Epidemiology of gout in Hong Kong: a population-based study from 2006 to 2016

Objective To determine the incidence and prevalence of gout in the general population and the utilisation of urate-lowering therapy (ULT) among patients with gout in Hong Kong. Methods A total of 2,741,862 subjects who attended any outpatient clinics or accident and emergency department (with or without hospitalisation) in 2005 and did not die before 2006 were identified from the Clinical Data Analysis and Reporting System (CDARS) of the Hospital Authority in Hong Kong. All subjects were followed until the end of 2016 or death. Demographics, diagnosis of gout, serum urate levels, and ULT prescriptions were retrieved from CDARS. Gout was defined by the diagnosis codes in CDARS. The serum urate levels achieved after prescribing ULT were the means of all serum urate levels measured 6 months after prescriptions. Results were analysed by R version 3.3.3 with package ‘prevalence’ version 0.4.0. Results The crude incidence of gout increased from 113.05/100,000 person-years (PY) in 2006 to 211.62/100,000 PY in 2016. The crude prevalence of gout increased from 1.56% in 2006 to 2.92% in 2016. Only 25.55% of patients with gout were prescribed ULT in 2016. 35.8% of patients treated with ULT were able to achieve the target serum urate level of < 6 mg/dL. Conclusions Population ageing as well as other risk factors contributed to an increase in the incidence and prevalence of gout in Hong Kong. In 2016, the crude prevalence of gout in Hong Kong was comparable to that in many western countries. However, only one in four patients with gout in Hong Kong was prescribed ULT.


Introduction
Gout is the most common inflammatory arthritis caused by deposition of monosodium urate crystals in peripheral joints. Although the prevalence of gout was estimated to be 0.08% globally [1], it has increased significantly over the last decade, especially in developed countries and Oceanic populations. The estimated prevalence of gout in the United States (US), UK, and European countries is 2-3% [2,3]. Asian countries and regions except Taiwan are considered to have a lower prevalence of gout due to differences in ethnicity and lifestyle [4,5].
The American College of Rheumatology (ACR) and European League Against Rheumatism (EULAR) recommend urate-lowering therapy (ULT) in all patients with recurrent flare, tophi, chronic kidney disease, and urolithiasis. In particular, EULAR also recommends initiation of ULT close to the time of first diagnosis in patients with earlyonset gout or with a very high baseline serum urate level. The target serum urate level should be maintained below 6 mg/dL but not less than 3 mg/dL in the long-term [6,7].
A local retrospective study of 279 subjects reported that 70% of patients with gout were prescribed ULT.
However, the study results could not be generalised to the population because it was conducted in a family medicine training centre [8]. Due to the lack of latest epidemiological data on gout in Hong Kong, we conducted this population-based study to determine the incidence and prevalence of gout as well as the utilisation of ULT among patients with gout in Hong Kong.

Methods
The Hospital Authority (HA) is the only public healthcare provider in Hong Kong. More than 90% of Hong Kong citizens utilise the public healthcare service [9].
The Clinical Data Analysis and Reporting System (CDARS) is a database managed and updated daily by the HA [10]. This database captures clinical parameters of each public healthcare service user. These parameters include patient demographics, death, diagnoses, hospital admissions, procedures performed, laboratory parameters, medication prescriptions, and dispensing histories. It has been used to conduct high-quality epidemiological studies of infectious and rheumatic diseases [9,11].
We included all subjects who attended any outpatient clinics or accident and emergency department (with or without hospitalisation) in 2005. Since all the data are  We defined the incidence of gout as the number of incident cases divided by the number of person-years (PY). The prevalence of gout was defined as the number of prevalent cases divided by the number of people included at the end of each calendar year. The prescription of ULT was defined as a continuous prescription of ULT for more than 30 days. Benzbromarone, sulfinpyrazone, and peticoglycase were not included because they were not available in the HA drug formulary [12]. Therefore, we only estimated the percentages of patients prescribed allopurinol, febuxostat, or probenecid. In patients treated with ULT, only the serum urate levels measured 6 months after commencement of ULT were included for analysis. Mean serum urate levels and percentages of patients achieving the treatment target were stratified by sex and the use of ULT.
Results were analysed by R version 3.3.3 (https://www. R-project.org/) with package 'prevalence' version 0.4.0. Incidence rate, prevalence, and their 95% confidence intervals (CIs) were estimated. Age-and sex-adjusted incidence and the prevalence of gout were adjusted based on the population data released by the Census and Statistics Department, The Hong Kong Special Administrative Region Government [13]. We used population data in 2006 as the reference population in this study. Ageand sex-specific incidence and prevalence, and male-tofemale incidence rate ratio and prevalence ratio were also estimated.
This study was approved by the Institutional Review Board of the University of Hong Kong/Hospital Authority Hong Kong West Cluster (IRB reference number: UW-533).  (Fig. 1). There was also a steady increase in the incidence of gout in both sexes. The age-and sex-adjusted incidence of gout tripled from 99. 41  The age-and sex-specific incidence and prevalence of gout in 2016 in Hong Kong were summarised in Supplementary Table S2 and S3, respectively. The age-specific incidence and prevalence of gout in both sexes increased in a non-linear fashion. 6.26% of the general population aged 80 years or older had gout in 2016. The incidence and prevalence of gout in male were higher than in female across all age groups. However, both the incidence ratio and prevalence ratio decreased with increasing age.

