The Impact of Uncontrolled Gout on Healthcare Utilization and Health Outcomes for United States Veterans Affairs Patients

Abstract

Introduction

Gout is an inflammatory arthritis that, when uncontrolled, can lead to chronic pain, disability, increased demand for healthcare, and poor health outcomes. This study sought to identify patients with gout and to describe the differences in epidemiology, pharmacotherapy, healthcare utilization, and outcomes for patients with controlled and uncontrolled gout in the United States (US) Veterans Affairs (VA) healthcare system.

Methods

This retrospective cohort study used electronic health record (EHR) data from VA patients from all US states and territories with gout from 1/1/2016 to 12/31/2022. The study included adult VA patients (18 +) with a diagnosis code for gout (ICD10 codes M10 or M1A) and two or more encounters 30 or more days apart. Uncontrolled gout was defined as one serum uric acid level (sUA) level > 8 mg/dl, tophi, or both in the study period.

Results

Of the 331,664 patients who met study criteria, 42% (138,068) were considered to have uncontrolled gout and 58% (193,596) were controlled. The uncontrolled group was younger (mean age 64 vs. 70 years, p < 0.01), and both groups were predominantly white non-Hispanic (58% and 70%) and male (99% and 99%). Specialist visits were more common in the uncontrolled group during follow-up: podiatry (38% vs. 30%, p < 0.01), rheumatology (24% vs. 9%, p < 0.01), and nephrology (24% vs. 12%, p < 0.01). Patients with uncontrolled gout were also significantly more likely to be seen in the emergency room (55% vs. 38%, p < 0.01) or admitted to the hospital (47% vs. 37%, p < 0.01) during follow-up.

Conclusions

Nearly half of VA patients with gout met criteria for uncontrolled gout, and these patients experienced greater healthcare utilization and worse health outcomes than patients with controlled gout. Patients with uncontrolled gout could benefit from additional/alternative approaches such as the adoption of a treat-to-target strategy and increasing referrals to a specialist.

Supplementary Information

The online version contains supplementary material available at 10.1007/s40744-025-00794-2.

Key Summary Points

Why carry out this study?

Gout affects about 5% of the population in the United States, about 6% of the Veterans Affairs (VA) population, and has been trending upwards since 2012.

This retrospective cohort study used Veterans Affairs (VA) electronic health record (EHR) data from all US states and territories to provide data on the impact of uncontrolled gout on healthcare utilization and health outcomes in VA patients with gout.

What was learned from this study?

Of the VA population, almost half of patients met criteria for uncontrolled disease, which was associated with higher healthcare utilization and worse outcomes, while 79% of all patients with gout were treated with urate-lowering therapy and only 24% were seen by a rheumatologist.

Patients with uncontrolled gout could benefit from additional/alternative approaches such as the adoption of a treat-to-target strategy and increasing referrals to specialist.

Introduction

From 2017 to 2018, gout affected 12.1 million individuals in the United States, representing approximately 5% of the population, which was an increase from 3.6% in 2012 [1]. Gout is even more prevalent among veterans. About 6% of those under Veterans Affairs (VA) care have been diagnosed, accentuating its significant impact on this vulnerable population [2]. Gout is a form of inflammatory arthritis that results from the buildup of uric acid crystals within the joints. It is characterized by gout flares—sudden, severe episodes of pain, accompanied by redness and swelling in one or multiple joints. Risk factors for gout include male sex-at-birth, obesity, high alcohol intake, and certain medical conditions such as high blood pressure and kidney disease. [3–5]. A universally accepted definition of uncontrolled gout (also referred to as advanced gout, refractory gout, and chronic gout) based on a specific serum uric acid level (sUA) and symptom criteria, has not been established. Studies have used sUA levels of > 6 mg/dl, > 7 mg/dl, or ≥ 8 mg/dl as a part of their definition of uncontrolled gout [6–8]. The presence of tophi or the need for an amputation may also indicate disease progression.

Medications are available to treat gout pain (NSAIDS and colchicine), block uric acid production (allopurinol and febuxostat), and improve uric acid excretion (probenecid). However, patients do not always respond adequately to their initial treatments, and some may persist with ineffective therapy for too long. Consequences include unnecessary pain, potential long-term complications from uncontrolled gout, and potential progression of disease severity. This is despite the availability of newer effective therapies, such as pegloticase, that have been approved for the treatment of gout that did not respond to first-line medications. Pegloticase significantly decreases uric acid in the bloodstream, which both prevents the formation of uric acid crystals in the joints and facilitates the dissolution of already-formed urate crystals and tophi [9].

Three Conflicting Guidelines

There are three guidelines used by healthcare professionals to determine treatment for patients with gout. They are written by three different groups; the American College of Physicians (ACP), the American College of Rheumatology (ACR), and the European League Against Rheumatism (EULAR) [10–12]. ACR generally recommends beginning urate lowering therapy (ULT) in patients with tophi, evidence of radiographic damage, in uncontrolled gout and uses a “treat-to-target” management strategy that aims for sUA levels of < 6 mg/dl. [11], ACP recommends a “treat-to-avoid-symptoms” approach and using a shared decision-making model [10]. EULAR states that ULT is indicated if patients have had frequent flares or urate arthropathy, they recommend all patients on ULT receive sUA monitoring, and has a goal sUA of < 6 mg/dl for patients on ULT [12].

Differences in recommendations across the three guidelines heighten the potential for confusion or uncertainty among clinicians. This jeopardizes patients' well-being by potentially increasing the risk of long-term complications from gout. A claims-based analysis of gout treatment in the US demonstrated that less than 80% of all uncontrolled patients received prescriptions for ULT and many patients with gout did not receive treatment according to published guidelines [6]. The authors also observed that patients seen by a rheumatologist instead of a primary care physician, at least once, had a significantly decreased number of emergency department visits (ED) for gout flares. The rheumatologists in this study were more likely to be following the ACR guidelines.

Impetus for this Study

The Veterans Affairs (VA) system is an attractive setting to study patients with gout because the VA primarily serves older male patients who are at greater risk due to baseline risk factors, medical conditions, and medications. Additionally, the VA is the largest vertically integrated healthcare system in the United States. It is an ideal setting to examine transitions of care from ambulatory, emergency, and inpatient settings; all within the same healthcare system.

This study sought to identify patients with gout and to describe the differences in epidemiology, pharmacotherapy, healthcare utilization, and outcomes for patients with controlled and uncontrolled gout in the United States (US) Veterans Affairs (VA) healthcare system.

