor medication adherence associated with the use of febuxostat (OR = 0.35; 95% CI 0.14–0.83; P = .02), uric acid levels exceeding 360 µmol/L (OR = 3.05; 95% CI 1.84–5.12; P = .00), moderate patient education (OR = 2.28; 95% CI 1.29–4.15; P = .01), moderate diet control (OR = 1.98; 95% CI 1.17–3.41; P = .01), and poor diet control (OR = 3.73; 95% CI 1.26–12.83; P = .02). The rate of T2T implementation in China is notably low among patients undergoing urate-lowering treatment of gout beyond 6 months. Importantly, medication adherence demonstrates a significant association with T2T outcomes.


Introduction
Gout, a prevalent arthritic condition, arises from the deposition of monosodium urate crystals.In the United States, gout affects approximately 3% to 4% of adults. [1]The purpose of gout treatment is to quickly and effectively relieve and eliminate acute symptoms, prevent the recurrence of acute arthritis, and reduce serum uric acid (sUA). [2][5] For patients with gout, urate-lowering treatment (ULT) is imperative, as it mitigates the frequency of gout attacks, reduces the involvement of joints, diminishes urate deposition, and safeguards articular cartilage and renal function. [6]Notably, the American College of Rheumatology (ACR) released guidelines in 2020 advocating for the treat-to-target (T2T) strategy in gout treatment. [7]These recent guidelines provide clinicians with directives on managing gout cases within the framework of a T2T approach.Specifically, the treatment guidelines endorse striving for a sUA target of <360 μmoI/L for all gout patients and <300 μmoI/L for those afflicted with tophaceous or severe gout. [8,9]However, a prevalence study revealed that only 22.3% of patients reached the stipulated target level, with the rate dwindling to a mere 11% for those on ULT for more than 12 months. [10]he European League Against Rheumatism guidelines emphasize the importance of patient education and lifestyle modifications as integral components of optimal long-term gout management. [11][14] The under-recognition of gout and the ensuing societal burden contribute to inadequately controlled sUA levels. [15,16]Consequently, cultivating patients' comprehension of the pivotal role of adhering to ULT and sustaining the targeted sUA level over the long term is imperative. [17]While interventions led by nurses, [18] pharmacists, [19,20] and physicians [7] to enhance T2T outcomes have been documented, other factors potentially influence the T2T strategy due to divergent medical policies across nations.Despite multiple studies shedding light on the impact of allopurinol on T2T, research focusing on the clinical uric acid-lowering effect of febuxostat remains limited.In recent times, increased attention towards T2T has surfaced in European and American countries, yet its exploration in Asia remains constrained. [21]To date, Asia's contribution to this realm primarily encompasses a single-center, short-term study in China, [22] and a small-sample study in the Philippines. [23]Given these gaps, we have embarked on a comprehensive, multi-center study on T2T.Our study aims to evaluate factors-such as demographic variables, disease-related conditions, comorbidities, and management approaches-that correlate with instances of failing to adhere to the T2T strategy in China.

Study design and setting
This cross-sectional study was conducted between July 2020 and May 2021 at 7 tertiary-level hospitals (chictr.org.cn ChiCTR2000034700).The study received approval from the Institutional Medical Ethics Committee of the Fourth Clinical Medical College of Guangzhou University of Chinese Medicine.

Participants
Informed written consent was procured from all participating individuals.The study enrolled male and female outpatients aged between 18 and 80 years.These patients fulfilled the criteria for acute gouty arthritis as approved by the ACR 1977 [24] or the ACR/European League Against Rheumatism gout classification of 2015.These criteria were based on the pattern of joint and bursa involvement during symptomatic episodes, serum urate levels, synovial fluid analysis from a symptomatic episode, and imaging evidence of urate deposition in affected joints or bursae. [25]Rheumatologists recommended the participation of patients who had undergone ULT for more than 6 months, were capable of independently completing the questionnaire, and were not afflicted with secondary gout or malignant tumors.
The state of well-controlled disease was defined as having reached the treatment target and experiencing no flare-ups or use of anti-inflammatory medication for a month. [26]A total of 468 gout patients were consecutively invited to partake, with 425 valid questionnaires collected for subsequent statistical analysis after excluding 43 patients who were ineligible, 10 patients who did not meet inclusion criteria, and 33 patients who met exclusion criteria.

