Background: This study analyzed visits for and factors associated with gout and gout medication treatment trends for the years 2007–2011 in the United States given the introduction of febuxostat, the first new treatment option for gout in over 40 years, which was introduced to the market in 2009. Methods: This study was a retrospective, cross-sectional, observational study of patients age 20 and older seen by providers who participated in the National Ambulatory Medical Care Survey (NAMCS), the National Hospital Ambulatory Medical Care Survey Outpatient Department (NHAMCS-OPD) or Emergency Department (NHAMCS-ED) in the United States. The outcome of interest was visits for gout diagnosis and visits where a gout medication was prescribed. Results: Approximately 1.2% of visits had a diagnosis of gout. There was a significant increase in the percentage of visits with a diagnosis of gout in years 2009–2011 compared to 2007–2008, which remained after adjusting for covariates of interest. Groups more likely to have a visit with gout included those ≥65 and 45–64 (both as compared to those 20–44), the African-American and ‘Other’ race groups (as compared to Caucasians) and those on a diuretic. Groups less likely to have a visit with gout included females, Hispanic/Latinos, those with insurance type of ‘Other’ and Medicaid (both as compared to private insurance) and visits to a hospital emergency setting (as compared to physician’s office visits). Conclusion: Although there was a significant increase in visits where gout is diagnosed across study years, the overall percentage of visits with a gout diagnosis is low in the US population. Treatment trends over the study years has remained consistent, with the introduction of febuxostat appearing to have little impact for the study years through 2011. Keywords: Gout, Febuxostat, NAMCS, NHAMCS-OPD, NHAMCS-ED Background increasing prevalence overall across all demographics Gout is a type of inflammatory arthritis associated with . Prior to 2009, pharmacological treatment options for the formation of urate crystals in the joints. It is esti- gout had not changed in many years. A new treatment mated that approximately 4% of Americans are affected, option, febuxostat, was approved by the Food and Drug but previous research has shown the prevalence of gout Administration in February 2009 . is increasing . Contributing factors to gout include in- No studies have evaluated the proportion of visits with crease in obesity, hypertension, and purine-rich diets . a gout diagnosis since 2009 when febuxostat was intro- The severity and progression of gout has been directly duced to the market. Previous studies have shown that correlated to an increase in age.  Gout is known to be an above normal BMI, hypertension, dyslipidemia, use of more predominant in males, but is seen in postmeno- a diuretic and older age are associated with an increased pausal women. The prevalence of gout is higher in Afri- risk of gout [5–10]. can Americans as compared to Caucasians, with With the first new gout therapy in over 40 years intro- duced to the market in 2009, this study sought to evalu- ate changes in gout-related ambulatory and emergency * Correspondence: email@example.com department visits for the years 2007 through 2011 (the Campbell University College of Pharmacy & Health Sciences, 180 Main Street PO Box 1090, Buies Creek, NC 27506, USA © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Castro et al. BMC Rheumatology (2018) 2:14 Page 2 of 9 most recent data available) as well as assess factors asso- yielding unbiased national annual estimates of visit oc- ciated with gout . Additionally, gout treatment currences, percentages, and characteristics . trends were plotted to determine the impact of febuxo- Because of the complex sample design, sampling er- stat on gout therapy since its introduction to the market. rors were determined using the SAS SURVEYFREQ and SURVEYLOGISTIC procedures which take into account Methods the clustered nature of the sample . The appropriate This retrospective, cross-sectional, observational study NOMCAR and DOMAIN statements/options were im- analyzed data collected in the National Ambulatory plemented in these procedures as recommended by the Medical Care Survey (NAMCS) and the National NCHS . Hospital Ambulatory Medical Care Survey Outpatient The dependent variable of interest was a diagnosis of Department & Emergency Department (NHAMCS- gout, where the denominator is the number of visits OPD/NHAMCS-ED) during the years 2007–2011. Hun- meeting the inclusion/exclusion criteria. A diagnosis of dreds of reports, manuscripts and books based on data gout was defined by the appropriate diagnosis codes from these widely utilized and respected surveys have found in any of the DIAG1-DIAG3 diagnosis fields or been published since the 1970s (https://www.cdc.gov/ appropriate gout medication codes for allopurinol, nchs/data/ahcd/namcs_nhamcs_publication_list.pdf). febuxostat, colchicine, probenecid, and colchicine- The NHAMCS is an annual, national probability sam- probenecid found in