Methods

Study Design

This retrospective cohort study used electronic health record (EHR) data from VA patients from all US states and territories with gout (ICD10: M10 or M1A) from January 1, 2016 to December 31, 2022. This study was an observational study and required no intervention or interferences with standard medical care, and thus it did not affect patient treatments. Historical data (medical history) were examined for 1 year prior to the study period. Index date was defined as the second encounter date that was coded for gout. Patients were followed until their last VA visit, loss to follow-up, record of death, or end of study period, whichever occurred earlier (Fig. 1).

Fig. 1.

Study design

Data Source

Data were extracted from the VA electronic health record. Comorbidities were identified using diagnosis codes included in Charlson Comorbidity Index. Patients came from all 18 Veterans Integrated Service Networks (VISNs), spanning all US states and territories. The VA is the largest integrated healthcare system in the United States, has healthcare facilities in all 50 states and Washington, D.C., and maintains an EHR, which includes structured administrative, clinical, laboratory, and pharmacy data repositories. These repositories include data from both hospital and clinic settings. Furthermore, the VA maintains a vital status file that enables investigators to determine patient mortality, even when it occurs outside the clinic or hospital.

Data on VA priority groups were collected. Groups (1–8) are assigned to veterans based on disabilities, service history, and income. The Department of Veterans Affairs utilizes the groups to determine eligibility and access to healthcare benefits.

Study Population

The team developed the cohort using structured electronic data. The study population consisted of adult VA patients (18 + years of age) with a diagnosis code for gout (ICD10 codes M10 or M1A) and two or more encounters 30 or more days apart, with the second date serving as the index date [2, 13], from 1/1/2016 to 12/31/2022 (Fig. 2).

Fig. 2.

Study inclusion flow chart

After meeting the initial study criteria, participants were then categorized as having controlled or uncontrolled gout. In the past, studies have varied in how they have defined poor serum urate (sUA) control. For example, Mikuls et al. and Chen-Xu et al. defined poor serum urate control as > 7 mg/dl, and Francis et al. used > 8 mg/dl for their definition [3, 13, 14]. A few other studies have used > 6 mg/dl, because that is the sUA cutoff mentioned in the ACR guidelines [10, 15]. Since there is no consistent definition of uncontrolled gout in the existing literature, we established our own definition after considering how other publications had defined uncontrolled gout. For our study, uncontrolled gout was defined as having at least one uric acid level > 8 mg/dl, tophi, or both during the study period. All other participants that met study criteria were considered to have controlled gout. We selected the > 8 mg/dl cutoff to identify patients with clearly elevated urate levels and a greater likelihood of being truly uncontrolled or symptomatic. Patients were excluded if they had a diagnosis code for drug-induced gout (M10.2) or tumor lysis syndrome (E88.3).

Health Outcomes

Health outcomes included days of follow-up (defined as the number of days from diagnosis to last follow-up, death, or end of study); presence and number of healthcare encounters, gout flares, sUA labs and levels, complications, and healthcare utilization (emergency room, urgent care, and hospital visits, plus number of hospital days); amputations; and death. The latter two (amputations and death) were studied because they were evaluated in prior literature [13, 14, 16]. To provide insight into fluctuating sUA levels over time, we reported the number of patients with sUA levels greater than 6 and 8 during the 12-month follow-up period, after patients had already been categorized as controlled or uncontrolled gout. Data were annualized to forecast outcomes at 12 months based on the mean observed during the follow-up period.

Data Analysis and Statistical Methods

Descriptive statistics were used to summarize baseline characteristics and treatments by calendar year. Bivariate statistical tests (chi-square, Fisher’s exact, Student’s t test, and Wilcoxon rank-sum test) were used to compare baseline characteristics, pharmacotherapy, healthcare utilization, and health outcomes for patients with controlled and uncontrolled gout and for uncontrolled patients who were treatment-naïve and experienced. Multivariable regression models were used to compare outcomes annualized at 12 months while controlling for age (< 65 vs. ≥ 65 years), sex (male vs. female), renal disease (presence vs. absence), diabetes with complications (presence vs. absence), and gout severity (uncontrolled vs. controlled). The models only included patients with gout (both controlled and uncontrolled). Adjusted odds ratios (AORs) with 95% confidence intervals were estimated to assess the associations. Analyses were conducted using R and SAS statistical software. An alpha level of < 0.05 was statistically significant. No adjustments were made for multiple comparisons.

Protection of Human Subjects

This observational study was performed in accordance with ethical principles that are consistent with the Declaration of Helsinki and the applicable legislation on non-interventional studies and/or observational studies. The final protocol of the observational study was approved by the University of Texas Health San Antonio Institutional Review Board (IRB) and the South Texas Veterans Health Care System Research & Development (HSR&D) committee. The IRB granted a Waiver of Informed Consent because it was not practical to carry out the research without the requested waiver.

Results

Of the 331,664 patients included in the study, 138,068 (42%) were uncontrolled and 193,596 (58%) were controlled. Both groups were predominantly white non-Hispanic (58% and 70%) and male (99% and 99%). The uncontrolled group was younger (mean age 64 vs. 70 years, p < 0.01), had more Black non-Hispanic patients (27% vs. 17%, p < 0.01), renal disease (29% vs. 21%, p < 0.01), and obese patients (BMI ≥ 30 kg/m²; 50% vs. 43%, p < 0.01) (Table 1, Fig. 3).

Table 1.

Cohort demographics for Veterans Affairs patients with uncontrolled and controlled gout

