Gout affects approximately 8.3 million people in the

RESEARCH Benefit Restrictions and Gout Treatment Sharon M. Wang, PharmD, MS; Gustavus A. Aranda, Jr., PharmD, MS; Sara Gao, MS; and Bimal V. Patel, PharmD, MS ABSTRACT BACKGROUND: Gout is a chronic rheumatic disease that can have serious sequelae, including persistent pain, nerve compression, joint destruction and deformities if left untreated. Febuxostat, initially introduced in the United States in 2009, was the first new treatment option for gout in over 40 years. With the introduction of a new drug into a therapeutic class that is composed of generically available options, utilization management will be a common strategy employed in an effort to contain cost; however, the effects of these strategies are not known for chronic gout treatment. OBJECTIVE: To evaluate the effect of utilization management strategies on chronic gout treatment. METHODS: This retrospective analysis examined claims data from a large, national pharmacy benefits manager with a client base that includes commercial HMOs, Medicaid, Medicare Part D, and self-insurers. The study population included patients aged 18 years or older who had at least 1 rejected claim for febuxostat in the 16-month identification period from March 1, 2009, through June 30, 2010. Outcomes of interest were the proportion of patients who filled a febuxostat prescription and proportion of patients who filled a prescription for another chronic gout treatment within 1 month of the febuxostat claim rejection date (the index date). Multivariate logistic regression models were used to assess factors affecting patient response to a rejected febuxostat claim. RESULTS: Of 1,034 patients with rejected febuxostat claims, 95% had claims denied due to utilization management: 36% due to step therapy, 25% due to lack of drug coverage, 18% due to quantity or other limits (i.e., fill limit exceeded, days supply exceeding benefit maximum, or maximum days supply limit exceeded), 16% due to prior authorization requirements, while 5% were due to other reasons unrelated to the utilization management strategies of interest. Other reasons included over 100 possible rejection reasons such as fill dispensed too soon, missing/invalid days supply, group not having benefit, physician not covered, non-matched group, not a network pharmacy, non-matched member, claim/member birth date not matching, and missing/invalid prescriber identifier. Subsequently, 474 (46%) of these patients filled a febuxostat prescription within 1 month of a rejected claim; 364 (35%) had not filled a prescription for any chronic gout medication within a month of the febuxostat claim rejection. Those filling a febuxostat prescription had higher pre-index total 6-month pharmacy costs than those not filling a chronic gout prescription ($1,718 vs. $988; P < 0.001). They also had a higher number of pre-index drug claims (25 vs. 18; P < 0.001). The regression model found the following variables to be statistically significant in positively influencing the likelihood of patients filling a febuxostat prescription within a month of a febuxostat claim rejection: (a) self-insured coverage (compared with commercial HMO coverage); (b) pre-index total prescription costs of at least $1,800; (c) claim rejection due to quantity or other limit (compared with lack of drug coverage); (d) claim rejection due to other reason (compared with lack of drug coverage); and (e) 1 and 1 pre-index colchicine claim. Patients with projected febuxostat copay of $100 to $149 were found to be less likely to fill febuxostat compared with patients with a projected copay of $0 to $19. CONCLUSION: Utilization management strategies likely result in gaps in gout treatment; 35% of patients with a denied febuxostat claim in this study population did not fill a prescription for any chronic gout therapy within a month of the claim denial. These findings are important in the consideration of benefit design in gout treatment. J Manag Care Pharm. 2013;19(9):773-82 Copyright 2013, Academy of Managed Care Pharmacy. All rights reserved. What is already known about this subject Gout is a chronic rheumatic disease with potentially severe sequelae. The introduction of febuxostat in 2009 provided the first new gout treatment option in over 40 years. Previously, drug therapies were limited to allopurinol, a xanthine oxidase inhibitor that reduces production of uric acid, and probenecid, a uricosuric agent that increases excretion of uric acid. Allopurinol has long been the mainstay of urate-lowering therapy, since probenecid is limited to patients with normal renal function. In current market situations, implementing utilization management strategies is likely to occur, with variable consequences. What this study adds A substantial proportion of gout patients with a denied febuxostat claim received no chronic gout therapy 1 month after the claim denial. Such findings are important to consider in structuring benefit designs for gout treatment. Gout affects approximately 8.3 million people in the United States 1 and is the most common cause of inflammatory arthritis in men over 40. 2 It is a rheumatic disease caused by hyperuricemia and the deposition of uric acid crystals in tissues and fluids within the body. Gout is characterized by recurrent attacks of swelling, erythema, warmth, and intense pain, which may persist for days to weeks. The development of gout typically progresses from asymptomatic hyperuricemia to acute gout attacks to intercritical gout to chronic tophaceous gout. 2 Advanced gout may cause persistent pain, nerve compression, ulceration of the skin, joint destruction and deformities, and more rarely, renal manifestations including urolithiasis. 2,3 The goals of treatment are to resolve acute gouty flares, prevent future attacks and the formation of tophi, and reduce www.amcp.org Vol. 19, No. 9 November/December 2013 JMCP Journal of Managed Care Pharmacy 773