any of the DRUGID1-DRUGID8 ple of ambulatory visits made to non-federal, general, medication fields. and short-stay hospitals in the US conducted by the Rao-Scott chi-square tests were used to analyze Centers for Disease Control and Prevention, National whether the proportion of visits with a diagnosis of gout Center for Health Statistics (NCHS). The multi-staged differs by year group (2007–2008 vs. 2009–2011) and sampling design is composed of four stages and includes whether any association exists between visits with a visits to both selected emergency care departments as diagnosis of gout and each of the following variables: well as hospital outpatient departments [12–14]. The age, sex, race, ethnicity, region (US geographic regions NAMCS is an annual, national probability sample of included Northeast, Midwest, South, and West), metro- visits made to the offices of non-federally employed phy- politan statistical area (MSA), insurance status (private, sicians classified by the American Medical Association Medicaid or Children’s Health Insurance Program or the American Osteopathic Association as “office- [CHIP]/State Children’s Health Insurance Program based, patient care” (excluding anesthesiologists, pathol- [SCHIP], Medicare, and other [worker’s compensation, ogists and radiologists) [12–14]. For more information self-pay, no charge/charity, other], setting type (phys- regarding the survey instruments, scope and sample de- ician office, hospital outpatient department, hospital sign, data collection and processing, estimation proce- emergency department) and diuretic use. These variables dures and reliability of survey estimates, go to http:// were grouped for analysis as shown in Table 1. Odds ra- www.cdc.gov/nchs/ahcd/ahcd_questionnaires.htm. tios (ORs), corresponding 95% confidence intervals (CIs) NAMCS, NHAMCS-OPD and NHAMCS-ED datasets and p-values were reported. covering five years (2007–2011) were included in this A multivariable logistic regression model was also study. Patients 20 or older from any of the three data- constructed in order to evaluate the predictive value bases were coded as included in the final analysis data- of all the independent variables of interest simultan- set. There were no exclusions for the study. Across all eously on visits with a diagnosis of gout, adjusting for five years a total of 128,734 raw records in the covariates of interest. As a primary model filter, only NHAMCS-OPD, 126,836 raw records in the NHAMCS- variables with an overall chi-square test of association ED and 126,651 in the NAMCS databases met the inclu- p-value < 0.2 were included in the multivariable sion criteria (382,221 combined). The study was submit- model (year group was included regardless). ORs with ted to the Campbell University Institutional Review corresponding 95% CIs and p-values for each level of Board and received an exemption due to the data each variable included in the model (in comparison sources used being publicly available and de-identified. to each variable’s reference group) were reported. No As such, since this research was based solely on the ana- collinearity issues between the independent variables lysis of previously collected, de-identified data, it com- included in the model were found. All analyses were plies with the Helsinki Declaration. generated using SAS software, version 9.3. Plots of The survey data were analyzed using the sampled visit the percentage of visits per year with gout medication weight that is the product of the corresponding sampling by drug class and individual drug were constructed to fractions at each stage in the sample design. The sam- descriptively assess gout treatment trends. pling weights have been adjusted by NCHS for survey Per NCHS recommendations, any variable with a sur- nonresponse as appropriate within each database, vey estimate based on either less than 30 records, a Castro et al. BMC Rheumatology (2018) 2:14 Page 3 of 9 Table 1 Demographics and Patient Characteristics (N = 382,221) relative standard error of more than 30%, or more than Characteristic No. (%) of Patient Visits 30% missing data was excluded from the analyses due to potential unreliability . As a result, the variables of Observation Period interest weight status, tobacco use, depression, hyperten- 2009–2011 588,791,594 (61) sion, and diabetes were excluded from all analyses due 2007–2008 375,839,802 (39) to not meeting one or more of the above listed criteria. Age (years) Missing values were treated as missing in the statistical Mean (SE) 53.9 (0.23) evaluation. No adjustments for multiple comparisons Age Group were made and p-values < 0.05 were considered statisti- ≥65 299,032,423 (31) cally significant. 