Data sources/measurement
To ensure representativeness, each participating physician was limited to assessing 5 consecutive patients.Electronic questionnaires were utilized for data collection.Gout patients were required to complete a standardized set of self-report questionnaires.The questionnaires were completed under the supervision of the doctor.These encompassed demographic variables like age, gender, body mass index categories (general: 18.5-23.9,overweight: 24-27.9,obesity: >28), education level, nonmanual labor, and family history.Disease-related conditions included disease duration, ULT duration, presence of tophi, medication usage, and the 8-item Morisky Medication Adherence Scale (poor medication adherence = <6, moderate medication adherence = 6-8, good medication adherence = 8). [27]Additionally, comorbid conditions encompassed hypertension, diabetes, hyperlipidemia, kidney stones, and coronary atherosclerotic heart disease.

Bias
The study adopted a sampling survey approach rather than a census, which might have introduced selection bias.Efforts were made to ameliorate recall bias by documenting each patient's case whenever feasible.

Key points
• The rate of T2T was low in gout patient.
• Medication adherence is a crucial determinant of T2T success.www.md-journal.com

Study size
The sample size was determined through nQuery Advisor software (Statistical Solutions Company, Ireland).Drawing on data from the Chinese Rheumatism Data Center 2016, where the rate of achieving target sUA levels over a 6-month period was 38.20%, assuming a rate of 0.4 with an acceptable error of 0.05, the estimated sample size was 369.

Statistical methods
Continuous variables were described using the median (from the 25th percentile to the 75th percentile), while categorical variables were presented as frequencies and percentages.The characteristics of subjects within the sUA > 360 µmol/L and sUA ≤ 360 µmol/L groups were compared employing independent t tests for normally distributed data and Mann-Whitney U tests for non-normally distributed data.Univariate analysis was conducted to identify variables associated with sUA > 360 µmol/L and well-controlled levels.For the purpose of identifying potential risk factors linked to poor compliance, a univariate logistic regression was performed individually for each variable (e.g., demographics, disease-related conditions, comorbidities, disease management).The independent variables displaying statistical significance were incorporated into the subsequent multivariate logistic regression.To circumvent the potential destabilizing impact of multicollinearity in the multivariate logistic regression model, we calculated the generalized variance inflation factor (GVIF) for each independent variable, and retained only those GVIF values [1/ (2×df)] below 1.54, signifying the absence of multicollinearity.Following this, we established a comprehensive model with the remaining variables and subsequently conducted a stepwise regression utilizing a backward selection approach to derive a simplified model with the lowest Akaike information criterion.Associations were conveyed as odds ratios (ORs) accompanied by their corresponding 95% confidence intervals (CIs).We assessed the performance of the nomogram in terms of discrimination and calibration.Specifically, we assessed its discrimination ability by estimating the area under the receiver operating characteristic curve.A bootstrapping approach with 1000 repetitions was employed to validate the model.All analyses were conducted exclusively on complete cases, without any imputation carried out.A P value of .05denoted statistical significance.The entirety of statistical analyses was executed utilizing R version 4.0.2(IDE-RStudio Company, America).

Characteristics of not well-controlled and wellcontrolled sUA
Even upon achieving the target sUA level, certain patients persisted in experiencing joint symptoms.As a result, we conducted The sample size for fitting the full model and simplified model was 425.BMI = body mass index, CI = confidence interval, OR = odds ratio.www.md-journal.coma further analysis to discern discrepancies between not wellcontrolled and well-controlled sUA levels.The particulars of patients are delineated in Table 3.The male patient ratio was notably higher in the not well-controlled group in contrast to the well-controlled group (99.2% vs 95.1%, P = .041

Characteristics of medication adherence for patients with gout
Medication adherence exhibited a robust association with the T2T approach.To delve further, subgroup analyses were conducted, comparing patient characteristics within the poor medication adherence group and the combined good + moderate medication adherence group (Table 6).Noteworthy The sample size for fitting the full model and simplified model was 425.BMI = body mass index, CI = confidence interval, OR = odds ratio.

The factors associated with poor medication adherence for patients with gout
Multivariate logistic regression models were utilized to evaluate factors associated with poor medication adherence in gout patients (Table 7).Our findings indicated that febuxostat utilization, uric acid >360 µmol/L, moderate patients' education, and moderate to poor diet control were linked with poor medication adherence in gout patients.