gout-related morbidity. 4 While treatment of acute gout may include medications to alleviate pain, inflammation, and swelling, those pharmacologic therapies do not treat the underlying cause. Current chronic drug therapies for treating gout include allopurinol, febuxostat, and probenecid. Allopurinol and febuxostat are xanthine oxidase (XO) inhibitors that decrease the production of uric acid, whereas probenecid lowers uric acid levels by increasing renal excretion of uric acid. Allopurinol has long been the mainstay of urate-lowering therapy, since probenecid is not recommended as first line therapy unless an XO inhibitor is contraindicated or not tolerated; probenecid is contraindicated in patients with nephrolithiasis and should be administered with caution in patients with renal disease or renal failure. 3,5 The introduction of the XO inhibitor febuxostat (Uloric) in 2009 provided the first new treatment option for gout in the United States in over 40 years. Febuxostat has been shown to be highly efficacious for reducing urate levels in randomized controlled trials. 6,7 With a new branded drug on the market in a therapeutic class mainly composed of generically available drugs, utilization management strategies such as prior authorizations, step therapy, and drug coverage exclusions may be implemented in an effort to contain costs by encouraging use of less expensive alternatives. Since there has not been a new drug for gout until the introduction of febuxostat, there are no studies looking at benefit design and use for gout treatment, particularly after the rejection of a claim. The objective of this study is to evaluate the effect of utilization management methods for chronic gout treatment. Methods Study Design and Data Source This retrospective claims database analysis assessed the impact of utilization management of chronic gout medication. Pharmacy claims data from MedImpact s database were used. MedImpact is a large private national pharmacy benefits company that provides pharmacy benefits management to a diverse client base, including commercial health maintenance organizations (HMOs), Medicaid, Medicare Part D, and self-insurers. The claims database contains detailed adjudication information, including the status of a claim (i.e., approved, rejected, and reversed) and denial reason for rejected claims. Subject Selection The target study population included members aged 18 and older who had at least 1 rejected prescription claim for febuxostat during the identification period from March 1, 2009, through June 30, 2010. Patients were required to have continuous eligibility during the 6-month pre- and 3-month post-index period, where the index date was defined as the date of first rejected prescription claim for febuxostat during the identification period. Patients who had a febuxostat claim prior to the identification period (i.e., pre-index period) were not included for analysis. Utilization Management This research focused on patients who had a rejected claim for febuxostat, where the rejection was attributed to utilization management such as (a) prior authorization, (b) drug not covered, (c) step therapy, and (d) quantity limits and other limits (i.e., fill limit exceeded, days supply exceeded, benefit maximum exceeded, or maximum days supply exceeded). There were 156 possible reasons why a claim may be rejected; however, if a claim s rejection reason was unrelated to 1 of the 4 utilization management strategies, then the rejected claim was classified as rejected due to other reason. Other reasons commonly included dispensed too soon, missing/invalid days supply, group did not have benefit, physician not covered, non-matched group, not a network pharmacy, non-matched member, claim/member birth date did not match, and missing/ invalid prescriber identifier. Patients with rejected claims due to drug was not covered had closed formularies, in which the drug was excluded from the formulary. Patients with open formularies do not experience a claim rejection for lack of drug coverage but instead may be exposed to paying a higher copay. However, patients with open formularies may nevertheless experience a claim rejection due to other utilization management policies or for other reasons. A claim may be rejected for multiple reasons. A hierarchy based on the stringency of the utilization management strategy was used to allow classification of rejected claims into mutually exclusive groups based on rejection reason. In the situation where the patient s claim had 2 rejections, we categorized rejections as (a) drug coverage, step therapy, quantity limit or other limit, or (b) other reason. The prior rejection reason was used for categorization rather than the latter ( other reason ). If the claim had multiple rejections not categorized as other reason, then prior authorization took precedence over step therapy or quantity limits/other limits for categorization. Between step therapy and quantity limits, step therapy took precedence. Reporting of claim rejection reason by the patient was based on the rejection reason for the index febuxostat rejected claim, where the index claim was the first rejected febuxostat claim in the identification period. Variables There were 2 main outcomes of interest measured in the postindex period: (a) proportion of patients who filled febuxostat within 1 month of the rejected claim, and (b) proportion of patients who used any chronic gout medication within 1 month of the rejected febuxostat claim. Chronic gout medications included the following agents used to reduce uric acid levels: febuxostat, allopurinol, probenecid, and colchicine/ probenecid. 774 Journal of Managed Care Pharmacy JMCP November/December 2013 Vol. 19, No. 9 www.amcp.org