45–64 351,209,142 (36) 20–44 314,389,831 (33) Results Sex During the study period (2007–2011), the NAMCS, Female 587,394,025 (61) NHAMCS-OPD, and NHAMCS-ED datasets include a Male 377,237,371 (39) total of 495,370 patient visits (unweighted, raw data). A Race total of 382,221 patient visits within this five-year period Other 33,516,436 (5) met the inclusion criteria and were included in this African-American 99,604,070 (14) study (Table 1). Most variables had no missing data, however, ethnicity, race and insurance status were miss- Caucasian 602,374,076 (82) ing 21, 16 and 5%, respectively. Gout was diagnosed in Ethnicity just over 1.2% of all patient visits. More patient visits oc- Hispanic/Latino 79,262,491 (11) curred during the more recent half of the study period Non-Hispanic/Latino 630,345,157 (89) (61% in the years 2009–2010). The mean age (SE) was Region 53.9 (0.23) years, with a similar percentage of patients in Northeast 189,690,837 (20) each of the three age groups. Of the patient visits Midwest 206,775,296 (21) included in the analyses, 61% were female, 82% were West 201,670,759 (21) Caucasian, 14% were African-American, and 11% were Hispanic/Latino. Nearly twice as many visits occurred in South 366,494,504 (38) the South (38%) as compared to the other regions (21% MSA in the Midwest, 21% in the West and 20% in the North- Non-MSA 120,846,147 (13) east). However, visits in metropolitan areas represented MSA 843,785,249 (87) 87% of the study total. Private insurance was presented Insurance Status at the majority of all patient visits (50%), with Medicare Other 98,736,670 (11) presented at 30% of visits and Medicaid at 10%. The vast Medicaid 88,374,398 (10) majority of visits occurred in a physician’s office (82%), reflecting the NAMCS survey data (collected in phys- Medicare 280,902,288 (30) ician offices). Only 7% of patient visits reported diuretic Private Insurance 461,730,141 (49) use. Setting Type The primary analysis showed a 30% relative increase in Hospital Emergency 94,712,528 (10) the percentage of patients who had a visit with a diagno- Hospital Outpatient 78,137,131 (8) sis of gout in the years 2009–2011 compared to 2007– Physician’s Office 791,781,738 (82) 2008 (1.3% vs. 1.0%, respectively; OR 1.28, 95% CI 1.10– Diuretic Use 1.49) (Table 2). The individual chi-square tests of the Yes 71,243,249 (7) other covariates of interest that comprised the first part of the secondary analysis, showed a significant associ- No 893,388,147 (93) ation between visits with a diagnosis of gout and the fol- Gout lowing variables: age, sex, ethnicity, race group, Yes 11,769,697 (1) insurance status, setting type, region and diuretic use. No 952,861,700 (99) See the univariable columns in Table 2 for the details of MSA Metropolitan Statistical Area these individual associations with visits for a diagnosis of Unweighted, raw study sample size gout. Survey weighting and clusters accounted for reflecting unbiased, national annual estimates of visit occurrences for the portion of the The weighted multivariable logistic regression model, population meeting the study inclusion/exclusion criteria allowing adjustment for the effect of potentially Castro et al. BMC Rheumatology (2018) 2:14 Page 4 of 9 Table 2 Gout Predictor Variables, Univariable and Multivariable Analyses* Data are given as number (%) of patients Univariable Multivariable Parameter Gout (%) No Gout (%) OR (95% CI) P OR (95% CI) P Observation Period 2009–2011 7,849,421 (1.3) 580,942,174 (98.7) 1.28 (1.10–1.49) 0.0010 1.24 (1.02–1.52) 0.0346 2007–2008 3,920,276 (1.0) 371,919,526 (99.0) Referent – Referent – Age Group (years) ≥ 65 7,113,034 (2.4) 291,919,369 (97.6) 8.64 (7.04–10.60) < 0.0001 4.94 (3.69–6.62) < 0.0001 45–64 3,772,290 (1.1) 347,436,851 (98.9) 3.85 (3.11–4.77) < 0.0001 2.52 (1.97–3.21) < 0.0001 20–44 884,372 (0.3) 313,505,459 (99.7) Referent – Referent – Sex Female 3,055,558 (0.5) 584,338,467 (99.5) 0.22 (0.20–0.25) < 0.0001 0.20 (0.18–0.24) < 0.0001 Male 8,714,139 (2.3) 368,523,233 (97.7) Referent – Referent – Race Other 574,306 (1.7) 32,942,130 (98.3) 1.43 (1.03–1.99) 0.0332 2.02 (1.47–2.77) < 0.0001 African-American 1,226,437 (1.2) 98,377,632 (98.8) 1.02 (0.82–1.27) 0.8513 1.33 (1.03–1.72) 0.0271 Caucasian 7,264,368 (1.2) 595,109,707 (98.8) Referent – Referent – Ethnicity Hispanic/Latino 511,614 (0.6) 78,750,878 (99.4) 0.49 (0.36–0.67) < 0.0001 0.64 (0.48–0.86) 0.0032 Non-Hispanic/Latino 8,252,946 (1.3) 622,092,211 (98.7) Referent – Referent – Region Northeast 2,019,734 (1.1) 187,761,102 (98.9) 0.90 (0.72–1.13) 0.3629 0.80 (0.62–1.03) 0.0823 Midwest 3,053,963 (1.5) 203,721,333 (98.5) 1.25 (1.04–1.52) 0.0169 1.16 (0.95–1.43) 0.1501 West 2,367,739 (1.2) 199,303,020 (98.8) 0.99 (0.80–1.24) 0.9580 0.86 (0.67–1.10) 0.2296 South 4,328,261 (1.2) 362,166,243 (98.8) Referent – Referent – MSA Non-MSA 1,709,090 (1.4) 119,137,057 (98.6) 1.18 (0.93–1.52) 0.1636 1.01 (0.77–1.32) 0.9462 MSA 10,060,607 (1.2) 833,724,642 (98.8) Referent – Referent – Insurance Status Other 407,967 (0.4) 98,328,703 (99.6) 0.43 (0.32–0.58) < 0.0001 0.51 (0.36–0.72) 0.0001 Medicaid 407,549 (0.5) 87,966,848 (99.5) 0.48 (0.35–0.66) < 0.0001 0.56 (0.39–0.80) 0.0014 Medicare 6,193,910 (2.2) 274,708,378 (97.8) 2.35 (2.07–2.68) < 0.0001 0.93 (0.74–1.15) 0.4846 Private Insurance 4,382,477 (0.9) 457,347,665 (99.1) Referent – Referent – Setting Type Hospital Emergency 353,199 (0.4) 94,359,329 (99.6) 0.28 (0.23–0.33) < 0.0001 0.42 (0.34–0.51) < 0.0001 Hospital