Discussion
In this study, the results obtained underscored a notable low rate of achieving T2T among patients undergoing ULT for more than 6 months.The enduring maintenance of the target sUA level over the long term emerged as a challenging pursuit.In alignment with established epidemiological trends, our study revealed a predilection for gout susceptibility among males, particularly those aged over 40 years. [33]he factors contributing to the inability to attain the target sUA level predominantly encompassed the economic burdens borne by patients, irregular medical visits, and the lack of guidance from general practitioners.Intriguingly, our findings unveiled that a higher educational level in some patients correlated with failure to achieve the desired sUA levels, possibly attributed to lifestyle facets like dietary habits and demanding work commitments.Furthermore, the documented significance of poor medication adherence as a gout risk factor accentuates its potential to lead to diminished T2T rates.Our investigations illuminated the pivotal role of medication adherence in the context of gout's T2T paradigm.As is often observed in chronic ailments, adherence to long-term gout treatment regimens remained suboptimal.Specifically, merely 26.82% of our participants exhibited a commendable level of adherence to ULT, aligning with analogous findings from various Western gout studies.However, the general picture of ULT adherence among gout patients remained unsatisfactory. [13]A comprehensive meta-analysis highlighted treatment adherence percentages spanning from 10% to 46%. [34]Our study, in particular, highlighted a higher prevalence of poor medication adherence within the sUA > 360 µmol/L group.Notably, the deficiency in medication adherence resonated with aspects like patients' education and diet control. [35]Equally striking was the consistent observation that older age and the presence of comorbidities such as hypertension or diabetes were linked with improved adherence. [13,34]n alignment with these findings, our study indicated inferior adherence among gout patients with concurrent hyperlipidemia.Conversely, another study illuminated that advanced age, higher body weight, the use of anti-hypertensive or colchicine medications, and the presence of conditions like dementia, diabetes, or dyslipidemia reduced the risk of non-persistence and nonadherence. [36]reater attention should be directed toward the management of chronic diseases to enhance medication adherence.Insufficient patient education can directly lead to a reduction in medication adherence. [35]Our study underscores that the absence of management from general practitioners constitutes a risk factor.Additionally, limited knowledge about gout contributes to patients' noncompliance with treatment regimens.The constraints on doctors' time often hinder the provision of thorough education on gout and ULT to their patients, consequently leading to acute gout flares and a lack of sustained adherence to ULT for effective, long-term sUA level control. [15]A minority of patients have been systematically educated on risk reduction strategies, addressing comorbidities, or personalized lifestyle recommendations pertinent

Figure 1 .
Figure 1.The nomogram illustrates the prediction model for high uric acid levels (>360 µmol/L) among patients with gout.The final model incorporates 7 covariates: gender, education, medication adherence, diet control, diabetes, economic burden, and management from general practitioners.

Figure 2 .
Figure 2. (A) The calibration plot showcases the agreement between predicted and observed probabilities, with the dashed line representing the ideal reference line.The points with error bars represent nomogram-predicted probabilities and corresponding 95% confidence intervals grouped across 4 quartile segments.(B) The ROC curve provides insight into the model's predictive accuracy for high uric acid risk.ROC = receiver operating characteristic.

Figure 3 .
Figure 3.The nomogram depicts the prediction model for the risk of not well-controlled uric acid levels among patients with gout.The final model integrates 5 covariates: gender, presence of tophi, medication adherence, patient's perception of gout, and the presence of established health records.

Table 1
Characteristics of the patients with gout between the sUA > 360 µmol/L and ≤360 µmol/L.

Table 1 .
Within the sUA > 360 µmol/L group, a larger proportion of patients (70.4%) possessed a higher level of education compared to the sUA ≤ 360 µmol/L group (59.6%).Notably, comorbidity with diabetes exhibited a statistically significant difference in the sUA ≤ 360 µmol/L group (1.6% vs 6.1%, P = .02).Significant variations were observed between the ≤360 µmol/L and sUA > 360 µmol/L groups in terms of medication adherence (good 7.4% vs 23.7%, general 30.5% vs 39.5%, poor 62.1% vs 36.8%,P<.001)andregularfollow-up(good54.3% vs 67.5%, general 26.4% vs 20.2%, Subsequently, a multivariable logistic regression was employed to formulate the risk model and select factors, adhering to both statistical significance (P < .05)andclinicalsignificance.We investigated factors linked to a sUA level exceeding 360 µmol/L among gout patients (Table2).The findings revealed that management from general practitioners, general and poor medicine adherence were significantly associated with sUA levels > 360 µmol/L (P < .05).A nomogram predicting elevated sUA levels > 360 µmol/L in gout patients is illustrated in Figure1.The nomogram had moderate discrimination with an area under the receiver operating characteristic curve of 0.716 (95% CI 0.663-0.769).To validate the model, a bootstrapping approach with 1000 repetitions was employed, yielding bias-corrected accuracy measures of a Brier score of 0.177, a calibration slope of 0.99, and a c-index of 0.726, as depicted in Figure2.

Table 2
Results of multivariate logistic regression models to assess the factors associated with uric acid > 360 µmol/L for patients with gout.

Table 3
Characteristics of the patients with gout between the not well-controlled and well-controlled.

Table 4
Results of multivariate logistic regression models to assess predictors of not well-controlled for patients with gout.

Table 5
Generalized variance inflation factors (GVIF) for candidate predictors of high uric acid > 360 µmol/L and well-controlled among patients with gout.