FIGURE 1 Data Construction Process All rejected febuxostat claims from March 1, 2009, to June 30, 2010 5,209 claims for 1,734 unique members Continuous enrollment in the same health plan 6-month pre- and 3-month post-index period 1,296 members Naive to febuxostat in the 6-month pre-index period 1,047 members Aged 18 years and had complete benefit information in MedImpact s database 1,034 members Had a claim rejected for utilization management reasons 981 members Had a claim rejected for other reasons 53 members Potential covariates included patient s age group, gender, payer business type (i.e., commercial, Medicaid, self-insured, and Medicare), geographical region of residence, and reason for the rejected febuxostat claim. Prescription claims history in the pre-index period for at least 1 agent used to treat hypertension, diabetes, dyslipidemia, and obesity was used as a proxy for disease-related comorbidity. This approximation of comorbidity used First DataBank s drug classification for those select conditions. RxRisk, a risk stratification model based on pharmacy claim data that is used for profiling and predicting health care utilization, was also assessed. 8 RxRisk uses the American Hospital Formulary Services (AHFS) drug classification for identifying drugs used for select diseases. Comorbidities are approximated based on prescription claims history for a drug that is contained within the AHFS category for a particular RxRisk disease category. Pre-index prescription claims history for select agents and RxRisk were considered as covariates because clinical guidelines for the management of gout suggest managing concomitant chronic diseases (e.g., obesity, diabetes, and metabolic disease) in addition to lifestyle modification and pharmacotherapy. 9-11 Since pharmacotherapy for acute gout management includes analgesic anti-inflammatory drugs (e.g., oral nonsteroidal anti-inflammatory drugs [NSAIDs] and cyclooxygenase-2 [COX-2] inhibitors), colchicine, corticosteroids, and opiate analgesics as adjunct therapy, pre-index claims for analgesics, colchicine, and corticosteroids were also considered as covariates. 9 Additional covariates included the total number of claims in the pre-index period and pharmacy costs in the pre-index period, which were measured and used to approximate the patient s burden of illness. Finally, the projected copay for the rejected febuxostat claim was calculated and included as a covariate. If a patient with a closed formulary decided to pay for the medication following a claim rejection, the patient would typically pay a discounted network price. Average wholesale price minus 15% was used as the copay amount for that scenario. For a patient with an open formulary, a best case scenario was assumed where the prior authorization requirement was met and the patient could receive febuxostat at a copay level for preferred brands. Statistical Analysis Demographics, clinical characteristics, and benefits information were assessed for patients with a rejected febuxostat claim and by patient response (i.e., if they filled nothing, filled febuxostat, filled allopurinol or probenecid or colchicine/probenecid). Since groups were not mutually exclusive (e.g., a patient could have filled allopurinol and febuxostat), groups were individually compared against patients who did not fill a chronic gout medication using t-tests. Similarly, total pharmacy costs and number of drug claims in the pre-index period were reported, and t-tests were used to compare patient response, using patients who filled nothing as the reference group. Outliers for total number of claims and costs in preindex variables were retained and included in the analysis. Multivariate logistic regression models were used to assess the factors that could affect patient response to a rejected febuxostat claim. The first model examined the proportion of www.amcp.org Vol. 19, No. 9 November/December 2013 JMCP Journal of Managed Care Pharmacy 775

FIGURE 2 Quality limits and other limits 503 (16%) Step therapy 1,051 (32%) Distribution of Rejected Febuxostat Claims by Rejection Reason (N = 3,259) Other reasons a 335 (10%) Prior authorization 659 (20%) Drug coverage 711 (22%) a Other reasons are unrelated to utilization management strategies of interest, with over 100 possible reasons (e.g., dispensed too soon, missing/invalid days supply, physician not covered, pharmacy not in network, etc.). patients who filled febuxostat within 1 month of the rejected claim, whereas the second model examined the proportion of patients who filled any chronic gout treatment within 1 month of the rejected claim. Both final models included the following covariates: age, payer business type, total pharmacy cost in the pre-index period, number of pharmacy claims in the preindex period, reason for febuxostat claim rejection, projected febuxostat copay, pre-index prescription claims history of analgesia (excluding colchicine), colchicine, and corticosteroids, and RxRisk category. RxRisk categories were included in the model when the prevalence was at least 5%. With the inclusion of RxRisk, pre-index claims history for select disease-related comorbidities based on First DataBank categorization was not included in the models to minimize redundancy. Geographical region of residence was not included as the regions were unbalanced and that factor did not lend itself to comparison. Distributions for all variables were assessed prior to dummy coding covariates. Correlation matrices and interaction terms (i.e., interaction between age and pre-index total pharmacy cost and interaction between age and pre-index claim count) were investigated before covariates were included in the final models. Statistical significance was based on P < 0.05. Results The study population included a total of 1,034 patients. Figure 1 shows the patient selection flow chart. Of these patients, 95% had rejected index febuxostat claims due to utilization management and 5% for other reasons: 36% (n = 369) of the rejections were attributed to step therapy, 25% (n = 259) to drug coverage, 18% (n = 183) to quantity limits and other limits, and 16% (n = 170) to prior authorization requirements. The total number of rejected febuxostat claims assessed for these patients was 3,259, with 32% (n = 1,051) and 22% (n = 711) of rejected claims attributed to step therapy and drug coverage, respectively (Figure 2). Table 1 provides the demographic and clinical characteristics of the patient population. Among these patients with a rejected febuxostat claim, most patients were male (73%); most were aged 40-64 (52%); and most were enrolled in a commercial HMO (63%). There were 31% of