Outpatient 892,347 (1.1) 77,244,784 (98.9) 0.86 (0.71–1.03) 0.0994 1.05 (0.88–1.26) 0.5718 Physician Office 10,524,151 (1.3) 781,257,587 (98.7) Referent – Referent – Diuretic Use Yes 3,575,442 (5.0) 67,667,807 (95.0) 5.71 (5.04–6.47) < 0.0001 4.00 (3.37–4.75) < 0.0001 No 8,194,254 (0.9) 885,193,893 (99.1) Referent – Referent – MSA Metropolitan Statistical Area * Survey weighting and clusters accounted for reflecting unbiased, national annual estimates of visit occurrences for the portion of the population meeting the study inclusion/exclusion criteria Note that per the model fitting criterion described in the methods section, no variables were excluded from the multivariable model important variables, demonstrated that significant asso- race group (‘Other’ vs. Caucasian as well as now ciations remained between visits with a gout diagnosis African-American vs. Caucasian), insurance status and age group (≥65 and 45–64 vs. 20–44), sex, ethnicity, (Other and Medicaid vs. private insurance, but no longer Castro et al. BMC Rheumatology (2018) 2:14 Page 5 of 9 Medicare vs. private insurance), hospital emergency de- showing the most common gout treatments can be seen partments (vs. physician office visits) and diuretic use. in Figs. 1 and 2. The percent of gout visits in each drug However, the associations between visits with a gout class appears similar across the years studied. Antigout diagnosis and region of the country and as well as gout (80–86%) and antihyperuricemic (67–75%) medications visits and Medicare (vs. private insurance) did not re- were consistently the most common treatment while main significant following adjustment for the effect of NSAIDs (14–20%) and steroids (6–12%) were a distant the other covariates of interest in the model. A diagnosis third and fourth, respectively. of gout remained more likely for patient visits in 2009– Among individual medications, allopurinol (67–74%) 2011 as compared to visits in 2007–2008 (OR 1.24, 95% is clearly the most common, three to four times as likely CI 1.02–1.52). Additionally, a diagnosis of gout remained as the next closest (colchicine). Febuxostat, probenecid more likely for visits with patients 65 and older and pa- and colchicine-probenecid were infrequently prescribed tients 45–64 as compare to visits for those 20–44 (OR 4. (all less than 5%). Allopurinol and colchicine dominate 94, 95% CI 3.69–6.62 and OR 2.52, 95% CI 1.97–3.21, the market, their use appears consistent across the years respectively). Gout was less likely to be diagnosed at studied, while probenecid’s use has decreased over time. visits for females as compared to males (OR 0.20, 95% Uptake does appear to be slow for febuxostat, the newest CI 0.18–0.24) and Hispanic/Latino patients as com- gout medication, as two years after entering the market, pared to Non-Hispanic/Latino patients (OR 0.64, 95% the percentage of patients with gout taking the drug re- CI 0.48–0.86), while gout was more likely to be diag- mains around 3%. nosed at visits for those on a diuretic as compared to those not on a diuretic (OR 4.00, 95% CI 3.37–4.75). Discussion Gout was less likely to be diagnosed at patient visits The proportion of patient visits with a diagnosis of gout with an insurance status of “Other” and Medicaid as increased between 2007 and 2011, and this increase was compared to those with private insurance (OR 0.51, significant. It is hypothesized that gout visits are on the 95% CI 0.36–0.72 and OR 0.56, 95% CI 0.39–0.80, re- rise due to increases in obesity, hypertension, and spectively). Patients whose visit was recorded in an purine-rich diets [1, 2]. Obesity increases the production emergency department were less likely (OR 0.42, 95% of serum urate (sUA) levels and also decreases urate ex- CI 0.34–0.51) to be diagnosed with gout as compared cretion while weight reduction has been associated with to those with a physician’s office visit (Table 2). Sec- uric acid level declination . Along with risk factors, ond order interaction terms were investigated, found there are many disease associations with gout including to contribute nothing significant to the understanding metabolic syndrome, hypertension, and cardiovascular of the overall results and were excluded from the disease . Metabolic syndrome has been strongly as- final reported model. sociated with gout; 60% of US population with gout also Finally, plots of the percentage of patient visits with has metabolic syndrome, a prevalence three times higher gout by year for medication class and individual drug in those with gout . Metabolic syndrome has likely Fig. 1 Gout Prescription Trends by Drug Class, 2007–2011 Castro et al. BMC Rheumatology (2018) 2:14 Page 6 of 9 Fig. 2 Gout Prescription Trends by Individual Drug, 2007–2011 increased over the study years, helping explain the rise the years of 1997 through 2012 according to a study in gout diagnoses . which utilized the Clinical Practice Datalink . An- The number of gout visits in this study was not as other study which looked at gout prevalence in the UK high as noted in a similar prior study,  