Baseline characteristics	Controlled (n = 193,596)	Uncontrolled (n = 138,068)	p value
Age, mean (SD)	–	–	–
Age (years)	70.08 (11.27)	64.44 (12.83)	< 0.001
Age group, %	–	–	–
< 65 years	49,219 (25%)	59,917 (43%)	< 0.001
65–74 years	81,962 (42%)	52,080 (38%)	< 0.001
> 74 years	62,415 (32%)	26,071 (19%)	< 0.001
Sex-at-birth, %	–	–	–
Male	191,309 (99%)	136,064 (99%)	< 0.001
Female	2287 (1%)	2004 (1%)	< 0.001
Race and ethnicity, %	–	–	–
Race Missing/Unknown	13,674 (7%)	8967 (6%)	< 0.001
Ethnicity Missing/Unknown	8263 (4%)	4950 (4%)	< 0.001
White non-Hispanic	135,175 (70%)	80,558 (58%)	< 0.001
Black non-Hispanic	33,701 (17%)	37,299 (27%)	< 0.001
Hispanic/Latinx	6397 (3%)	6585 (5%)	< 0.001
Other non-Hispanic	6305 (3%)	6543 (5%)	< 0.001
Charlson Score, mean (SD)	1.46 (1.52)	1.62 (1.74)	< 0.001
Charlson Age Score, mean (SD)	4.04 (2.07)	3.67 (2.45)	< 0.001
Comorbidities, %	–	–	–
Myocardial infarction	6080 (3%)	5414 (4%)	< 0.001
HF	21,350 (11%)	22,706 (16%)	< 0.001
Peripheral vascular disease	19,037 (10%)	15,315 (11%)	< 0.001
CVA or TIA	14,228 (7%)	9379 (7%)	< 0.001
Dementia	6323 (3%)	2591 (2%)	< 0.001
COPD	32,023 (17%)	24,114 (17%)	< 0.001
Connective tissue disease	4292 (2%)	3506 (3%)	< 0.001
Peptic ulcer disease	1634 (< 1%)	1311 (< 1%)	0.001
Mild liver disease	9675 (5%)	10,121 (7%)	< 0.001
Moderate or severe liver disease	862 (< 1%)	850 (< 1%)	< 0.001
Diabetes mellitus without complications	73,205 (38%)	48,523 (35%)	< 0.001
Diabetes with complications	32,738 (17%)	25,489 (18%)	< 0.001
Hemiplegia and paraplegia	1181 (< 1%)	793 (< 1%)	0.188
Renal disease	39,957 (21%)	39,737 (29%)	< 0.001
Solid tumor	20,483 (11%)	12,912 (9%)	< 0.001
Leukemia	1188 (< 1%)	1014 (< 1%)	< 0.001
Lymphoma	1116 (< 1%)	804 (< 1%)	0.826
Cancer	18,607 (10%)	12,144 (9%)	< 0.001
Metastatic carcinoma	1317 (< 1%)	837 (< 1%)	0.009
HIV/AIDS	536 (< 1%)	812 (< 1%)	< 0.001
Gout risk factors
Alcohol dependence	7885 (4%)	7994 (6%)	< 0.001
Hypertension	151,954 (78%)	103,715 (75%)	< 0.001
CKD
Stage 1	701 (< 1%)	648 (< 1%)	< 0.001
Stage 2	3757 (2%)	4174 (3%)	< 0.001
Stage 3	2580 (1%)	3632 (3%)	< 0.001
Stage 4	4587 (2%)	7162 (5%)	< 0.001
Stage 5	1657 (< 1%)	1504 (1%)	< 0.001
BMI group, %
Missing	45,492 (23%)	29,012 (21%)	< 0.001
< 20 kg/m2	1149 (< 1%)	678 (< 1%)	< 0.001
20 to < 25 kg/m2	14,795 (8%)	8309 (6%)	< 0.001
25 to < 30 kg/m2	48,697 (25%)	30,760 (22%)	< 0.001
≥ 30 kg/m2	83,463 (43%)	69,309 (50%)	< 0.001
VA priority group, %
Missing	645 (< 1%)	354 (< 1%)	< 0.001
Group 1	72,770 (38%)	59,593 (43%)	< 0.001
Group 2–6	82,301 (43%)	57,877 (42%)	< 0.001
Group 7–8	37,880 (20%)	20,244 (15%)	< 0.001
Current medications that induce gout, %
Diuretics	83,247 (43%)	75,737 (55%)	< 0.001
Salicylates	59,740 (31%)	50,175 (36%)	< 0.001
Immunosuppressants	5223 (3%)	4786 (3%)	< 0.001
Tuberculosis anti-infectives	68 (< 1%)	60 (< 1%)	0.228
Hormone replacement	3790 (2%)	3313 (2%)	< 0.001

HF heart failure, COPD chronic obstructive pulmonary disease, CVA cerebrovascular accident, SD standard deviation, TIA transient ischemic attack, CKD chronic kidney disease, BMI body mass index, VA Veterans Affairs

Fig. 3.

Age distribution in Veterans Affairs patients with uncontrolled and controlled gout. ^aSD standard deviation

Patients with uncontrolled gout saw specialists more often during follow-up: podiatry (38% vs. 30%, p < 0.01), rheumatology (24% vs. 9%, p < 0.01), and nephrology (24% vs. 12%, p < 0.01). During follow-up, patients who were grouped as uncontrolled gout were significantly more likely to have sUA levels above 6 mg/dl (92% vs. 54%, p < 0.01) and 8 mg/dl (74% vs. 3%, p < 0.01), and had more healthcare utilization in the emergency room (55% vs. 38%, p < 0.01) and hospital (47% vs. 37%, p < 0.01) (Tables 2 and 3).

Table 2.

Annualized outcomes at 12 months for Veterans Affairs patients with gout

Outcomesa	Controlled (n = 193,596)	Uncontrolled (n = 138,068)	p value
Length of follow-up, %
Patients with 12 months or more of follow-up	187,626 (97%)	134,558 (97%)	< 0.001
Healthcare utilization, %
Any clinic visit	193,594 (100%)	138,068 (100%)	0.771
Primary care visit	191,175 (99%)	136,898 (99%)	< 0.001
Podiatry visit	58,933 (30%)	52,071 (38%)	< 0.001
Rheumatology visit	17,578 (9%)	33,730 (24%)	< 0.001
Nephrology (with dialysis) visit	23,988 (12%)	33,374 (24%)	< 0.001
Symptoms, %
Gout flare	8338 (4%)	19,888 (14%)	< 0.001
sUA labs and levels, %
sUA lab drawn during follow-up	131,323 (68%)	128,841 (93%)	< 0.001
sUA level > 6 during follow-up	70,847 (54%)	119,028 (92%)	< 0.001
sUA level > 8 during follow-up	3999 (3%)	94,698 (74%)	< 0.001
Complications and outcomes, %
Amputation	1605 (< 1%)	1653 (1%)	< 0.001
Emergency room visit	74,383 (38%)	75,413 (55%)	< 0.001
Urgent care visit	14,705 (8%)	13,699 (10%)	< 0.001
Hospital visit	70,769 (37%)	64,362 (47%)	< 0.001
Death	18,770 (10%)	13,571 (10%)	0.201

^asUA serum uric acid

Table 3.