patients who had pre-index claims for allopurinol, and 36% had claims for colchicine. RxRisk categories approximated that 25% of patients had diabetes; 55% had heart disease/hypertension; and 47% had hyperlipidemia, based on pre-index pharmacy claims history. During the pre-index period, the mean total pharmacy cost per patient was $1,390 (standard deviation [SD] = $2,028) and mean number of drug claims per patient was 22 (SD = 19; Table 2). Among the study population, 46% (n = 474) filled a febuxostat prescription within a month following a rejected febuxostat claim while 35% (n = 364) had not filled any chronic gout medication prescription within a month after a claim rejection (Table 1). Patient characteristics and pre-index utilization for patients by response to a claim rejection are also provided in Table 1. Pre-index mean total pharmacy costs for patients who filled febuxostat were higher than for patients who did not fill any chronic gout medication ($1,718 [SD = $2,353] vs. $988 [SD = $1,543]; P < 0.001; Table 2). The number of drug claims in the pre-index period was also higher in patients who subsequently filled febuxostat compared with no chronic gout medication (25 [SD = 20] vs. 18 [SD = 15]; P < 0.001; Table 2). Projected febuxostat copay could only be calculated for 1,027 patients due to missing benefits information for some patients. Mean projected copay was $84 (SD = $54). Table 3 provides the final regression model that identifies factors associated with patients who filled febuxostat prescriptions within 1 month of a rejected claim. Interaction terms for age with pre-index total pharmacy cost and age with pre-index claims count were not found to be statistically significant (P > 0.05) and were not included in the final model. Patients with pre-index total prescription costs of at least $1,800 were more likely to have filled a febuxostat prescription compared with patients with pre-index total prescription costs of $0 to $199 (odds ratio [OR] = 2.12; P = 0.020). Compared with claims rejected for drug coverage, claims rejected due to quantity limit or other limit were 13 times as likely to have resulted in a febuxostat fill, and rejections for other reason were 3.2 times as likely (OR = 13.18; P < 0.001 and OR = 3.22; P = 0.001, respectively). Patients with a projected febuxostat copay of $100 to $149 were less likely to have filled a febuxostat prescription than when the copay amount was $0 to $19 (OR = 0.48; P = 0.008). Patients with 1 claim for colchicine in the pre-index period were more likely to have filled a febuxostat prescription than patients without colchicine (OR = 1.37; 776 Journal of Managed Care Pharmacy JMCP November/December 2013 Vol. 19, No. 9 www.amcp.org

TABLE 1 Demographic, Clinical Characteristics, and Febuxostat Rejection Reasons of the Patient Population All Patients Filled Nothing Within 30 Days a Filled Febuxostat Within 30 Days a Filled Allopurinol Within 30 Days a Filled Probenecid or Colchicine/ Probenecid Within 30 Days a Member Count N % N % N % P Value b N % P Value b N % Member count 1,034 100.00 364 35.20 474 45.84 208 20.12 10 0.97 Gender Female 275 26.60 90 24.73 132 27.85 0.311 55 26.44 0.650 3 1.09 Male 759 73.40 274 75.27 342 72.15 0.311 153 73.56 0.650 7 0.92 Age 18-39 77 7.45 31 8.52 29 6.12 0.191 19 9.13 0.802 1 1.30 40-64 536 51.84 193 53.02 239 50.42 0.456 110 52.88 0.975 3 0.56 65 421 40.72 140 38.46 206 43.46 0.146 79 37.98 0.910 6 1.43 Payer type Commercial-HMO 647 62.57 210 57.69 307 64.77 0.037 136 65.38 0.071 8 1.24 Medicaid 73 7.06 34 9.34 24 5.06 0.020 18 8.65 0.784 - - Part D 254 24.56 101 27.75 106 22.36 0.073 49 23.56 0.274 2 0.79 Self-insured 60 5.80 19 5.22 37 7.81 0.128 5 2.40 0.075 - - Geographical location Northeast 75 7.25 24 6.59 43 9.07 0.182 9 4.33 0.239 1 1.33 Midwest 702 67.89 245 67.31 326 68.78 0.652 143 68.75 0.723 5 0.71 South 37 3.58 16 4.40 13 2.74 0.208 7 3.37 0.533 2 5.41 Mountain 32 3.09 10 2.75 19 4.01 0.311 4 1.92 0.521 - - Pacific 188 18.18 69 18.96 73 15.40 0.174 45 21.63 0.441 2 1.06 Pre-index utilization ( 1 Rx claim) for c Chronic gout c Allopurinol 318 30.75 100 27.47 163 34.39 0.033 67 32.21 0.231 2 0.63 Probenecid 37 3.58 9 2.47 22 4.64 0.087 2 0.96 0.155 9 24.32 Probenecid/colchicine 19 1.84 4 1.10 10 2.11 0.239 3 1.44 0.730 1 5.26 Colchicine 377 36.46 113 31.04 191 40.30 0.006 77 37.02 0.145 5 1.33 Analgesia NSAIDs 376 36.36 143 39.29 149 31.43 0.018 85 40.87 0.711 5 1.33 Narcotics 119 11.51 34 9.34 57 12.03 0.210 29 13.94 0.107 3 2.52 Non-narcotic salicylates 44 4.26 18 4.95 15 3.16 0.202 11 5.29 0.857 - - Other 9 0.87 3 0.82 4 0.84 0.975 2 0.96 0.866 - - Corticosteroids 379 36.65 119 32.69 194 40.93 0.015 69 33.17 0.907 5 1.32 Hypertension/cardiovascular disease 728 70.41 242 66.48 345 72.78 0.049 150 72.12 0.164 10 1.37 Dyslipidemia 462 44.68 162 44.51 220 46.41 0.583 85 40.87 0.399 7 1.52 Diabetes 202 19.54 62 17.03 97 20.46 0.210 47 22.60 0.104 4 1.98 Obesity 8 0.77 1 2.75 4 8.44 0.258 3 1.44 0.182 - - RxRisk category c Anxiety and tension 172 16.63 53 14.56 83 17.51 0.252 37 17.79 0.309 - - Coronary/peripheral vascular 176 17.02 46 12.64 91 19.20 0.009 40 19.23 0.043 4 2.27 Depression 162 15.67 42 11.54 87 18.35 0.005 36 17.31 0.065 1 0.62 Diabetes 259 25.05 74 20.33 128 27.00 0.023 65 31.25 0.005 4 1.54 Epilepsy 96 9.28 26 7.14 50 10.55 0.082 24 11.54 0.092 2 2.08 Gastric acid disorder 261 25.24 71 19.51 136 28.69 0.002 56 26.92 0.040 4 1.53 Heart disease/hypertension 563 54.45 189 51.92 264 55.70 0.278 117 56.25 0.319 4 0.71 Hyperlipidemia 486 47.00 163 44.78 229 48.31 0.310 99 47.60 0.516 7 1.44 Hypertension 669 64.70 208 57.14 323 68.14 0.001 148 71.15 0.001 10 1.49 Irritable bowel syndrome 88 8.51 27 7.42 45 9.49 0.281 16 7.69 0.905 1 1.14 Thyroid disorder 142 13.73 40 10.99 78 16.46 0.021 24 11.54 0.841 3 2.11 Benefits information Febuxostat on formulary 517 50.00 154 42.31 279 58.86 < 0.001 95 45.67 0.436 6 1.16 Febuxostat rejection reason Prior authorization 170 16.44 62 17.03 66 13.92 0.215 43 20.67 0.280 2 1.18 Drug coverage 259 25.05 108 29.67 80 16.88 < 0.001 68 32.69 0.452 2 0.77 Step therapy 369 35.69 156 42.86 138 29.11 < 0.001 80 38.46 0.305 4 1.08 Quantity limits and other limits 183 17.70 19 5.22 159 33.54 < 0.001 13 6.25 0.607 2 1.09 Other reasons 53 5.13 19 5.22 31 6.54 0.418 4 1.92 0.029 - - a Categories not mutually exclusive. b T-tests were used to compare groups with the Filled Nothing reference group; however, a t-test for Filled Probenecid or Colchicine/Probenecid was not conducted due to small sample size. c Categories not mutually exclusive. NSAID = nonsteroidal anti-inflammatory drug; Rx = prescription. www.amcp.org Vol. 19, No. 9 November/December 2013 JMCP Journal of Managed Care Pharmacy 777