which could and Germany from 2000 to 2005 with the IMS Disease be attributed to several different factors. Although not Analyzer found a prevalence of 1.4% . A study of the specified in their methods, the Krishnan and Chen study Canadian province of British Columbia from 2000 to appears to have used aggregated estimates for the years 2012 using PopulationDataBC found a prevalence of 3. and databases studied, rather than the average annual 8% in 2012, and there was a noted increase over the estimates used in this study. Additionally, Krishnan and study period . A Swedish study examined gout Chen only utilized NAMCS and NHAMCS-OPD, trends from 2002 to 2012 and found a prevalence of 1. whereas this study also utilized NHAMCS-ED. Despite 8% in 2012 as well as an increase over the study period. approximately 95 million visits attributed to the  A study in Taiwan utilizing the National Health In- NHAMCS-ED, visits for gout in the ED were less likely. surance Research Database found a higher prevalence This likely increased the total number of overall visits rate of 6.24% over the study period of 2005 to 2010 . without adding a commensurate number of gout-specific With the exception of the Taiwan study , all studies visits to the numerator. Further, while only 31% of the were consistent with this study’s findings with regards to study population was aged 65 or older, 61% was female gout prevalence and increasing prevalence over the and the prevalence of gout is known to both increase years. with age and be more prevalent in males . This study demonstrated an association between age, Another prior study with higher gout estimates by sex, race group and gout visits, with an increased pro- Zhu, et al. was based on participant reported data from portion among older age groups (≥45 years of age), the National Health and Nutrition Examination Survey males, African American and ‘Other’ race groups. These (NHANES), which lends itself to estimating true preva- finding are consistent with previous studies [1, 3, 8, 10] lence . This study is based on provider reported am- showing that the risk of developing gout is age-related, bulatory, outpatient and emergency visits, therefore [1, 8] and that estrogen is protective in premenopausal limiting the ability to estimate the true prevalence of women due to its uricosuric effect . ‘Other’ race gout. This database distinction helps explain this differ- within the databases consists of Asian, Native Hawaiian ence in gout estimates. It is worth noting that NHANES or other Pacific Islander, American Indian or Alaska Na- as well as the data sources used for this study are all tive, or more than one race reported. A higher preva- population-based surveys. lence of gout is well known in Asians and Pacific The study results are consistent with several inter- Islanders, as well as African Americans with genetics national epidemiology studies which examined the playing a role due to hyperuricemia-associated DNA se- prevalence of gout [19–23]. The prevalence of gout has quence variations [24, 25]. However, diet and the pres- increased significantly in the United Kingdom (UK) over ence of co-morbidities cannot be ruled out. Castro et al. BMC Rheumatology (2018) 2:14 Page 7 of 9 Hispanic/Latino individuals were found to be less likely years. Antigout and antihyperuricemic medication clas- to have a visit with gout than Non-Hispanic/Latinos. One ses remained the two most commonly prescribed treat- possible explanation is related to diet. A previous study ments, while NSAIDs and steroids were used less. It is showed that Non-Hispanic/Latinos consume more red worth noting that due to drug class coding within the meat and seafood when compared to Hispanic/Latinos. databases some medications could have been coded in  Diets rich in red meat and seafood are widely known both the antigout and antihyperuricemic class (i.e., allo- to be associated with the production of uric acid . purinol and febuxostat) since drugs may be coded in as Given that Hispanic/Latino diets are typically more heav- many as four different medication classes. This might ily based on grains and beans along with fresh fruits and explain why the antigout percentage is greater than the vegetables, Hispanic/ Latinos may produce less uric acid antihyperuricemics. However, the findings in this study resulting in a lower incidence of gout . are consistent with the prior NAMCS and NHAMCS- Patient visits with Medicaid and ‘Other’ insurance OPD study which looked at gout treatment trends up were less likely to have a diagnosis of gout. Despite the through 2009 . These treatment trends can also be ex- lack of statistically significant interaction effects in the plained by typical prescribing patterns for a gouty attack multivariable model, this could be attributed to the role versus prophylaxis treatment to prevent gout flare. of age with the risk of developing gout [1, 8]. ‘Other’ in- NSAIDs and steroids are typically only used for gouty at- surance consisted of worker’s compensation, self-pay, no tacks and patients are treated prophylactically