Annualized outcomes annualized at 12 months for Veterans Affairs patients with gout

Outcomesa	Controlled (n = 193,596)	Uncontrolled (n = 138,068)	p value
Length of follow-up, mean (SD)
Days of follow-up	N/A	N/A
Healthcare utilization, mean (SD)
Number of clinic visits	34.63 (48.12)	44.29 (53.76)	< 0.001
Number of primary care visits	7.61 (6.88)	8.92 (7.70)	< 0.001
Number of podiatry visits	1.69 (2.52)	1.67 (2.77)	0.031
Number of rheumatology visits	1.51 (1.85)	1.70 (1.82)	< 0.001
Number of nephrology (with dialysis) visits	7.27 (27.31)	5.77 (19.08)	< 0.001
UA labs and levels, mean (SD)
Number of UA labs drawn during follow-up	0.74 (0.85)	1.26 (1.47)	< 0.001
Symptoms, mean (SD)
Number of gout flares	0.35 (0.35)	0.37 (0.37)	< 0.001
Complications and outcomes, mean (SD)
Number of amputations	0.60 (0.95)	0.52 (0.73)	< 0.001
Number of emergency room visits	1.56 (2.43)	1.74 (2.38)	< 0.001
Number of urgent care visits	0.85 (1.37)	0.97 (1.47)	< 0.001
Number of hospital visits	1.16 (1.98)	1.21 (1.88)	< 0.001
Number of hospital days	12.26 (34.86)	11.63 (29.62)	< 0.001

^aSD standard deviation, UA uric acid

Overall, patients with uncontrolled gout in the VA were primarily prescribed allopurinol, of which use increased from 60% in 2016 to 70% in 2022 and febuxostat where use decreased from 7% in 2016 to 4% in 2022. (Figs. 4, 5, and 6) We did not collect medication doses and therefore cannot speak to whether or not ULT dosing was optimized.

Fig. 4.

Therapies for Veterans Affairs patients with uncontrolled gout: All lines, n = 138,068. ^aPatients might have received multiple therapies each year, either concurrently or consecutively. ^bSulfinpyrazone, pegloticase, and pegloticase + methotrexate were not given during study period

Fig. 5.

Therapies for Veterans Affairs patients with uncontrolled gout: First-line, n = 138,068. ^aPatients might have received multiple therapies each year, either concurrently or consecutively. ^bSulfinpyrazone, pegloticase, and pegloticase + methotrexate were not given during study period

Fig. 6.

Most prescribed medications for Veterans Affairs patients with uncontrolled gout in 2016 and 2022

Multivariable Models to Predict Healthcare Utilization and Health Outcomes

Multivariable logistic regression models were used to determine if gout severity (uncontrolled vs. controlled) was associated with healthcare utilization and health outcomes at 12 months. The covariates were selected based on prior literature, which suggested differences in gout control based on these characteristics [2, 3, 7, 14], and included patient age (< 65 vs. ≥ 65 years), sex (male vs. female), diabetes with complications (presence vs. absence), and renal disease (presence vs. absence). Results reflect adjusted odds ratios (AOR) and 95% confidence intervals (95% CI) (Table 4).

Table 4.

Multivariable logistic regression to determine if gout severity (uncontrolled vs. controlled) was associated with healthcare utilization and health outcomes at 12 months

Multivariable logistic regression for predictors of:	Estimate	Odds ratio (95%CI)	p value
Follow-up
Age (< 65 vs. > / = 65)	0.90	2.46 (2.32–2.61)	< 0.001
Sex (male vs. female)	– 0.06	0.94 (0.76–1.14)	0.56
Renal disease (yes vs. no)	– 1.07	0.34 (0.33–0.36)	< 0.001
Gout severity (uncontrolled vs. controlled)	0.20	1.22 (1.17–1.27)	< 0.001
Primary care visits
Age (< 65 vs. > / = 65)	0.73	2.07 (1.94–2.21)	< 0.001
Sex (male vs. female)	1.09	2.97 (2.57–3.42)	< 0.001
Renal disease (yes vs. no)	– 0.63	0.53 (0.51–0.56)	< 0.001
Gout severity (uncontrolled vs. controlled)	0.13	1.14 (1.08–1.20)	< 0.001
Podiatry visits
Age (< 65 vs. > / = 65)	– 0.22	0.80 (0.78–0.81)	< 0.001
Sex (male vs. female)	– 0.37	0.69 (0.65–0.74)	< 0.001
Renal disease (yes vs. no)	0.54	1.72 (1.69–1.76)	< 0.001
Gout severity (uncontrolled vs. controlled)	0.28	1.32 (1.29–1.34)	< 0.001
Rheumatology visits
Age (< 65 vs. > / = 65)	0.19	1.21 (1.18–1.23)	< 0.001
Sex (male vs. female)	– 0.65	0.52 (0.48–0.56)	< 0.001
Renal disease (yes vs. no)	0.44	1.55 (1.51–1.59)	< 0.001
Gout severity (uncontrolled vs. controlled)	1.08	2.94 (2.87–3.01)	< 0.001
Nephrology visits
Age (< 65 vs. > / = 65)	0.13	1.14 (1.11–1.18)	< 0.001
Sex (male vs. female)	– 0.52	0.59 (0.54–0.66)	< 0.001
Renal disease (yes vs. no)	3.23	25.39 (24.61–26.20)	< 0.001
Gout severity (uncontrolled vs. controlled)	0.72	2.06 (2.00–2.11)	< 0.001
Emergency room visits
Age (< 65 vs. > / = 65)	0.31	1.36 (1.34–1.38)	< 0.001
Sex (male vs. female)	– 0.30	0.74 (0.70–0.79)	< 0.001
Renal disease (yes vs. no)	0.65	1.92 (1.89–1.96)	< 0.001
Gout severity (uncontrolled vs. controlled)	0.54	1.72 (1.70–1.75)	< 0.001
Urgent care visits
Age (< 65 vs. > / = 65)	0.20	1.22 (1.18–1.27)	< 0.001
Sex (male vs. female)	– 0.04	0.96 (0.84–1.12)	0.62
Renal disease (yes vs. no)	0.18	1.20 (1.15–1.25)	< 0.001
Gout severity (uncontrolled vs. controlled)	0.38	1.47 (1.42–1.52)	< 0.001
Hospital visits
Age (< 65 vs. > / = 65)	– 0.09	0.91 (0.89–0.93)	< 0.001
Sex (male vs. female)	– 0.23	0.80 (0.74–0.86)	< 0.001
Renal disease (yes vs. no)	1.01	2.75 (2.70–2.80)	< 0.001
Gout severity (uncontrolled vs. controlled)	0.27	1.31 (1.29–1.34)	< 0.001
Gout flares
Age (< 65 vs. > / = 65)	0.60	1.82 (1.75–1.89)	< 0.001
Sex (male vs. female)	– 0.14	0.87 (0.76–1.01)	0.05
Renal disease (yes vs. no)	0.02	1.02 (0.98–1.07)	0.35
Gout severity (uncontrolled vs. controlled)	1.13	3.10 (2.98–3.23)	< 0.001
Amputations
Age (< 65 vs. > / = 65)	0.01	1.01 (0.88–1.16)	0.88
Sex (male vs. female)	0.95	2.58 (1.27–6.51)	0.02
Renal disease (yes vs. no)	1.46	4.31 (3.80–4.88)	< 0.001
Gout severity (uncontrolled vs. controlled)	0.16	1.17 (1.04–1.32)	0.01
Death
Age (< 65 vs. > / = 65)	1.18	3.27 (3.05–3.51)	< 0.001
Sex (male vs. female)	– 0.13	0.88 (0.70–1.09)	0.27
Renal disease (yes vs. no)	– 1.22	0.29 (0.28–0.31)	< 0.001
Gout severity (uncontrolled vs. controlled)	0.14	1.15 (1.10–1.20)	< 0.001