TABLE 2 Total Pharmacy Costs and Number of Drug Claims in 6-Month Pre-Index Period (N = 1,034) a Member Count Total pharmacy costs in the pre-index period Number of drug claims in the pre-index period All Patients Filled Nothing Within 30 Days Filled Febuxostat Within 30 Days N Mean Median SD N Mean Median SD N Mean Median SD P Value b 1,034 $1,390 $714 $2,028 364 $988 $510 $1,543 474 $1,718 $878 $2,353 <0.001 1,034 23 18 19 364 18 16 15 474 25 20 20 <0.001 Filled Allopurinol Within 30 Days Filled Probenecid or Colchicine/Probenecid Within 30 Days Member Count N Mean Median SD P Value b N Mean Median SD Total pharmacy costs in 208 $1,293 $712 $1,784 0.040 10 $3,049 $2,696 $2,633 the pre-index period Number of drug claims in the pre-index period 208 26 20 22 <0.001 10 39 38 15 a Groups not mutually exclusive (i.e., a patient may have filled 1 chronic gout agent within 30 days of a febuxostat rejected claim). b T-tests were used to compare groups with the Filled Nothing reference group; however, a t-test for Filled Probenecid or Colchicine/Probenecid was not conducted due to small sample size. SD = standard deviation. P = 0.043). Other variables were not found to be statistically significant. Table 4 provides the final regression model for whether patients filled any chronic gout medication within 1 month of a rejected febuxostat claim. Interaction terms were not found to be statistically significant and were not included in the final model. Patients were 2.7 times as likely to have filled a prescription for a chronic gout agent if they had at least 40 drug claims in the pre-index period compared with patients with 0 to 9 claims (OR = 2.71; P = 0.0108), but less likely if they had a projected febuxostat copay of $40 to $59 (OR = 0.51; P = 0.0366) or $100 to $149 (OR = 0.408; P = 0.0028) compared with a projected copay of $0 to $19. Of the comorbidities included in the model, only hypertension was identified as statistically significant, with increased likelihood of filling a chronic gout medication prescription (OR = 1.56; P = 0.0143). Other variables were not found to be statistically significant. Discussion The results of this analysis show that while 46% (n = 474) of patients ultimately filled a febuxostat prescription within 30 days after having one febuxostat rejection, 35% (n = 364) of patients did not receive any chronic gout medication within a month of the rejected claim. Within the study population, 31% was not using febuxostat as first-line therapy in chronic gout management, as they had had at least 1 prescription claim for allopurinol in the 6-month pre-index period (Table 1). These patients may have had inadequate urate-lowering effect with prior chronic gout medication, were intolerant to gout medications, or may have received inadequate allopurinol doses and were subsequently prescribed febuxostat. Utilization management strategies, such as preferred drug lists, prior authorization and step therapy, are intended to shift use of medications from more costly agents to less expensive alternatives for pharmacy cost savings; 12 however, a third of patients who experienced benefit restrictions on febuxostat did not receive either febuxostat or a less expensive alternative, such as allopurinol. Rejected febuxostat claims comprise 37% of the total submitted febuxostat claims volume for MedImpact in the 1-year period following its market availability; thus, a significant proportion of patients prescribed febuxostat do not receive chronic gout treatment [data on file]. Benefit restrictions can be overcome and nonpreferred medications can be available to patients. Prior authorization, step therapy, and other policies encourage more judicious use of costly, new, or potentially toxic drugs; however, patients who do not know that their prescriptions requires prior authorization until they arrive at the pharmacy might give up attempting to fill the prescription instead of seeking authorization. 13 These patients may continue to fill their prescriptions for acute gout, such as colchicine, analgesics, and corticosteroids. Because acute treatments do not treat the underlying cause of hyperuricemia and crystal formation, asymptomatic hyperuricemia can progress to acute and recurrent gout flares and then to chronic tophaceous gout, which can cause persistent pain and nerve compression syndrome, skin ulcerations, and can lead to joint destruction and deformities. 2 The literature on medication utilization following a claim rejection is sparse, but it appears that there is variability in response to benefit restrictions depending on the type of medication, disease, or unknown factors related to the pharmacy or adjudication process. The percentage of patients who did not receive any chronic gout medication following a febuxostat claim rejection (35%) is lower than the 46% of patients who did not initiate smoking cessation pharmacotherapy within 6 months of having a rejected claim for varenicline, as seen in the Zeng et al. (2010) study, 14 and substantially higher than the proportion of patients who did not receive any antihypertensive medications (7%) within 12 months of having a rejected claim for an angiotensin receptor blocker, as seen in 778 Journal of Managed Care Pharmacy JMCP November/December 2013 Vol. 19, No. 9 www.amcp.org

