Plain Language Summary
Gout is a common, inflammatory arthritis caused by high uric acid levels in the blood. Serum uric acid (sUA) levels should remain below 6.0 mg/dl to reduce the number of gout flares, tophi development (urate build-ups on bones and in joints), and long-term joint and organ damage. This study examined insurance claims (2007–2016 Humana Research Database) to see if there were differences between patients with uncontrolled (sUA ≥ 8.0 mg/dl) and controlled (sUA < 6 mg/dl) gout in comorbidities, medication use, and reasons for seeking medical care. Patients who had a gout diagnosis, information for at least 6 months before and after gout diagnosis, and at least 90 days of urate-lowering therapy within 1 year of gout diagnosis were included. A total of 5473 and 1358 patients made up the controlled and uncontrolled groups, respectively. Both groups commonly had high blood pressure, high amounts of blood lipids (includes cholesterol), diabetes, cardiovascular disease, and chronic kidney disease (CKD). However, patients with uncontrolled gout had a higher prevalence of diabetes, CKD, and cardiovascular disease (including heart failure and atrial fibrillation). Specifically, CKD was more advanced in patients with uncontrolled gout (34.6% of patients had advanced CKD [Stage 4–5] compared to 22.2% of patients with controlled gout). Overall opioid use was higher in uncontrolled gout patients. This study found major differences between controlled and uncontrolled gout patients that contribute to higher disease burden for uncontrolled patients.
Introduction
Gout is a progressive, systemic inflammatory arthritis that is caused by hyperuricemia and is estimated to affect 3.9% of the population in the United States (8.3 million people in 2008 [
1], 9.2 million people in 2016 [
2]) and between 0.9 and 2.5% of the population in Europe [
3‐
5]. Uric acid has a serum solubility limit of 6.8 mg/dl and current guidelines recommend that gout patients maintain a serum uric acid (sUA) level of < 6 mg/dl [
6]. Chronically elevated sUA levels can lead to monosodium urate crystal deposition in joints, soft tissues, cartilage, and organs [
7,
8]. As a result, patients who do not meet target sUA levels have increased rates of gout flares and more persistent visible tophi. Therefore, patients with severe disease, as determined by the presence of visible tophi, chronic arthropathy, and/or frequent gout attacks, have a lower target sUA of < 5 mg/dl [
9]. Unfortunately, some patients have refractory gout, defined as the persistence of active disease symptoms and hyperuricemia (sUA > 6 mg/dl) despite the use of urate-lowering therapies. It has been estimated that 2–6% of gout patients have hyperuricemia because of refractory disease or urate-lowering therapy contraindications or intolerability [
10].
Gout results in chronic inflammation throughout the body, even when patients are in asymptomatic, flare-free periods. As a result, ongoing systemic and joint damage can occur [
11,
12], particularly in patients with an sUA > 6 mg/dl [
13]. Gout has been previously associated with renal disease [
14,
15], cardiovascular complications [
15,
16], and death [
16,
17], and, more recently, a higher premature death rate [
14,
18]. Gout-associated comorbidities generally include hypertension [
19‐
21], type 2 diabetes [
22,
23], hyperlipidemia [
19], cardiovascular disease [
24], coronary heart disease [
24‐
26], heart failure [
27,
28], atrial fibrillation [
28,
29], and stroke [
30].
Comorbidities of gout have been well studied in comparison to non-gouty populations. However, differences between controlled and uncontrolled gout patients are not well reported or understood. The current study examined and compared a relatively large population of controlled and uncontrolled gout patients that underwent at least 90 days of urate-lowering therapy. De-identified patient data were obtained from the Humana Research Database (2007–2016) and were specifically examined for comorbidities, medication usage, and reasons for seeking medical care.
Methods
The PearlDiver software (PearlDiver, Inc., Colorado Springs, CO, USA) was used to identify gout patients in the Humana Research Database (2007–2016), which contains data of both private-pay and Medicare patients. This study used only summarized de-identified patient data that was obtained from an existing database and did not involve the collection, use, or transmittal of individually identifiable data. Therefore, institutional review board approval for this study was not needed.