CI confidence interval

Patients with uncontrolled gout had more primary care visits (OR: 1.14, 1.08–1.20), podiatry visits (1.32, 1.30–1.35), rheumatology visits (2.93, 2.86–3.00), nephrology visits (2.05, 1.99–2.10), ER visits (1.73, 1.64–1.71), urgent care visits (1.47, 1.42–1.52), and hospital visits (1.30, 1.28–1.33), as well as experience more gout flares (3.10, 2.98–3.23), amputations (1.13, 1.00–1.28), and a higher mortality (1.17, 1.12–1.22). All reported values were statistically significant.

Younger patients (< 65 years) had more follow up (2.39, 2.25–2.54), primary care visits (OR: 2.03, 1.91–2.17), rheumatology visits (1.22, 1.19–1.25), nephrology visits (1.22, 1.18–1.26), urgent care visits (1.26, 1.21–1.31), ER visits (1.42, 1.40–1.45), gout flares (1.84, 1.77–1.91), amputations (1.30, 1.13–1.49), and a higher mortality (3.15, 2.93–3.38). Older patients (65 +) had more hospital visits (0.97, 0.95–0.99) and podiatry visits (0.91, 0.89–0.93). All reported values were statistically significant.

Men with uncontrolled gout, compared to women, were more likely to undergo amputation (OR: 2.82, 1.38–7.12), had more primary care visits (2.96, 2.56–3.40). Women were more likely to have rheumatology visits (0.51, 0.48–0.56), podiatry visits (0.71, 0.66–0.76), nephrology visits (0.60, 0.54–0.66), hospital visits (0.81, 0.75–0.87), and ER visits (0.75, 0.71–0.80). All reported values were statistically significant.

Renal disease was also associated with higher odds of amputation (OR: 2.05, 31.80–2.34) and more rheumatology visits (1.48, 1.45–1.52), nephrology visits (21.23, 20.6–21.93), urgent care visits (1.10, 1.05–1.14), and ER visits (1.67, 1.64–1.71). All reported values were statistically significant.

Finally, diabetes with complications was associated with higher odds of podiatry visits (4.36, 4.27–4.45), rheumatology visits (1.18, 1.15–1.22), nephrology visits (2.18, 2.12–2.24), ER visits (1.74, 1.71–1.78), urgent care visits (1.40, 1.33–1.46), hospital visits (2.16, 2.11–2.20), gout flares (1.16, 1.10–1.21), and amputations (9.92 (8.61–11.45). All reported values were statistically significant.

An alternate multivariable logistic regression model, Table 5, was used to determine if gout severity (uncontrolled vs. controlled) was associated with healthcare utilization and health outcomes at 12 months. For this analysis, more variables from Table 1 were included as covariates in the model. The overall conclusions were consistent for eight of ten dependent variables. Namely, patients with uncontrolled gout had more podiatry visits (1.23, 1.20–1.25), rheumatology visits (2.78, 2.71–2.85), nephrology visits (1.72, 1.67–1.77), ER visits (1.58, 1.55–1.61), urgent care visits (1.35, 1.30–1.40), and hospital visits (1.18, 1.15–1.21), as well as experienced more gout flares (2.84, 2.72–2.97) and a higher mortality (1.18, 1.12–1.25). Two of the dependent variables were no longer statistically significant in the alternate model: primary care visits (OR: 1.05, 0.98–1.13) and amputations (0.99, 0.86–1.13).

Table 5.

An alternate multivariable logistic regression model to determine if gout severity (uncontrolled vs. controlled) is associated with healthcare utilization and health outcomes at 12 months