TABLE 3 Logistic Regression Model to Identify Factors Associated with Patients Who Filled a Febuxostat Claim Within 1 Month of Rejection Dependent Variable: Filled Febuxostat (Yes/No) Independent Variable Co-efficient P Value (ChiSq) Odds Ratio OR Lower 95% OR Upper 95% Female 0.033 0.851 1.034 0.731 1.461 Age group 18-39 -0.125 0.704 0.883 0.463 1.681 40-64 -0.267 0.161 0.765 0.527 1.112 65 (reference) Business type Commercial (reference) Medicaid 0.045 0.891 1.046 0.553 1.976 Self-insured 0.642 0.049 1.9 1.002 3.603 Medicare -0.177 0.432 0.838 0.539 1.303 Total Rx costs in the pre-index period $0-$199 (reference) $200-$699 0.122 0.629 1.13 0.688 1.855 $700-$1,099-0.006 0.985 0.994 0.533 1.853 $1,100-$1,799 0.318 0.328 1.374 0.727 2.599 $1,800 0.752 0.020 2.122 1.128 3.99 Number of Rx claims in the pre-index period 0-9 (reference) 10-19 0.067 0.791 1.069 0.651 1.756 20-29 0.331 0.277 1.393 0.767 2.53 30-39 0.104 0.766 1.109 0.559 2.2 40 0.399 0.301 1.491 0.7 3.175 Rejection reason Drug coverage (reference) Prior authorization 0.383 0.113 1.467 0.913 2.358 Quantity limit and other limits 2.578 <.0001 13.176 7.472 23.233 Step therapy 0.139 0.503 1.149 0.765 1.726 Other 1.171 0.001 3.224 1.649 6.303 Projected rejected febuxostat copay a $0-$19 (reference) $20-$39 0.196 0.503 1.217 0.685 2.162 $40-$59-0.567 0.057 0.567 0.316 1.018 $60-$99-0.675 0.159 0.509 0.199 1.303 $100-$149-0.733 0.008 0.48 0.281 0.823 $150-1.623 0.160 0.197 0.021 1.896 Pre-index use of analgesia Any analgesic -0.048 0.885 0.953 0.496 1.829 NSAIDs -0.430 0.177 0.65 0.348 1.214 Narcotics -0.085 0.770 0.918 0.519 1.624 Pre-index use of colchicine 0.314 0.043 1.369 1.011 1.854 Pre-index use of corticosteroids 0.265 0.091 1.304 0.958 1.774 RxRisk score by category (with 5% prevalence) Anxiety and tension -0.136 0.514 0.873 0.581 1.312 Coronary/peripheral vascular -0.248 0.251 0.78 0.511 1.192 Depression 0.260 0.227 1.296 0.851 1.974 Diabetes -0.108 0.559 0.898 0.625 1.289 Epilepsy 0.001 0.997 1.001 0.597 1.677 Gastric acid disorder 0.086 0.635 1.09 0.763 1.557 Heart disease and hypertension -0.091 0.609 0.913 0.646 1.292 Hyperlipidemia -0.098 0.563 0.907 0.651 1.264 Hypertension 0.237 0.206 1.268 0.877 1.832 Irritable bowel syndrome 0.008 0.976 1.008 0.607 1.674 Thyroid disease 0.269 0.220 1.309 0.852 2.011 a Projected rejected copay indicates whether the member has a closed formulary. Average wholesale price minus 15% was used as the copay amount; if the member had an open formulary then a best case scenario was assumed in which prior authorization restrictions were met and the member was assumed to have received the medication for a copay level for preferred brand copay. ChiSq = chi-square; NSAID = nonsteroidal anti-inflammatory drug; OR = odds ratio; Rx = prescription. www.amcp.org Vol. 19, No. 9 November/December 2013 JMCP Journal of Managed Care Pharmacy 779

TABLE 4 Logistic Regression Model to Identify Factors Associated with Patients Who Filled Any Chronic Gout Medication Within 1 Month of Rejection Dependent Variable: Filled Chronic Gout Medication (Yes/No) Independent variable Coefficient P Value (ChiSq) Odds Ratio OR Lower 95% OR Upper 95% Female -0.020 0.912 0.980 0.690 1.392 Age group 18-39 -0.032 0.921 0.969 0.518 1.812 40-64 -0.225 0.250 0.799 0.545 1.171 65 (reference) Business type Commercial (reference) Medicaid -0.438 0.152 0.645 0.354 1.175 Self-insured 0.115 0.722 1.122 0.595 2.113 Medicare -0.418 0.068 0.658 0.420 1.032 Total Rx costs in the pre-index period $0-$199 (reference) $200-$699-0.040 0.867 0.961 0.604 1.530 $700-$1,099-0.051 0.867 0.951 0.525 1.721 $1,100-$1,799-0.013 0.967 0.987 0.533 1.829 $1,800 0.251 0.429 1.285 0.690 2.392 Number of Rx claims in the pre-index period 0-9 (reference) 10-19 0.031 0.897 1.031 0.650 1.637 20-29 0.352 0.232 1.422 0.798 2.531 30-39 0.549 0.108 1.731 0.886 3.383 40 0.996 0.011 2.707 1.258 5.825 Rejection reason Drug coverage (reference) Prior authorization 0.092 0.694 1.097 0.692 1.738 Quantity limit and other limits 1.547 <0.001 4.697 2.617 8.433 Step therapy -0.238 0.229 0.788 0.535 1.161 Other 0.253 0.463 1.288 0.655 2.533 Projected rejected febuxostat copay a $0-$19 (reference) $20-$39-0.024 0.942 0.977 0.516 1.848 $40-$59-0.672 0.037 0.510 0.272 0.959 $60-$99-0.506 0.292 0.603 0.235 1.545 $100-$149-0.896 0.003 0.408 0.227 0.734 $150-1.424 0.130 0.241 0.038 1.522 Pre-index use of analgesia Any analgesic 0.124 0.719 1.132 0.575 2.230 NSAIDs -0.327 0.326 0.721 0.375 1.385 Narcotics -0.033 0.912 0.967 0.538 1.739 Pre-index use of colchicine 0.279 0.074 1.322 0.973 1.794 Pre-index use of corticosteroids 0.014 0.928 1.014 0.746 1.379 RxRisk score by category (with 5% prevalence) Anxiety and tension -0.057 0.788 0.945 0.626 1.428 Coronary/peripheral vascular -0.021 0.926 0.979 0.631 1.520 Depression 0.293 0.195 1.341 0.861 2.087 Diabetes 0.075 0.689 1.078 0.745 1.560 Epilepsy -0.013 0.962 0.987 0.571 1.705 Gastric acid disorder 0.123 0.508 1.131 0.786 1.628 Heart disease and hypertension -0.184 0.289 0.832 0.591 1.170 Hyperlipidemia -0.150 0.374 0.861 0.619 1.197 Hypertension 0.443 0.014 1.558 1.093 2.221 Irritable bowel syndrome -0.080 0.765 0.923 0.547 1.559 Thyroid disease 0.177 0.441 1.193 0.761 1.871 a Projected rejected copay indicates whether the member has a closed formulary. Average wholesale price minus 15% was used as the copay amount; if the member had an open formulary then a best case scenario was assumed in which prior authorization restrictions were met and the member was assumed to have received the medication for a copay level for preferred brand copay. ChiSq = chi-square; NSAID = nonsteroidal anti-inflammatory drug; OR = odds ratio; Rx = prescription. 780 Journal of Managed Care Pharmacy JMCP November/December 2013 Vol. 19, No. 9 www.amcp.org