Adult patients who were enrolled in Humana for at least 6 months before and after the first documentation of a gout ICD-9/10 diagnosis code (274.*, M10.*, M1A.*) were included. The Humana claims database was examined for patients who had controlled (sUA < 6.0 mg/dl [
9]) or uncontrolled (sUA ≥ 8.0 mg/dl) gout based on sUA levels measured at least 90 days after beginning urate-lowering therapy. Uncontrolled gout was conservatively defined as an sUA of ≥ 8 mg/dl to make differences between groups more distinct and allow for better detection of both renal [
31] and non-renal comorbidities. Patients excluded were those who (1) had fewer than two sUA measurements, (2) had inconsistent sUA measurements (i.e., not all sUA measurements < 6.0 mg/dl or ≥ 8 mg/dl), (3) had sUA levels between 6.0 and 8.0 mg/dl, or (4) had not received ≥ 90 days of urate-lowering therapy.
Comorbidities and medication use following gout diagnosis were examined and compared between groups to better understand the influence of chronically elevated sUA on gout and other conditions. Further, we identified the reasons why gout patients sought medical care. This included diagnosis codes entered during hospital inpatient, hospital outpatient, emergency room—hospital, inpatient psychiatric facility, and comprehensive outpatient rehabilitation facility visits. Smokers were identified using tobacco-specific ICD-9 (305.1, V15.82) and ICD-10 (F17.2*, T65.2*, Z71.6, Z72.0, and Z87.891) diagnosis codes.
Data are presented as mean ± standard deviation when possible. Differences between groups were examined using Student’s t tests for continuous variables. Fisher’s exact tests or odds ratios (ORs) were used to compare categorical variables between groups (controlled gout group served as reference for OR calculations). Statistical significance was defined as p < 0.05.
Discussion
Clear, but sometimes discrepant treatment guidelines from multiple organizations exist [
6,
9,
32‐
36], but gout remains an undertreated and often poorly or under-managed disease [
37‐
41]. A recent population-based study indicated that only 33% of gout patients in the United States were on urate-lowering therapies between 2007 and 2014 [
2]. Though gout management that specifically targets lowering sUA levels improves gout sequelae (e.g., flares, joint damage, tophi) [
42], gout patients have higher levels of hypertension [
17‐
19], cardiovascular disease [
24,
43,
44], and kidney disease [
15,
27,
45] than patients without gout. However, less has been reported or is known about differences between the controlled and uncontrolled gout populations. The current study of nearly 7000 gout patients was designed to identify any such differences and improve our understanding of gout sequelae, especially with respect to the consequences of not achieving target sUA levels.
We found significantly higher rates of renal disease prevalence and severity, as well as cardiovascular disease and diabetes, in uncontrolled (≥ 8.0 mg/dl) gout patients as compared with controlled (< 6.0 mg/dl) gout patients. Generally, these differences manifested in increased rates of seeking medical care in uncontrolled gout patients. These findings agree with prior studies that found associations between elevated sUA levels and increased occurrence of cardiovascular disease [
46], and events [
46,
47], and death [
48] and between gout and diabetes [
24]. In comparison to the controlled gout group, CKD was more prevalent in the uncontrolled gout group, with a higher proportion having stages 3–5. Previous studies have also shown correlations between sUA levels and risk of renal failure [
49] and end-stage renal disease [
50]. Smoking was ruled out as a potential confounder in these assessments as approximately 2% of both the controlled and uncontrolled groups were smokers (based on previously validated ICD-9 smoking codes [
51]). However, not all physicians use ICD codes to document smoking status and the reported proportion is likely an underestimate for both study groups.
The uncontrolled gout group had nearly double the number of hospitalizations as the controlled gout group. The higher prevalence of multiple comorbidities in the uncontrolled gout population likely underlies this observation. Previously, patients with gout have been reported to have higher rates of hospitalization for cardiorenal complications than patients without gout [
52]. However, further research is needed to better understand contributing factors to the increased hospitalization rate observed here in uncontrolled gout patients compared with controlled gout patients.