Multivariable logistic regression for predictors of:	p value
Follow-up
Age (< 65 vs. > / = 65)	< 0.001
Sex (male vs. female)	0.27
Renal disease (yes vs. no)	0.93
Gout severity (uncontrolled vs. controlled)	< 0.001
White non-Hispanic (yes vs. no)	0.79
Hispanic/Latinx (yes vs. no)	0.13
Charlson Age Score	< 0.001
Alcohol dependence (yes vs. no)	< 0.001
Hypertension (yes vs. no)	< 0.001
CKD (yes vs. no)	< 0.001
BMI (> = 30 vs. < 30)	< 0.001
Diuretics (yes vs. no)	0.73
Salicylates (yes vs. no)	< 0.001
Immunosuppressants (yes vs. no)	< 0.001
Hormone replacement (yes vs. no)	< 0.001
Diabetes with complications (yes vs. no)	< 0.001
Primary care visits
Age (< 65 vs. > / = 65)	< 0.001
Sex (male vs. female)	< 0.001
Renal disease (yes vs. no)	< 0.001
Gout severity (uncontrolled vs. controlled)	0.21
White non-Hispanic (yes vs. no)	< 0.001
Hispanic/Latinx (yes vs. no)	< 0.01
Charlson Age Score	< 0.001
Alcohol dependence (yes vs. no)	0.89
Hypertension (yes vs. no)	< 0.001
CKD (yes vs. no)	0.65
BMI (> = 30 vs. < 30)	< 0.001
Diuretics (yes vs. no)	< 0.001
Salicylates (yes vs. no)	0.58
Immunosuppressants (yes vs. no)	0.03
Hormone replacement (yes vs. no)	< 0.001
Diabetes with complications (yes vs. no)	< 0.001
Podiatry visits
Age (< 65 vs. > / = 65)	< 0.001
Sex (male vs. female)	< 0.001
Renal disease (yes vs. no)	< 0.001
Gout severity (uncontrolled vs. controlled)	< 0.001
White non-Hispanic (yes vs. no)	< 0.001
Hispanic/Latinx (yes vs. no)	< 0.001
Charlson Age Score	< 0.001
Alcohol dependence (yes vs. no)	< 0.001
Hypertension (yes vs. no)	0.06
CKD (yes vs. no)	0.85
BMI (> = 30 vs. < 30)	< 0.001
Diuretics (yes vs. no)	< 0.001
Salicylates (yes vs. no)	< 0.001
Immunosuppressants (yes vs. no)	< 0.001
Hormone replacement (yes vs. no)	< 0.001
Diabetes with complications (yes vs. no)	< 0.001
Rheumatology visits
Age (< 65 vs. > / = 65)	< 0.001
Sex (male vs. female)	< 0.001
Renal disease (yes vs. no)	< 0.001
Gout severity (uncontrolled vs. controlled)	< 0.001
White non-Hispanic (yes vs. no)	< 0.001
Hispanic/Latinx (yes vs. no)	< 0.001
Charlson Age Score	< 0.001
Alcohol dependence (yes vs. no)	0.20
Hypertension (yes vs. no)	< 0.001
CKD (yes vs. no)	0.05
BMI (> = 30 vs. < 30)	0.04
Diuretics (yes vs. no)	< 0.001
Salicylates (yes vs. no)	< 0.001
Immunosuppressants (yes vs. no)	< 0.001
Hormone replacement (yes vs. no)	< 0.001
Diabetes with complications (yes vs. no)	< 0.001
Nephrology visits
Age (< 65 vs. > / = 65)	< 0.001
Sex (male vs. female)	< 0.001
Renal disease (yes vs. no)	< 0.001
Gout severity (uncontrolled vs. controlled)	< 0.001
White non-Hispanic (yes vs. no)	< 0.001
Hispanic/Latinx (yes vs. no)	< 0.001
Charlson Age Score	< 0.001
Alcohol dependence (yes vs. no)	< 0.001
Hypertension (yes vs. no)	< 0.001
CKD (yes vs. no)	< 0.001
BMI (> = 30 vs. < 30)	< 0.001
Diuretics (yes vs. no)	< 0.001
Salicylates (yes vs. no)	< 0.001
Immunosuppressants (yes vs. no)	< 0.001
Hormone replacement (yes vs. no)	0.14
Diabetes with complications (yes vs. no)	< 0.001
Emergency room visits
Age (< 65 vs. > / = 65)	< 0.001
Sex (male vs. female)	< 0.001
Renal disease (yes vs. no)	< 0.001
Gout severity (uncontrolled vs. controlled)	< 0.001
White non-Hispanic (yes vs. no)	< 0.001
Hispanic/Latinx (yes vs. no)	0.04
Charlson Age Score	< 0.001
Alcohol dependence (yes vs. no)	< 0.001
Hypertension (yes vs. no)	< 0.001
CKD (yes vs. no)	< 0.001
BMI (> = 30 vs. < 30)	< 0.001
Diuretics (yes vs. no)	< 0.001
Salicylates (yes vs. no)	< 0.001
Immunosuppressants (yes vs. no)	< 0.001
Hormone replacement (yes vs. no)	< 0.01
Diabetes with complications (yes vs. no)	< 0.001
Urgent care visits
Age (< 65 vs. > / = 65)	< 0.001
Sex (male vs. female)	0.66
Renal disease (yes vs. no)	< 0.01
Gout severity (uncontrolled vs. controlled)	< 0.001
White non-Hispanic (yes vs. no)	0.89
Hispanic/Latinx (yes vs. no)	< 0.001
Charlson Age Score	< 0.001
Alcohol dependence (yes vs. no)	< 0.001
Hypertension (yes vs. no)	0.81
CKD (yes vs. no)	0.88
BMI (> = 30 vs. < 30)	0.53
Diuretics (yes vs. no)	< 0.001
Salicylates (yes vs. no)	< 0.001
Immunosuppressants (yes vs. no)	0.24
Hormone replacement (yes vs. no)	< 0.01
Diabetes with complications (yes vs. no)	< 0.01
Hospital visits
Age (< 65 vs. > / = 65)	< 0.001
Sex (male vs. female)	< 0.001
Renal disease (yes vs. no)	< 0.001
Gout severity (uncontrolled vs. controlled)	< 0.001
White non-Hispanic (yes vs. no)	< 0.001
Hispanic/Latinx (yes vs. no)	0.01
Charlson Age Score	< 0.001
Alcohol dependence (yes vs. no)	< 0.001
Hypertension (yes vs. no)	< 0.001
CKD (yes vs. no)	< 0.001
BMI (> = 30 vs. < 30)	< 0.001
Diuretics (yes vs. no)	< 0.001
Salicylates (yes vs. no)	< 0.001
Immunosuppressants (yes vs. no)	< 0.001
Hormone replacement (yes vs. no)	0.54
Diabetes with complications (yes vs. no)	< 0.001
Gout flares
Age (< 65 vs. > / = 65)	< 0.001
Sex (male vs. female)	0.50
Renal disease (yes vs. no)	0.01
Gout severity (uncontrolled vs. controlled)	< 0.001
White non-Hispanic (yes vs. no)	< 0.001
Hispanic/Latinx (yes vs. no)	0.24
Charlson Age Score	0.32
Alcohol dependence (yes vs. no)	< 0.001
Hypertension (yes vs. no)	< 0.001
CKD (yes vs. no)	< 0.01
BMI (> = 30 vs. < 30)	< 0.001
Diuretics (yes vs. no)	< 0.001
Salicylates (yes vs. no)	< 0.001
Immunosuppressants (yes vs. no)	0.02
Hormone replacement (yes vs. no)	0.35
Diabetes with complications (yes vs. no)	0.42
Amputations
Age (< 65 vs. > / = 65)	< 0.001
Sex (male vs. female)	0.03
Renal disease (yes vs. no)	0.33
Gout severity (uncontrolled vs. controlled)	0.91
White non-Hispanic (yes vs. no)	< 0.01
Hispanic/Latinx (yes vs. no)	0.65
Charlson Age Score	< 0.001
Alcohol dependence (yes vs. no)	< 0.01
Hypertension (yes vs. no)	< 0.01
CKD (yes vs. no)	< 0.01
BMI (> = 30 vs. < 30)	0.02
Diuretics (yes vs. no)	0.38
Salicylates (yes vs. no)	< 0.001
Immunosuppressants (yes vs. no)	0.07
Hormone replacement (yes vs. no)	0.45
Diabetes with complications (yes vs. no)	< 0.001
Death
Age (< 65 vs. > / = 65)	< 0.01
Sex (male vs. female)	0.10
Renal disease (yes vs. no)	0.33
Gout severity (uncontrolled vs. controlled)	< 0.001
White non-Hispanic (yes vs. no)	0.57
Hispanic/Latinx (yes vs. no)	0.25
Charlson Age Score	< 0.001
Alcohol dependence (yes vs. no)	< 0.001
Hypertension (yes vs. no)	0.04
CKD (yes vs. no)	< 0.001
BMI (> = 30 vs. < 30)	< 0.001
Diuretics (yes vs. no)	< 0.001
Salicylates (yes vs. no)	< 0.001
Immunosuppressants (yes vs. no)	< 0.001
Hormone replacement (yes vs. no)	< 0.001
Diabetes with complications (yes vs. no)	< 0.001