a study by Yokoyama et al. (2007). 15 Analysis results indicate that the likelihood of filling a chronic gout prescription was not different for denials based on prior authorization or step therapy, compared with a drug not being covered, after adjusting for several patient and clinical characteristics (P = 0.694 and P = 0.229). Typically, these utilization management strategies require that the patient demonstrate a medical necessity for the nonpreferred agent (e.g., tried and failed a preferred agent or cannot tolerate a preferred agent). This practice is typically considered more stringent than quantity limits. Quantity limits and other limits were found to positively influence the fill of a chronic gout medication compared with a drug coverage rejection (OR = 4.697; P < 0.001). Studies on prior authorization and step therapy show that these utilization management strategies are effective in reducing pharmacy expenditure; however, whether the savings are offset by increased medical expenses can vary and may largely depend on the disease in question and available treatments. 16-21 A claims analysis reviewing step therapy for antidepressants showed that the days supplied and medication cost decreased, but overall and mental health-related inpatient and emergency room utilization and costs increased for employer plans. 16 A 1998 study found that common utilization management strategies for decreasing drug expenditures (e.g., restrictiveness of formulary) may be associated with higher severity-adjusted resource utilization for select conditions and that this pattern is more pronounced in the elderly. 17 However, studies for COX-2 inhibitors show savings in pharmacy expenditure without increases in medical expenditure. 18-21 To our knowledge, no published studies have examined the impact of utilization management strategies on the use of chronic gout treatment or its effect on overall health care utilization and costs. This is understandable, since treatment options for chronic gout management are limited, and until the introduction of febuxostat, all medications had long been available as generics. Utilization management strategies tend to encourage use of medications that have been available longer and are typically available as generics. 22 Studies suggest that health care costs were higher in patients with gout compared with patients without gout, 23 controlling for other patient and clinical characteristics, and that higher serum urate levels were associated with increased frequency and risk of gout episodes and higher health care cost and utilization. 24 However, it is not known whether utilization management strategies may result in savings in overall health care expenditures; further research is needed. Our results did not find pre-index use of analgesia or colchicine to influence the fill of a chronic gout treatment following a rejection of febuxostat (P = 0.719 and P = 0.074). Colchicine is considered a mainstay in acute gout treatment. 2,9-11 A study by Lauterio et al. (2010), which collected Internet responses from 160 physicians (primary care physicians [PCPs] = 107 and rheumatologists = 53) who treat gout patients, found that for inactive gout, urate-lowering therapy was more commonly initiated by rheumatologists than PCPs (66% vs. 47%, respectively), and PCPs tended to use high-dose regimens of colchicine rather than approved low-dose regimens. 25 The disease severity or pain intensity for patients in our study were unknown, and the prescriber s specialty was not assessed but may be a contributing factor of whether a patient fills a chronic gout medication subsequent to a claim rejection. Additionally, patients with at least 40 prescription claims in the 6 months prior to a rejected claim were more likely to fill a prescription for urate-lowering gout medication compared with those with 0 to 9 prescriptions (OR = 2.707; P = 0.011). These patients have more experience with prescription medications and may be more knowledgeable about utilization management strategies and thus more motivated to overcome benefit restrictions or may have more comprehensive insurance coverage. As there is large variability in the effect of different utilization management strategies based on the type of disease, treatment, and alternative options in addition to patient responses, these are important considerations in designing policy. 13 Limitations This study has several notable limitations. This is an observational analysis using pharmacy claims data and only included outcomes and variables that were measured. Thus, there may be unobserved factors that can potentially confound the analysis. Disease-related comorbidities may not be adequately included in the analysis as they were approximated based on use of prescription medications. Disease severity and prior medical use were not available for analysis. Also, it is not known whether these patients with a claim rejection without a subsequent fill of a chronic gout medication may have filled febuxostat but paid out of pocket for the medication or received office samples, in which case a claim for the medication would not be captured. As NSAIDs at lower dosage strengths are widely available over the counter and typically not covered by non-medicaid payers, this analysis underestimates use of these agents as adjunct therapy. Lastly, rejected febuxostat claims comprise 37% of the total submitted febuxostat claims volume for MedImpact in the 1-year period following its market availability; thus, this study does not assess use by all patients receiving febuxostat. Conclusion This study identified a treatment gap in the use of chronic gout therapy in the presence of utilization management methods, in that 35% of patients with a denied febuxostat claim did not fill a prescription for any chronic gout medication within a month of the claim denial. These findings are important considerations for structuring benefit design for gout treatment. www.amcp.org Vol. 19, No. 9 November/December 2013 JMCP Journal of Managed Care Pharmacy 781