Differences in medication use between controlled and uncontrolled gout patients were notable. Urate-lowering therapies were used by all patients in both groups. Allopurinol was used by a large proportion of patients in both groups (> 91%), but the average dose of the controlled group was higher than that of the uncontrolled group (262.3 vs. 173.4 mg). The relatively low allopurinol dose in both groups may be representative of allopurinol under-utilization, a known issue that arises from both physician under-prescribing and patient non-compliance [
53]. The lower dose in the uncontrolled group may also be representative of claims timing (2007–2016). Prior to 2017, it was not well known that allopurinol dose could be safely increased in patients with renal impairment [
54]. Further, allopurinol use with furosemide, which was used in the uncontrolled group more often, can result in an increased sUA via drug–drug interaction [
55]. Both colchicine (34.2 vs. 10.4%) and febuxostat (8.9 vs. 3.9%) were used more often in uncontrolled gout patients, but at similar daily doses. These differences were not surprising and likely reflect insufficient allopurinol dosing, poor treatment response, and/or intolerance to higher allopurinol doses. The increased use of colchicine likely reflects increased gout flare severity and frequency in the uncontrolled group. This finding is consistent with prior studies that have shown an increased frequency of gout flares in patients with sUA levels > 6 mg/dl [
42,
56‐
58].
Overall usage of pain and anti-inflammatory medication (corticosteroids, NSAIDs, and opioids) was higher in the uncontrolled group than in the controlled group. Opioid use was significantly higher in uncontrolled gout patients, but only by a narrow margin (59.9 vs. 55.5%). Corticosteroid use was also higher in uncontrolled gout patients (prednisone: 37.2 vs. 27.4%, methylprednisolone: 26.7 vs. 21.3%; both p < 0.001). Overall NSAID use was similar between patient groups (uncontrolled: 43.8%, controlled: 41.2%), with the exception of indomethacin, which was used more often in uncontrolled patients (19.4 vs. 5.3%, p < 0.001). Because indomethacin, corticosteroids, and opioids are all used to manage pain associated with gout flare, these findings are likely attributable to higher gout flare severity and frequency in the uncontrolled gout group. Increased medication use for gout flares may impact overall patient health, particularly in patients with kidney disease (NSAID nephrotoxicity) and diabetes (corticosteroid-induced hyperglycemia). However, the high use of pain medications in the controlled group indicates that gout patients generally experience high levels of pain (related or unrelated to gout), even when sUA levels meet the urate-lowering therapy target.
Anti-hypertensive medication use was high in both controlled and uncontrolled gout patients. This finding supports the well-known and long-established link between both hyperuricemia and hypertension [
59‐
62] and between gout and hypertension [
20,
21,
44]. Lisinopril and amlodipine were used most commonly and use of both agents was similar in both groups. Hydralazine and isosorbide were used more often in uncontrolled patients, likely reflective of hard to control hypertension and, more speculatively, increased cardiovascular disease, particularly heart failure. All diuretics examined were used significantly more often in the uncontrolled gout population than in the controlled gout population. This may be reflective of the observed higher rates of heart failure and/or the more severe kidney disease in the uncontrolled group.
Our study had several limitations. First, causality of the identified differences between controlled and uncontrolled gout groups cannot be determined with claims data. Second, our gout groups were not age- or sex-matched. This was intentional, and allowed differences between controlled and uncontrolled gout patients, including demographic parameters, to be more easily identified. Both groups had a similar proportion of men and women, but the controlled group was slightly older (72.5 vs. 69.1 years) and had a longer duration in the medical plan (2.67 vs. 1.85 years). This could have skewed our results toward a higher comorbidity/medication use prevalence in the controlled gout group. However other than hyperlipidemia, our uncontrolled gout patients had higher comorbidity rates. Our data were analyzed using both ICD-9 and ICD-10 codes. Therefore, multiple diagnosis codes had to be “bucketed” to accurately obtain the number of patients with a certain condition. It is possible that more obscure ICD codes were unintentionally omitted from our analyses, resulting in an underestimation of some comorbidities. However, our proportion of uncontrolled patients was as expected. Approximately 80% of included patients had a sustained sUA < 6 mg/dl after at least 90 days of oral urate-lowering therapy. This proportion is in agreement with a Veteran’s Administration gout registry study showing approximately 75% of patients achieved target sUA [
63]. Lastly, it is possible that successful sUA-lowering therapy is a proxy for medical care access and/or quality of medication compliance, both of which would confound our findings. We hope future research will clarify these limitations and hypotheses.