after an charge/charity, and other, while Medicaid also included initial gout attack to prevent future attacks [2, 31]. In the Children’s Health Insurance Program. Patients with addition, the risk of side effects with NSAIDs such as ‘Other’ insurance or Medicaid are less likely to be older, gastrointestinal bleeds, renal failure, and hypertension the age group at the highest risk for gout. likely impacted their use in treatment, especially in the Individuals were significantly less likely to have a visit case when chronic treatment is warranted [32, 33]. with a diagnosis of gout in a hospital emergency setting As evident from the figure showing the percentage of than they were in a physician’s office. Patients are more visits by year for individual gout drugs, allopurinol con- likely to visit a provider’s office for routine check-ups tinues to be the most prescribed treatment with colchicine and for chronic conditions like gout. While individuals second, a finding also consistent with Krishnan and Chen may visit a hospital for an initial or particularly severe . Allopurinol dominated the market as the only medica- attack of gout, they are presumably more likely to visit tion to reduce uric acid synthesis until the introduction of their provider when simply attempting to help keep their febuxostat in 2009 [2, 4, 11]. As expected, the percentage gout under control. Furthermore, patients with gout are of visits with febuxostat increased following its approval. much more likely to visit their provider if they are being Despite this, allopurinol and colchicine use changed little prescribed gout prophylaxis medication. from 2009 through 2011, evidence that febuxostat intro- A study by Garg, et al. looked at gout-related health care duction had minimal impact on the treatment trends for utilization in US emergency departments utilizing the Na- the study years. Probenecid use has declined over the tional Emergency Department Sample (NEDS) from 2006 years which can be explained by its potential for drug- to 2008 . The Garg study found approximately 0.7% of drug interactions as well as less favorable side effect pro- ED visits to be gout-related, slightly higher than 0.4% file, including risk of urolithiasis [2, 11, 34]. found in this study . A similar study by Jinno, et al. The previously mentioned international studies showed also utilized NEDS and examined gout ED visits from similar treatment trends. Allopurinol was prescribed for 2006 to 2012 . This study found 0.19% of visits with a most patients in UK and Germany at 89 and 93% respect- primary diagnosis of gout . Although not exactly com- ively; while colchicine use was only around 15–16% for parable to this study, which includes non-ED databases in both . Probenecid use was minimal (< 1%), but addition to the NHAMCS-ED, similar findings with both NSAIDs were utilized 80–90% for prophylaxis. . Allo- of these studies include gout-related ED visits being more purinol was also most commonly prescribed in British likely with men, and increasing age; and less likely with Columbia, Canada, with less than 1% use of febuxostat and different insurance types [28, 29]. probenecid . Colchicine and steroid use increased in Diuretic use was four times more likely to be associated British Columbia over the study period, while NSAID use with a gout visit. Previous research has shown that indi- declined by 31% . A study in Australia in 2005 found viduals who have high blood pressure and are also taking allopurinol to comprise 98.4% of all urate lowering therapy a diuretic have an increased risk for acquiring gout . with probenecid at < 1% . There was a common theme The diuretics’ mechanism of action is thought to contrib- from these studies of the overall underutilization of urate- ute to gout, increasing uric acid reabsorption . lowering treatment for gout [19–23, 35]. The graph of gout medication class by year showed The study is not without limitations. The observational, consistency in use among the drug classes over the cross-sectional nature of the study design limited the Castro et al. BMC Rheumatology (2018) 2:14 Page 8 of 9 authors to statements of association between visits with Poster presentations 2016 Wiggins Academic Sympo- gout diagnosis and the factors of interest. No claims of sium, Campbell University, Buies Creek, NC. causality can be made. Furthermore, the cross-sectional 2017 Interprofessional Health Sciences Research Sym- nature of the data sources used did not allow for repeated posium, Campbell University, Buies Creek, NC. measurements on patients over time. Several variables of interest, including alcoholism, Parkinson’s disease, depres- Endnotes sion, hypertension, weight status, tobacco use, and losar- The NAMCS and NHAMCS surveys define the four tan use had to be excluded from all analyses due to regions that comprise this variable as follows:Northeast: missing data and/or reliability issues. This is particularly Connecticut, Maine, Massachusetts, New Hampshire, unfortunate for variables such as hypertension and weight New Jersey, New York, Pennsylvania, Rhode