Results
Despite the increase in incidence and prevalence of gout in Hong Kong, the utilisation of ULT remained low. In 2016, only 25.5% of patients with gout were prescribed ULT. Female patients with gout were less often prescribed ULT than male patients. The most commonly prescribed ULT was allopurinol, and its average doses are summarised in Table 1. Among patients treated with  (Fig. 3).
There was a steady decrease in the mean serum urate levels in patients with gout from 2006 to 2016 as illustrated in Fig. 4. Female patients and patients treated with ULT had lower serum urate levels. In 2016, the mean serum urate levels in patients with or without ULT prescriptions were 7.88 mg/dL and 7.13 mg/dL respectively.
The estimated proportions of patients who were able to achieve the treatment target (< 6 mg/dL) are illustrated in Fig. 5. In general, patients treated with ULT were more likely to achieve the treatment target and the percentage also increased from 19.5 to 28.1% from 2006 to 2016. Although less female patients were prescribed ULT, they were more likely to achieve the treatment target than male patients.

Discussion
This is the latest population-based study that investigates the incidence and prevalence of gout in Hong Kong.  [14]. As population ageing continues, we expected the crude incidence and prevalence of gout will further increase.
Our study also showed a non-linear increase in agespecific incidence and prevalence of gout. This phenomenon was also shown in a population-based study in Taiwan [4]. The incidence of gout was highest at the age ≥ 80 because this age group had the highest number of incident cases but the lowest PY.
The latest incidence and prevalence of gout in Hong Kong were similar to those in the western countries, including the UK [2,15], New Zealand [16], Denmark [17], Australia [18], and the US according to the US National Health Nutrition and Examination Survey in 2015-2016 [18]. However, the prevalence of gout did not increase in the US.
An increase in both the age-and sex-adjusted incidence and prevalence suggested that other risk factors, including alcohol consumption, obesity, and diabetes, might contribute to this observation. The Population  [14,19]. Results of this survey also showed that there was an increase in regular alcohol consumption. The proportion of population with regular alcohol consumption increased from 9.5% in 2003/2004 to 11.1% in 2013/2014, while the prevalence of self-reported diabetes increased from 3.8 to 5.5%. Although CDARS does not capture body mass index and alcohol consumption, results of this survey may explain the increase in the incidence and prevalence of gout beyond population ageing.
The management of gout remained suboptimal in Hong Kong because the utilisation of ULT was insufficient [4]. The prescription of febuxostat was less than 1% because it was not subsidised by the government. Although allopurinol was fully subsidised, only 24.5% of patients with gout received this medication [12].
The efficacy of ULT in reducing serum urate levels has been demonstrated in many randomised controlled trials. In addition, the use of xanthine oxidase inhibitors in patients with gout is associated with cardiovascular and renal benefits. Epidemiological studies suggested that allopurinol might decrease mortality in patients with congestive heart failure and the risk of myocardial infarction [20]. However, Fig. 5 a Percentages of patients with gout who can achieve the treatment target (< 6 mg/dL) from 2006 to 2016. b Percentages of patients with gout who can achieve the treatment target (< 6 mg/dL) stratified by sex allopurinol is associated with an increased risk of severe cutaneous adverse reaction in patients carrying the HLA-B*5801 allele. According to a population-based study conducted in Taiwan, among Han Chinese who carry the HLA-B*5801 allele are 580 times more likely to develop allopurinol-induced severe cutaneous adverse reaction than those who do not carry the allele [21]. Although febuxostat is more effective than allopurinol, it is associated with increased cardiovascular and all-cause mortality compared to allopurinol [22]. It is worth noting that the results have been criticised due to high drop-out and many other methodological issues. Nevertheless, many studies have confirmed that hyperuricaemia is associated with increased mortality, and therefore, ULT should be considered seriously in patients with gout.

Limitation
This study is not without limitations. We used the diagnosis codes in CDARS; therefore, the diagnostic accuracy depends on physician's coding. Compliance to ULT cannot be captured in CDARS. Therefore, the effect of compliance to ULT on serum urate levels among patients with gout cannot be assessed in this study. In addition, not all patients with gout had a serum urate level measurement. Therefore, the mean serum urate levels cannot represent all the patients with gout in this study. Although we covered more than 90% of the general population in Hong Kong, CDARS does not include patients receiving medical care in the private sector. This social bias could not be remedied as socioeconomic data are not included in the database.

Conclusion
Population ageing contributes to an increase in the incidence and prevalence of gout in Hong Kong. The crude prevalence of gout in Hong Kong was 2.92% in 2016, which is comparable to that reported in other western countries. Despite an increase in the prevalence of gout, the utilisation of ULT remained low. Only one in four patients with gout was prescribed urate-lowering agents.
Additional file 2: Supplementary Table S1. Age-and sex-adjusted incidence and prevalence of gout in Hong Kong from 2006 to 2016. Supplementary Table S2. Age-and sex-specific incidence of gout in Hong Kong in 2016. Supplementary Table S3. Age-and sex-specific prevalence of gout in Hong Kong in 2016.