Authors SHARON M. WANG, PharmD, MS, is Manager, Specialty Clinical Pharmacy Services; SARA GAO, MS, is Strategic Business Analyst; and BIMAL V. PATEL, PharmD, MS, is Director, Health Outcomes Research, MedImpact Healthcare Systems, Inc., San Diego, California. GUSTAVUS A. ARANDA, Jr., PharmD, MS, is Director, Health Services, Takeda Pharmaceuticals USA, Deerfield, Illinois. AUTHOR CORRESPONDENCE: Sharon M. Wang, PharmD, MS, MedImpact Healthcare Systems, Inc., 10181 Scripps Gateway Ct., San Diego, CA 92131. Tel.: 858.790.6196; E-mail: sharon.wang@medimpact.com. DISCLOSURES This study was sponsored by Takeda Pharmaceuticals USA, Inc. Aranda was an employee of Takeda, and Gao was an employee of MedImpact Healthcare Systems, Inc. at the time the research was conducted and the manuscript was prepared. Wang and Patel are employees of MedImpact Healthcare Systems. Wang and Aranda were responsible for study concept and design, and Gao did most of the data collection. Data interpretation and the first manuscript came primarily from Wang, assisted by Aranda and Patel. The manuscript was revised by Wang, Aranda, Gao, and Patel. Wang, Gao, and Patel had full access to the complete data in the study and take full responsibility for the integrity of the data and the accuracy of the data analysis. The manuscript was written in its entirety by the authors. ACKNOWLEDGMENTS The authors wish to thank Feng Zeng, PhD, Senior Health Economist at MedImpact, for his contribution to the concept and design of this study and Amanda Qiu, MS, Health Data Analyst III at MedImpact, for her assistance with data analysis for this study. References 1. Zhu Y, Pandya BJ, Choi HK. Prevalence of gout and hyperuricemia in the US general population: the National Health and Nutrition Examination Survey. Arthritis Rheum. 2011;63(10):3136-41. 2. Kim KY, Schumacher HR, Hunsche E, Wertheimer AI, Kong SX. A literature review of the epidemiology and treatment of acute gout. Clin Ther. 2003;25(6):1593-617. 3. Khanna D, Fitzgerald JD, Khanna PP, et al. 2012 American College of Rheumatology guidelines for management of gout. Part 1: systematic nonpharmacologic and pharmacologic therapeutic approaches to hyperuricemia. Arthritis Care Res (Hoboken). 2012;64(10):1431-46. 4. U.S. Centers for Disease Control and Prevention. Gout. Available at: http://www.cdc.gov/arthritis/basics/gout.htm. Accessed July 28, 2013. 5. Probenecid. Available at: http://www.drugs.com/pro/probenecid.html. Accessed July 28, 2013. 6. Schumacher HR Jr, Becker MA, Wortmann RL, et al. Effects of febuxostat versus allopurinol and placebo in reducing serum urate in subjects with hyperuricemia and gout: a 28-week, phase III, randomized, double-blind, parallel-group trial. Arthritis Rheum. 2008;59(11):1540-48. 7. Becker MA, Schumacher HR Jr, Wortmann RL, et al. Febuxostat compared with allopurinol in patients with hyperuricemia and gout. N Engl J Med. 2005;353(23):2450-61. 8. Fishman PA, Goodman MJ, Hornbrook MC, Meenan RT, Bachman DJ, O Keefe Rosetti MC. Risk adjustment using automated ambulatory pharmacy data: the RxRisk model. Med Care. 2003;41(1):84-99. 9. Jordan KM, Cameron JS, Snaith M, et al. British Society for Rheumatology and British Health Professionals in Rheumatology guideline for the management of gout. Rheumatology (Oxford). 2007;46(8):1372-74. 10. Zhang W, Doherty M, Pascual E, et al. EULAR evidence-based recommendations for gout. Part II. Management: report of a task force of the EULAR Standing Committee for International Clinical Studies Including Therapeutics (ESCISIT). Ann Rheum Dis. 2006;65(10):1312-24. 11. U.S. Department of Health and Human Services. Agency for Healthcare Research and Quality. National Guideline Clearinghouse. Guideline Summary NGC-7365: management of initial gout in adults. Available at: http://www.ngc.gov/content.aspx?id=14867. Accessed July 28, 2013. 12. Goldman DP, Joyce GF, Zheng Y. Prescription drug cost-sharing: associations with medication and medical utilization and spending and health. JAMA. 2007;298(1):61-69. 13. Soumerai SB. Benefits and risks of increasing restrictions on access to costly drugs in Medicaid. Health Aff (Millwood). 2004;23(1):135-46. 14. Zeng F, Chen CI, Mastey V, Zou KH, Harnett J, Patel BV. Utilization management for smoking cessation pharmacotherapy: varenicline rejected claims analysis. Am J Manag Care. 2010;16(9):667-74. 15. Yokoyama K, Yang W, Preblick R, Frech-Tamas F. Effects of a steptherapy program for angiotensin receptor blockers on antihypertensive medication utilization patterns and cost of drug therapy. J Manag Care Pharm. 2007;13(3):235-44. Available at: http://www.amcp.org/workarea/ DownloadAsset.aspx?id=7566. 16. Mark TL, Gibson TM, McGuigan K, Chu BC. The effects of antidepressant step therapy protocols on pharmaceutical and medical utilization and expenditures. Am J Psychiatry. 2010;167(10):1202-09. 17. Horn SD, Sharkey PD, Phillips-Harris C. Formulary limitations and the elderly: results from the Managed Care Outcomes Project. Am J Manag Care. 1998;4(8):1105-13. 18. Gleason PP, Williams C, Hrdy S, Hartwig SC, Lassen D. Medical and pharmacy expenditures after implementation of a cyclooxygenase-2 inhibitor prior authorization program. Pharmacotherapy. 2005;25(7):924-34. 19. Smalley WE, Griffin MR, Fought RL, Sullivan L, Ray WA. Effect of a prior-authorization requirement on the use of nonsteroidal anti-inflammatory drugs by Medicaid patients. N Engl J Med. 1995;332(24):1612-17. 20. Hartung DM, Touchette DR, Ketchum KL, Haxby DG, Goldberg BW. Effects of a prior-authorization policy for celecoxib on medical service and prescription drug use in a managed care Medicaid population. Clin Ther. 2004;26(9):1518-32. 21. Carroll NV, Smith JC, Berringer RA, Oestreich GL. Evaluation of an automated system for prior authorization: a COX-2 inhibitor example. Am J Manag Care. 2006;12(9):501-08. 22. Lichtenberg FR. The effect of access restrictions on the vintage of drugs used by Medicaid enrollees. Am J Manag Care. 2005;11:SP7-SP13. 23. Wu EQ, Patel PA, Yu AP, et al. Disease-related and all-cause health care costs of elderly patients with gout. J Manag Care Pharm. 2008;14(2):164-75. Available at: http://www.amcp.org/workarea/downloadasset.aspx?id=7826. 24. Wu EQ, Patel PA, Mody RR, et al. Frequency, risk, and cost of goutrelated episodes among the elderly: does serum uric acid level matter? J Rheumatol. 2009;36(5):1032-40. 25. Lauterio TJ, Cummings T, Davis MW. Recent prescribing habits of primary care physicans and rheumatologists for their patients with gout [conference poster abstract]. J Manag Care Pharm. 2010;16(7):545-46. Available at: http://www.amcp.org/workarea/downloadasset.aspx?id=8382. 782 Journal of Managed Care Pharmacy JMCP November/December 2013 Vol. 19, No. 9 www.amcp.org