CKD chronic kidney disease, BMI body mass index

Supplementary Material

Supplementary Table S1 contains an alternate model (bivariate regression) to compare demographics for VA Patients with uncontrolled and controlled gout. All conclusions from the alternate model were consistent with those reported above.

The complete table of Table 5 can be found in the supplementary material in Table S2, with all estimates, odds ratios, and confidence intervals for all covariates.

Discussion

This study aimed to identify differences in epidemiology, pharmacotherapy, healthcare utilization, and health outcomes of patients in the VA with different stages of gout. The variations in treatment strategies presented in clinical guidelines have likely contributed to the inconsistencies observed in gout treatment.

Demographics and Pre-existing Conditions

In the VA, 331,664 patients with gout were identified over a 7-year study period (2016–2022), with 42% (138,068) having uncontrolled gout (Fig. 2). Helget and colleagues investigated gout incidence, prevalence, and burden in the VA over a 10-year study period (2005–2014), which preceded our study period. Overall, the sample size of VA patients with gout was similar between our study and the study by Helget et al. (331,664 vs. 326,668, respectively). However, it is important to note that our study, despite spanning fewer years (7 vs. 10), still had a larger sample size. Finally, the study by Helget and colleagues did not differentiate between patients with controlled and uncontrolled gout, so our study provides important additional information about those subgroups.

A different non-VA claims-based study by Francis-Sedlak et al. identified differences between patients with controlled and uncontrolled gout. They used a smaller uncontrolled sample (138,068 in our study vs. 6831 in the Francis-Sedlak study), the uncontrolled population had fewer male patients (ours: 99% vs. Francis-Sedlak: 69%), and most of the patients were white non-Hispanic patients (ours: 59% vs. Francis-Sedlak: 60%). Their study likely had a smaller sample size because their definition of gout was stricter (sUA of 8 mg/dl, 90 + days of ULT, and 2 + sUA labs). Although the severity of disease in our sample was not as advanced compared to the Francis-Sedlak study, our uncontrolled group had a higher prevalence of renal disease, more severe renal impairment (CKD stage > 3), and increased emergency room and hospital visits. This suggests that even with less severe gout, inadequate disease management is still associated with greater renal disease burden and higher healthcare utilization.

Healthcare Utilization

In our study, gout-related healthcare utilization was noted to be higher in the uncontrolled group when compared to the controlled group. Consistent with previous work, we found higher emergency care, urgent care, and hospital visits in the uncontrolled group [6, 13–15]. Typically, if patients are seeking emergency care, it is because they are experiencing an acute gout attack and severe pain. This suggests that when gout treatment is not managed effectively, patients are more likely to need higher-acuity care.

Previously, it was found that if a patient with gout was seen by a rheumatologist, beneficial differences were observed in health outcomes related to gout. Edwards et al. conducted a retrospective claims-based study that investigated the types of providers seeing and treating patients with gout from 2015 to 2018. They found the number of patients with uncontrolled gout that were seen by a rheumatologist was almost half, which is less than with what we found (ours: 24% vs. theirs: 52%). Our study extends further to 2022, and shows that there has not been any improvement with specialist involvement, even in the VA. This is important to consider because Edwards and colleagues determined that those seen by a rheumatologist experienced better disease management, which resulted in less emergency room visits for gout.

Treatment Patterns

Allopurinol is used to lower uric acid levels for the prevention and management of gout and is commonly prescribed first-line for gout. In line with other studies, we found allopurinol was the most prescribed medication for gout in the VA [3, 14]. Chen-Xu conducted a retrospective study examining gout prevalence using NHANES data. They found that only 33% of all patients with gout were prescribed ULT, whereas we found that 79% of all patients with gout in the VA system were treated with ULTs. There are several possible reasons for this big difference. One factor may be the period in which the two studies took place (2007–2016 vs. 2016–2022), ours being the most recent. Additionally, the increase in ULT prescriptions might have been influenced by the fact that the VA is a closed system with good medication access, by guideline updates from EULAR in 2017 and ACR in 2020, as well as an overall rise in gout prevalence over the years [3, 17].

Gout-Related Healthcare Utilization and Health Outcomes

Our findings indicate that certain risk factors, such as uncontrolled gout, younger age, male sex-at-birth, diabetes with complications, and renal disease, are associated with worse outcomes and increased healthcare utilization. Additionally, after adding additional covariables (CKD, BMI, hypertension, drugs that induce gout, etc.) that may influence disease management, patients with uncontrolled gout were still predicted to have higher healthcare utilization, except for primary care visits and amputations. These findings stress the importance of a more unified and proactive approach to gout management across providers, specialties, and guidelines. Proper management and care could help to prevent complications, improve patients’ quality of life, and reduce total healthcare costs. This is supported by Flores and colleagues who compared patients with uncontrolled gout, controlled gout, and patients without gout costs and burdens [15]. They found that patients with uncontrolled gout had higher costs than patients with controlled gout and without gout. Additionally, the only notable difference between patients with controlled gout and non-gout controls was increased activity impairment, proposing that well-managed gout can substantially reduce the clinical and economic burden.