Island, status, both known to be significantly associated with Vermont; Midwest: Illinois, Indiana, Iowa, Kansas, gout. All of the databases utilized are limited to three diag- Michigan, Minnesota, Missouri,Nebraska, North Dakota, noses. The NAMCS and NHAMCS-OPD include a data Ohio, South Dakota, Wisconsin; South: Alabama, Arkansas, field to collect other specific disease states (includes hyper- Delaware, District of Columbia, Florida, Georgia, Kentucky, tension, diabetes, and depression), but NHAMCS-ED does Louisiana, Maryland, Mississippi, North Carolina,Okla- not and only collects the diabetes variable of interest in homa, South Carolina, Tennessee, Texas, Virginia, West their other specific disease field. This likely contributed to Virginia; West: Arizona, California, Colorado, Idaho, the missing data for such highly prevalent conditions like Montana, Nevada, New Mexico, Oregon, Utah, Washington, hypertension and diabetes. The NAMCS and NHAMCS Wyoming, Alaska, Hawaii databases do not include federal offices or hospitals, includ- Abbreviations ing Veterans Affairs facilities where gout can be prevalent. CIs: Confidence intervals; MSA: Metropolitan statistical area; NAMCS: National In addition, the databases do not provide a true prevalence Ambulatory Medical Care Survey; NCHS: National Center for Health Statistics; of gout, but rather a surrogate via visits for gout based on NEDS: National Emergency Department Sample; NHAMCS-OPD/NHAMCS- ED: National Hospital Ambulatory Medical Care Survey Outpatient diagnostic codes from the three recorded diagnoses and Department & Emergency Department; NHANES: National Health and gout medications prescribed at the visits. It is not uncom- Nutrition Examination Survey; ORs: Odds ratios; sUA: Serum urate mon for epidemiological studies to rely on diagnostic codes for estimating prevalence. Those studies which have relied Availability of data and materials The datasets used and/or analysed during the current study are available on such codes have shown good accuracy. In addition, al- from the corresponding author on reasonable request. Note that the survey though gout medications were also used to identify gout data from which the analysis dataset was constructed can be found at the visits, there is a chance that medications like colchicine and CDC website https://www.cdc.gov/nchs/ahcd/index.htm. probenecid were used for conditions other than gout. How- Authors’ contributions ever, such alternative uses are rare. Study strengths include KEC, KDC, DR, and MH wrote the Background, Discussion, Results, and the use of nationally representative, population-based sur- Conclusions sections of the manuscript. MJ conducted all analyses and veys which allow for generalizing findings to the portion of wrote the Methods section of the manuscript. All authors read and approved the final manuscript. the US population that is commensurate with the study population. Further, the databases are provider reported Ethics approval and consent to participate data which allows for more reliability of results as com- The study was submitted to the Campbell University Institutional Review Board and received an exemption due to the data sources used being pared to patient reported data. This is the first study known publicly available and de-identified. As such, since this research was based to the authors to investigate febuxostat in the treatment of solely on the analysis of previously collected, de-identified data, it complies gout since its approval in 2009 . with the Helsinki Declaration. Competing interests Conclusion The authors declare that they have no competing interests. This study found that the proportion of visits with a diagnosis of gout continues to increase, although the Publisher’sNote overall percentage of gout remains low in the US popu- Springer Nature remains neutral with regard to jurisdictional claims in lation. Individuals who are male, aged 45–64 or 65 and published maps and institutional affiliations. older, non-Hispanic/Latino, African American or of Received: 8 January 2018 Accepted: 24 April 2018 ‘Other’ race, use private insurance, present to a physi- cian’s office, or use a diuretic are more likely to have a visit with a diagnosis of gout. Treatment trends over the References study years by medication class and individual gout 1. Zhu Y, Pandya BJ, Choi HK. Prevalence of gout and hyperuricemia in the US general population. Arthritis Rheum. 2011;63:3136–41. medications have remained consistent, with the intro- 2. Rymal E, Rizzolo D. Gout a comprehensive review. JAAPA. 2014;27:26–31. duction of febuxostat having little impact for the study 3. Sunkureddi PS, Nguyen-Oghalai TU, Karnath BM. Clinical signs of gout. years through 2011. Hospital Physician. 2006;42:39–42,47. Castro et al. BMC Rheumatology (2018) 2:14 Page 9 of 9 4. U.S. Food and Drug Administration. FDA Approved Drug Products. http:// 30. Pascual E, Perdiguero M. Gout, diuretics and the kidney. Ann Rheum Dis. www.accessdata.fda.gov/scripts/cder/drugsatfda/index.cfm?fuseaction= 2006;65:981–2. Search.DrugDetails. Accessed 10 Aug 2016. 31. Rees F, Hui M, Doherty M. Optimizing current treatment of gout. Nat Rev 5. Choi HK, Atkinson K, Karlson EW, Curhan G. Obesity, weight change, Rheumatol. 2014;10:271–83. hypertension, diuretic use, and risk of gout in men: the health professionals 32. Moon KW, Kim J, Kim JH, Song R, Lee EY, Song YW, et al. Risk factors for follow-up study. Arch Intern Med. 2005;165:742–8. acute kidney injury by non-steroidal anti-inflammatory drugs in patients 6. Primatesta P, Plana E, Rothenbacher D. Gout treatment and comorbidities: a with hyperuricemia. Rheumatology. 2011;50:2278–82. retrospective cohort study in a large US managed care population. BMC 33. Madhok R, Wu O, McKellar G, Singh G. Non-steroidal anti-inflammatory Musculoskelet Disord. 2011;12:103. drugs-changes in prescribing may be warranted. Rheumatology. 2006;45: 1458–60. 7. Juraschek SP, Miller ER, Gelber AC. Body mass index, obesity, and prevalent 34. Reinders MK, van Roon EN, Jansen TL, Deising J, Griep EN, Hoekstra M, et al. gout in the United States in 1988-1994 and 2007-2010. Arthritis Care Res. Efficacy and tolerability of urate-lowering drugs in gout: a randomized 2013;65:127–32. controlled trial of benzbromarone versus probenecid after failure of 8. Krishnan E, Chen L. Trends in physician diagnosed gout and gout therapies allopurinol. Ann Rheum Dis. 2009;68:51–6. in the US: results from the national ambulatory health care surveys 1993 to 35. Chung Y, Lu CY, Graham GG, Mant A, Day RO. Utilization of allopurinol in 2009. Arthritis Res Ther. 2013;15:R181. the Australian community. Int Med J. 2008;38:388–95. 9. Lee SJ, Hirsch JD, Terkeltaub R, Khanna D, Singh JA, Sarkin A, et al. Perceptions of disease and health-related quality of life among patients with gout. Rheumatology (Oxford). 2009;48:582–6. 10. Bhole V, de Vera M, Rahman M, Krishnan E, Choi H. Epidemiology of gout in women: fifty-two-year followup of a prospective cohort. Arthritis Rheum. 2010;62:1069–76. 11. Baker JF, Schumacher HR. Update on gout and hyperuricemia. Int J Clin Pract. 2010;64:371–7. 12. McCaig LF, Burt CW. Understanding and interpreting the National Hospital Ambulatory Medical Care survey: key questions and answers. Ann Emerg Med. 2012;60:716–21. 13. Sample text for describing NHAMCS in a research article. https://www.cdc. gov/nchs/data/ahcd/Sample_Text_for_Describing_NHAMCS_in_Research_ Article.pdf. Accessed 1 Mar 2018. 14. National Center for Health Statistics. Scope and sample design. http://www. cdc.gov/nchs/ahcd/ahcd_scope.htm#namcs_scope. Accessed 1 Mar 2018. 15. SAS Institute Inc. SAS/STAT 9.3 User’s Guide. Cary: SAS Institute Inc; 2011. 16. Saag KG, Choi H. Epidemiology, risk factors, and lifestyle modifications for gout. Arthritis Res Ther. 2006;8(1):S2. 17. Choi HK, Ford ES, Li C, Curhan G. Prevalence of the metabolic syndrome in patients with gout: the third National Health and nutrition examination survey. Arthritis Rheum. 2007;57:109–15. 18. Aguilar M, Bhuket T, Torres S, Liu B, Wong RJ. Prevalence of the metabolic syndrome in the United States, 2003-2012. JAMA. 2015;313:1973–4. 19. Kuo CF, Grainge MJ, Mallen C, Zhang W, Doherty M. Rising burden of gout in the UK but continuing suboptimal management: a nationwide population study. Ann Rheum Dis. 2015;74:661–7. 20. Annemans L, Spaepen E, Gaskin M, Bonnemaire M, Malier V, Gilbert T, et al. Gout in the UK and Germany: prevalence, comorbidities and management in general practice 2000-2005. Ann Rheum Dis. 2008;67:960–6. 21. Rai SK, Avina-Zubieta JA, McCormick N, De Vera MA, Shojania K, Sayre EC, et al. The rising prevalence and incidence of gout in British Columbia, Canada: population-based trends from 2000 to 2012. Semin Arthritis Rheum. 2017; 46:451–6. 22. Dehlin M, Drivelegka P, Sigurdardottir V, Svard A, Jacobsson LTH. Incidence and prevalence of gout in western Sweden. Arthritis Res Ther. 2016;18:164. 23. Kuo CF, Grainge MJ, See LC, Yu KH, Luo SF, Zhang W, et al. Epidemiology and management of gout in Taiwan: a nationwide population study. Arthritis Res Ther. 2015;17:13. 24. Reginato AM, Mount DB, Yang I, Choi HK. The genetics of hyperuricaemia and gout. Nat Rev Rheumatol. 2012;8:610–21. 25. Yang Q, Köttgen A, Dehghan A, Smith AV, Glazer NL, Chen MH, et al. Multiple genetic loci influence serum urate levels and their relationship with gout and cardiovascular disease risk factors. Circ Cardiovasc Genet. 2010;3:523–30. 26. Winkleby MA, Albright CL, Howard-Pitney AB, Lin J, Fortmann SP. Hispanic/ white differences in dietary fat intake among low educated adults and children. Prev Med. 1994;23:465–73. 27. Diet.com. Hispanic and latino diet. http://www.diet.com/g/hispanic-and- latino-diet. Accessed 1 Mar 2018. 28. Garg R, Sayles HR, Yu F, Michaud K, Singh J, Saag KG, et al. Gout-related health care utilization in US emergency departments, 2006 through 2008. Arthritis Care Res. 2013;65:571–7. 29. Jinno S, Hasegawa K, Neogi T, Goho T, Dubreuil M. Trends in emergency department visits and charges in the United States between 2006 and 2012. J Rheumatol. 2016;43:1589–92.
– Springer Journals
Published: May 30, 2018