Newer Treatment Options for Gout

Pegloticase is a newer gout treatment with an FDA indication for uncontrolled gout when patients on maximum doses of oral ULTs still have active manifestations of gout—elevated sUA, repeated flares, or visible tophi. Pegloticase directly converts circulating uric acid to allantoin, a benign molecule easily removed by the kidneys. Long-term safety studies have found that a sUA < 6 mg/dl was achieved by 55% of patients in 6 months, and that 61% of all tophi had resolved after 1 year of pegloticase treatment [18]. Pegloticase can be given in combination with methotrexate to prevent anti-drug antibodies (ADAs) from developing and to prolong pegloticase effectiveness. A recent study demonstrated that 71% of patients on pegloticase plus methotrexate continued to have a positive response at 6 months, compared to 38.5% on pegloticase alone [19]. Pegloticase was not widely used in the VA (%) during the study period, but it should be considered as an approved option for patients with gout who have not experienced success with initial treatment options.

Limitations

There are potential limitations to consider. The VA population with gout consists primarily of older, white, male patients. Thus, our study might not be generalizable to other populations because of differences in racial and ethnic compositions of the VA and because of differences in access to care. Replication of findings in larger populations of women with gout is warranted. EHR data is not maintained for research purposes and might contain errors. There might be variation in the extent of physician reporting of patients’ comorbidities. The observational retrospective study design increases the possibility of missing information and classifying the sample population incorrectly. To combat this, we only included patients with two or more encounters that had a gout diagnosis code. And the use of ICD-10 codes to form our sample presents the possible limitation that not all patients with gout were included if an incorrect ICD-10 code was used. Use of ICD10 code M10.2 (“drug-induced gout”) did not exclude patients on current medications that might induce gout. Furthermore, medication compliance and dosing for the gout medications are important, but were not collected for this study.

The definition of uncontrolled gout in this study was informed by prior studies, which used a single UA level above a cut-point (usually 6, 7, or 8 mg/dl) to define uncontrolled gout. It is important to note that this strategy might overestimate the true burden of uncontrolled gout because patients might occasionally have a higher UA level if they temporarily stopped their ULT. It might be better for future studies to look for multiple UA levels above a certain cut-point. That said, retrospective studies like these often lack data on serial UA levels, so a prospective clinical trial might be a better study design to obtain serial UA levels, even when patients are asymptomatic, to better describe the fluctuation of UA levels and the impact of those on the definition of uncontrolled gout. We believe using the highest cut-point identified in the literature (> 8 mg/dl) and collecting data on UA levels during follow-up reduced the likelihood that we overestimated the true burden of uncontrolled gout. However, our study is still susceptible to the issue of occasional, single, high UA levels.

Strengths

Nevertheless, this study has important strengths. The VA is the largest integrated healthcare system in the US and is national in scope. The VA maintains a comprehensive computerized system that contains patient and provider demographics, inpatient and outpatient data, laboratory data, and pharmacy data. This enables us to use large populations of controlled and uncontrolled patients with gout to study many relevant variables that are readily obtained in the normal course of a patient’s clinical evaluation and enables us to follow patients through transitions of care in a setting were all patients have the same access to care. In addition to this, the VA population incorporates a variety of patients from different socioeconomic backgrounds with different medical conditions. The VA presents us with real-world evidence that is applicable to a wide variety of patients.

Conclusions

Despite the high prevalence of ULT use, nearly half of VA patients with gout met criteria for uncontrolled gout, and those patients experienced greater healthcare utilization and worse health outcomes than patients with controlled gout. Patients with uncontrolled gout would benefit from the adoption of a treat-to-target strategy and increased specialist referrals.

Supplementary Information

Below is the link to the electronic supplementary material.

Author Contributions

Study concept and design: Adriana Vargus, Corbyn M. Gilmore, Jim M. Koeller, Grace C. Lee, and Christopher R. Frei. Data and statistical analysis: Adriana Vargus, Corbyn M. Gilmore, Xavier F. Jones, and Christopher R. Frei. Interpretation of data: Adriana Vargus, Corbyn M. Gilmore, Jim M. Koeller, Grace C. Lee, Haridarshan Patel, Brian LaMoreaux, Xavier F. Jones, Christopher R. Frei. Drafting of the manuscript: Adriana Vargus and Christopher R. Frei. Critical revision of the manuscript for important intellectual content: Adriana Vargus, Corbyn M. Gilmore, Jim M. Koeller, Grace C. Lee, Haridarshan Patel, Brian LaMoreaux, Xavier F. Jones, Christopher R. Frei. Study supervision: Christopher R. Frei.

Funding

Funding for the study was provided by Amgen Inc. (formerly Horizon Therapeutics) as a research grant to the Foundation for Advancing Veterans’ Health Research, a non-profit entity within the Audie L. Murphy Veterans Hospital, San Antonio, TX. Dr. Frei was supported, in part, by an NIH Clinical and Translational Science Award (National Center for Advancing Translational Sciences, UM1 TR004538) while the study was being conducted. The journal’s Rapid Service Fee was funded by Amgen Inc.

Data Availability

The data belong to the U.S. Department of Veterans Affairs. Federal regulations prohibit the sharing of data. The data are not publicly available, and require specific authorization through the VA.

Declarations

Conflict of Interest

Amgen Inc. (formerly Horizon Therapeutics) provided money to the Foundation for Advancing Veterans’ Health Research, for Jim M. Koeller, Grace C. Lee, Xavier F. Jones, and Christopher R. Frei to perform this research. Xavier F. Jones’ and Christopher R. Frei’s institutions have also received grant money for them to perform research, from AstraZeneca, in the last 3 years. Haridarshan Patel and Brian LaMoreaux are employees of Amgen Inc. Adriana Vargus and Corbyn M. Gilmore declare no conflicts of interest.

Ethical Approval

This observational study was performed in accordance with ethical principles that are consistent with the Declaration of Helsinki and the applicable legislation on non-interventional studies and/or observational studies. The final protocol of the study was approved by the University of Texas Health San Antonio Institutional Review Board (IRB) and the South Texas Veterans Health Care System Research & Development (HSR&D) committee. The IRB granted a Waiver of Informed Consent because it was not practical to carry out the research without the requested waiver.